One antigen, one gold? A quantitative analysis of immunogold labeling of plasma membrane 5'-nucleotidase in frozen thin sections

We present an evaluation of the efficiency of immunogold labeling for a low abundance plasma membrane protein. Several independent methods were used to determine the density of 5'-nucleotidase on the plasma membrane of the Fao cell. These methods include morphometry in combination with either enzymology or cell surface radiometric assay. Immunocytochemistry of frozen thin sections with either single or double layers of antibody and visualized with protein A complexed with 5 nm colloidal gold was used to estimate the same density. The application of a balance sheet to immunogold labeling demonstrates that the labeling is never quantitative. For example, labeling of the cell surface is always greater than labeling on the section. We show that departures from the "one antigen, one gold" ideal are systematic, so that an efficiency can be calculated and quantitative results can be obtained. The ability to obtain reliable quantitative results from immunogold labeling extends the utility of this already powerful technique.     

Optimization of immunogold labeling TEM: an ELISA-based method for evaluation of blocking agents for quantitative detection of antigen.

We developed an ELISA-based method for rapid selection of optimal blocking agents to be used in antigen quantification by immunogold labeling electron microscopy. Casein, skim milk, BSA from two sources, acetylated BSA, fish skin gelatin, horse serum, and goat serum were tested for their ability to block nonspecific binding of antibody to recombinant Vitreoscilla hemoglobin (VHb) antigen expressed in Escherichia coli cells by ELISA and the results were confirmed by quantitative immunogold labeling transmission electron microscopy (TEM). Ability to minimize NSB was also evaluated by dot-blot and Western blotting methods. The results demonstrated that ELISA was most accurate in predicting the most efficient blocking agent for TEM. Existing methods could not provide an accurate picture of the ability of various reagents to suppress background labeling. The sensitivity of detection of antigens by immunoelectron microscopy depends on the assay procedure being optimized to obtain the highest possible signal along with as low a background (noise) as possible. Our study indicated that an ELISA-based evaluation of various blocking agents could help in the rapid selection and optimization of a suitable protocol for immunogold localization and quantification of antigens by TEM.     

Optimization of immunogold labeling TEM. An ELISA-based method for rapid and convenient simulation of processing conditions for quantitative detection of antigen.

We developed an ELISA-based method for rapid optimization of various tissue processing parameters in immunogold labeling for electron microscopy. The effects of aldehyde fixation, tannic acid, postfixation, dehydration, temperature, and antigen retrieval on antibody binding activity of Vitreoscilla hemoglobin (VHb) expressed in E. coli cells were assayed by ELISA and the results confirmed by quantitative immunogold labeling transmission electron microscopy (TEM). Our results demonstrated that low concentrations (0.2%) of glutaraldehyde fixation caused minimal loss in total binding compared to higher concentrations. Dehydration in up to 70% ethanol resulted in some distortion of cellular ultrastructure but better antibody binding activity compared to dehydration up to 100%. Postfixation or incorporation of tannic acid in the primary fixative caused almost total loss of activity, whereas antigen retrieval of osmium-postfixed material resulted in approximately 90-100% recovery. The sensitivity of detection of proteins by immunogold labeling electron microscopy depends on the retention of antibody binding activity during tissue processing steps, e.g., fixation and dehydration. Our study indicated that an ELISA-based screening method of various tissue processing procedures could help in rapid selection and optimization of a suitable protocol for immunogold localization and quantification of antigen by TEM.     

Use of immunogold electron microscopy and monoclonal antibodies in the identification of nuclear substructures

A cytochemical technique for the ultrastructural localization of unique nuclear antigens is reported. Using a post-embedding indirect immunogold labeling procedure, nuclear antigens in electron-dense regions of the nucleus are localized with a minimum of nonspecific staining. Using this technique and indirect immunofluorescence, a panel of antinuclear monoclonal antibodies is shown to recognize preferentially cell cycle-dependent nuclear substructures. The antigenic domains recognized include specific regions in condensed chromatin, interchromatin granules, euchromatin, and chromosomes. The specificity of antigen recognition is demonstrated with qualitative and quantitative immunogold electron microscopy and immunoblot analysis. These results reveal the existence of previously undefined supramolecular organization within the nucleus and demonstrate the utility of the immunogold procedure when monoclonal antibodies are used.     

An artificial test substrate for evaluating electron microscopic immunocytochemical labeling reactions

We describe an artificial substrate system for optimization of labeling parameters in electron microscope immunocytochemical studies. The system involves use of blocks of glutaraldehyde-polymerized BSA into which a desired antigen is incorporated by a simple soaking procedure. The resulting antigen-impregnated artificial substrate can then be fixed and embedded identically to a piece of tissue. The BSA substrate can also be dried and then sectioned for immunolabeling with or without chemical fixation and without exposing the antigen to dehydrating agents and embedding resins. The effects of various fixation and embedding procedures can thus be evaluated separately. Other parameters affecting immunocytochemical labeling, such as antibody and conjugate concentration, can also be evaluated. We used this system, along with immunogold labeling, to determine quantitatively the optimal fixation and embedding conditions for labeling of hepatitis B surface antigen (HbsAg), human IgG, and horseradish peroxidase. Using unfixed and unembedded HBsAg, we were able to detect antigen concentrations below 20 micrograms/ml. We have shown that it is not possible to label HBsAg within resin-embedded cells using conventional aldehyde fixation protocols and polyclonal antibodies.     

Quantitative aspects of immunogold labeling in embedded and in nonembedded sections

We tried to control immunolabeling conditions so that information about antigen concentration could be achieved by quantifying labeling patterns. Working with immunogold labeling procedures in ultrathin cryosections, we observed that differential penetration of immunoreagents causes considerable differences in labeling efficiency between various cell structures. Therefore, in these nonembedded sections, labeling densities can only be used to measure variations in antigen concentration within one cell structure. After embedding the tissue in 30% polyacrylamide (PAA), differences in penetration were negated. The equalizing effect of PAA on the labeling efficiency enabled us to design a simple immunocytochemical method by which concentrations of a protein can be measured in situ at subcellular levels, provided that no variations in the protein's structural conformation occur that would affect its immunoreactivity. In spite of a higher sensitivity observed for Ig-gold, we preferred to use protein A-gold in our system because of the low nonspecific labeling and the more precise antigen detection by the latter immunomarker.     

Quantitation in immunocytochemistry: correlation of immunogold labeling to absolute number of membrane antigens

Baby hamster kidney cells infected with Semliki Forest virus were used as a model system for quantitative immunocytochemical labeling studies. In this system, a well-characterized membrane protein complex is present in different concentrations in three separate locations. Using immunogold labeling of cryosections, we compared the number of gold particles labeling the membranes of endoplasmic reticulum, Golgi stack, and fully formed virions at the plasma membrane to the biochemically determined concentrations. The efficiency of labeling was 40, 13, and 14% for the three structures, respectively. In a comparative study, Lowicryl K4M sections were found to give significantly lower levels of labeling.     

Ultrastructural study of protein kinase C-betaII localization during the cell cycle of human glioma cells

Transmission electron microscopy and immunogold labeling were used to determine how PKC-betaII is localized at stages in the cell cycle of the human glioma cell line U-373MG. Results show that immunogold particles in both dimethylsulfoxide (DMSO) and calphostin C (0.5 microM)-treated cells were mainly located in the cytoplasm. The concentration of gold particles in the nucleus was relatively small and constant throughout the cell cycle of both DMSO and calphostin C treated cells. Micrographs revealed changes in PKC-betaII during the cell cycle. The concentration of gold particles in the DMSO-treated cells was constant until 8 h. Subsequently, cytoplasmic PKC-betaII oscillated with an increased at 10 h, a rapid decrease at 12 h, and a rise at 14 h. The concentration of the gold particles then gradually decreased. In contrast, immunogold labeling in calphostin C-treated cells increased gradually up to 10 h. Subsequently, the pattern of PKC-betaII labeling in calphostin C-treated cells recapitulated those of control cells as seen by the rapid decline of PKC-betaII labeling at 12 h and its re-accumulation at 14 h. Additionally, there was a rapid increase at 20 h. Western blots of PKC-betaII showed constant PKC-betaII immunoreactivity throughout the cell cycle. In comparison to Western blots, in-situ immunogold labeling revealed changes in PKC-II immunoreactivity at 10 h and 14 h. This technique may represent intracellular immunoreactivity of PKC-betaII. The results from the immunogold labeling technique suggest that binding of calphostin C to the regulatory domain of PKC-betaII provokes a conformation change in PKC-betaII, preventing its activation and degradation.     

Use of immunogold preembedding technique to detect hepatitis A viral antigen in infected cells

Localization of virus and viral antigen in cell cultures infected with a rapidly replicating isolate of strain HM-175 of hepatitis A virus (HAV; pHM-175) was accomplished by using immunogold probes. Cells infected under one-step growth curve conditions were prefixed with 2% paraformaldehyde and 0.1-0.001% saponin at appropriate times postinfection for detection of maximum virus and viral antigen. An indirect labeling technique was employed using monoclonal antibody to HAV followed by 5 nm gold-antimouse IgG conjugate. Cells were then fixed by standard electron microscopy techniques and thin sectioned. This prefixation technique allowed penetration of the immunogold probes and moderate preservation of ultrastructure. Within infected cell cytoplasm, numerous antigenic sites were labeled with six to 200 gold particles. Two types of cells were infected with HAV and somewhat different results were obtained with the two cell types. In BS-C-1 cells, where a cytopathic effect (CPE) was not observed, myelin figures were immunogold labeled or frequently were located near immunogold-labeled sites. Vesicles containing viruslike particles (14-22 nm) were also observed. A significant observation in infected FRhK-4 cells was the presence of multivesicular bodies labeled with immunogold. Microfilaments were commonly seen near the multivesicular bodies. Our results demonstrate that the choice of prefixation method for immunogold labeling should be empirically determined for the cell type and condition.     

Detection of Cell Surface Antigens in Tissue Sections by Means of Pre-Embedding Immunogold Staining

The immunogold technique has been used in electron microscopy to detect cytoplasmic and extracellular antigens by postembedding techniques. It has also been used to detect plasma-membrane-associated molecules on suspended cells and, recently, to visualise cell surface antigens in ultrathin sections of Lowicryl embedded specimens. In the present study, cell surface antigens of rat kidney and human skin were identified in tissue sections by using pre-embedding immunogold labeling. Brush border microvillar antigens and dermal lymphocyte antigens both bound numerous gold particles. The immunogold staining described here has the advantage over immunoperoxidase procedures that it is not subject to diffusion or reabsorption artifacts, and allows estimation of the antigen density on labeled cells. Furthermore, this pre-embedding immunogold technique is ideally suited to detecting cell surface-associated antigens since it preserves antigenicity, allows gold particle penetration and enhances cell membrane profiles.     

Detection of cell surface antigens in tissue sections by means of pre-embedding immunogold staining

The immunogold technique has been used in electron microscopy to detect cytoplasmic and extracellular antigens by postembedding techniques. It has also been used to detect plasma-membrane-associated molecules on suspended cells and, recently, to visualize cell surface antigens in ultrathin sections of Lowicryl embedded specimens. In the present study, cell surface antigens of rat kidney and human skin were identified in tissue sections by using pre-embedding immunogold labeling. Brush border microvillar antigens and dermal lymphocyte antigens both bound numerous gold particles. The immunogold staining described here has the advantage over immunoperoxidase procedures that is not subject to diffusion or reabsorption artifacts, and allows estimation of the antigen density on labeled cells. Furthermore, this pre-embedding immunogold technique is ideally suited to detecting cell surface-associated antigens since it preserves antigenicity, allows gold particle penetration and enhances cell membrane profiles.     

Immunogold detection of co-localized neuropeptides: methodological aspects.

Whatever the protocol used, electron microscopic immunogold detection still suffers from a lack of sensitivity. In rat supraoptico-posthypophyseal neurons, unlabeled secretory granules are always detectable after electron microscopic immunocytochemistry, and their real status remains questionable. To improve the sensitivity of this approach, we assessed a protocol to visualize either one or the other of co-localized neuropeptides, i.e., vasopressin or galanin, after two successive rounds of immunogold with the same primary antibody performed on both faces of the grid. The use of different-sized gold particles enabled us to visualize the respective contribution of each face of the section to the final labeling. Our results showed a moderate but significant increase in both the proportion of labeled granules and the labeling intensity. Although limited, this improvement of immunogold detection strengthens the relevance of quantitative studies at the electron microscopic level, likely to reveal fine variations of the neuron peptidergic content. However, this enhancement depended on the peptide studied. The present data confirmed a progressive decrease of vasopressin immunoreactivity, already suggested by the single-staining procedure, all along the hypothalamo-posthypophyseal tract. In contrast, labeling intensity for galanin remained steady. Finally, our double-face labeling supported a preferential routing of galanin-containing secretory granules towards dendrites.     

Quantitative studies of keratin expression with the post-embedding immunogold labeling method

Immunoreactivity of the 56.5 KD acidic (type I) keratin was localized ultrastructurally and quantified in normal human epidermis using the specific monoclonal antibody KL1 and post-embedding immunogold labeling. The protein was detected in keratin intermediate filament bundles of all suprabasal keratinocytes. Keratohyalin granules and desmosomal plaques were labeled only on the periphery, in regions where keratin filaments penetrate these structures. The 56.5 KD keratin immunoreactivity increased from the first suprabasal layer onwards and reached its maximum in the outmost spinous layer. A subsequent abrupt decrease of the specific immunogold labeling was observed in the granular layer. This low reactivity, which persisted also in the horny layer, may be partially explained by either protein degradation or masking of the antigenic sites by a filament-aggregating material occurring at these stages of keratinocyte terminal differentiation. Statistical comparison of the quantitative results obtained in various cell and tissue compartments revealed no significant differences between the background labeling levels observed in the basal layer of epidermis with KL1, a control monoclonal antibody, or the immunogold conjugate alone. Our results confirm the specificity of 56.5 KD keratin for terminally differentiating suprabasal keratinocytes and demonstrate the importance of appropriate control studies when a post-embedding immunogold labeling method is employed.     

Immunocytochemical assays of amylase and chymotrypsinogen in rat pancreas secretory granules. Efficacy of using immunogold-labeled ultrathin cryosections to estimate relative protein concentrations

An exploration was conducted as to whether the relative concentration of two intracellular proteins could be evaluated quantitatively from their labeling densities in ultrathin cryosections labeled with the immunogold technique. As a model rat pancreatic cells were used in which the content of amylase (Am) and chymotrypsinogen (Ch) was experimentally altered. Rats were fed either normal laboratory chow or food containing soybean trypsin inhibitor (STI), which affects the Am/Ch ratio in the tissues. The changes in Am and Ch protein levels and enzyme activities were measured biochemically in cell suspension homogenates or in zymogen granule fractions. Within 5 days a maximal change in the Am/Ch was observed as a result of adaptation to the STI diet. The Am/Ch ratio determined biochemically was compared with that from counts of gold particles bound to the respective protein in immunogold-labeled cryosections. The two data sets matched fairly well, indicating that the intensity of the immunoreaction is a reliable reflection of antigen concentration in this system.     

Quantitative immunogold localization of protein phosphatase 2B (calcineurin) in Paramecium cells.

For immunogold EM labeling analysis, we fixed Paramecium cells in 4% formaldehyde and 0.125% glutaraldehyde, followed by low-temperature embedding in unicryl and UV polymerization. We first quantified some obvious but thus far neglected side effects of section staining on immunogold labeling, using mono- or polyclonal antibodies (Abs) against defined secretory and cell surface components, followed by F(ab)(2)- or protein A-gold conjugates. Use of alkaline lead staining resulted in considerable rearrangement and loss of label unless sections were postfixed by glutaraldehyde after gold labeling. This artifact is specific for section staining with lead. It can be avoided by staining sections with aqueous uranyl acetate only to achieve high-resolution immunogold localization of a protein phosphatase on unicryl sections. In general, phosphatases are assumed to be closely, although loosely, associated with their targets. Because the occurrence of protein phosphatase 2B (calcineurin) in Paramecium has been previously established by biochemical and immunological work, as well as by molecular biology, we have used Abs against mammalian CaN or its subunits, CaN-A and CaN-B, for antigen mapping in these cells by quantitative immunogold labeling analysis. Using ABs against whole CaN, four structures are selectively labeled (with slightly decreasing intensity), i.e., infraciliary lattice (centrin-containing contractile cortical filament network), parasomal sacs (coated pits), and outlines of alveolar sacs (subplasmalemmal calcium stores, tightly attached to the cell membrane), as well as rims of chromatin-containing nuclear domains. In other subcellular regions, gold granules reached densities three to four times above background outside the cell but there was no selective enrichment, e.g., in cilia, ciliary basal bodies, cytosol, mitochondria, trichocysts (dense-core secretory organelles), and non-chromatin nuclear domains. Their labeling density was 4- to 8.5-fold (average 6.5-fold) less than that on selectively labeled structures. Labeling tendency was about the same with Abs against either subunit. Our findings may facilitate the examination of molecular targets contained in the selectively labeled structures. (J Histochem Cytochem 48:1269-1281, 2000)     

Apical to basolateral surface area ratio and polarity of MDCK cells grown on different supports.

We have established conditions under which Madin-Darby canine kidney cells develop a well-polarized monolayer on polycarbonate filters and on transparent filters. These filters have biochemical and mechanical advantages over the nitrocellulose filters which have been widely used. Transepithelial resistance was established 10 h after plating and stabilized after 24 h. The distribution of protein antigens was followed by surface immunofluorescence and quantitated by a surface immunoassay that we developed. Uvomorulin was localized to the lateral membrane, with low amounts detectable on the basal membrane. The 58-kDa antigen was distributed over the entire basolateral domain, including cell processes extending into the filter pores. This distribution was confirmed by immunogold labeling of frozen sections. The 114-kDa antigen was found to be present at similar surface densities on both the apical and the basolateral domain. The support used for growth had profound effects on the cell morphology. A morphometric analysis of the plasma membrane of both strains of the cell line showed an increase in the number and size of the microvilli, and a smoother basal membrane as compared to published data on nitrocellulose filters. The apical to basolateral surface area ratio was therefore modified.     

Toxoplasma gondii: immunocytochemistry of four immunodominant antigens with monoclonal antibodies

Four monoclonal antibodies in which diagnostic usefulness has been observed, concerning congenital, acquired, and reactivated toxoplasmosis, were raised against Toxoplasma gondii tachyzo?tes in order to localize immunodominant antigens. On immunoblots, it appears that McAb IV47, McAB GII9, McAb II38, and McAb IE10 identify families of proteins with estimated molecular weights of 28-30 kDa, 30 kDa, 45-50 kDa, and 66-70 kDa, respectively. By immunogold preembedding techniques one can observe an homogeneous labeling of the outer pellicle of the tachyzo?tes with the McAb GII9 and IV47 and a light labeling with the McAb II38 and IE10. The three-dimensional observation of cell surface antigens is performed by applying a modified metal extraction replica method, i.e., A plasma polymerization method of glow discharge by Tanaka (1979). By immunogold preembedding techniques [with saponin permeabilization (0.1%)], and by immunogold postembedding techniques, a labeling of the rhoptries is observed with McAb GII9 and McAb IV47 but essentially all label is found with McAb II38 and IE10. With McAb GII9 a uniform labeling is observed on the cell surface. By immunoenzymatic techniques (peroxidase) a cell surface labeling is observed with the four McAb. Intracellular Toxoplasma, the outer pellicle, and the vesicles of the network (elaborated by Toxoplasma in parasitophorous vacuole) are also labeled with McAb IE10. These results indicate that McAb GII9 recognizes antigens of the antigen family (P 30) located on the cell surface and in the rhoptries. The antigen recognized by McAb IV47 is essentially located on and beneath the Toxoplasma cell surface membrane, and McAb II38 and IE10 identify preferentially rhoptry proteins.     

Usefulness of the immunogold technique in quantitation of a soluble protein in ultra-thin sections

We used a model system to study whether measurements of absolute local antigen concentrations at the electron microscopic level are feasible by counting immunogold labeling density in ultra-thin sections. The model system consisted of a matrix of a variable concentration of gelatin, which was mixed with given concentrations of rat pancreas amylase and fixed according to various fixation protocols. With a relatively mild fixation, there was no clear proportionality between anti-amylase gold labeling and amylase concentration in ultra-thin cryosections. This was presumably due to uncontrolled loss of amylase from the sections. After stronger fixation with 2% glutaraldehyde for 4 hr, labeling density reflected the amylase concentration very well. We observed that matrix (gelatin) density influenced labeling density. A low gelatin concentration of 5% allowed penetration of immunoreagents into the cryosection, resulting in a high and variable labeling density. In gelatin concentrations of 10% and 20%, labeling density was lower but proportional to amylase concentration. To establish an equal (minimal) penetration of immunoreagents, we embedded model blocks with different matrix densities in polyacrylamide (PAA). In ultra-thin cryosections of these PAA-embedded blocks, anti-amylase labeling was proportional to amylase concentration even at a low (5%) gelatin concentration. Anti-amylase labeling in ultra-thin sections from Lowicryl K4M low temperature-embedded blocks was higher than in PAA sections, but the results were less consistent and depended to some extent on matrix density. These results, together with the earlier observation that acrylamide completely penetrates intracellular compartments (Slot JW, Geuze HJ: Biol Cell 44:325, 1982), demonstrate that it is possible to measure true intracellular concentrations of soluble proteins in situ using ultra-thin cryosections of PAA-embedded tissue.     

Statistical evaluation of colocalization patterns in immunogold labeling experiments

The ultrastructural localization of various antigens in a cell using antibodies conjugated to gold particles is a powerful instrument in biological research. However, statistical or stereological tools for testing the observed patterns for significant clustering or colocalization are missing. The paper presents a method for the quantitative analysis of single or multiple immunogold labeling patterns using interpoint distances and tests the method using experimental data. The clustering or colocalization of gold particles was detected using various characteristics of the distribution of distances between them. Pair correlation and cross-correlation functions were used for exploratory analysis; second order reduced K (or cross-K) functions were used for testing the statistical significance of observed events. Confidence intervals of function values were estimated by Monte Carlo simulations of the Poisson process for independent particles, and results were visualized in histograms. Furthermore, a suitability of K functions modified by censoring or weighting was tested. The reliability of the method was assessed by evaluating the labeling patterns of nascent DNA and several nuclear proteins with known functions in replication foci of HeLa cells. The results demonstrate that the method is a powerful tool in biological investigations for testing the statistical significance of observed clustering or colocalization patterns in immunogold labeling experiments.     

Pemphigus vulgaris antigen, a desmoglein type of cadherin, is localized within keratinocyte desmosomes

Pemphigus vulgaris antigen (PVA) is a member of the desmoglein subfamily of cadherin cell adhesion molecules. Because autoantibodies in this disease cause blisters due to loss of epidermal cell adhesion, and because desmoglein is found in the desmosome cell adhesion junction, we wanted to determine if PVA is also found in desmosomes. By immunofluorescence, PV IgG bound, in a dotted pattern, to the cell surface of cultured human keratinocytes induced to differentiate with calcium, suggesting junctional staining. However, by preembedding, immunogold electron microscopic studies only slight labeling could be detected in desmosomes, presumably because of difficulty in gold penetration of intact desmosomes. We therefore treated the keratinocytes with 0.01% trypsin in 1 mM calcium, conditions known to preserve cadherin antigenicity but that caused slight separation of desmosomes, before immunogold staining. In this case there was extensive labeling of the extracellular part of desmosomes but not of the interdesmosomal cell membrane which was stained with anti-beta 2- microglobulin antibodies. To confirm the specificity of this binding we showed that antibodies raised in rabbits against the extracellular portions of PVA also bound desmosomes in these cultures. In intact mouse epidermis we could also show slight, but specific, immunogold desmosomal labeling with PV IgG. Furthermore, neonatal mice injected with PV IgG affinity purified on PVA showed desmosomal separation with the IgG localized to desmosomal cores. These results indicate that PVA is organized and concentrated within the desmosome where it presumably functions to maintain the integrity of stratifying epithelia.