Indian Journal of Ophthalmology

CURRENT OPHTHALMOLOGY
Year
: 1998  |  Volume : 46  |  Issue : 1  |  Page : 3--13

Molecular biologic techniques in ophthalmic pathology


B Rajeev, J Biswas 
 Department of Ophthalmic Pathology, L.V. Prasad Eye Institute, Hyderabad, India

Correspondence Address:
B Rajeev
Department of Ophthalmic Pathology, L.V. Prasad Eye Institute, Hyderabad
India

Abstract

The polymerase chain reaction (PCR) and nucleic acid hybridization assays are recently introduced molecular techniques that allow for the identification of extremely small quantities of specific nucleic acids. These techniques have significant advantages over more conventional laboratory techniques, but also have some limitations. They are bound to have tremendous potential in diagnostic ophthalmic pathology and also in investigative pathology for deciphering the pathophysiology of ocular diseases. Despite their increased sensitivity and specificity, the results will still have to be co-related with clinical findings for maximum impact. For an ophthalmologist to derive maximum benefit, knowledge of these techniques, and their advantages, and limitations is essential. This article describes the basic concepts of molecular biology and the techniques of PCR, nucleic acid hybridization, and immunohistochemistry.



How to cite this article:
Rajeev B, Biswas J. Molecular biologic techniques in ophthalmic pathology.Indian J Ophthalmol 1998;46:3-13


How to cite this URL:
Rajeev B, Biswas J. Molecular biologic techniques in ophthalmic pathology. Indian J Ophthalmol [serial online] 1998 [cited 2024 Mar 28 ];46:3-13
Available from: https://journals.lww.com/ijo/pages/default.aspx/text.asp?1998/46/1/3/14984


Full Text

Molecular biology is a recent scientific development that has already had a profound impact on medicine including basic and clinical research. In the past decade these advances have been applied to diagnostic pathology in almost all medical subspecialties including ophthalmology. These include processing of biopsy material with newer techniques such as immunohistochemistry, in situ hybridization, and polymerase chain reaction. These techniques now enable a definitive diagnosis in cases which were previously labeled as idiopathic despite histopathological evaluation. Coupled with this are the advances in ophthalmic surgical techniques which have enabled removal of tissue (biopsy, fine needle aspiration) with minimal complications or functional loss. This combination of improved tissue collection and optimal utilization with newer techniques provides a high diagnostic yield when compared to conventional techniques. The conventional techniques such as routine histopathology, histochemistry, electron microscopy, microbial cultures, and antibody estimation, have some limitations when compared to the newer molecular techniques but will continue to have an important role. The small size of ophthalmic biopsy specimens such as corneal tissue, aqueous, vitreous, and chorioretinal tissue, demands a high level of co-ordination between the ophthalmologist and ophthalmic pathologist to decide how the specimen can be processed and also an understanding of the advantages and limitations of all available techniques.

The aim of this article is to highlight the principles of molecular biology for the practicing ophthalmologist and to summarize the contributions this technology has made to diagnostic ophthalmic pathology. Although immunohistochemistry is not strictly a 'molecular' technique, it is included in this article to complete the list of recent advances in diagnostic pathology.

 Basic Concepts in Molecular Biology[1],[2]



A few concepts in molecular biology are vital for understanding the techniques described later. Genetic information is contained in deoxyribonucleic acid (DNA) which consists of a series of nucleotides (bases) with a specific sequence necessary to encode a particular protein (the genetic code). The bases found in DNA are the purines adenine (A) and guanine (G), and the pyrimidines thymine (T) and cytosine (C). The DNA molecule in chromosomes consists of two long complementary strands of nucleotides held together in a double helix by specific hydrogen bonding between the nucleotides. The expression of this genetic information is via the processes of transcription (DNA to messenger RNA [mRNA]) occurring in the nucleus, and translation (mRNA to protein) in the cytoplasm. New DNA strands are synthesized by copying a preexisting strand according to the rules of complementary base pairing (A always binds to T and G binds to C). During DNA replication, the two strands of the double helix separate and each strand acts as a template for synthesis of a new complementary strand.

Recombinant DNA technology, which is an important part of molecular biology, consists of a variety of laboratory techniques that enable the isolation, duplication, manipulation, amplification, and analysis of DNA in vitro. An array of enzymes are available that allow intricate manipulation of DNA. These enzymes include restriction endonucleases, ligases, polymerases, and reverse transcriptases. Restriction endonucleases (usually obtained from bacteria) recognize and cleave DNA molecules only at or near a very specific site, usually at a certain sequence of 4-6 nucleotide bases. For example, the enzyme EcoR1 cleaves DNA only at the exact base sequence GAATTC. The fragments that result from cleavage by a restriction endonuclease are characteristic and specific for that DNA molecule. These fragments can be sorted by size using gel electrophoresis and a "map" or "fingerprint" can be obtained. DNA polymerase catalyzes DNA synthesis by adding nucleotides to an existing DNA strand such as the primer used in the polymerase chain reaction. Reverse transcriptases (obtained from retroviruses) use a single stranded RNA as a template for synthesis of the complementary DNA (cDNA) molecule. This enzyme is useful in obtaining a DNA sequence from a messenger RNA molecule.

The technique of cloning involves multiplying a DNA sequence of interest in bacteria such as E. coli by using bacterial plasmids or bacteriophages. This technique provides abundant, pure, and defined sequences of DNA which can be used as probes for DNA hybridization assays. During cloning markers (labels) can be incorporated in the DNA to enable detection in hybridization assays. Sequencing involves techniques used to determine the exact sequences of bases in a cloned DNA fragment.

The double-stranded DNA molecule can be denatured (unwound) into single-stranded DNA by heating or with alkaline pH. These procedures break hydrogen bonds which hold the two complementary DNA strands together but do not affect the covalent bonds, thus preserving the base sequence of the relatively stable single-stranded DNA molecule. The two complementary single strands will anneal (renature, rewind) by reforming the hydrogen bonds when the temperature is lowered with appropriate pH and ionic strength conditions. Single strands of DNA will locate and renature with their complementary sequence even in the presence of large numbers of noncomplementary DNA molecules. The base sequence of two DNA molecules need not be exactly complementary for annealing to occur; however, the more exactly complementary the base sequence is, the more stable are the reformed double strands. Using this property, one can accurately detect a specific DNA fragment from within the entire genome using the technique of hybridization with a labeled DNA probe.

 Immunohistochemistry[3]



The practice of pathology began with routine histology involving the examination of tissue sections with the conventional hematoxylin and eosin staining. In this era interpretation and diagnosis was predominantly based on tissue and cellular morphology and in situations where histological diagnosis was in doubt, the final resort was the opinion of a senior colleague or comparison with published descriptions and photomicrographs. Subsequently the field of histochemistry, based on the chemical reactions with cellular or tissue components, further enhanced the pathologist's interpretive ability. In the early 1940s the more reliable and objective technique of immunohistochemistry was introduced. This technique enabled the localization and identification of cellular or tissue constituents (antigens) by means of antigen-antibody interactions. Since its inception this technique has been standardized and simplified for routine use in a pathology laboratory.

 Technique



Immunohistochemistry uses the property of specific binding between an antigen and an antibody. An antibody labeled with a visible marker is used to detect an antigen (usually a protein) in tissue sections (cellular components or microorganisms). It is possible to prepare specific antibodies against any tissue or cellular element and microorganisms that have an antigenic determinant. These antibodies then can be used as specific probes to localize the corresponding antigen within tissue and cell preparations.

Antibodies are named according to their corresponding antigen or specificity (for example, ant-keratin). Other parameters used to describe an antibody include the host from which the antibody is derived, monoclonal or polyclonal nature of the antibody, and whether the antibody is labeled with a marker. For example, biotinylated goat anti-human CD4 antibody means this antibody is labeled with biotin, is derived from goat serum (following sensitization with the human CD4 antigen) and is specific to the human CD4 molecule (on helper T-cells). In the lymphoid system the CD nomenclature is used to describe the antigens on the cell surface and to identify cells of the lymphoid system. Antibodies (immunoglobulin molecules) can serve both as antibodies, binding specifically to tissue antigens, and also as antigens, providing antigenic determinants against which (secondary) antibodies may be raised. This is exploited in immunohistologic techniques to achieve the remarkable specificity, versatility, and simplicity that immunohistologic methods enjoy. Monoclonal antibodies are preferred as they are more specific than polyclonal antibodies which can bind nonspecifically to unrelated tissue antigens and give false positive results.

The simplest immunohistochemical method is to conjugate a label directly to the antibody against a given antigen. The conjugate (labeled antibody) is then applied to the tissue section in a single step, the direct technique [Figure:1]. Although simple, this is not a sensitive technique and does not allow the antisera to be used in other methods. To improve this, an indirect technique [Figure:1] can be employed. Here the tissue section is incubated with an unlabelled (primary) antibody, followed by a secondary antibody conjugated with the label of choice, raised against the species of the primary antibody. This allows a single labeled antibody to be used against a large number of primary antibodies provided they are from the same species. One practical disadvantage of the direct technique is that for the detection of different antigens it is necessary to conjugate separately the appropriate primary antibodies. Also, with regard to expensive antibodies, the direct conjugate procedure usually demands that the primary antibody be utilized at a relatively high concentration in comparison with the indirect technique. Additionally, since in the secondary antisera a number of antibodies will bind to the primary antibody, there is an increase in sensitivity. The advantages of the indirect system include, increased versatility since a single conjugated antibody can be used with several different primary antibodies, the conjugation process has to be applied only to the secondary antibody, the primary antibody can often be used at a higher working dilution to achieve successful staining, the secondary antibody is easily prepared reducing costs, and it allows additional specificity controls providing a valuable assessment of the validity of any staining pattern observed.

Once antibody molecules bind to the corresponding antigen in tissues, they cannot be seen with the light microscope or even with the electron microscope unless they are labeled by some potentially visible marker. A variety of labels are available including fluorescent compounds (immunofluorescence), enzymes that can be visualized by virtue of their property of catalyzing a colored reaction product from a suitable substrate system, and proteins such as biotin [Table:1].

The immunofluorescence method needs a fluorescence microscope for visualization of the antigen-antibody complex, does not allow for prolonged storage, and formalin fixed tissue may not be ideal for this technique due to autofluorescence of formalin fixed tissue. The sections must be photographed to make a permanent record. This technique is therefore only uncommonly used in diagnostic pathology and is largely restricted to frozen sections with poor morphologic resolution.

The immunoperoxidase technique [Figure:2] which uses antibodies labeled with the enzyme horseradish peroxidase, is much more convenient to use on ophthalmic specimens as it can be done on formalin fixed paraffin embedded tissue. This technique does not require fluorescence microscope and the stained sections can be stored for a long period of time. Currently horseradish peroxidase in conjunction with diaminobenzidine and hydrogen peroxide as the substrate (produces a dark brown color) is being widely used to demonstrate at least some antigens in routinely processed (formalin-paraffin) tissue. Due to the presence of melanin in some of the ocular tissue, another substrate, AEC (3-amino-6-ethyl carbazole) which yields a brick red color is preferred so that it can be differentiated from melanin without difficulty.

In the avidin-biotin conjugate (ABC) technique [Figure:3], the secondary antibody is labeled with biotin (primary antibody is labeled in the direct technique). The antigen-antibody complex is then detected by using the strong affinity between biotin and avidin. The complex of avidin, biotin and the enzyme peroxidase attaches to the biotin molecule on the antibody. The product is visualized by adding a chromogen as in the immunoperoxidase technique. As is obvious from [Figure:3], the ABC complex serves to localize several molecules of horseradish peroxidase at the site of the antigen thus increasing the sensitivity of this technique.

Compared to routine histopathology staining immunohistochemistry is a more complex technique and therefore prone to more errors. Immunohistochemistry results must therefore always be interpreted along with appropriate controls. Usage of both positive controls (tissues known to possess the antigen under investigation) and negative controls (using non-immune serum instead of primary antibody) makes the procedure very precise. The positive control validates not only the steps of the procedure but also confirms the preservation of the antigen under investigation. Some antigens may be denatured by fixation and processing and rendered non detectable. As far as possible the positive control section should be taken from tissue processed in a manner identical to that of the tissue under study.

 Variations



Two or more antigens can be simultaneously detected in the same tissue section using antibodies with different specificity (no cross reaction) and with different substrates (colour). Using this technique multiple simultaneous infections can be detected in ocular tissue.[4] Similarly, immunohistochemistry can be combined with in situ hybridization.

Immunoelectron microscopy combines immunostaining with ultrastructural evaluation of tissue. This is achieved by rendering the end products electron-dense by using appropriate reagents or using labels that are visible directly by electron microscopy, such as gold.

 Advantages



Compared to the conventional techniques, immunohistochemistry permits greater diagnostic specificity and accuracy at a reasonable cost. Antigens including nuclear, cytoplasmic, and membrane bound substances can be exactly localized and simultaneously evaluated along with morphology. This advantage is not apparent with the newer molecular techniques such as PCR and in situ hybridization.

 Limitations



Immunohistochemistry although a quite sensitive and specific histologic method in diagnostic pathology, has certain limitations. It is important to correlate the immunohistochemical results with the tumor morphology based on haematoxylin- eosin and other special stains. Several factors are known to interfere with a immunohistochemical study which include tissue fixation (type, duration of fixation), processing (time, temperature, reagents), concentration, specificity and sensitivity of the primary antibody and the detection system. In addition other factors such as extensive necrosis or crushing of the tissue can give false positive results. Also, several artifacts can occur on immunostaining including staining at the edge (edge artifact) due to higher concentration of the antibody, tissue drying and improper fixation. Frozen unfixed tissue can pose an infection hazard.

 Ophthalmic applications



The immunohistochemical studies described above can be done on biopsy specimens, smears obtained from aqueous or vitreous aspirates and impression cytology specimens. This technique can also be used on fine needle aspiration biopsy smears for individual cell recognition. These types of cell preparations can be subjected to minimal or carefully controlled fixation, and are therefore particularly suitable for demonstration of a wide range of antigens by immunocytochemical methods.

It is important to note that fixation and processing techniques can alter some antigens and render them non detectable. It is therefore important to liase with the ophthalmic pathologist to identify the best method for detecting the antigen in question. Antibodies which can be used on fixed tissue allow better morphological evaluation and also permit study of archival material. For formalin fixation, it is best to use the minimum fixation time that is just enough to fix the tissue and then transfer the tissue to a suitable holding medium. In situations where faint or negative staining is seen, some of the antigenicity may be restored by incubating the paraffin sections with proteolytic enzymes, or by the use of the antigen retrieval techniques to unmask the antigens. Cryostat sections (frozen tissue) are employed when the antigen under investigation is formalin-labile, when tissue morphology is not at a premium, and occasionally when a rapid technique (immunofluorescence) is required. Immunofluorescence technique on cryostat sections is commonly used to detect immunoglobulin deposition in ocular cicatricial pemphigoid.

In ophthalmic pathology, immunohistochemistry[5] is most commonly used to diagnose and classify tumors [Figure:4] and [Figure:5]. The antibody L26 for B cells and.UCHL1 for T cells can be used to aid the diagnosis of ocular B cell lymphoma [Figure:6]. Antibodies to immunoglobulin light chains (kappa and lambda) can be used to identify the monoclonal status of lymphocytes in an orbital lymphoid lesion. Since each lymphocyte produces immunoglobulins with either kappa or lambda light chains, positive staining with antibodies to either kappa or lambda only would indicate a monoclonal proliferation and hence a lymphoma [Figure:7]. A reactive lymphoid lesion will show equal staining with both antibodies. Immunohistochemistry can also be used to diagnose ocular infections including viral retinitis [Figure:8] and [Figure:9].[6,7] Uncommon applications that have been described include localization of herpes viral antigens in herpetic keratitis,[8] and correlating corneal amyloidosis with clinical manifestations using antibodies specific to the various types of amyloid proteins, including amyloid A protein, amyloid P, and amyloid from immunoglobulin light chains.[9,10] A list of antibodies used commonly in the ophthalmic pathology laboratory is provided in [Table:2].

 Polymerase Chain Reaction (PCR)



PCR[11,12] is a new molecular biology technique which involves enzymatic amplification of a specific sequence of DNA. This technique was first described by Kary Mullis and co-workers in 1985.

 Technique



The first step in PCR involves the identification of the target DNA sequence to be amplified. Before PCR can be performed the target DNA sequence must be known so that a set of complementary nucleotide primers (single stranded) that flank the target sequence can be designed and synthesized. These primers are usually 18-22 nucleotides in length and their sequence is very critical for specificity. Each primer is complementary to one of the two strands of the target DNA. The tissue which is to be analyzed for the presence or absence of the target DNA is processed and all the DNA present in it is extracted. This sample DNA is added to a small tube along with the two primers, the enzyme Taq polymerase (obtained from Thermus aquaticus, a thermophilic bacterium), and all four deoxyribonucleotides (dATP, dGTP, dTTP, and dCTP) in an appropriate amplification buffer. The process consists of a set of cycles (denaturation, annealing, elongation) each of which doubles the number of desired molecules in the reaction [Figure:10]. Thus, the number of molecules increase geometrically and after 20 cycles over one million product molecules have been formed for each molecule of template DNA in the starting mixture.

The tube is placed in a thermo-cycler which is programmed to cyclically increase and decrease the temperature for a specified number of cycles. PCR begins when the temperature is raised to 94� C. At this temperature the double standard DNA from the original sample denatures. The temperature is then lowered (40-60� C) to allow the primers to anneal to a specific site on the target DNA dictated by the sequence of the primer. Since the primer is added in excess quantity, this annealing is preferred compared to the annealing of the original double strands. The primers flank the target sequence and are so oriented that DNA synthesis proceeds across the regions between the primers. The DNA polymerase enzyme in the mixture begins elongation (at 72� C) of the primer to produce a complementary copy of the target DNA. This cycle is then repeated and the DNA strand synthesized from the primer itself acts as a template for another primer to bind in the next cycle. On an average, the cycles are repeated about 30 times to produce millions of copies of the target sequence (between the primers) originally present in the sample. In the absence of the specific target sequences, amplification of the DNA will not occur.

To detect and confirm the specificity of the amplified PCR product, the reaction mixture is usually electrophoresed through a agarose gel (and stained with ethidium bromide) to separate it from the the remaining sample and extraneous DNA that may be amplified during the PCR. This non-specific amplification can usually be differentiated from the targeted DNA by identification of a band on the gel at the molecular weight predicted by the length of the targeted DNA segment. More specific verification can be achieved by blotting the gel onto a membrane of nitrocellulose or nylon for hybridization with a known probe (Southern blot). This can then be photographed to provide permanent documentation of the result.

A positive control and a negative buffer control is always included in each reaction. Positive control should ideally be a DNA segment that is expected to exist in the sample tissue.

 Variations



A number of variations of the PCR reaction have been developed to maximize its utility in specific applications. Nested PCR is of particular value in detection of very small quantities of target DNA in clinical samples. In nested PCR an inner and an outer set of primers are used. PCR is first performed using the outer set. A portion of the first amplification is then reamplified using the inner set of primers. This technique increases sensitivity and specificity. In another modification, mRNA sequences can be amplified using the PCR. First, DNA is obtained from the mRNA by using reverse transcriptase (RT). This DNA is then amplified using the regular PCR technique. This modification is known as RT-PCR. Presence of mRNA indicates activity of that corresponding gene or infective agent. Due to the transient and labile nature of mRNA this technique tends to be more difficult. RT-PCR can be used to differentiate active from latent infection as in herpes simplex keratitis.[13]

Although PCR can be used to amplify DNA even from samples of formalin-fixed, wax-embedded tissue, a limitation of PCR, especially for the histopathologist, is that it has not been possible to localize the amplified DNA sequences in cells or tissue sections. The technique of in situ PCR combines PCR on tissue sections with in situ hybridization and permits the localization of specific amplified DNA segments within isolated cells and sections of tissue.

 Advantages



The major advantage of the PCR is to detect the DNA of the micro-organisms which cannot be cultured easily (such as viruses) or take time (for example, mycobacterium). [Table:3] lists some of the microorganisms that can currently be detected by this technique. Cultures still have a role to play since they confirm viability and can be followed by antibiotic sensitivity testing.

Compared to antibody estimation in serum or ocular fluids, PCR has the advantage that is detects the microbial DNA itself, and is much more sensitive and specific (unlike cross reactions with antibody estimation). However, unlike PCR, serial antibody titers can be used to monitor the patient's response to treatment.

DNA retrieved from archived tissue fixed in formalin and stored for several years can be used to identify biological agents (viral, bacterial, parasitic) or alterations in cellular DNA.

 Limitations



The extreme sensitivity of PCR also constitutes its greatest drawback, namely, the problem of generating false-positive results owing to sample contamination. With proper precautions, this problem can usually be avoided. For this reason PCR is likely to be restricted to well equipped laboratories with trained personnel. PCR does not distinguish viable from nonviable organisms and is not yet truly quantifiable. It is therefore difficult to assess the relevance of a positive PCR especially in locations such as the conjunctival sac which normally harbors bacteria. Similarly, PCR cannot differentiate between active and latent infection. To a certain extent RT-PCR to detect mRNA may solve this problem.

Unlike with microbial cultures, with PCR and the other techniques described in this article, one gets only what is looked for. A few culture media will suffice to detect and grow most microorganisms, but with PCR, the technique will have to be repeated for each microorganism that is suspected. One way of minimizing this drawback is to use primers against DNA sequences that are conserved across species (for example, herpes group) and once detected, use primers against the non-conserved portions to differentiate within this group (herpes simplex virus [HSV], herpes zoster virus [HZV], cytomegalovirus [CMV], Epstein-Barr virus [EBV]).

Lastly, PCR can detect only those agents for which the DNA sequence and primers are known. Also, it does not provide cellular morphology and localization.

 Ophthalmic applications



The tissues which can be submitted for PCR include intraocular fluid (aqueous and vitreous), tears, any fresh ocular tissue, formalin fixed or paraffin embedded tissue, and even stained or unstained cytology slide or tissue sections from which DNA can be extracted. PCR for identifying infectious agents can be performed on as little as 1 μL of aqueous, vitreous, or tear fluid. Therefore, 0.1 mL (100 μL) of specimen obtained by anterior chamber tap or paracentesis, and vitreous biopsy can be used to perform PCR for multiple organisms.

The most common applications of PCR include the identification of pathogenic organisms that are difficult to culture or where tissue available is small. PCR can also add to the current understanding of the pathophysiology of ocular diseases. Using PCR a possible link between the iridocorneal endothelial syndrome and the herpes simplex virus has been reported.[14]

Improvements in antibiotic and antiviral therapy for specific organisms and in their delivery techniques makes it imperative that PCR be used more often for diagnosis in patients with uveitis and retinitis that present atypically or that do not respond to therapy based on clinical judgment. The use of PCR in such situations has been reported, including detection of herpes group of viruses,[15] toxoplasmosis,[16],[17] and mycobacterium tuberculosis.[18]

Since the PCR is more sensitive than microbial cultures and special histological stains for microorganisms, it may be the only method of identifying the presence of an infectious organism. Hykin et al[19] showed that nested PCR could detect Propionibacterium acnes in cases with endophthalmitis where culture and Gram's stain were reported negative.

A more likely potential application is to detect microbial agents in eyes with infective uveitis with non-specific clinical features especially panuveitis. Possible candidates would include leptospira, borellia, and brucella species.

 Nucleic Acid Hybridization



Nucleic acid hybridization is a relatively new technique that uses a segment of DNA (a DNA probe) to detect viral, bacterial, or human DNA of interest in a specimen.

 Technique



The principle involved in hybridization is explained in the section on basic concepts in molecular biology. When labeled single stranded DNA probe molecules are added to denatured DNA in a sample, annealing may result in "hybrid" double strands with one strand consisting of labeled probe DNA and the other strand from the original DNA in the specimen. By using assay conditions that are unfavorable for all hybrid molecules except the most stable, the hybridization reaction can be made highly specific for the target DNA.

Like immunoassays, all hybridization assays have common basic features and steps, and many variations. The clinical sample (tissue or fluids) may be treated to purify and extract the DNA (dot/blot hybridization), the hybridization assay can be done on tissue sections (in situ hybridization),[12] or the Southern blot technique can be used to detect DNA sequences that are integrated into host DNA.

For dot-blot assays [Figure:11], the entire extracted DNA is denatured, usually by heating; the single strands are attached to a solid phase, most commonly done at present by "dotting" the DNA solution onto nitrocellulose filter paper. Baking the filter permanently immobilizes the DNA on the filter. The labeled DNA probe is then added to the filter. If target DNA is present, the labeled probe molecules will locate and renature with their complementary sequences forming stable, double stranded hybrids in the filter. These can be detected because one of the strands contains label. In the absence of target DNA, the labeled single stranded probe molecules are not retained in the filter and are washed away. This technique can be designed to be either qualitative or quantitative. Dots are less informative and are mainly used to screen large numbers of specimens or to demonstrate changes with time by running multiple samples (at different times) side by side. Dots can be quantitated by densitometry.

In situ hybridization (ISH) utilizes specific labeled DNA sequences known as DNA probes, which are complimentary to the target DNA. Both paraffin-wax embedded as well as cryostat tissue sections can be used. The technique involves several steps: (i) denaturation of the DNA or RNA by exposing to high temperature or by alteration of pH, (ii) denatured DNA or RNA is then hybridized to DNA probe, and (iii) detection of the probe. These probes are labeled by radioactive isotopes, biotin, or immunoreactive groups so that the hybridized product can be detected by autoradiography, avidin-biotin enzymes, or colorimetric enzyme substrate system respectively. Nonradioactive probes are easier to use and therefore commonly adopted for diagnostic ophthalmic pathology. The tissue sections can be counter stained with hematoxylin for morphological evaluation. As with the previous two techniques, appropriate negative and positive controls are mandatory. Probe specificity should be confirmed with Southern blot analysis to ensure that the probe hybridizes to a sequence of the expected size. Nonspecific binding of the probe can be controlled by using a probe lacking homology with the target sequence or by using tissue without the target sequence as a negative control. A positive control can be tissue containing the target sequence. ISH is technically more demanding than dot-blot hybridization but permits tissue localization.

In the Southern blot technique, DNA is separated from homogenized tissue and cleaved with restriction endonucleases generating fragments of DNA of various sizes. These fragments are separated according to size by gel electrophoresis, and then transferred from the gel to a filter paper for hybridization assay.

Probes with as few as 20 nucleotide bases are statistically unique. In other words, if a sequence of 20 bases from a viral genome is selected, that exact sequence is highly unlikely to be found anywhere else in the human genome. DNA probes are obtained in two ways. First, the sequence can be obtained by cloning. Probes of essentially any length may be obtained in this manner. Second the probe can be synthesized (DNA synthesizer) in vitro from individual nucleic acid bases (synthetic oligonucleotide probes). Limitation of current synthetic procedures restricts the number of nucleotide bases of synthetic probes. For diagnostic pathology short DNA probes are preferred due to several advantages. The synthesis is simpler than the cloning procedure, shorter probes require shorter assay times, and short probes are more reliable in detection of point mutations in DNA. As with the PCR, the exact base sequence of the target DNA of interest must be known for in vitro synthesis of DNA probes. Longer probes increase the sensitivity since more label can be attached to the DNA molecule. Knowledge of the exact base sequence is not required to clone a DNA sequence.

DNA probes can be labeled in ways similar to antibodies in immunohistochemistry, such as with radioisotopes, biotin, or fluorescent molecules. Radioisotope probes require autoradiography or scintillation counting for detection and are generally the most sensitive methods. The radioisotope label is generated by enzymatically inserting purine and pyrimidine bases containing 32P into the probe. The disadvantages of radio-labeled probes are those of any isotopic assay-instability, short shelf life, and inconvenience of handling and disposing of radioactive molecules. Probes labeled with biotin are most widely used in clinical laboratories. Nucleotide bases covalently linked to biotin are enzymatically inserted into the probe. The biotin label is detected with avidin linked enzyme complexes resulting in a color reaction as described with the immunohistochemistry technique.

 Variations



Variations in the hybridization assays may include changes in probe synthesis method, probe length, probe label, probe specificity, etc. In the in situ hybridization technique, messenger RNA, which may be present in many more copies than DNA of a give gene, may be used as the target of the probe to increase the sensitivity. In situ hybridization technique to localize DNA or RNA, can be combined with immunohistochemical technique to localize a protein antigen on the same tissue section. Thus it is possible to anatomically localize the sites of peptide synthesis and distribution in tissues.

 Advantages



DNA hybridization assays have an overall advantage over the conventional techniques. These include yield of unique information not obtainable by other methods (for example, identification of antibiotic resistance genes), greater sensitivity and specificity, more rapid and economical, and less stringent requirements for specimen handling. Also, in some cells mRNA may be detected by ISH but the protein it encodes for may not be detectable by immunohistochemistry. ISH technique works on fixed and embedded tissue and may show better morphology than immuno-histochemistry on frozen sections.

Sensitivity may be advantageous in infections (for example, Epstein-Barr virus, Mycobacterium) for which the sensitivity of immunoassay (ELISA) is currently inadequate. For certain viruses, the sensitivity of hybridization assays is greater than viral culture. In addition to this, these assays can provide information relating to carrier status.

For viral culture, specimens must contain viable virus and therefore require careful handling, especially the fragile viruses such as cytomegalovirus. For immunoassays, antigenic structures (protein) must be maintained intact. DNA, in contrast, is stable and may persist in a specimen even after virus particles are no longer viable and viral proteins no longer retain their antigenic structure. DNA, even if fragmented in a degenerating specimen, may contain a long enough unique sequence to be recognized by a probe.

As with the PCR, hybridization provides the ability to select the DNA sequence for use as a probe depending on the desired specificity of the assay. In screening for a virus, a segment of the viral genome which is conserved among all strains may be used.

 Limitations



False-positive results owing to non-specific binding of the probe is possible. For in situ hybridization, fixation and tissue preparation (protein digestion to enable probe penetration) are critical steps and need standardization and stringent quality control. Both these can be minimized by using appropriate controls.

As with PCR, and unlike with microbial cultures, one gets only what is looked for. A few culture media will suffice to detect and grow most microorganisms, but with hybridization, a separate probe will have to be used for each microorganism that is suspected. One way of minimizing this drawback is to use probes against DNA sequences that are conserved across species (for example, herpes group) and once detected, use probes against the non-conserved portions to differentiate within this group (HSV, HZV, CMV, EBV).

Like PCR, hybridization assays can be used for only those agents for which the DNA sequence is known.

 Ophthalmic applications



DNA probes have been used to detect an ever growing list of microorganisms in clinical specimens. Assays on fluid specimens, including aqueous, vitreous, and serum are technically similar to those used to detect organisms in tissue. At present, the list of organisms for which hybridization assays have been developed includes the hepatitis B virus, CMV, EBV, HSV, adenovirus, human immunodeficiency virus, mycobacteria and chlamydia. In ocular infections, hybridization technique can be used as a diagnostic tool, to study the natural life cycle and spread of a virus, as well as detect a latent status. In situ hybridization has been used in the diagnosis of viral retinitis form retinal biopsy specimens.[20] DNA of human papilloma virus has been demonstrated in dysplastic and malignant lesions of the conjunctiva and cornea.[12]

Hybridization techniques have also been used extensively to study DNA in tissues. Southern blot analysis for DNA in tissues are being used for detection of integrated viral genomes in host tissue for diagnosis of latent viral infections, malignant transformation due to viral integration, detection of gene rearrangement for diagnosis and classification of lymphocyte malignancies, and detection of oncogenes in tissue.

In orbital lymphomas, DNA probes are used to detect clonal gene rearrangements of the immunoglobulin genes characteristic of B-cell lymphomas and clonal rearrangements of the T-receptor molecule characteristic of T-cell lymphomas.[21] The use of antibodies, especially monoclonal, to lymphocyte surface antigens has been of great value in diagnosing and classifying lymphoid neoplasms. However, interpretation of surface marker studies is sometimes difficult due to the possible presence of a large population of reactive cells. Use of hybridization probes to detect lymphocyte clones that have a single type of gene rearrangement for immunoglobulin/ T-cell receptor offers several advantages. The clonal rearrangement may be present before the characteristic surface antigen is expressed. Hybridization techniques are sensitive enough to detect as few as 1-3% tumor cells in a mixture of normal cells, a sensitivity that far exceeds that of surface marker studies.

Other applications include early prenatal and preclinical diagnosis of several metabolic/genetic diseases including retinitis pigmentosa, and easier and more accurate HLA typing.

References

1Della NG. Molecular biology in ophthalmology: a review of principles and recent advances. Arch Ophthalmol 1997;114:457-63.
2Watson JD, Gilman M, Witkowski J, Zoller M. Recombinant DNA. 2nd ed. New York: Scientific American Books;1992.
3Taylor CR. Principles of immunomicroscopy. In: Taylor CR, Cote RJ, editors. Immunomicroscopy: A Diagnostic Tool for the Surgical Pathologist. Philadelphia: WB Saunders Company: 1994. p 1-20.
4Skolnik PR, Pomerantz RJ, de la Monte, Lee SF, Hsuing GD, Foos RY. Dual infection of retina with human immunodeficiency virus type I and cytomeglovirus. Am J Ophthalmol1989;107:361-72.
5Messme EP, Font RL. Application of immunohistochemistry to ophthalmic pathology. Ophthalmology 1984;91:701-7.
6Pepose JS. Infectious retinitis:diagnostic modalities. Ophthalmology 1986;93:570-73.
7Biswas J, Madhavan HN, Badrinath SS. Demonstration of herpes simplex virus from lens aspirate in healed acute retinal necrosis (ARN) syndrome. Br J Ophthalmol 1997;81:802-3.
8Holbach LM, Font RL, Wilhelmus KR. Recurrent herpes simplex keratitis with concurrent epithelial and stromal involvement:immunohistochemical and ultrastructural observations. Arch Ophthalmol 1991;109:692-95.
9Loeffler KU, Edward DP, Tso MOM. An immunohistochemical study of gelsolin immunoreactivity in corneal amyloidosis. Am J Ophthalmol 1992;113:546-54.
10Matta CS, Tabbara KF, Cameron JA, Hidayat AA, al Rajhi AA. Climatic droplet keratopathy with corneal amyloidosis. Ophthalmology 1991;98:192-95.
11Adleberg JM, Wittwer C. Use of the polymerase chain reaction in the diagnosis of ocular disease. Curr Opinion Ophthalmol 1995;6:80-85.
12Garcia-Ferrer FJ, Blatt AN, Laycock KA, Pepose JS. Molecular biologic techniques in ophthalmic pathology. Ophthalmol Clin North Am 1995;8:25-36.
13Kaye SB, Lynas C, Patterson A, Risk JM, McCarthy K, Hart CA. Evidence for herpes simplex viral latency in the human cornea. Br J Ophthalmol 1991;75:195-200.
14Alvarado JA, Underwood JL, Green WR, Sarah WU, Murphy CG, Hwang DG, et al. Detection of herpes simplex viral DNA in the iridocorneal endothelial syndrome. Arch Ophthalmol 1994;112:1601-9.
15Biswas J, Mayr AJ, Martin WJ, Rao NA. Detection of human cytomeglovirus in ocular tissue by polymerase chain reaction and in situ DNA hybridization. Graefe's Arch Clin Exp Ophthalmol 1993;231:66-70.
16Elkins BR, Holland GN, Opremcak EM, Dunn JP, Jabs DA, Johnston WH, et al. Ocular toxoplasmosis misdiagnosed as cytomegalovirus retinopathy in immunocompromised patients. Ophthalmology 1994;101:499-507.
17Manners RM, O'Connell S, Guy EC, Joynson DH, Canning CR, Etchells DE. Use of the polymerase chain reaction in the diagnosis of acquired ocular toxoplasmosis in an immunocompetent adult. Br J Ophthalmol 1994;78:584-86.
18Kotake S, Kimura K, Yoshikawa K, Sasamoto Y, Matsuda T, Nishikawa T, et al. Polymerase chain reaction for the detection of mycobacterium tuberculosis in ocular tuberculosis. Am J Ophthalmol 1994;117:805-6.
19Hykin PG, Tobal K, Mclntyre G, Matheson MM, Towler HM, Lightman SL. The diagnosis of delayed postoperative endophthalmitis by polymerase chain reaction of bacterial DNA in vitreous samples J Med Microbiol 1994;40:408-15.
20Freeman WR, Wiley CA. In situ nucleic acid hybridization. Surv Ophthalmol 1989;34:187-92.
21Margo CE. Orbital and ocular adnexal lymphoma: evolving concepts. Ophthalmol Clin North Am 1995;8:167-77.