evidence-based practices in applied ...

A HISTORICAL VIEW OF IMMUNOHISTOLOGY, WITH AN ORIENTATION TOWARDS QUALITYASSURANCE

Mark R. Wick, M.D.

Wick 2 Immunohistochemistry (IHC) has existed as an area of scientific inquiry for 70 years, and it truly has changed the way in which anatomic pathology is practiced during that span of time (1,2). Until the availability of IHC, scientists and physicians were limited in their ability to identify cellular products in-situ. Conventional histochemistry antedated IHC by almost a century, and had itself been a huge technological breakthrough (3-6). Nevertheless, purely-histochemical methods were limited in their capacity to identify many cellular products that may have diagnostic, prognostic, or predictive value in the practice of Medicine. The same comment can be made for another adjunctive histomorphological procedure, transmission electron microscopy (TEM) (7-10), which had been introduced by Ernst Ruska—a physicist—in 1931 (11). Nonetheless, there is little doubt that TEM likewise extended the analytic potentials of light microscopy and histochemistry significantly. Surprisingly, and rather inexplicably, both TEM and IHC were only slowly integrated into the clinical (hospital-based) practice of pathology. Ultrastructural analysis was still being “introduced” as a useful procedure for patient care 50 years after its inception (8), and IHC did not enjoy widespread interest or application by practitioners until around 1980 (2). Perhaps because of this protracted evolution, relatively little attention was given, until relatively recently, to the role of quality-assurance in either TEM or diagnostic IHC (DIHC). In particular, pathologists and other physicians especially tended to have a naïve expectation that immunostains were merely formulaic—in other words, if one used appropriate reagents and followed prescribed procedural steps, an optimal result was expected to obtain. That attitude likely derived from experience with histochemistry, where such provisions would typically produce the expected outcome. As we shall consider shortly, DIHC is anything but mechanical. Many biological and chemical factors have a meaningful impact on the final results of this method, and these must be addressed individually wherever possible. The concept of adding a detectable chemical “tag” to target-specific reagent antibodies seems today to be a straightforward, if not simple, idea. Nevertheless, such a conclusion is purely contextual. In 1940, the structure of antibodies was only rudimentarily understood, and the notion of attaching a visible chromophore to them was completely novel. A 27 year-old medical resident from Boston, MA—Dr. Albert Hewett Coons—developed the idea while on vacation in Europe (12). At least one German colleague—Dr. Kurt Apitz, of Charite’ Hospital in Berlin—thought little of it, for good reasons (12). The necessary process of joining chemicals to antibodies had never been attempted, and synthesis of the chemical “tags” that Coons had in mind—fluorescent molecules—also was a fledgling area. Finally, no microscope capable of visualizing fluophores then existed. Undeterred, Coons returned to Boston to work out each of these problems. By 1941, he and his colleagues had demonstrated not only the feasibility but the applicability of fluorescent immunohistology for the localization of particular protein targets in human tissues (13,14). That development lauched the entire scientific discipline of immunohistochemistry and earned Coons the prestigious Albert Lasker Award in 1959 (15). Probably because immunofluorescence-microscopy does not allow for a simultaneous appreciation of morphological detail, the procedure did not enjoy widespread clinical use and was principally regarded as a research tool. There were exceptions to that statement, however, principally represented by the diagnostic use of fluophores in renal pathology and dermatopathology (16-19). Fundamentally, and especially in the practice of hospital-based pathology, an expanded application of DIHC depended on further development of visible chemical “partners” for reagent antibodies. The next step in this evolution was taken in the area of TEM, where it was realized that certain electrondense (and therefore visible) chemical moieties—such as ferritin, osmium, and gold salts—could be bound directly to reagent antibodies, as fluorescein isocyanate had been (20-23). Hence, those reagents provided another method for localizing protein targets in substrate tissues, but at an ultrastructural level. An additional development involved the use of a gold-protein-A adduct as an indicator in TEM, preceded by incubation of target tissues with unlabeled reagent antibody (n.b.: protein-A is a proteinaceous product of Staphylococcus aureus, and is capable of binding to the Fc portion of all immunoglobulins) (24). Once again, however, ultrastructural IHC was impractical for most practicing anatomic pathologists, because they did not have access to electron microscopes and the technique in question was quite tedious. Finally, in the late 1960s, Ludwig Sternberger & colleagues developed an effective immunohistological procedure that could be used with formalin-fixed, paraffinized tissue sections and the light microscope (25,26). Several molecules of horseradish peroxidase were bound as a bridge to the Fab portions of 2 antibodies, forming a “head-to-head” peroxidase-antiperoxidase (PAP) complex. The latter had 2 free Fc fragments, either one of which could link to the Fab portion of a secondary “bridge” antibody that was, in turn, also able to recognize a specific tissue-bound reagent antibody. Antibodies comprising the PAP complex were raised in the same animal hosts as those which produced the primary reagent antibodies, whereas the secondary “bridge” antibody derived from

Wick 3 another species. The most common early example of such a construct was a rabbit primary antibody-sheep antirabbit “bridge” antibody-rabbit PAP complex (26). Peroxidases are redox enzymes that catalyze reactions between electron donors and recipients, according to the following equations: ROOR' + electron donor (2 e-) + 2H+ → ROH + R'OH, or Acceptor + H2O2 → oxidized acceptor + H2O In the classical PAP technique, hydrogen peroxide (H2O2) is used as the electron donor, and 3-3’-diaminobenzidine tetrahydrochloride-—which forms a colored precipitate when oxidized—is the electron-recipient. The final result is a light-microscopic preparation in which brown-black labels mark the sites of specific primary antibody binding, where target proteins reside in the tissue. That construct is responsible for the slang-term “brown stains,” in reference to DIHC. Finally, pathologists and other scientists had a practical, relatively rapid (24 hr.), and ecumenical alternative technique to immunofluorescence microscopy that could be used in every-day practice. It seemingly remained only for an increasing number of specific primary antibodies to be raised and marketed, before the entire panoply of human proteins could be localized in tissue sections. Realities and limitations of the PAP technique soon lessened that grand expectation. It became evident that some proteinaceous targets existed in only low densities in various tissues. Moreover, the standard process of formalin-fixation and paraffin-embedding appeared to denature, mask, or cross-link some proteins in such a way that primary antibodies could not bind to them (27,28). Depending upon the particulars of tissue-procurement and processing, PAP stains for any given target in any given specimen might be strongly reactive, weakly positive, or altogether negative, in an unpredictable way. This chain of events brings us to a crucial watershed in the development of DIHC as a method, and philosophies about how the technique should be used. For practicing pathologists, the aim of immunohistology was, and is, to visualize molecular constituents of tissue that have diagnostic importance. Inexplicable variability in staining intensity, as mentioned in the previous paragraph, threatens that goal and was quickly recognized as a serious potential source of interpretative error. For example, if S100 protein were to be found in a metastatic undifferentiated large-cell malignancy, in the absence of keratin, the probable diagnosis would be one of metastatic melanoma. However, keratin might actually be present in the tumor cells but missed because of technical problems in tissue processing or immunohistochemical procedure. Keratin-positive, S100-positive neoplasms are represented by carcinomas that originate in selected sites (29). Therefore, failure to detect keratin—or other similarlydispositive markers—in such lesions would produce a significant mistake in the generation of categorical data. It is important to realize that quantification of results in diagnostic IHC is meaningful only in a binary context—i.e., immunostains are ideally either positive or negative. As a consequence of that premise, a cardinal objective for diagnostic pathologists became the maximization of specific immunolabeling through a variety of methods, while at the same time maintaining minimal background “noise” in IHC preparations (30). To a large extent, that intent and practice remains in place today. In line with the earlier comment that masking, degradation, or cross-linking of available epitopes may occur in formalin-fixed tissue, compensatory efforts at signal-amplification were two-pronged. One mode of attack on the problem was to devise ever more sensitive IHC techniques, in the hope of recognizing low levels of an available protein target. The best known of higher-sensitivity alternatives to the PAP procedure was developed in the late 1970s by Hsu and colleagues (31-33); namely, the avidin-biotin-peroxidase complex (ABC) procedure. That innovative technique capitalized on the ability to attach biotin molecules to secondary antibodies, and also the capacity to build large indicator-complexes which include avidin, biotin, and horseradish peroxidase. The latter composites can be attached to the biotinylated secondary antibody, which is, in turn, bound to the Fc portion of a specific primary reagent antibody. The result compounds the number of peroxidase molecules that are associated with any one protein target, far beyond the biochemical capability of the PAP technique. Therefore, an amplification of the immunostaining signal is the predictable outcome. Later variations on that theme included labeled streptavidin-biotin-peroxidase (LSAB), ABPAP (seriallycombined PAP and ABC procedures), and alkaline phosphatase-antialkaline phosphatase (APAAP) methods (3438), and, more recently, a paradigm in which approximately 20 secondary antibodies from more than one animal source are attached polymerically to a dextran backbone that also carries >100 peroxidase molecules (37,39). That approach, called dextran-polymer-based (EnvisionR) IHC, obviates the need for separate labeled secondary antibodies from differing animals. At the same time, it greatly increases final immunostaining intensity.

Wick 4 All of those approaches for signal maximization center on the notion of increasing the numbers of signal molecules that are bound to a target protein in tissue. They are all effective in visualizing low densities of antigens whose epitopes are still at least partially- open to bind to primary antibodies. However, what could be done about desired targets with completely “masked” or cross-linked epitopes? Trading, perhaps, on pathologists’ experiences in immunohematology—where it has been known for decades that controlled enzymatic digestion could unmask certain antigens on erythrocytes (40,41)—the same procedure was applied to paraffin sections in DIHC in the late 1970s. Pepsin, trypsin, proteinase-3, ficin, pronase, papain, and bromelain were, and still are, employed in this setting (42-47). Predictably, the results demonstrated that different enzymes affected various targets differently. In other words, one catalyst might enhance immunoreactivity for protein “A” but decrease labeling for protein “B.” Another offshoot of this work was the realization that certain classes of tissue constituents were routinely masked by formalin fixation; a prime example is represented by the intermediate-filament proteins, including keratin, vimentin, desmin, neurofilament, and glial fibrillary acidic protein (28). Those markers can only be visualized optimally using some type of unmasking procedure. The same statement applies to virtually all intranuclear proteins (48). The next major advance in this area of DIHC occurred in the early 1990s. Empirical experience showed that the controlled heating of paraffin sections, when immersed in ionic solutions in a microwave oven or a steamer, could accomplish the same results as proteolytic unmasking methods (49-55). Thus, the term “heat-induced epitope retrieval” (HIER) was coined. Today, this process is a de rigueur element of practical immunohistology. In similarity to enzymatic digestion, HIER augments the intensity of immunolabeling for some markers and decreases it for others, vis-à-vis IHC procedures that omit an unmasking step (55). But, what, exactly, is being “undone” by proteolysis or HIER? To this day, the answer to that question is still vague. Several hypotheses have been advanced to account for epitope-unmasking. These include the breakage of fixation-induced coupling of “irrelevant” but sterically-interfering large proteins to peptide epitopes; the abrogation of electrostatic, van der Waal-like charges between epitopes and Fab fragments of reagent antibodies; dissolution of cagelike calcium complexes around epitope sequences; and a reversal of Mannich reactions between proteins (56-59). The latter are organic reactions featuring the amino- alkylation of acidic protons, placed next to carbonyl groups during formaldehyde fixation (58). Other tissue fixatives have been evaluated as alternatives to formalin, including solutions based on methyl alcohol, ethyl alcohol, acetone, or combinations thereof (60-65). These are either expensive or unwieldy for use in routine hospital pathology; moreover, they produce their own peculiar alterations in epitope preservation and are not necessarily any “kinder” to certain target proteins than formalin. All of these considerations may seem arcane, but they are, in fact, key preanalytical and intra-analytical elements. It is not possible to obtain perfect control of the concentrations and pH of fixatives and buffers, the duration of fixation, the dimensions of tissue blocks and histologic sections used for IHC, the biological activities of various proteolytic enzymes, and the nature of heat-distribution during HIER procedures. We do not mean to say that pathologists must not try to accomplish that task, but a more realistic goal is to aim for an end-result of consistent and functionally-binary (positive or negative) results in DIHC. Undeniably, an element of artificiality accompanies that approach, because one externally controls the ranges of reactivity in any immunostaining procedure. Nevertheless, as long as technical parameters are maintained within narrow confines, even an artificial system can be effectively used diagnostically. The Canadian Association of Pathologists has recently published a set of guidelines that are useful in this specific context (66). They also include a discussion of proper “positive” and “negative” controls in DIHC, as well as a consideration of cross-validating techniques (see http://ajcp.ascpjournals.org/content/133/3/354.full). Furthermore, a series of other papers, written over a 20-year period, has also outlined methods for quality-assurance in diagnostic immunohistology (67-73).

Wick 5

REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18.

19. 20. 21. 22. 23. 24. 25. 26.

27.

Taylor CR: Immunohistological approach to tumor diagnosis. Oncology 1978; 35: 187-197. Taylor CR: Immunoperoxidase techniques: practical and theoretical aspects. Arch Pathol Lab Med 1978; 102: 113-121. Pearse AGE: A review of modern methods in histochemistry. J Clin Pathol 1951; 4: 1-36. Pearse AGE: Histochemistry and its application to the basic sciences. Lect Sci Basis Med 1955; 4: 358386. Lillie RD: Problems of fixation in histochemistry. J Histochem Cytochem 1958; 6: 301-302. Bennett HS: A perception of histochemistry. J Histochem Cytochem 1983; 31(Suppl): 127-130. Rosai J, Rodriguez HA: Application of electron microscopy to the differential diagnosis of tumors. Am J Clin Pathol 1968; 50: 555-562. Kuzela DC, True LD, Eiseman B: The role of electron microscopy in the management of surgical patients. Ann Surg 1982; 195: 1-11. Fisher C, Ramsay AD, Griffiths M, McDougall J: An assessment of the value of electron microscopy in tumor diagnosis. J Clin Pathol 1985; 38: 403-408. Ordonez NG, Mackay B: Electron microscopy in tumor diagnosis: indications for its use in the immunohistochemical era. Hum Pathol 1998; 29: 1403-1411. Ruska E: Ernst Ruska: Autobiography, Nobel Foundation Press, Stockholm, 1986. McDevitt HO: Albert Hewett Coons, 1912-1978, National Acadamies Press, New York, 1979. Coons AH, Creech HJ, Jones R: Immunological properties of an antibody containing a fluorescent group. Proc Soc Exp Biol 1941; 47: 200-202. Coons AH: The beginnings of immunofluorescence. J Immunol 1961; 87: 499-503. Lasker Foundation: 1959 Winners—Albert Lasker Basic Medical Research Award. (http://www.laskerfoundation.org/awards/1959basic.htm) (Accessed 7-12-2010). Beutner EH, Jordon RE: Demonstration of skin antibodies in sera of pemphigus vulgaris patients by indirect immunofluorescent staining. Proc Soc Exp Biol Med 1964: 117: 505-510. Tan EM, Kunkel HG: An immunofluorescence study of the skin lesions in systemic lupus erythematosus. Arthritis Rheum 1966; 9: 37-46. Nagasawa T, Miyakawa Y, Shibata S: New fluorescent staining method for renal biopsy: introduction of anti-glomerular basement membrane labeled antibody. Saishin Igaku (Modern Medicine) 1968; 23: 26562663. Wilson CB, Dixon FJ, Fortner JG, Cerilli GJ: Glomerular basement membrane-reactive antibody in antilymphocyte globulin. J Clin Invest 1971; 50: 1525-1535. Hall CE, Nisonoff A, Slayter HS: Electron microscopyc observations of rabbit antibodies. J Biophys Biochem Cytol 1959; 6: 407-412. Singer SJ, Schick AF: The properties of specific stains for electron microscopy prepared by the conjugation of antibody molecules with ferritin. J Biophys Biochem Cytol 1961; 9: 519-537. Slot JW, Posthuma G, Chang LY, Crapo JD, Geuze HJ: Quantitative aspects of immunogold labeling in embedded and in nonembedded sections. Am J Anat 1989; 185: 271-281. Sternberger LA: Electron microscopic immunocytochemistry: a review. J Histochem Cytochem 1967; 15: 139-159. Roth J, Heitz PU: Immunolabeling with the protein A-gold technique: an overview. Ultrastruct Pathol 1989; 13: 467-484. Sternberger LA, Cuculis JJ: Method for enzymatic intensification of the immunocytochemical reaction without use of labeled antibodies. J Histochem Cytochem 1969; 17: 190. Sternberger LA, Hardy PH Jr, Cuculis JJ, Meyer HG: The unlabeled antibody-enzyme method of immunohistochemistry: preparation and properties of soluble antigen-antibody complex (horseradish peroxidase-anti-horseradish peroxidase) and its use in identification of spirochetes. J Histochem Cytochem 1970; 18: 315-333. Tornehave D, Folkersen J, Teisner B, Chemnitz J: Immunohistochemical aspects of immunological crossreaction and masking of epitopes for localization studies on pregnancy-associated plasma protein A. Histochem J 1986; 18: 184-188.

Wick 6 28. DeLellis RA, Kwan P: Technical considerations in the immunohistochemical demonstration of intermediate filaments. Am J Surg Pathol 1988; 12(Suppl): 17-23. 29. Drier JK, Swanson PE, Cherwitz DL, Wick MR: S100 protein immunoreactivity in poorly-differentiated carcinomas: immunohistochemical comparison with malignant melanoma. Arch Pathol Lab Med 1987; 111: 447-452. 30. Anonymous: Signal-to-nose ratio. (http://en.wikipedia.org/wiki/signal-to-noise_ratio) (Accessed 7-122010). 31. Hsu SM, Raine L, Fanger H. The use of antiavidin antibody and avidin-biotin-peroxidase complex in immunoperoxidase technics. Am J Clin Pathol 1981; 75: 816-821. 32. Hsu SM, Raine L, Fanger H: Use of avidin-biotin-peroxidase complex (ABC) in immunoperoxidase techniques: a comparison between ABC and unlabeled antibody (PAP) procedures. J Histochem Cytochem 1981; 29: 577-580. 33. Guesdon JL, Ternynck T, Avrameas S: The use of avidin-biotin interaction in immunoenzyme techniques. J Histochem Cytochem 1979; 27: 1131-1136. 34. Bratthauer GL: The avidin-biotin complex (ABC) method and other avidin-biotin bindings methods. Methods Mol Biol 2010; 588: 257-270. 35. Miller RT, Groothuis CL; Improved avidin-biotin immunoperoxidase method for terminal deoxyribonucleotidyl transferase and immunophenotypic characterization of blood cells. Am J Clin Pathol 1990; 93: 670-674. 36. Elias JM, Margiotta M, Gaborc D: Sensitivity and detection efficiency of the peroxidase-antiperoxidase (PAP), avidin-biotin-peroxidase complex (ABC), and peroxidase-labeled avidin-biotin (LAB) methods. Am J Clin Pathol 1989; 92: 62-67. 37. Mokry J: Versatility of immunohistochemical reactions: comprehensive survey of detection systems. Acta Medica 1996; 39: 129-140. 38. Swanson PE, Kagen KA, Wick MR: Avidin-biotin-peroxidase-antiperoxidase (ABPAP) complex: an immunocytochemical method with enhanced sensitivity. Am J Clin Pathol 1987; 88: 162-176. 39. Kammerer U, Kapp M, Gassel AM, et al.: A new rapid immunohistochemical staining technique using the Envision antibody complex. J Histochem Cytochem 2001; 49: 623-630. 40. Masouredis SP, Sudora E, Mahan L, Victoria EJ: Quantitative immunoferritin microscopy of Fya, Fyb, Jka, U, and Dib antigen site numbers on human red cells. Blood 1980; 56: 969-977. 41. Ripoche J, Sim RB: Loss of complement receptor type 1(CR1) on aging of erythrocytes: studies of proteolytic release of the receptor. Biochem J 1986; 235: 815-821. 42. Andrade RE, Hagen KA, Swanson PE, Wick MR: The use of proteolysis with ficin for immunostaining of paraffin sections: a study of lymphoid, mesenchymal, and epithelial determinants in human tissues. Am J Clin Pathol 1988; 90: 33-39. 43. Hajdu I: The immunohistochemical detection of J-chain in lymphoid cells in tissue sections: the necessity of trypsin digestion. Cell Immunol 1983; 79: 157-163. 44. Dell’Orto P, Viale G, Colombi R, Braidotti P, Coggi G: Immunohistochemical localization of human immunoglobulins and lysozyme in epoxy-embedded lymph nodes: effect of different fixatives and of proteolytic digestion. J Histochem Cytochem 1982; 30: 630-636. 45. Miller RT, Swanson PE, Wick MR: Fixation & epitope retrieval in diagnostic immunohistochemistry: a concise review with practical considerations. Appl Immunohistochem Mol Morphol 2000; 8: 228-235. 46. Hiort O, Lwan PW, DeLellis RA: Immunohistochemistry of estrogen receptor protein in paraffin sections: effects of enzymatic pretreatment and cobalt chloride intensification. Am J Clin Pathol 1988; 90: 559-563. 47. Pileri S, Serra L, Martinelli G: The use of pronase enhances sensitivity of the PAP method in the detection of intracytoplasmic immunoglobulins. Basic Appl Histochem 1980; 24: 203-207. 48. Taylor CR, Shi SR, Chaiwun B, et al.: Strategies for improving the immunohistochemical staining of various intranuclear prognostic markers in formalin-paraffin sections: androgen receptor, estrogen receptor, progesterone receptor, p53 protein, proliferating cell nuclear antigen, and Ki-67 antigen revealed by antigen retrieval techniques. Hum Pathol 1994; 25: 263-270. 49. Shi SR, Key ME, Kalra KL: Antigen retrieval in formalin-fixed paraffin-embedded tissues: an enhancement method for immunohistochemical staining based on microwave oven heating of tissue sections. J Histochem Cytochem 1991; 39: 741-748. 50. Suurmeijer AJH: Micdrowave-stimulated antigen retrieval: a new method facilitating immunohistochemistry of formalin-fixed, paraffin-embedded tissue. Histochem J 1992; 24: 597.

Wick 7 51. Gown AM, deWever N, Battifora H: Microwave-based antigenic unmasking: a revolutionary new technique for routine immunohistochemistry. Appl Immunohistochem 1993; 1: 256-266. 52. Norton AJ, Jordan S, Yeomans P: Brief high temperature heat denaturation (pressure cooking): a simple and effective method of antigen retrieval for routinely-processed tissues. J Pathol 1994; 173: 371-379. 53. Cattoretti G, Suurmeijer AJH: Antigen unmasking on formalin-fixed paraffin-embedded tissues using microwaves: a review. Adv Anat Pathol 1995; 2: 2-9. 54. Miller RT, Estran C: Heat-induced epitope retrieval with a pressure cooker: suggestions for optimal use. Appl Immunohistochem 1995; 3: 190-193. 55. Pileri S, Roncador G, Ceccarelli C, et al.: Antigen retrieval techniques in immunohistochemistry: comparison among different methods. J Pathol 1997; 183: 116-123. 56. Bogen SA, Vani K, Sompuram SR: Molecular mechanisms of antigen retrieval: antigen retrieval reverses steric interference caused by formalin-induced cross-links. Biotech Histochem 2009; 84: 207-215. 57. Boenisch T: Heat-induced antigen retrieval: what are we retrieving? J Histochem Cytochem 2006; 54: 961-964. 58. Leong TY, Leong ASY: How does antigen retrieval work? Adv Anat Pathol 2007; 14: 129-131. 59. Werner M, Von Wasielewski R, Komminoth P: Antigen retrieval, signal amplification, and intensification in immunohistochemistry. Histochem Cell Biol 1996; 105: 253-260. 60. Puchtler H, Meloan SN: On the chemistry of formaldehyde fixation and its effects on immunohistochemical reactions. Histochemistry 1985; 82: 201-204. 61. Paterson DA, Reid CP, Anderson TJ, Hawkins RA: Assessment of estrogen receptor content of breast carcinoma by immunohistochemical techniques on fixed and frozen tissue and by biochemical ligandbinding assay. J Clin Pathol 1990; 43: 46-51. 62. Fisher CJ, Gillett CE, Vojtesek G, Barnes DM, Millis RR: Problems with p53 immunohistochemical staining: the effect of fixation and variation in the methods of evaluation. Br J Cancer 1994; 69: 26-31. 63. Battifora H, Kopinski M: The influence of protease digestion and duration of fixation on the immunostaining of keratins: a comparison of formalin and ethanol fixation. J Histochem Cytochem 1986; 34: 1095-1100. 64. Elias JM, Gown AM, Nakamura RM, et al.: Quality control in immunohistochemistry: report of a workshop sponsored by the Biological Stain Commission. Am J Clin Pathol 1989; 92: 836-843. 65. Werner M, Chott A, Fabiano A, Battifora H: Effect of formalin tissue fixation and processing on immunohistochemistry. Am J Surg Pathol 2000; 24: 1016-1019. 66. Torlakovic EE, Riddell R, Banerjee D, et al.: Canadian Association of Pathologists-Associatio canadienne des pathologists National Standards Committee/Immunohistochemistry: best practice recommendations for standardization of immunohistochemistry tests. Am J Clin Pathol 2010; 133: 354-365. 67. Reynolds GJ: External quality assurance and assessment in immunocytochemistry. Histopathology 1989; 15: 627-633. 68. Wick MR: Technologic anarchy? Am J Clin Pathol 1989; 91(Suppl): S1. 69. Swanson PE: HIERanarchy: the state of the art in immunohistochemistry. Am J Clin Pathol 1997; 107: 139-140. 70. Seidal T, Balaton A, Battifora H: Interpretation and quantification of immunostains. Am J Surg Pathol 2001; 25: 1204-1207. 71. Wick MR, Mills SE: Consensual interpretive guidelines for diagnostic immunohistochemistry. Am J Surg Pathol 2001; 26: 1208-1210. 72. Wick MR, Swanson PE: Targeted controls in clinical immunohistochemistry: a useful approach to quality assurance. Am J Clin Pathol 2002; 117: 7-8. 73. Shi SR, Liu C, Pootrakul L, et al.: Evaluation of the value of frozen tissue sections used as “gold standards” for immunohistochemistry. Am J Clin Pathol 2008; 129: 358-366.

DIAGNOSTIC ELECTRON MICROSCOPY: A HISTORICAL PERSPECTIVE

Mark R. Wick, MD

University of Virginia Medical Center, Charlottesville, VA ([email protected]; www.markwickmd.com)

The principal discoveries that led to the construction of the transmission electron microscope by Knoll and Ruska in 1931 were those of J.J. Thomson, who demonstrated that cathode rays are negatively charged electrons that could be deflected by an electric field, and of Louis de Broglie, who showed that a moving electron has an associated wavelength which is inversely proportional to its momentum. The optical resolution attainable by transmission electron microscopy (TEM) is much greater than that of a light microscope; it is 0.2 nm for TEM as opposed to 200 nm for a first-class light microscope. In 1934, Ladislos Marton of Brussels first published a paper that included an electron micrograph of a biological specimen, a section of a Drosera intermedia leaf. However,high resolution electron microscopic images of biological cells and tissues were not obtainable until better instruments and better methods of specimen preparations were developed during the decades from 1940 to 1970. These advances included the introduction of osmium tetroxide and glutaraldehyde; methacrylate and epoxy resins; glass and diamond knives; and "electrondense" lead and uranium salts for staining of TEM grids. After the first TEM description of a neoplastic cell in a rat by Porter and Thompson in 1947, several isolated reports on the ultrastructural features of human and animal tumors appeared from from 1953 to 1965. During that period, virions also were discovered in neoplastic cells from a variety of animals and in tumor cells from leukemic human patients. Although ultrastructural analyses of selected specific neoplasms were anecdotally performed in North America & Europe in the early 1960s, no published articles appeared in the general surgical pathology literature on this topic until 1968. At that time, Rosai & Rodriquez described several tumors in which TEM helped to resolve difficult diagnostic interpretations. Subsequent to that article in the American Journal of Clinical Pathology, a number of additional papers, reviews, and book chapters expounded on the application of TEM to tumor diagnosis. The first book entirely devoted to that subject was authored by Ghadially in 1980, followed by an illustrated monograph written by Erlandson in 1981. In addition to being an adjunct to the diagnosis of human tumors, electron microscopy has proven useful in the evaluation of various nonneoplastic human diseases such as glomerulonephritides, myopathies, viral infections, and storage disorders. Indeed, the latter conditions now account for the great bulk of cases for which ultrastructural studies are currently obtained. CONTRIBUTIONS AND LIMITATIONS OF DIAGNOSTIC ELECTRON MICROSCOPY Contributions 1. Today, transmission electron microscopy is most often used to help resolve difficult differential diagnoses. For example, is a high grade sarcoma a malignant fibrous histiocytoma, a monophasic synovial sarcoma, a leiomyosarcoma, or a malignant peripheral nerve sheath tumor? 2. Ultrastructural studies can still be useful in the evaluation of tumors with uninterpretable immunohistological findings, or those which for technical reasons are unsuitable for reliable immunohistochemical evaluation. 3. In several selected medical diseases such as myopathies, glomerulonephritides, storage diseases, and others, TEM continues to be the best implement for definitive diagnosis. Limitations 1. Many tumors cannot be examined adequately by TEM because of extensive degenerative changes in them prior to fixation. Improper sampling for ultrastructural study is also a problem. 2. In practical terms, only a small portion of any given tumor can be processed for TEM and therefore sampling bias is a definite issue. A limited amount of time also is available for examining specimens. Diagnostic electron microscopy requires a level of patience and diligence that unfortunately is lacking in many individuals, especially busy surgical pathologists. That is why diagnostic TEM often is relegated to inexperienced pathology residents, PhDs, or technicians. 3. Transmission electron microscopy has been disappointing as a means for establishing the lineage of differentiation for a number of neoplasms, such as alveolar soft part sarcoma, Ewing's sarcoma, epithelioid sarcoma, and synovial sarcoma.

REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36.

Azar HA. Pathology of human neoplasms. An atlas of diagnostic electron microscopy and immunohistochemistry. New York: Raven Press, 1988. Bernard W. The detection and study of tumor viruses with the electron microscope. Cancer Res 1960;20:712-727. Billingham ME, Bristow MR, Mason JW, Joseph U. Endomyocardial biopsy. In: Braunwald E, Mock MB, Watson JT, eds. Congestive heart failure. New York: Grune & Stratton, 1982:237-251. Bonikos DS, Bensch KG, Kempson RL. The contribution of electron microscopy to the differential diagnosis of tumors. Beitr PathoI1976;158:417-444. Damjanov I. Ultrastructural pathology of human tumors. Annual Research Reviews. Montreal: Eden Press, vol I, 1979;voI2, 1980. Dickersin GR. Diagnostic electron microscopy. A text/atlas. New York: Igaku-Shoin, 1988. Erlandson RA. Diagnostic transmission electron microscopy of human tumors. The interpretation of submicroscopic structures in human neoplastic cells. New York: Masson, 1981. Erlandson RA. Application of transmission electron microscopy to human tumor diagnosis: an historical perspective. Cancer Invest 1987; 5:487-505. Fine J-D. Epidermolysis bullosa. Clinical aspects, pathology, and recent advances in research. Int J DermatoI1986;25:143-157. Fisher C. The value of electron microscopy and immunohistochemistry in the diagnosis of soft tissue sarcomas: a study of 200 cases. Histopathology 1990; 16:441-454. Fisher C, Ramsay AD, Griffiths M, McDougall J. An assessment of the value of electron microscopy in tumor diagnosis. J Clin PathoI1985;38:403-408. Ghadially FN. Diagnostic electron microscopy of tumors. London: Butterworths, 1980; 2nd ed, 1985. Gillespie H. Ultrastructural diagnosis of large cell "undifferentiated" neoplasia. Diag HistopathoI1982;5:33-51. Gyorkey F, Min K-W, Krisko I, Gyorkey P. The usefulness of electron microscopy in the diagnosis of human tumors. Hum PathoI1975;6:421-441. Jenis EH, Lowenthal DT. Kidney biopsy interpretation. Philadelphia: FA Davis, 1977. Low FN, Freeman JA. Electron microscopic atlas of normal and leukemic human blood. New York: McGraw-Hili, 1958. Luse S. Ultrastructural characteristics of normal and neoplastic cells. In: Homburger F, ed. Progress in experimental tumor research, vol 2. New York: S. Karger, 1961: 1-35. Mackay B, ed. Introduction to diagnostic electron microscopy. New York: Appleton-Century-Crofts, 1981. Mackay B, Brunner JM, Ordonez NG. Electron microscopy in surgical pathology: I. Lab M ed 1988; 19(1): 13-17. Mackay B, Brunner JM, Ordonez NG. Electron microscopy in surgical pathology: II. Lab Med 1988; 19(2):78-83. Mackay B, Osborne BM. The contribution of electron microscopy to the diagnosis of tumors. Pathobiol Annu 1978; 8:359-405. Mackay B, Silva EG. Diagnostic electron microscopy in oncology. Pathol Annu 1980; 15(2):241-270. Mackay B, Srigley JR. Diagnostic electron microscopy: review of current applications. Lab Med 1983; 14:639-643. Marton L. Electron microscopy of biological objects. Nature 1934; 133: 911 Miller SE. Electron microscopy in rapid viral diagnosis. EMSA Bull 1989; 19(1):53-59. Palmer EL, Martin ML. Electron microscopy in viral diagnosis. Boca Raton, FL: CRC Press, 1988. Peven DR, Gruhn JD. The development of electron microscopy. Arch Pathol Lab Med 1985; 109:683-691. Rosai J, Rodriquez HA. Application of electron microscopy to the differential diagnosis of tumors. Am J Clin Pathol 1968; 50:555-562. Ruska E. The emergence of the electron microscope. Connection between realization and the first patent application, documents of an invention. J Ultrastruct Mol Struct Res 1986;95:3-28. Russo J, Tait L, Russo IH. Current basis for the ultrastructural clinical diagnosis of tumors: a review. J Electron Microsc Tech 1985;2:305-351. Schochet SS J r. Electron microscopy of skeletal muscle and peripheral nerve biopsy specimens. In: Mackay B, ed. Introduction to diagnostic electron microscopy. New York: Appleton-Century-Crofts, 1981:131-169. Silva FG, Pirani CL. Electron microscopic study of medical diseases of the kidney: update - 1988. Mod Patho11988; 1:292-315. Spargo BH, Seymour AE, Ordonez NG. Renal biopsy pathology with diagnostic and therapeutic implications. New York: John Wiley & Sons, 1980. Trump BF, Jones RT, eds. Diagnostic electron microscopy. New York: John Wiley & Sons, vol I, 1978; vol 2, 1979; vol 3, 1980; vol 4,1983. Wick MR, Patterson JW: Multimodal pathologic diagnosis of malignant melanoma: integration of morphology, histochemistry, immunohistology, and electron microscopy. J Histotechnol 2003; 26: 253-258. Williams MJ, Uzman BG. Uses and contributions of diagnostic electron microscopy in surgical pathology; a study of 20 Veterans Administration Hospitals. Hum Patho11984; 15:738-745.

HISTOCHEMISTRY: A HISTORICAL REVIEW

Mark R. Wick, MD

Divisions of Surgical Pathology & Cytopathology and Autopsy Pathology, University of Virginia Medical Center, Charlottesville, Virginia, USA (E-mail: [email protected] = www.markwickmd.com ).

Wick Page 2

One can defensibly argue that biochemistry and histology originated from the same human interest; that is, a desire to know the basic structure and composition of living things. From the beginning of time, a series of observations—both scientific and fanciful—accrued in an effort to inform that topic. Such “data” emanated from several and diverse sources, such as hunter-gatherers, alchemists, mathematicians, abbatoir workers, physicians, anatomists, morticians, astrologers, sorcerers, philosophers, and theologians (1-4). In ancient Greece, Hippocrates theorized that diseases were caused by imbalances in four basic body-substances, called “humors:” phlegm, blood, black bile and yellow bile (4,5). Astoundingly, variations on that mechanistic scheme were accepted as dogma until the nineteenth century. Diets designed to "cleanse putrefied juices" were therapeutically joined with purging, or venesection, or both, to reestablish a balance between the 4 humors (6). Theophrastus Phillippus Aureolus Bombastus von Hohenheim [1493-1541] (also known as Paracelsus) was among the first to challenge such views and practices (4). He believed that illness was induced by factors originating without, rather than within, the body, and that it resulted—at least partly—from imbalances of indigenous chemicals and minerals (7). As a corollary to that premise, Paracelsus encouraged investigations of the compounds and elements that comprised plant and animal tissues. Moreover, he opened the door to the bona fide practice of pharmacy, in which prescribed external substances were taken into the body and targeted to the presumed sources of biochemical aberration or deficiency (4). That approach clearly affected the primary focus of Medicine in the Middle Ages, which was nonmorphological and primitively centered on biological chemistry. Physiological mechanisms and anatomic structure were regarded as relatively inconsequential during that period of history. Therefore, no disadvantage was attached to destructive (digestive) analysis of plants and animals, in efforts to discern their chemical constitution (8). Beginning in the 16th century, and through the efforts of Andreas Vesalius, William Harvey, Anton van Leeuwenhoek, and others, the study of anatomic structures grew steadily at gross and microscopic levels (1,3). Botany was the principal scientific discipline in which such activities evolved; early textbooks on the subject of plant histochemistry included Essai de Chimie Microscopique Appliquee a la Physiologie and Nouveau Systeme de Chimie Organique, both by Francois-Vincent Raspail (1830 & 1833) (9,10); Lehrbuch der physiologischen Chemie by Karl Gotthelf Lehmann (1842) (11); and Handbuch der Experimental Physiologie der Pflanzen by Julius von Sachs (1865) (12). Interestingly, botanists retained a basic interest in the cellular chemical processes that were illumined by histochemistry, whereas zoology-oriented histologists and histochemists used microscopy and staining techniques primarily to further the development of microanatomy, taxonomy, and nosology, more or less in vacuo. The latter situation led to an interesting, Darwinesque competition. Physiological-cellular chemists began to disparage the efforts of histologists in the second half of the 19 th century, as unworthy of true scientific respect. Morphologists were regarded as little more than clerks and scribes who recorded their visual observations without correlating them to chemical findings (13). That perception was furthered by the tendency of many histologists to embrace new stains and dyes as a means to an end (i.e., morphological discrimination), rather than as ligands for cellular chemicals that had yet to be delineated. A.G.E. Pearse framed this picture well, in saying “although diagnostic significance was attached to many of the new color reactions, no attempt was made to put them on a physical or chemical basis” (8). Hence, in the era introduced by August Bencke in the 1860s and 1870s, with aniline dyes and similar reagents in hand, histology-based histochemists broke ranks, in philosophical and heuristic terms, with cellular biochemists (14). A rancorous dichotomy persisted between the two groups well into the 20th century; indeed, as late as 1962, P.R. Lewis stated that “the decay of histology as a science can be traced to the introduction by Bencke, in 1852, of the aniline dyes which were in general use by 1880, followed closely by the development of paraffin-sectioning and photomicrography as routine techniques. Thus, by the end of the [19th] century, a fashion was set in histology which even today has not been completely supplanted. In retrospect, it may seem strange that the attention of histologists was concentrated for so long upon descriptive morphology without their making any serious attempt to study the chemistry of the structures they were staining… it is the use of a staining procedure with known chemical specificity that distinguishes a histochemical from a histological technique” (15). The foregoing material sets the stage for a discussion of the three main categories into which “histochemists” can be assigned over the past 150 years. These are (1) investigators who were chemistryoriented but not concerned with morphology; (2) those with contemporaneous interests in physiological chemistry, histology, and technology; and (3) applied (diagnostic) histochemists (histopathologists).

Wick Page 3

“Histochemists” Who Were Indifferent to Morphology As mentioned earlier, one, rather extreme, view of living organisms was that their structure is only important as a way of partitioning inorganic and organic substances, or chemical reactions. This was the credo of “pure” biochemists, who typically subscribed to “destructive” or “digestive” histochemistry. In such a context, the possible affinity of tissue for chemical laboratory reagents had “meaning” only if it illuminated the biochemical constitution of the substrate (8). An example is represented in an early analysis, by Francois-Vincent Raspail, of starch in plant tissues, using the binding of iodine solution as an indicator (9). The amylose in plant carbohydrates enters a colloidal suspension in water, and comprises long polymeric chains of glucose units that are interconnected by alpha-acetal linkages, forming a threedimensional coil. Iodine molecules can intercalate with the amylose coil, yielding a blue moiety (10). The latter property obtains regardless of whether the target is ground-up plant material studied in a test tube, intact amylose-rich organs that are infused with potassium iodide, or histological sections of tissue that are “stained” with iodine and visualized with a microscope. To histochemists belonging to the pragmatic group under discussion, it would not matter—it would be sufficient to know that the target tissue did indeed contain starch, explaining its iodinophilia. Similar comments can be made regarding iron deposits (hemosiderin) in plants and animals. Max Perls was among the first to show that acidified potassium ferrocyanide solution binds to iron in tissue, forming a relatively-insoluble blue-purple precipitate with the chemical formula Fe7(CN)18(H2O)x, where 14 ≤ x ≤ 16 (17). Again, in an egalitarian sense, it might be regarded as immaterial whether the iron was demonstrated in a glass beaker, an intact organism, or a microscope slide. Analogous models include the demonstrations of peroxidase in pus by Edwin Klebs, using tincture of guaiac in 1868 (18); Paul Ehrlich’s’ detection of cytochrome oxidase (originally called “Nadi oxidase”) in 1885 by intravenous injection of alpha-naphthol and p-phenylenediamine into animals (19); Rudolf Heidenhain’s discovery that selected cells in the gastric mucosa would turn brown when exposed to chromic acid (“chromaffinity”) (20); and Friedrich Miescher’s identification of deoxyribonucleic acid (DNA) in cellular nuclei through its selective binding to methyl green (21). All of these assessments provided new information, but, from the perspective of current-day morphologists, they would be unsatisfying because the particular cellular locations of the chemical substances in question were not addressed. In that vein, microanatomy as a discipline was advancing in the 1800s as well, despite its being regarded as a “non-science” by biochemists of the period. The first attempt at a comprehensive textbook of histology was published in 1841 by Friedrich Gustav Henle (22), followed by a succession of additional works by Albert Donne, Arthur Hill Hassall, Rudolph Albert von Kolliker, Lionel Beale, Gottlieb Gluge, John Scott Burdon-Sanderson, Georg Eduard von Rindfleisch, Andre-Victor Cornil & Louis-Antoine Ranvier, Edward Albert Schafer, Phillip Stohr, and other authors in the latter half of the nineteenth century (23). The topical approach in several of those publications was to combine microanatomy with physiology, stressing both structure and function simultaneously. That orientation led to development of the next tier of histochemists, whose work occupied much of the twentieth century. “In-Situ” Biochemical Histochemists Given the availability of dyes that evolved during the mid-1800s, histologists of the period began to use them enthusiastically. Nevetheless, as cited earlier, that practice was often unaccompanied by a clear understanding of exactly what was being labeled by tissue staining procedures. Moreover, the detailed chemical mechanisms for such techniques commonly went unstudied as well. Dyes that had entered into biological use included “cochineal” agents such as mucicarmine (24); aniline dyes (25); hematoxylin and its congeners (26); precipitatable silver solutions (27); Schiff-base derivatives (28); colloidal suspensions of metal ions (29); phthalocyanines (30); cotton dyes (e.g., Congo red; Pagoda red) (31); methyl & ethyl green (32); and others. An international group of investigators increasingly focused on the biochemical processes and targets that were associated with the use of such reagents, in the 1890s and beyond. They included—but were not limited to—individuals such as Paul Ehrlich, Santiago Ramon y Cajal, Karl Weigert, Pio del RioHortega, Joseph von Gerlach, Paul Mayer, Friedrich Miescher, Alfred Fischer, Gustav Mann, Robert Feulgen, Julius von Kossa, Frank Burr Mallory, Lucien Lison, Clyde Mason, Maffo Vialli, Emile Chamot, David Glick, George Gomori, Anthony Guy Everson Pearse, and Ralph Lillie (8). Some early explanations for the cellular affinities of dyes were scientifically-infantile; for example, Fischer suggested in 1899 that all stains were merely absorbed passively by tissue (33). Conversely, Ehrlich and Miescher correctly believed that specific chemical coupling was responsible (21,34). And, in his text entitled “Physiological Histology”—published in 1902—Mann stated that “it is not sufficient to content ourselves with using acid

Wick Page 4

and basic dyes and speculating on the basic or acid nature of the tissues, or to apply color radicals with oxidizing or reducing properties… we must endeavor to find staining reactions which will indicate not only the presence of certain elements such as iron or phosphorus, but the presence of organic complexes such as the carbohydrate groups, the nucleins, protamines, and others” (35). Most scientists in the above-listed group took that directive to heart, as did others after them. Indeed, many microscopists became so engrossed by a characterization of in-situ chemical reactions that the practical and diagnostic uses of histochemistry were given short shrift. In other words, the admonitions of Mann (35), and Lewis after him (15), became the marching orders of the day. Histochemical textbooks written by Lison in 1936 (36); Glick in 1949 (37); Gomori in 1952 (38); A.G.E. Pearse in 1953 (39); Lillie in 1954 (40); John Bancroft in 1967 (41); John Kiernan in 1981 (42); and Barbara Sumner in 1988 (43) were devoted largely to the chemistry of tissues as seen under the microscope. As a result, knowledge of cellular biochemistry grew exponentially during the twentieth century. By 2000, Coleman (44) was able to say confidently that “histochemistry and cytochemistry… allow precise analysis of the chemistry of cells and tissues in relation to structural organization.” He also went on to state that histochemistry was still a useful and productive field of study, and that “there are…few other disciplines in experimental biology or medicine that can make a similar claim.” Histochemists with a Diagnostic Orientation As mentioned earlier in this discussion, philosophical tension has existed between “basic” and “applied” histochemists for well over 100 years. This is not a novel situation, and, in fact, it has applied to every one of the “translational” scientific techniques used in morphology-oriented areas of laboratory medicine. Electron microscopy, immunohistology, in-situ hybridization, polymerase chain-reaction-based procedures, and other “blotting” technologies have served as comparable battlegrounds for purists and practitioners (45). In 1955, Jonas Friedenwald—a “basic” researcher in ophthalmology at Johns Hopkins University—published a review of applied histochemistry, including in it several maxims that are still true (46). In regard to criticisms that focused on the “nonspecificity” or crudeness of some histochemical reactions, he said “criteria of specificity [in histochemistry] are similar to those in qualitative chemistry in general… [they] can be very much enhanced if two or more different reactions can be applied and compared.” Presciently, he went to opine that “analysis in-situ is a non-quantitative procedure. Sensitivity, therefore, merely concerns the limits at which the reaction is discernible.” The latter comments apply equally well today, in reference to modern attempts at “quantitative” immunohistochemistry (47). Three years earlier, Robert Stowell—chair of pathology at the University of California-Davis— had suggested that “fundamental, critical research on new cytochemical techniques will do more to advance our eventual understanding of normal tissues and neoplasia than the application of the relatively few and often none-too-satisfactory histochemical and cytochemical techniques now available” (48). In counterpoint, A.G.E. Pearse—arguably the most well-versed histochemist of all—responded thusly to Dr. Stowell: “in medicine the new and imperfect remedy does not await perfection by the research of groups of collaborating investigators in the pure sciences. It is applied forthwith to…patients by the practitioners of medicine, and it is often by their observations and research that real advancement in the use of the remedy, and in knowledge of its mechanism and meaning, is brought about. I believe very strongly, therefore, that the methods of modern histochemistry, despite their imperfections, should be applied by all practitioners in the biological, cytological, and pathological sciences” (49). After the passage of another 20 years, Pearse could further state that “histopathology can be transformed by the application of any technology which confers upon its observations an increase in objectivity. Foremost in the field comes histochemistry, for a variety of reasons. These include sheer breadth of scope and overwhelming numerical superiority in respect of techniques” (50). Pearse (39), Bancroft & Stevens (51), and Filipe & Lake (52) took such tenets and built textbooks around them in the latter part of the twentieth century. With such perspectives by experienced hospital pathologists, the place of histochemistry as a valuable clinical method was solidified. Despite refinement and flux in the nosological categorization of some human diseases, histochemical analysis continues to offer important information in regard to histopathological diagnosis and differential diagnosis. The Nexus of Histochemistry with Immunology & Molecular Biology Dr. Albert Coons was still a house-officer at Massachusetts General Hospital when he conceived a simple but revolutionary idea. His thought was to label antibodies with a chemical tag, so that their binding to predefined antigens in tissue could be visualized microscopically. Despite the fact that antibody structure was only primitively understood at that time, and the lack of a proven technique for artificially binding

Wick Page 5

other substances to them, Coons pursued the concept doggedly over several years (53). Eventually, specific antibodies were produced in vivo in animal hosts that were specific for particular proteins. They were coupled successfully with fluorescein isocyanate, and proved to be effective in localizing polypeptide targets in histologic sections that were illuminated by ultraviolet light with a special microscope (54,55). Hence, the field of immunofluorescence-based histochemistry was thereby established by Coons, who won the prestigious Albert Lasker award in 1959 for that contribution (56). Today, many different chemical tags can be linked with a plethora of antibody reagents that are clinically relevant in pathology. In addition, the facet of molecular biology known as in-situ hybridization (ISH) is predicated on a similar construct. However, in place of antibody probes to which chemical indicators can be joined, the basic investigative tools in ISH are specific sequences of nucleic acid that are complementary to DNA or ribonucleic acid (RNA) targets of interest in tissue sections (57,58). Conclusions Histochemistry has had a long history, as well as a broad interface with many of the other life sciences. Because the in-situ chemical reactions it concerns have been thoroughly studied over many years, histochemistry is now one of the most objective methods in biology and medicine. That fact should not be forgotten in the current fervor over “new” techniques in pathology, nor should one slight the practical utility of histochemistry in a variety of clinical differential diagnostic settings. The rapidity, reproducibility, and relatively low expense attached to this form of biomedical analysis continue to recommend it as a valuable enterprise, after nearly 200 years of existence.

Wick Page 6

References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20.

21. 22. 23. 24. 25. 26. 27. 28. 29. 30.

Persaud TVN: The Early History of Human Anatomy: from Antiquity to the Beginning of the Modern Era, CC Thomas Books; Springfield, IL, 1984. Dyer GS, Thorndike ME: Quidne mortui vivos docent? The evolving purpose of human dissection in medical education. Acad Med 2000; 75: 969-979. McLachlan JC, Patten D: Anatomy teaching: ghosts of the past, present, and future. Med Educ 2006; 40: 243-253. Nuland S: The Mysteries Within, Simon & Schuster, New York, 2000; pp. 50-130. Gill NS: Hippocratic method and the four humors in medicine. http://ancienthistory.about.com/cs/hippocrates/a/hippocraticmeds.htm. Accessed 1-14-2010. Shryock RH: Medicine & Society in America: 1660-1860. Cornell University Press, Ithaca, 1972; pp. 69-72. Anonymous: Paracelsus. http://en.wikipedia.org/wiki/Paracelsus. Accessed 1-4-2010. Pearse AGE: The history of histochemistry. In: Histochemistry (Pearse AGE, Ed), 2nd Ed., Churchill, London, 1960; pp. 1-12. Raspail F-V: Essai de Chimie Microscopique Appliquee a la Physiologie, Meihac Publishers, Paris,1830. Raspail F-V : Nouveau Systeme de Chimie Organique, JB Bailliere Publishers, Paris, 1833. Lehmann KG : Lehrbuch der physiologischen Chemie, Engelmann Publishers, Leipzig, 1842. von Sachs J: Handbuch der Experimental Physiologie der Pflanzen, Engelmann Publishers, Leipzig, 1865. von Bunge GB: Lehrbuch der physiologischen und pathologischen Chemie, Vogel Publishers, Leipzig, 1887. Kornhauser SI: The history of staining: the development of cytological staining. Stain Technol 1930; 5: 117-125. Lewis PR: Histochemistry in biology. In: Viewpoints in Biology (Garthy JD, Ed), Butterworths, London, 1962; pp. 49-88. Yu X, Houtman C, Atalla RH: The complex of amylose and iodine. Carbohydrate Res 1996; 22: 129-141. Perls M: Nachweis von Eisenoxyd in gewissen pigmenten. Virchows Arch 1867; 39: 42-48. Klebs E: Die pyrogene substanz. Z Med Wiss 1868; 6: 417-437. Ehrlich P: Das sauerstoffbedurfnis des Organismus: Eine farbenanalytische Studie. Hirschwald Publishers, Berlin, 1865; pp. 364-432. Drozdov I, Modlin IM, Kidd M, Goloubinov VV: From Leningrad to London: the saga of Kulchitsky and the legacy of the enterochromaffin cell. Neuroendocrinology 2009; 89: 109120. Miescher F: Die Spermatozoen einiger Wirbeltiere: ein Beitrag zur Histochemie. Verh Natforsch Ges Basel 1874; 6: 138-208. Henle FGJ: Allgemeine Anatomie, Voss Publishers, Leipzig, 1841. Bracegirdle B: The history of histology: a brief survey of sources. Hist Sci 1977; 15: 77-101. Conn HJ, Kornhauser SI: The history of staining: cochineal dyes. Biotechnic Histochem 1928; 4: 110-121. Johnston WT: The discovery of aniline and the origin of the term “aniline dye.” Biotechnic Histochem 2008; 83: 83-87. Titford M: The long history of hematoxylin. Biotechnic Histochem 2005; 80: 73-78. Heinz TR: Evolution of the silver and gold stains in neurohistology. Biotechnic Histochem 2005; 80: 211-222. Kasten F: Schiff-type reagents in cytochemistry. I. Theoretical & practical considerations. Histochemie 1959; 1: 466-509. Hale CW: Histochemical demonstration of acid polysaccharides in animal tissue. Nature 1946; 157: 802-810. Scott JE: The molecular biology of histochemical staining by cationic phthalocyanin dyes: the design of replacements for Alcian blue. J Microsc 1980; 119: 373-381.

Wick Page 7

31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47.

48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58.

Yanagihara M, Mehregan AH, Mehregan DR: Staining of amyloid with cotton dyes. Arch Dermatol 1984; 120: 1184-1185. Pollister AW, Lauchtenberger C: The nature of the specificity of methyl green for chromatin. Proc Natl Acad Sci USA 1949; 35: 111-116. Fischer A: Fixierung, Farbung, und Bau des Protoplasmas, Gustav Fischer Publishers, Leipzig, 1899. Ehrlich P: Encyclopadie der Mikroskopischen Technik, Urban & Schwarzenberg Publishers, Berlin, 1903. Mann G: Physiological Histology: Methods & Theory, Oxford Press, London, 1902. Lison L: Histochimie Animale, Gautier-Villars Publishers, Paris, 1936. Glick D: Techniques of Histo- & Cyto-chemistry, Interscience Publishers, New York, 1949. Gomori G: Microscopic Histochemistry, Chicago University Press, Chicago, 1952. Pearse AGE: Histochemistry, Churchill Publishers, London, 1953. Lillie RD: Histopathologic Technic & Practical Histochemistry, Blakiston Co., New York, 1954. Bancroft JD: An Introduction to Histochemical Technique, Butterworths, New York, 1967. Kiernan JA: Histological & Histochemical Methods: Theory & Practice. Pergamon Press, Oxford, 1981. Sumner BEH: Basic Histochemistry, John Wiley & Sons, Hoboken, NJ, 1988. Coleman R: The impact of histochemistry—a historical perspective. Acta Histochem 2000; 102: 5-14. Wick MR: Unpublished observations, 1978-2009. Friedenwald JS: Histochemistry—a review. Pharmacol Rev 1955; 7: 83-96. Aitken SJ, Thomas JS, Langdon SP, Harrison DJ, Faratian D: Quantitative analysis of changes in ER, PR, and HER2 expression in primary breast cancers and paired nodal metastases. Ann Oncol 2009; Nov. 2009 (E-pub ahead of print). Stowell RE: Use of histochemical and cytochemical technics in problems in pathology. Lab Invest 1952; 1: 210-230. Pearse AGE: The place of histochemistry today. Postgrad Med J 1953; 29: 536-537. Pearse AGE: The role of histochemistry in increasing objectivity in histopathology. Postgrad Med J 1975; 51: 708-710. Bancroft JD, Stevens A (Eds): Histopathologic Technic & Practical Histochemistry, McGraw-Hill, New York, 1965. Filipe MI, Lake BD (Eds): Histochemistry in Pathology, Churchill-Livingstone, Edinburgh, 1983. Karnovsky MJ: Dedication to Albert H. Coons, 1912-1978. J Histochem Cytochem 1979; 27: 1117-1118. Coons AH, Creech HJ, Jones RN: Immunological properties of an antibody containing a fluorescent group. Proc Soc Exp Biol Med 1941; 47: 200-202. Coons AH, Creech HJ, Jones RN, Berliner E: The demonstration of pneumococcal antigen in tissues by the use of fluorescent antibody. J Immunol 1942; 45: 159-170. Anonymous: The Albert Lasker awards for 1959. Am J Public Health Nations Health 1959; 49: 1686-1690. Shipley J: Putting the colors into chromogenic in-situ hybridization (CISH). J Pathol 2006; 210: 1-2. Isola J, Tanner M: Chromogenic in-situ hybridization in tumor pathology. Methods Mol Med 2004; 97: 133-144.

pg. 1

Tissue In Situ Hybridization 40 years and still cooking… History of Pathology Society USCAP 2012 Companion Meeting Mark H. Stoler, MD University of Virginia Introduction The history of in situ hybridization very much parallels the history of molecular diagnostics. Most techniques in the field of molecular diagnostics are hybridization-based assays directed at the detection of specific DNA or RNA sequences that have been extracted from a cellular sample. Implicit in these analyses are assumptions regarding the cellular composition of the sample. If one is looking for a virus trophic for a specific cell type or an oncogene expressed in a particular type of tumor cell, those cells have to be present in the sample under analysis. Paradoxically, modern diagnostic methods like endoscopic biopsy and fine needle aspiration cytology are forcing pathologists to do "more with less". With a limited sample, prioritization of analyses and maximizing yield are important daily considerations. In anatomic pathology, morphologic analysis is usually the highest priority, as morphology provides the fundamental diagnostic context for the interpretation of other clinical and laboratory data. The technique of in situ hybridization (ISH) permits the detection of nucleic acid targets within a microanatomic context. Like other powerful nucleic acid based techniques, it capitalizes on the potentially high specificity and sensitivity inherent in directly probing for DNA/RNA sequences. Unlike other biochemical methods, it preserves morphology, thus permitting accurate localization of signal source. Thus, in situ hybridization uniquely facilitates the optimal synthesis of histopathologic and molecular biologic data. This presentation will review a bit of history but focus on principles of in situ hybridization that support its increasing usage in daily practice. A Bit of History During the early 1970s the modern era of molecular biology was ushered in with the advent of recombinant DNA technology. The isolation of restriction enzymes that cut DNA at a defined sequence, the ability to clone DNA sequences in bacteria, and the development of enzymatic methods for labeling DNA as well as the reverse transcription of RNA into complementary DNA, formed the foundation of all hybridization methods. In situ hybridization actually predated some of these developments, making it one of the oldest molecular diagnostic techniques. Described in the late 1960s, ISH was first used to detect amplified DNA targets in the cell nuclei of cytologic preparations. Shortly thereafter the method was applied to chromosomes to study the production of ribosomal RNA. In the same year, ISH was applied to histologic sections. The first description of the use of ISH for the study of an infectious agent came from France where this new method was applied to the study of Shope rabbit papillomavirus DNA replication in tissue sections. Epstein Barr virus and adenoviruses were also the subject of early ISH studies. Parenthetically, Harald zurHausen widely known for the recent Nobel prize for HPV, but he performed the first EBV ISH in 1972!. The first histologic messenger RNA (mRNA) studies were also described about the same time. During the late

Stoler/ISH pg. 2

1970s and early 1980s the ability to design and label improved high specific activity DNA and more importantly, single stranded RNA probes heralded the development of more sensitive methods for detecting low copy number viral sequences for the study of latent infections of the CNS. This was quickly followed by the application of quantitative ISH techniques for the detection of messenger RNA in morphologically distinct cell populations. The development of nonradioactive detection systems for ISH presaged a time of easier and more widespread clinical implementation of this technology. Polymerase chain reaction-based amplification techniques have been coupled to ISH as far back as the 1990s, in an effort to increase method sensitivity. Other amplification methods, both target-based and signal-based continue in development Relationship of ISH to other molecular methods All nucleic acid hybridization tests are based on the fact that two antiparallel single-stranded nucleic acid molecules will recognize one another and bind to each other (hybridize) on the basis of hydrogen bonding and the hydrophobic base-stacking interactions of complementary base pairs {(A)denine binds to (T)hymine(or (U)racil if RNA; (G)uanine binds to (C)ytosine}. Thus, DNA:DNA, DNA:RNA, and RNA:RNA duplexes are possible. While a full discussion of the factors affecting hybridization is beyond the scope of this presentation, repetition of some of these basic principles as applied to ISH is essential to an appreciation of probe and protocol selection for in situ studies. Recall that the interaction between duplex strands of nucleic acid is a temperature dependent phenomenon. The reference point for this interaction is the hybrid melting temperature (Tm), defined as the temperature at which 50% of a population of duplex molecules dissociates, i.e., melts, into single strands. The major factors affecting duplex stability include the ionic strength of the reaction, the mole percentage of G/C base pairs in the probe, the probe length, the percentage of non-complementary bases between the probe and target (% mismatch) and the formamide percentage (a duplex destabilizing agent) in the solution. Increasing the ionic strength, probe length, and G/C content increases duplex stability whereas increasing temperature, formamide percentage, and mismatch percentage decreases duplex stability. The stringency of a hybridization reaction refers to the degree to which the reaction conditions favor duplex dissociation; ie., high stringency can be attained with high temperature, low salt, or high formamide concentrations, or with combinations of all three. Duplexes formed when the two strands have a high degree of base homology will better withstand a high stringency wash than will duplexes of lesser homology. Most hybridization reactions are carried out at a relatively low stringency of Tm - 25oC. This temperature has been empirically determined to produce an optimal rate of hybridization, but does allow for the formation of imperfect, i.e., less than perfectly complementary, hybrids. Rather than totally controlling specificity with the hybridization step, most ISH procedures use a series of post-hybridization washes at higher stringencies to dissociate the imperfect hybrids leaving only specifically bound probe on the target. Another factors to consider in any hybridization are the type of nucleic acids in the probe and target. For instance, DNA:DNA hybrids are approximately 10-15oC less stable than RNA:RNA hybrids with RNA:DNA hybrids having an intermediate Tm. Thus, in comparing protocols and data, a detailed knowledge of the hybridization conditions, may be the only way of comparing data even on identical probes, as stringency, together with the base homology between probe and target, controls the specificity of the hybridization. The ability to localize specific DNA and messenger RNA sequences in histologic preparations has made in situ hybridization an extremely valuable research tool that is finally becoming applicable for routine clinical diagnosis. Analogous to the now ancient Southern vs.

Stoler/ISH pg. 3

northern blots, ISH assays can be divided into those directed at the detection of DNA vs. RNA. When probing for a DNA target the probe may be directed at host or foreign DNA. Foreign DNA in human systems most often is of microbial, usually of viral origin. Since all of an organism's cells have essentially the same DNA content, detection of cellular DNA may or may not be of utility in probing for functional differences among cells. Probes for single point mutation in DNA are unlikely to function in situ. Most genes are present in only one to a few copies per chromosome, and may be below the limit of sensitivity. However in many neoplasms gene amplifications or translocations may occur and if the number of amplified copies is high enough and sensitive methods are used, direct evidence of amplification or translocation may well be detectable in situ. This is especially applicable in the interphase nuclei of most pathology specimens, but applies equally well to chromosomal preparations. In situ hybridization for messenger RNA complements both DNA in situ hybridization and immunocytochemistry. Only immunological methods and in situ hybridization for mRNA allow identification of individual cells expressing specific genes, as opposed to cells merely carrying a gene. While immunological detection of a protein or hybridization detection of the protein's messenger RNA provides similar information, the time or site of protein synthesis as well as protein localization may be different than the time or site of transcription of its messenger RNA. Some of the limitations inherent in probing cellular DNA can be overcome by probing for the presence or absence of the corresponding mRNA. For example, for a period of time before more sensitive methods were developed, the N-myc oncogene in neuroblastoma was in select cases amplified in tumor cells, but the size of the then available probes and the number of amplified copies made detection of this prognostically important abnormality undetectable at the DNA level, whereas it was readily apparent when the over transcribed mRNA is probed for. Yet with progress the direct DNA amplication site can now be routinely visualized. Relationship to immunohistochemistry Unlike other biochemical methods, but like immunohistochemistry (IMH), ISH preserves morphology, thus permitting accurate localization of signal source. Should in situ hybridization replace immunohistochemistry as the molecular diagnostic of choice in the anatomic pathology laboratory? For one technology to replace another should imply the newer technique enjoys some kind of clear superiority. It may be more sensitive or more specific. It may be equivalent or even of lower sensitivity/specificity, but easier or faster or cheaper. It may provide critical clinical information that other currently available techniques do not. Depending upon the questions being asked, in situ hybridization may "beat " immunohistochemistry or be "beaten" by it. At the present time, "complement" rather than "replacement" seems a better choice of words. The arguments for an in situ method of nucleic acid detection have in part been made by analogy to immunohistochemistry. If analyzing biochemical information outside of the context of morphology was sufficient, than all such analyses would be done in the clinical chemistry, immunology and wet molecular biology labs. Since we are focused on history, the evolution of the use of estrogen receptor (ER) analyses in breast cancer makes a relevant example. Biochemical assays of ER content on pieces of breast cancer tissue were in the 1970s and 80s found to provide useful prognostic or therapeutic information. But there were problems with these assays. Technical issues included having adequate amounts of tissue, handling the tissue such that the ER is preserved ina functional state, knowing that the tissue is representative of the neoplastic process, and finally, the assay used in many labs was often analytically complex and of poor precision. With time and the widespread impact of mammographic screening, the tumor size has shrunk and just having a big enough piece of tumor increasingly dominated the tension between making a diagnosis and then the ancillary studies. With the advent of monoclonal

Stoler/ISH pg. 4

antibodies against the ER protein, immunohistochemistry clearly demonstrated the nature of many of the variables. Tissues are clearly heterogeneous in terms of their cellular content as well as their patterns of ER expression, even within morphologically similar cell types. This clear demonstration of cellular heterogeneity in gene expression forced re-evaluation of the traditional analytic methods. What also drove the case for an in situ type of analysis is the fact that the nature of biopsy material is changing. Today most specimens are inadequate for traditional biochemical analysis. The tumor is too small, may not even be grossly visible, or may be a just a cytologic preparation. With small samples, it is unacceptable to compromise the diagnosis by ordering a test destructive of the morphology. In situ methods eliminate the "how to divide the tissue" conflict. Single adjacent sections can provide serial molecular analyses and signal localization is assured. The question is then reduced to which in situ method is the one of choice. What was true for breast cancer is now being re-lived with lung cancer, colon cancer, melanoma, etc. driven by the availability pathway specific therapy As noted above, only immunohistology for antigens and in situ hybridization for mRNA allow identification of individual cells expressing specific genes, as opposed to cells merely carrying a gene. In some cases, small genes or genes of low copy number even though they may be amplified, are below the sensitivity of DNA detection. In these cases the amplification provided by transcription of the corresponding mRNA often allows localization of the related sequence. While immunological detection of a protein or hybridization detection of the message encoding a protein provides similar information, the time or site of protein localization may be different than the time or site of transcription of its mRNA. It is not uncommon to have a cell that contains the message for a protein, but in which protein is not detectable immunologically. This may reflect differences in technical sensitivity or may have biological significance in, for instance, suggesting translational regulation of protein levels. How to choose between IMH and ISH? Clearly, immunohistochemistry is a much more familiar and relatively standardized methodology, so if an IMH assay for a target of interest is functioning in your lab don't change. As a guide, when contemplating a choice between ISH and immunohistochemistry several questions can be addressed: Does an antibody exist for the gene product of interest and, if so, what is its specificity? Does the antibody effectively interact with its target under the available conditions of specimen fixation and processing? Is it relevant to distinguish site of synthesis from site of antigen localization? Is the concentration of antigen sufficient for detection given the sensitivity of the assay? All of these questions factor into the choice between immunologic and hybridization methods. Indeed, much of the IMH and ISH literature of the last few years is centered on newer generation antibodies, technical improvements, antigen retrieval, etc. forcing one to continually reevaluate previous convictions. Assays that didn't work before, are now routine. Tumors that were defined by a clear pattern of antigenicity, now have many exceptions to the rules. Thus, the opinions of today are subject to evolution as technology progresses. Methodologic Issues Sensitivity As should be clear, the ultimate sensitivity of ISH is a function of multiple interrelated steps of a given probe-target combination and protocol. The sensitivity of in situ hybridization depends upon the signal-to-noise ratio that can be achieved. The signal depends upon the fraction of target in the tissue that is retained and accessible for hybridization, the mass of probe relative to the target (saturation), and the complexity and specific activity of the probe as visualized through a defined detection system. Noise, which can be limiting, depends upon the extent to which the probe nonspecifically binds to the background and is proportional to the

Stoler/ISH pg. 5

probe concentration. While some assays claim single copy sensitivity, the truth is often in the details of the methods Another measure of sensitivity is a relative comparison of different detection methods for a similar target. There are many ways to skin the proverbial cat. Do I do ISH or PCR or both? Much of the practical work in our field entails validating and cross-referencing the different methods. And of course the answers will change and evolve. Specificity and Controls One of the true advantages of ISH is the specificity of the information derived from this type of combined morphologic and molecular analysis. If the distribution of the signal is not in the appropriate cell type, the specificity of the probe for the target may be in question or alternatively the biology is not what you thought! Obviously when first evaluating a new probe, it is ideal to have known positive and negative control cells or tissues processed in a manner identical to the clinical or experimental material. The specificity of the hybridization is controlled by sequence homology and stringency. If the stringency is very low, nonspecific hybridization will be favored. Historically this was sometimes used to select for related sequences, potentially enhancing sensitivity at the cost of specificity. HPV 16 as an example, was discovered using low stringency hybridizations with HPV 6/11 probes. Conversely, if the stringency is set too high there may be loss of specific signal, i.e. a false negative result. Thus, sensitivity and specificity are in part, reciprocally related. Asymmetric RNA probes, depending upon probe orientation and tissue denaturation, also allow specific control of the hybridization to DNA or mRNA. If a gene is transcribed in only one direction, the case for most but not all eukaryotic genes, the sense orientation probe provides an ideal nonhomologous control for background signals during mRNA detection. Additionally, one can verify that the signal is derived from RNA or DNA by pretreating the section with RNase or DNase, thereby causing a loss of signal. Alternatively others have used competitive hybridization between labelled and unlabeled probe to prove specificity. An extremely useful control, particularly when one is trying to make comparisons between multiple specimens, is to probe for a ubiquitously expressed message (eg. actin) and to normalize signal interpretations to the level of this target. Specimens with significant loss of the control signal may be unevaluable, especially if one is probing for mRNAs that are only moderately expressed. The converse is that some highly expressed messages may be detectable despite considerable mRNA loss. An example might be the EBER RNAs of Epstein Barr virus which may be present at a level of 106-107 copies/cell. If most messages are expressed at a level of 103 copies /cell than 99.9% of the mRNA could be lost and the target cell could still look like a high expresser.

Applicability Over the last forty years in situ hybridization has evolved from a technically demanding research technique in “the hands of the few” or “the one”, to what is finally, thanks to simplification and advances in automation, a technique that is clinically usable in the “hands of the many”. The methods described potentially have clear-cut applications in nearly all areas of pathology. The primary applications in use today fall into three major categories: Detection of infectious agents:

Stoler/ISH pg. 6

Gene probes are available for the identification of the genetic material of numerous bacteria, Mycobacteria, and viruses. The harder or more dangerous it is to culture an organism the more likely it is that a molecular detection format will be preferred. For some organisms, molecular detection and analysis has essentially defined the classification and/or life cycle of the organism in the absence of routinely available microbial culture. In situ hybridization methods major advantage over other methods is the highly specific localization of the organism within a specific cell type or histologic lesion, thus providing an essential pathogenic link between the presence of an organisms genetic material and the presence of disease. Cytogenetics: ISH on chromosomes is the method of choice for localizing a gene to a specific locus. Nucleic acid probes are commercially available for large portions of every human chromosome as well as to specific gene rearrangements associated with certain forms of neoplasia. Fluorescent in situ hybridization (FISH) is the major growth field in cytogenetics today has wide applicability in diagnostics, particularly when applied to interphase nuclei. Gene Expression. The ongoing analysis of the regulation of cellular gene expression in health and disease is what translational pathology is all about. The sequential analysis of cellular DNA, RNA and protein through a combined ISH/IMH approach continues to lead to a much better understanding of the interaction of how cellular biochemical events impact on pathogenesis, refine classifications, predict prognosis and now most importantly, help us help our clinician colleagues select therapy. Still Cooking: However, tempering my enthusiasm for all thing in situ, are some observations that have persisted too long and will continue to impact widespread adoption - it is why we are “still cooking” and not done. The current lack of method standardization is probably the leading impediment to widespread clinical implementation of ISH. There is no such thing as “standard in situ hybridization”. Inter-laboratory comparison data make this reality quite clear. Issues of technical sensitivity and specificity need to be compared in terms of clinical utility based on the population under study and the question to be asked. Having an assay that can detect a few hundred copies of a gene sequence in a few hours is irrelevant if the targets under study only express ten copies. Similarly, assays that can detect very low levels of gene expression may also be of limited utility if they cannot be performed within a clinically useful time frame. Issues of prevalence, predictive value, turnaround time and cost are all important considerations As noted above, competing methodologies may provide similar data and choosing among them is definitely not trivial. And yet after 40 years, TISH (or dish / rish / fish / cish, etc) is ready to serve. Some “Amplified” Reading: some classic, some not Angerer LM, Stoler MH, Angerer RC. In situ hybridization with RNA probes - An annotated recipe. In: Valentino K, Eberwine J, Barchus J, eds. In-Situ Hybridization: Applications to Neurobiology. New York: Oxford University Press; 1987:42-70. DeLellis RA. In situ hybridization techniques for the analysis of gene expression: applications in tumor pathology. [Review]. Human Pathology 1994; 25:580-5. Gall JG, Pardue ML. Formation and detection of RNA-DNA hybrid molecules in cytological preparations. Proceedings of the National Academy of Sciences of the United States of America 1969; 63:378-83.

Stoler/ISH pg. 7 Grady-Leopardi EF, al e. Detection of N-myc oncogene expression in human neuroblastoma by in situ hybridization and blot analysis: relationship to clinical outcome. Cancer Research 1986; 46:3196-3199. Gray JW, Lucas J, Kallioniemi O, et al. Applications of fluorescence in situ hybridization in biological dosimetry and detection of disease-specific chromosome aberrations. Progress In Clinical & Biological Research 1991; 372:399-411. Leary JJ, Brigati DJ, Ward DC. Rapid and sensitive colorimetric method for visualizing biotin-labeled DNA probes hybridized to DNA or RNA immobilized on nitrocellulose: Bio-blots. Proceedings of the National Academy of Sciences of the United States of America 1983; 80:4045-9. Lloyd RV, Jin L, Bonnerup MK. In situ hybridization in diagnostic pathology. Mayo Clinic Proceedings 1994; 69:597-8. Moorman, A.F., et al., Practical aspects of radio-isotopic in situ hybridization on RNA. Histochemical Journal, 1993. 25(4): p. 251-66. Myerson, D. and K.W. Henne, Automation of in situ hybridization. American Journal of Clinical Pathology, 1992. 98(4 Suppl 1): p. S39-46. Nuovo GJ, Gallery F, MacConnell P, Becker J, Bloch W. An improved technique for the in situ detection of DNA after polymerase chain reaction amplification. American Journal of Pathology 1991; 139:1239-44. Orth G, Jeanteur P, Croissant O. Evidence for and localization of vegetative viral DNA replication by autoradiographic detection of RNA-DNA hybrids in sections of tumors induced by Shope papilloma virus. Proceedings of the National Academy of Sciences of the United States of America 1971; 68:1876-80. Pardue ML, Gall JG. Chromosomal localization of mouse satellite DNA. Science 1970; 168:1356-8. Pardue ML, Gall JG. Molecular hybridization of radioactive DNA to the DNA of cytological preparations. Proceedings of the National Academy of Sciences of the United States of America 1969; 64:600-4. Segal GH, Shick E, Fishleder AJ, Tubbs RR, Stoler MH. In situ hybridization analysis of lymphoproliferative disorders: assessment of clonality by immunoglobulin light chain messenger RNA expression. Diagnostic Molecular Pathology 1994; 3:170-177. Stoler MH, Broker TR. In situ hybridization detection of human papillomavirus DNAs and messenger RNAs in genital condylomas and a cervical carcinoma. Human Pathology 1986; 17:1250-8. Stoler MH, Eskin TA, Benn S, Angerer RC, Angerer LM. Human T-cell lymphotropic virus type III infection of the central nervous system. A preliminary in situ analysis. Jama 1986; 256:2360-4. Stoler MH, Ratliff NB. Potential and problems of the in situ molecular detection of viral genomes [editorial; comment]. American Journal of Clinical Pathology 1990; 93:714-6. Stoler MH. In situ hybridization. Clinics in Laboratory Medicine 1990; 10:215-36. Stoler MH. Tissue In Situ Hybridization, Chapter 61. In: Henry JB, ed. Clinical Diagnosis and Management by Laboratory Methods. 19th ed: W. B. Saunders; 1996:1400-1412. Stoler, M.H., In situ hybridization. A research technique or routine diagnostic test?. Archives of Pathology & Laboratory Medicine, 1993. 117(5): p. 478-81. Tubbs, RR and Stoler, MH, eds Cell and Tissue Based Molecular Pathology (a volume in the series Foundations in Diagnostic Pathology, Goldblum, JR, ed.) Churchill Livingston, 2009. Valentino KL, Eberwine JH, Barchas JD. In Situ Hybridization: Applications to Neurobiology. New York: Oxford University Press, 1987. zur Hausen H, Schultz-Holthausen H. Detection of Epstein-Barr viral genomes in human tumor cells by nucleic acid hybridization. In: Biggs PM, de-The G, Payne LN, eds. Oncogenesis and Herpes. Lyon: I.A.R.C., 1972:321-325.

Tissue In Situ Hybridization 40 years and still cooking… History of Pathology Society USCAP 2012 Companion Meeting Mark H. Stoler, MD University of Virginia

COI Dr. Stoler has consulted with Ventana Medical Systems on ISH methodology

In Situ Molecular Detection • History • Principles • Examples

History-1 Late 60s-early 70s Amplified DNA in nuclear preps Ribosomal RNA off chromosomes First histologic section for DNA Viral Targets: CRPV, EBV, Adeno

History-2 1970s to 1980s Improved probe labeling methods Single stranded RNA probes mRNA expression studies Oligonucleotides Non radioactive labeling of probes parallels to IMH

History-3 1990s to 2000 Improved probe labeling methods other in situ enzymatic amplification methods: pcr, transcription, CSA AUTOMATION clinical applications proficiency testing

History-4 2000s-Today Improved probe labeling methods Refined protocols AUTOMATION commercialization clinical applications FDA approvals Therapeutic implications

Flow of Genetic Information Flow of Genetic Information

DNA

transcription

RNA

translation

Protein

Molecular Detection Methods • • • • •

Southern blot northern blot dot blot polymerase chain reaction in situ hybridization

In Situ Molecular Detection • • • •

in situ hybridization in situ reverse transcription in situ polymerase chain reaction immunohistochemistry

In Situ Hybridization vs. Other Molecular Techniques • uniquely facilitates the optimal synthesis of histopathologic and molecular biologic data – – – –

tissue conservation tissue cellular heterogeneity cellular heterogeneity of gene expression subcellular compartmentalization

In Situ Hybridization: RNA-DNA

In Situ Hybridization: RNA - mRNA

In Situ Hybridization: RNA - ( RNA + DNA)

In Situ Hybridization: negative control

Probe Detection Patterns • PROBE ORIENTATION DETECTS – – – –

antisense / no denaturation antisense / with denaturation sense / with denaturation sense / no denaturation

RNA RNA + DNA DNA no signal

ISH vs. Immunochemistry • well established methodology • simpler technology • Questions – – – – – –

does an antibody exist how well is the specificity of the antibody defined is antigencity maintained is there antigenic cross reactivity what is the abundance of protein vs message is the site of protein localization equivalent to the site of protein synthesis

In Situ Hybridization • Infectious Disease • Cytogenetics • Gene Expression

Infectious Disease • Target organisms – Organisms that are difficult, dangerous or impossible to culture • • • •

Viruses Bacteria Parasites Fungi

Cytogenetics • chromosomal • interphase

Gene Expression • • • • •

oncogenes tumor markers hormones / receptors enzymes structural markers

Advantages of In Situ Hybridization • • • • • • • •

conserves tissue preserves morphology sensitive and specific single cell /subcellular analysis archival material potentially rapid potentially very quantitative correlates with pathology

In situ hybridization factors • • • • • • • • • • • •

probe type probe complexity probe size probe specific activity probe concentration hybridization conditions stringency target type target abundance target availability tissue processing detection systems

Probe Factors • • • • •

probe type probe complexity probe size probe specific activity probe concentration

Types of Probes • • • •

Double-stranded DNA Single-stranded DNA RNA oligonucleotide

Signal vs. Probe concentration

Technical Effects on Signal Technical Effects on Signal proteolysis

stringency

kinetics exposure/development

probe size

Detection Systems • autoradiography • nonradioactive

Labelling Methods • Isotopic – – – – –

3H 35S 32P 14C 125I

Nonisotopic biotin fluorochromes enzymatic coupling immunologic targets enzyme cascades

Isotopic Methods • Advantages – increased sensitivity – control of specific activity – easily quantitated

• Disadvantages – probe instability – autoradiography – biohazard

Nonisotopic methods • Advantages – – – –

probe stability variety of detection methods rapid turnaround no biohazard

• Disadvantages – – – –

relative sensitivity probe / detection system penetration variety of detection methods reagent nuclease contamination

Disadvantages of In Situ Hybridization • • • • • •

SENSITIVITY labor intensive possibility of sampling error special fixation / handling exposure times protocol variability

Clinical vs Research • competing methodologies • laboratory reliability • research as to clinical utility

EXAMPLES

In Situ Hybridization