Los Angeles

Date:
1981
Reference:
PP/CRI/E/1/29/18
Catalogue details

Licence: In copyright

Credit: Los Angeles. Source: Wellcome Collection.

    31/41
    he number of photographs i in the section is much The methods described above have been very broadly applied not only to light microscopic problems, but to electron microscopic problems as well. There are two commonly used methods for EM visualization. Pre- embedmenl staining (see references for excellent work by V. Pickel and colleagues) involves light glutaraldehyde fixation, 20 /im vibratome sectioning, PAP immunocy- tochemistry as described above, followed by more fixa tion, osmication, embedding, and visualization. This is an excellent method and allows for the demonstration of most antigens (especially peptides and proteins; see Figure 4). The second method is known as postembed- ment staining (Moriarty and Halmi, 1972; see references for G. Pelletier's careful methods). This approach starts with fixed, thin-sectioned, plastic embedded tissue. The surface of the plastic is etched away, followed by the PAP immunocytochemistry as described above. This is sections (i.e., for localization of two substances in the same cell). (See Figure 5 for a postembedment micro graph.) Both methods work for most peptides. In general, immunocytochemical techniques have been very widely applied in the neurosciences. A wide range of substances and systems have been studied with these methods (i.e., tyrosine hydroxylase, dopamine-/?- hydroxylase, glutamic acid decarboxylase, serotonin, leucine enkephalin, substance P, thyrotropin-releasing hormone, cholecystokinins, ß-endorphin, ACTH, soma tostatin, angiotensin II, vasoactive intestinal peptide, leuteinizing hormone-releasing hormone, etc.). Many laboratories have added to the list and expanded it to in- ; tù - Figure 4. Enkephalin-positive terminals in rat neostria tum (arrows). Tissue prepared by preembedment stain ing according to Pickel et al. (1980). Electron micro graph kindly provided by V. Pickel, Cornell University. postfixed in ,î Figure 5. /3-Lipotropin-positive fiber i leus. The tissue was fixed in 2.5% glutaraldehyde and ).5% OsO„. Positive fibers are indicated by lined processes by U. Prepared by post- itaining according to Pelletier et al. (1981). Electron micrograph kindly provided by G. Pelletier, Laval University. TRIALS AND TRIBULATIONS Despite all the substances and neuronal systems stud ied with immunocytochemistry, there are many prob lems. Some are merely procedural and others are fun damental (and therefore potentially fatal). The following is a partial list of such factors. Purity of immunizing antigen equals antibody specific ity. This may well be the issue in immunocytochemistry. An impure antigen used in immunization results in im pure antisera. If one is purifying a protein, it is likely iresent when the rabbit is •a agaii it their against the desired protein. Peptidi are often synthetically produced, resulting in a much- Radioimmunoassay vs. immunocytochemistry. In ra dioimmunoassay the antiserum can contain several an tibody populations directed against a large number of antigens and still function very well as an assay. In ra dioimmunoassay one deals only with the antibody pop ulation that binds to the labeled ligand (e.g., [ 125 I]/3-en- dorphin). In striking contrast, immunocytochemistry sees all (or most) antigens in tissue for which a sig nificant antibody population exists. Thus the antiserum against three or four neuronal antigens may be very sensitive and specific in radioimmunoassay, but very confusing for immunocytochemistry (it may show all four antigens!). Immunocytochemistry therefore re quires an antiserum with a single major antibody population. Controls for immunocytochemistry. There are several classes of controls for the specificity of an immunocyto chemical demonstration. The most critical and thor ough one involves adding 100- to 1000-fold molar ex cess of synthetic antigen to the antibody solution to prevent (or block) binding of the antibody to the tissue. In the ideal situation all originally stained structures will be unstained. It is rational to infer that the syn thetic antigen added to block the antiserum is identi cal—or at least structurally very similar—to the antigen bound in the tissue. There are other controls used in systems where synthetic antigens are scarce (i.e., most protein antibodies). These controls use preimmune or nonimmune serum to contrast with the primary im mune serum. This is generally a weak control because it leaves the question of specificity wide open. (It is un clear how the preimmune and specific serum differ. Is it antibodies against the antigen in question, or is it an tibodies against an unrelated antigen?) The set of technical controls involves use of the sec ond and third antisera (i.e., GAR, DAB, etc.) on sec tions without the use of a specific primary antiserum. Affinity purification of antisera. One approach to im proving the specificity of an antiserum involves the use of a column into which the antigen is linked in order to purify IgG molecules with significant affinity toward the antigen (March et al., 1974). This method allows the initial preparation of crude antisera, followed by an ex cellent clean-up step (i.e., an affinity column). Its suc cess intimately depends on pure antigen for the affinity This approach has the added benefit of concentrating specific antibodies while reducing the nonspecific an tibody content. In immunocytochemistry we are often faced with a good antigen-antibody reaction (perhaps at low titer) in the presence of high tissue background staining. Removal of the nonspecific IgG molecules re duces nonspecific binding to tissue and greatly reduces background. Thus the visible signal is greatly enhanced. Monoclonal antibodies. Most antisera are produced in whole animals, involving their entire immune system. It is currently believed that a single specific IgG (anti body) population comes from a group of genetically identical lymphocytes. This antibody population is called monoclonal because it is putatively derived from a tibody preparation with a very large number of lym phocyte clones (therefore called polyclonal). Serum preparations derive much of their power and many of their problems (i.e., cross reactivity) from their polyclo nal nature. For example, on a statistical basis a serum antibody is likely to have many middle-affinity popula tions and a few high-affinity populations. Often it is the high-affinity populations that are the important ones in radioimmunoassay and immunocytochemistry. A recent approach (cf. Meldere et al., 1978) to anti body production involves the immunization of a rat or mouse with the subsequent extraction of its lympho cytes. These cells are then fused with a cancerous cell line and grown in vitro. They are plated to separate in dividual clones and screened by many methods for their ability to produce specific antisera. The lymphocyte clones that produce the desired IgG population are subcloned and cultured in order to obtain large amounts of their specific IgG products, i.e., monoclonal With an impure antigen, after proper screening and subcloning, it is possible to produce an IgG species that reacts with a single molecular determinant. Thus the problems of multiple antibody populations are avoided. Yet the issues of genuine cross reactivity are not. That is, if two antigen molecules contain the same sequence, they should cross-react with the same IgG molecules (even if those are monoclonal in origin). A problem with monoclonal preparations is their affinity. As noted above, high-affinity antibody populations are much less common than middle-affinity ones. Yet most radioimmunoassay and immunocytochemical systems depend on the high-affinity component of a serum (i.e., a polyclonal preparation). On a purely probabilistic basis, most lymphocytes and their clones produce mid dle-affinity, not high affinity, IgG populations. The usual IgG screening strategy is for specificity, not affin ity. Thus one may have solved the cross reactivity prob- have the high affinity-high capacity seen in serum preparations. Families of antigens. A major source of confusion in immunocytochemistry of brain is the problem of fami lies of antigens cross-reacting with apparently specific antisera. To date, the best example can be found in the active sequence Tyr-Gly-Gly-Phe-Met (or Leu) seen in all five known opioid peptides. Another example is the common melanocyte-stimulating hormone (MSH) se quence in the ß-endorphin/a-MSH precursor that oc curs three times in this molecule (Nakanishi et al., 1979), as well as at least one other neuronal system in brain, the a-2 neuronal system (Watson and Akil, 1980). Since many antibodies against peptides are raised against fragments containing the active site, these anti- sera will often see any molecule using this active peptide sequence. For example, an antibody against a-neo-end- orphin (Kangawa et al., 1981; Tyr-Gly-Gly-Phe- ieu-Arg-Lys-Tyr-Pro-Lys) might well react with dynorphin (Goldstein et al., 1979; Tyr-Gly-Gly-Phe- Lew-Arg-Arg-Ile-Arg-Pro-Lys-Leu-Lys . . .) or even ^-endorphin (Li and Chung. 1976; Tyr-Gly-Gly-Phe- Afer-Thr-Ser-Glu-Lys . . .). Immunocytochemistry with crude antisera (polyclo nal) needs to be carefully studied for these complex cross reactivities. At times a simple competition-block-