Chapter 3 Identifying Different Types of Chromatin Using Giemsa Staining Juan C. Stockert, Alfonso Blázquez-Castro, and Richard W. Horobin Abstract Mixtures of polychrome methylene blue-eosin Y (i.e., Giemsa stain) are widely used in biological staining. They induce a striking purple coloration of chromatin DNA (the Romanowsky-Giemsa effect), which contrasts with the blue-stained RNA-containing cytoplasm and nucleoli. After specific prestaining treatments that induce chromatin disorganization (giving banded or harlequin chromosomes), Giemsa staining produces a differential coloration, with C- and G-bands appearing in purple whereas remaining chromosome regions are blue. Unsubstituted (TT) and bromo-substituted (BT) DNAs also appear purple and blue, respectively. The same occurs in the case of BT and BB chromatids. In addition to discussing the use of Giemsa stain as a suitable method to reveal specific features of chromosome structure, some molecular processes and models are also described to explain Giemsa staining mechanisms of chromatin. Key words Chromatin, Chromosome banding, DNA staining, Eosin Y, Harlequins, Light microscopy, Thiazine dyes
1
Introduction Dyes are widely used in biochemical and biological investigations, as well as in medical diagnostic procedures. In particular they are applied as analytical tools for the localization and identification of cell and tissue structures [1, 2]. Molecular approaches to the design of dyes [3, 4] and staining procedures [5] have a considerable history. Efforts to achieve a molecular understanding of the mechanisms of such biological staining are indeed of even longer standing [6–9]. Various aspects of these approaches have been recently reviewed [10–12]. Given this considerable literature, it is both unfortunate and puzzling that biomedical professionals often do not understand, or are not interested in, the physical and chemical aspects of biological staining methods. The contents of this chapter, and others in this volume, attempt to help improve this situation.
Juan C. Stockert et al. (eds.), Functional Analysis of DNA and Chromatin, Methods in Molecular Biology, vol. 1094, DOI 10.1007/978-1-62703-706-8_3, © Springer Science+Business Media New York 2014
25
26
Juan C. Stockert et al.
Fig. 1 Chemical structures of azure B and eosin Y. In the latter dye the planes of the xanthene chromophore and that of the pendant phenyl group are nearly perpendicular to each other
It is known that certain staining procedures used for microscopy give differential coloration to chromatin. Such selective reactions depend on a variety of mechanisms. These include ortho- and metachromatic colors due to binding of dye monomers (intercalation) or dye aggregates (stacking) to DNA, physical phenomena which arise both with acridines [13] and thiazines [14, 15]. Other processes, such as nuclear staining by cuprolinic blue [16] and ruthenium red [17], involve chemical redox reactions. Finally, multi-dye systems such as Giemsa staining [18, 19] are also used. Although the mechanisms of differential staining by these dyes are not yet precisely understood at the molecular level, probably involving both physical and chemical parameters, it is likely that differential chromatin reactivity—“stainability”—also results from specific prestaining treatments. Multi-dye staining mixtures involving azure B plus eosin Y, the topic of this chapter, have been described by a diversity of names. For an amusing summary of this, see the title of a paper by Krafts [20]. The present paper uses the most common generic term for such stains, namely, “Giemsa.” Note however that the purple staining of chromatin characteristic of this stain has been termed the “Romanowsky-Giemsa effect.” 1.1 Giemsa Staining Mechanism
Giemsa stains usually contain polychrome methylene blue plus eosin Y. The “polychrome” epithet describes the presence of methylene blue plus its demethylated homologues, in particular azure B; see Fig. 1 for key dye structures. A recent review on the history and application of these dye mixtures has been published [21].
Differential Giemsa Staining of Chromatin
27
Giemsa stains are currently applied in bacteriology, cytogenetics, cytology, hematology, histopathology, and parasitology [12, 22, 23] to reveal basophilic cytoplasm, chromatin, ortho- and metachromatic cytoplasmic granules, and nucleoli in different colors. Typically, condensed chromatin DNA (in nuclei and chromosomes) gives rise to a purple (magenta) color, whereas RNA-containing basophilic cytoplasm and nucleoli appear in blue [24, 25]. The Giemsa stain is an important polychromatic staining method for peripheral blood, bone marrow, cervical smears, Plasmodium, Leishmania, and Trypanosoma parasites, as well as for banded and harlequin metaphase chromosomes. Some of the diverse end users routinely use the terms Jenner, Leishman, MayGrünwald, or Wright stains rather than Giemsa. Properties and applications of such dye mixtures, as well as their individual thiazine and eosin components, have been described [1, 26, 27]. The principles and methodology of Giemsa staining for cytogenetics, cytopathology, and hematology—with particular reference to equipment, reagents, and troubleshooting—have been recently reviewed [28–30]. In the present chapter, only the differential Giemsa staining of chromatin will be considered. Differential coloration of chromatin by thiazine dyes and Giemsa stain is influenced by dye concentration, staining time, local availability of binding sites, and different diffusion rates of dyes [12, 25, 31, 32]. The purple color arising during Giemsa chromatin staining, with an absorption peak around 550 nm, originates from a complex between cationic thiazines and eosin Y, which is formed preferentially on certain polyanions such as the DNA present in chromatin [12, 25]. Wittekind et al. [22, 33] demonstrated directly that a Giemsa-type stain could be prepared by mixing two pure dyes, namely, azure B (AB) and eosin Y (EY). It is known that the purple color of Giemsa-stained nuclei and chromosomes (whether banded or not) arises from the precipitation of a hydrophobic thiazine eosinate complex with a molar ratio 2:1 [31, 34]. This complex would be the same compound that slowly precipitates from aqueous Giemsa solutions, in which strong hydrophobic interactions occur between stacked thiazine and eosin dyes. Interestingly, after extraction from chromatin and DNA, the ratio thiazine/eosin Y is 2.3:1 (chromatin) and 2.8:1 (DNA) [31]. The thiazine dye present in excess of the 2:1 ratio probably corresponds to additional DNA intercalated dye (which would be 3:1 at saturation). The precise structure of the DNA-AB-EY complex remains unknown, but possible structures can be considered. Early views, with eosin Y attached directly to two intercalated thiazines [31], are now disregarded [35]. In any event, the sequence of the process is as shown in Fig. 2. First, DNA intercalation (a, b) and external (c) dye binding occur, followed by formation of a 2:1 AB-EY complex on DNA (d) with an orthogonal relation between
28
Juan C. Stockert et al.
Fig. 2 Schematic representation of DNA, DNA-AB, and AB2-EY complexes (AB: azure B, EY: eosin Y). Thick vertical lines are negatively charged deoxyribosephosphate chains. (a) DNA (bp: base pairs). (b) DNA with intercalated thiazines (iT). (c) Intercalated DNA with additional external thiazines (eT). (d) DNA-AB2-EY complex based on that described by Friedrich et al. [35], with the suggested orthogonal position of xanthene (X) and thiazine rings; ph: EY phenyl group. (e) AB2-EY complex after dissociation (arrow) from DNA, showing stacked coplanar xanthene and thiazine chromophores. In (c–e) only one side of external DNAdye interactions is shown for clarity
chromophores [35]. It must be noted that in this case, a DNA-AB complex with 1:1 phosphate-AB stoichiometry was first obtained and subsequently stained by EY. However the purple compound formed by one eosin Y and two thiazine dye molecules does not necessarily remain attached to DNA, and purple AB-EY that precipitates on its own might cause the color reaction of chromatin [25]. If this were so, two thiazine dyes should first bind to DNA phosphates at the correct distance apart, in agreement with early studies on thiazine binding to phosphate groups; see Bergeron and Singer [36]. Subsequently they would combine with one EY to form the purple AB-EY complex that precipitates, thereby freeing DNA phosphates to bind more thiazines and EY again [25, 34] (Fig. 2e). This cyclic self-catalytic process (DNA → DNA-AB2 → DNAAB2-EY → DNA + AB2-EY) would be responsible for the intensification of the purple staining of chromatin DNA (and other polyanions) by a “template effect” [12]. The continuous accumulation of precipitate by self-assembly of the AB2-EY
Differential Giemsa Staining of Chromatin
29
Fig. 3 Frontal (a) and lateral view (b) of a molecular model of two units of a proposed azure B-eosin Y (AB2-EY) complex
complex on a nanostructured polyanionic scaffold can explain the color enhancement effect which facilitates the visualization of certain small purple stained structures such as centromeres and chromosome bands. To illustrate the possible molecular structure of the AB2-EY complex, a simple and plausible model was constructed (Fig. 3). Geometry optimization was performed with the molecular mechanics force field (MM+) method using HyperChem 8 software and Polak-Ribière algorithm converged at a conjugate gradient of 0.1 kcal/(Ǻ mol). Fig. 3a, b shows two face-to-face related AB2-EY complexes in which each unit is a sandwich of two AB dyes (with S atoms in cis orientation) and one EY (with a heterocyclic electronegative O atom stacked between the electropositive S atoms). In this model strong electrostatic, hydrophobic, and π-electron interactions occur between the planar tricycle rings of AB and EY chromophores, which show identical aromatic size as assessed by the largest conjugated fragment values of 18 (see Horobin, ref. 8) and opposite partial charges. Although this is the optimal geometry, it may be noted that other AB-EY arrangements are possible. One wonders if other alternative complexes might be responsible for the historically well-known [7] “carmine” or “red” (i.e., not purple) nuclear coloration of Giemsa-stained malarial parasites. As to why this might occur, the Plasmodium falciparum genome is known to contain the unusual feature of “super AT-rich regions” [37]. 1.2 Differential Giemsa Staining
Native DNA is the substrate required to induce purple Giemsa staining in normal chromatin. However following various specific treatments some regions of chromatin DNA become modified (mainly disorganized), resulting in differential staining. Purple and
30
Juan C. Stockert et al.
blue chromatin colors were first reported after staining with Giemsa at high pH [38, 39] or following specific pretreatments [18, 19]. Although a differential color reaction in chromosomes seems to be a simple model for Giemsa staining, no clear relation to the previous treatment and staining conditions is apparent. Note, for instance, the contrasting results in the case of differentially bromosubstituted chromatids when subjected to other procedures such as direct staining with alkaline Giemsa solution [40], treatment with hot perchloric or hydrochloric acid [41], or DNase digestion after Hoechst 33258 plus UV treatment [42], which give a reverse staining pattern in comparison with that produced by the conventional fluorescence-plus-Giemsa (FPG) method. 1.2.1 G- and C-bands
It is known that G- and C-banding are related in a sequential manner to the progressive disorganization (e.g., decondensation, dispersion, degradation) of chromatin [43, 44]. Formation of the purple complex was originally postulated to explain the staining of chromosome G-bands [31, 34]. However, G-bands can be seen in the absence of eosin Y, and most thiazine dyes when applied alone give a metachromatic G-banding pattern [32]. In these cases, G-bands appear highly stained without formation of a purple eosin Y containing compound. After appropriate pretreatments and Giemsa staining, C- and G-bands appear in purple, whereas the remaining (non-banded) chromosome arms are blue. However, there is no significant difference in DNA content between both regions [31, 34]. After a staining time long enough to produce purple bands, other regions are still blue [18]. This suggests that changes in the chromatin compaction and DNA strandedness have occurred, with an overall decrease in chromatin packing. Changes in color and staining intensity of C-bands from mouse chromosomes are shown in Fig. 4a, b. It is known that treatments that produce C-bands cause considerable swelling of chromatin, with concomitant protein extraction and DNA denaturation [18, 31]. Highly repetitive adenine-thymine-rich DNA located in the centromeric heterochromatin of mouse chromosomes (see ref. 45) reassociates very rapidly [13], and then the occurrence of purple and blue staining at a specific time would indicate differential DNA strandedness [18]. Interestingly, the centromeric heterochromatin of non-banded and C-banded chromosomes shows an opposite Giemsa staining reaction. It is pale violet-blue or blue in nonbanded chromosomes, but becomes purple in C-bands. The contrary occurs in the chromosome arms. The highly compact centromeric heterochromatin would not allow the rapid formation of the purple compound, and at short staining times it stains blue mainly by thiazine monomers. After procedures involving protein extraction, DNA denaturation, and loss of chromatin compaction, C-bands stain purple. In this context, it is known that sperm nuclei
Differential Giemsa Staining of Chromatin
31
Fig. 4 (a, b) Giemsa-stained mouse metaphase chromosomes before (a) and after (b) C-banding. Arrows indicate weakly stained centromeric heterochromatin (a) and purple C-bands (b). Giemsa stain in (a) was removed with 5 % acetic acid in ethanol before production of C-bands by heating (b). (c) Harlequin banding in Allium cepa roots showing sister chromatids exchanges (SCE). Procedure involved BrdU labeling of chromosomes, followed by photosensitization (10−7 M meso-tetra-(4-pyridyl)-porphine, red light irradiation), and FPG staining [53]. Note the purple and blue TT and BT chromatids, respectively; the large arrow shows a “dot” SCE. Scale bars: 5 μm (b) and 2 μm (c)
are stained blue by toluidine blue but they become swollen and violet after reduction of disulfide bridges [15]. Thus not only a differential compaction but a differential reactivity would occur between C-bands and chromosome arms, with a conspicuous inversion of colors and staining intensity (Fig. 4a, b) (see spectral profiles in Fig. 5a). The compact chromatin DNA in the centromeric heterochromatin would limit the external thiazine binding to DNA phosphates, and therefore the purple complex cannot be formed if no external thiazines are available to complex with eosin Y. 1.2.2 Sister Chromatid Differentiation
When using selective Giemsa staining to differentiate between sister chromatids, purple and blue colors are again produced. The method involves treatment of cells with the thymidine analogue 5-bromo2′-deoxyuridine (BrdU) and incorporation of bromouracil (instead thymine) during DNA replication, followed by treatment with a photosensitizing agent and then UV irradiation [19, 46]. Using this FPG staining technique [47], normal chromatids (thymine– thymine (TT) in each DNA chain) and bromo-substituted chromatids (BT, BB) can be easily distinguished. Purple and blue colors are seen in TT/BT chromatids, respectively, and the same occurs in the case of BT/BB chromatids, but here the staining time must be somewhat prolonged to produce the purple color. When sister chromatid exchanges (SCE) occur, Giemsa-stained “harlequin”
32
Juan C. Stockert et al.
Fig. 5 Scheme (left side) of differential Giemsa-stained C-band (a), harlequin chromosome with three SCE (b), and nuclear halo (c), with their respective spectral profiles (right side). C C-band, E euchromatin, TT and BT thymine/thymine- and bromo/thymine-containing chromatids, respectively, CC compact chromatin, H chromatin halo, NM nuclear membrane remnant. Note the contrasting absorption curves from purple- (P) and blue (B)-stained structures
chromosomes show TT and BT chromatids with purple and blue colors, respectively (Fig. 4c). The corresponding spectral profiles are seen in Fig. 5b. It is significant that single-stranded DNA and undercondensed chromatin with loss of histones (H2B) occur in bromo-substituted chromatids [46].
Differential Giemsa Staining of Chromatin 1.2.3 Chromatin Halos
2
33
Analogous selective color reactions occur in other chromatin model systems (Fig. 5c). After depletion of histones by polyanions such as heparin and dextran sulfate, chromatin expands to form a halo composed of naked DNA fibers [48]. To further confirm that purple and blue colors arise from compact and disperse chromatin, respectively, nuclear halos were induced in nucleated erythrocytes from unfixed chicken blood smears by treatment with 2 mg/ml dextran sufate and 0.1 % Nonidet P-40 in PBS for 30 min [48]. Compact nuclear remnants formed by DNA-nonhistone proteins and surrounding DNA halos show different colors and spectral profiles (Fig. 5c).
Materials Suitable metaphase chromosome preparations can be obtained using any of the current cytogenetic methods (short-term blood cultures, bone marrow, cultured cell lines, lymphoid tissues, or solid animal tumors), using metaphase arrest, adequate hypotonic treatment, fixation in freshly made methanol-glacial acetic acid (3:1, v/v), and the conventional air- or flame-drying techniques. These preparative procedures are used routinely, and have been described in detail [44, 49–51].
3
Methods In this section, only the Giemsa staining method for chromatin will be described.
3.1 C-bands, G-bands, and Harlequin Chromosomes
Banding is a well-known method in cytogenetics. C- and G-bands can be induced by a variety of procedures, such as those using alkali, hot saline solutions, proteolytic enzymes (e.g., trypsin), reduction of disulfide bonds, urea, or combined treatments [18, 43, 44, 49, 50, 52]. In the case of harlequins, DNA labeling with BrdU and the precise FPG procedure for plant and animal cells are well documented and described elsewhere [19, 46, 51, 53]. In this volume, chapters entitled “FISH Methods in Cytogenetic Studies,” “Atomic Force Microscopy for Analyzing Metaphase Chromosomes. Comparison of AFM Images with Fluorescence Labeling Images of Banding Pattern,” and “Ultrastructural and Immunofluorescent Methods for the Study of the XY Body as a Biomarker” also describe the preparative methods for microscopical studies on chromosomes. Regarding certain chromosome preparative steps before Giemsa staining, see Note 1.
3.2
Practical procedures for Giemsa staining have been described many times, and well-known variants are described by Clark [54] and
Giemsa Staining
34
Juan C. Stockert et al.
Kiernan [55]. The staining method is relatively easy and rapid, although a few important points must be taken into account to achieve successful results. Prepare the staining solution as follows: 1. Filter the stock solution of Giemsa stain prior to use (see Note 2). 2. Dilute with 0.03 M Sørensen phosphate buffer pH 6.4–6.8 or 0.03 M HEPES buffer pH 6.5 (Giemsa/buffer 1:10, v/v) (see Note 3). A crucial factor is the freshness of the aqueous Giemsa solution. 3. Place the slide with chromosomes under study on a horizontal staining rack over a sink and flood the slide with the diluted Giemsa solution. Leave for 10–15 min at room temperature (see Note 4). 4. Wash preparations twice with tap water for 0.5–1 min, air-dry, and mount directly in DePeX (Serva, Heidelberg, Germany) (see Note 5). 3.3
4
Results
After Giemsa staining, chromosome regions that remain relatively compact following specific banding and harlequin procedures are purple. In contrast, disorganized or disperse chromatin should show a pale blue or blue–gray color (see Note 6). Purple chromosome regions correspond to C- and G-bands (see Fig. 4b). Purple chromatin corresponds either to TT chromatids (as opposed to BT ones, see Fig. 4c) or to BT chromatids (as opposed to BB ones).
Notes 1. Chromosome preparation procedures. (a) Aldehyde fixation of chromosome preparations (e.g., using formaldehyde or glutaraldehyde) must be avoided because it prevents banding [13, 43]. (b) Chromosome pretreatments (banding, harlequin). Differential Giemsa staining is strongly dependent on the chromatin disorganization produced by specific pretreatments. Therefore, variations in the staining time are required, depending on the fixing and spreading conditions, as well as the specific preparative methods applied for chromosome analysis. If banding and harlequin pretreatment methods are more vigorous than necessary, more chromatin disorganization occurs and more prolonged Giemsa staining is needed to achieve good color contrast. 2. Generally, adequate Giemsa stain batches can be purchased from any of the current suppliers (e.g., Sigma-Aldrich,
Differential Giemsa Staining of Chromatin
35
St. Louis, MO, USA; Merck, Darmstadt, Germany; Fluka, Buchs, Switzerland; BDH, Poole, UK; Polysciences, Warrington, PA, USA). Using stains which are Biological Stain Commission certified is an additional and simple way to obtain effective batches. Commercial Giemsa solutions contain thiazine dyes and eosin Y dissolved in nonaqueous solvents, generally methanol and also glycerol, which increase the stability of the diluted Giemsa solution. The stock Giemsa solution should not be used if it contains a visible precipitate. 3. The purple coloration is sensitive to the pH and ionic strength of the aqueous Giemsa solution. A buffer solution (pH 6.4– 6.8) is currently used, but tap water, if not too alkaline, will serve quite adequately. Distilled water is usually too acid (pH 5.5). At pH 6.5–8 the color is intense, but it decreases at pH 5–6 and is abolished at pH 3.5. Giemsa in concentrated salt solutions gives no purple reaction, because inorganic cations prevent thiazine binding to phosphates. To achieve correct purple (or blue) staining, pretreatment solutions and Giemsa stain must not contain di- or trivalent cations (e.g., Mg2+, Ca2+, Al3+), as these block phosphate groups. However, washing the stained material with salt solutions gives better purple staining and banding patterns because stacked thiazines are removed. 4. Staining time with Giemsa must be assessed for each specimen depending on pretreatments, fixation, and thickness of the sample. This final factor is not usually important since cell monolayers and metaphase chromosome smears are very thin microscopic samples. 5. Preparations should be not dehydrated because ethanol removes xanthene and thiazine dyes and abolishes the differential staining. 6. Staining colors and intensity. (a) True Romanowsky-Giemsa effect in untreated cells gives purple nuclei (and chromosomes) and blue cytoplasms. Formation of the purple complex is a gradual process. Depending on the type of material, Giemsa staining for 1 min gives blue chromatin, after 2–3 min color is blue– violet, and purple coloration usually occurs only after 5–7 min. Therefore, if nuclei stain blue, the staining time usually needs to be extended. (b) Purple vs. blue chromatin is not easily observed, and may simply be overlooked. It is often assumed that successful differential Giemsa staining is achieved when banded (C, G) vs. non-banded chromosome arms are well contrasted, e.g., purple++ vs. purple + or blue++ vs. blue+. Thus banded and non-banded regions are simply described as “dark-” and “light-stained” chromosome segments.
36
Juan C. Stockert et al.
The same occurs in the case of unsubstituted vs. differentially bromo-substituted chromatids, TT and BT being described as dark- and light-stained chromatids, respectively. BT and BB are also dark- and light-stained chromatids. When no correct purple/blue ratio is obtained, the best strategy is to modify the staining time. Obviously, in photomicrographs obtained through a filter to improve contrast (e.g., green), color differences are abolished. (c) EY is a fluorescent dye. EY emission is easily observed in hematoxylin-eosin-stained preparations [56]. However, no EY fluorescence occurs in the purple complex on account of the strong quenching effect of thiazines.
Acknowledgements We thank M. Cañete, J. Espada, and A. Villanueva for valuable collaboration. This work was supported by a grant (CTQ201020870-C03-03) from the Ministerio de Ciencia e Innovación, Spain. References 1. Horobin RW, Kiernan JA (2002) Conn’s biological stains. A handbook of dyes, stains and fluorochromes for use in biology and medicine, 10th edn. Bios Scientific Publishers, Oxford 2. Zollinger H (2003) Color chemistry. Synthesis, properties, and applications of organic dyes and pigments, 3rd edn. VHCA, Zürich, WileyVCH, Weinheim 3. Scott JE (1973) Affinity, competition and specific interactions in the biochemistry and histochemistry of polyelectrolytes. Biochem Soc Trans 1:787–806 4. Scott JE (1980) The molecular biology of histochemical staining by cationic phthalocyanine dyes: the design of replacements for Alcian blue. J Microsc 119:373–381 5. Horobin RW (1988) Understanding histochemistry: selection, evaluation and design of biological stains. Horwood, Chichester 6. Mann G (1902) Physiological histology. Clarendon, Oxford 7. Baker JR (1958) Principles of biological microtechnique: a study of fixation and dyeing. Methuen, London 8. Horobin RW (1982) Histochemistry. An explanatory outline of histochemistry and biophysical staining. Gustav Fischer, Butterworths, Stuttgart, London
9. Stockert JC (1985) Cytochemistry of nucleic acids: binding mechanisms of dyes and fluorochromes. Microsc Electron Biol Celular 9:89–131 10. Dapson RW (2005) Dye-tissue interactions: mechanisms, quantification and bonding parameters for dyes used in biological staining. Biotech Histochem 80:49–72 11. Stockert JC, Abasolo MI (2011) Inaccurate chemical structure of dyes and fluorochromes found in the literature can be problematic for teaching and research. Biotech Histochem 86:52–60 12. Horobin RW (2011) How Romanowsky stains work and why they remain valuable ̶ including a proposed universal Romanowsky staining mechanism and a rational troubleshooting scheme. Biotech Histochem 86:36–51 13. Stockert JC, Lisanti JA (1972) Acridine-orange differential fluorescence of fast- and slowreassociating DNA after in situ DNA denaturation and reassociation. Chromosoma 37:117–130 14. Armas-Portela R, Cañete M, Stockert JC (1981) Orthochromatic and metachromatic staining reactions of chromatin by thiazine dyes. Biol Zentralbl 100:673–678 15. Barrera C, Mazzolli AB, Pelling C, Stockert JC (1993) Metachromatic staining of human
Differential Giemsa Staining of Chromatin
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26. 27.
28. 29.
30.
sperm nuclei after reduction of disulphide bonds. Acta Histochem 94:141–149 Juarranz A, Cañete M, Stockert JC (1987) Colour differences in the chromatin staining by Cuprolinic blue. Z Mikrosk Anat Forsch 101(532–536):1987 Gutiérrez-Gonzálvez MG, Armas-Portela R, Stockert JC (1987) Differential staining of biological structures by ruthenium red. J Microsc 145:333–340 Lisanti JA, Stockert JC (1974) Color differences in heated and Giemsa stained mouse chromosomes. Stain Technol 49:253–256 Hazen MJ, Villanueva A, Juarranz A et al (1985) Photosensitizing dyes and fluorochromes as substitutes for 33258 Hoechst in the fluorescence–plus−Giemsa (FPG) chromosome technique. Histochemistry 83:241–244 Krafts K (1993) The Ehrlich-ChenzinskyPlehn-Malachowski-Romanowsky-NochtJenner-May-Grünwald-Leishman-ReuterWright-Giemsa-Lillie-Roe-Wilcox stain. The mystery unfolds. Clin Lab Med 13:759–771 Krafts K, Hempelmann E, Oleksyn BJ (2011) The color purple: from royalty to laboratory, with apologies to Malachowski. Biotech Histochem 86:7–35 Wittekind DH (1983) On the nature of Romanowsky-Giemsa staining and its significance for cytochemistry and histochemistry: an overall view. Histochem J 15:1029–1047 Barcia JJ (2007) The Giemsa stain: its history and applications. Int J Surg Pathol 15: 292–296 Wittekind DH (1979) On the nature of Romanowsky dyes and the RomanowskyGiemsa effect. Clin Lab Haematol 1:247–262 Horobin RW, Walter KJ (1987) Understanding Romanowsky staining. I: The RomanowskyGiemsa effect in blood smears. Histochemistry 86:331–336 Power KT (1982) The Romanowsky stains: a review. Am J Med Technol 48:519–523 Green FJ (1990) The Sigma-Aldrich handbook of stains, dyes and indicators. Aldrich Chemicals Co., Milwaukee, Wisconsin Dolan M (2011) The role of the Giemsa stain in cytogenetics. Biotech Histochem 86:94–97 Dunning K, Safo AO (2011) The ultimate Wright-Giemsa stain: 60 years in the making. Biotech Histochem 86:69–75 Krafts KP, Pambuccian SE (2011) Romanowsky staining in cytopathology: history, advantages and limitations. Biotech Histochem 86:82–93
37
31. Sumner AT, Evans HJ (1973) Mechanisms involved in the banding of chromosomes with quinacrine and Giemsa. II. The interaction of the dyes with the chromosomal components. Exp Cell Res 81:223–236 32. Comings DE (1975) Mechanisms of chromosome banding. IV. Optical properties of the Giemsa dyes. Chromosoma 50:89–110 33. Wittekind DH, Kretschmer V, Löhr W (1976) Kann Azur B-Eosin die May-GrünwaldGiemsa-Färbung ersetzen? Blut 32:71–78 34. Sumner AT (1980) Dye binding mechanisms in G-banding of chromosomes. J Microsc 119: 397–406 35. Friedrich K, Seiffert W, Zimmermann HW (1990) Romanowsky dyes and RomanowskyGiemsa effect. 5. Structural investigations of the purple DNA-AB-EY dye complexes of Romanowsky-Giemsa staining. Histochemistry 93:247–256 36. Bergeron JA, Singer M (1958) Metachromasy: an experimental and theoretical reevaluation. J Biophys Biochem Cytol 4:433–457 37. Woynarowski JM, Krugliak M, Ginsburg H (2007) Pharmacogenomic analyses of targeting the AT-rich malaria parasite genome with AT-specific alkylating drugs. Mol Biochem Parasitol 154:70–81 38. Bobrow M, Madan K, Pearson PL (1972) Staining of some specific regions of human chromosomes, particularly the secondary constriction of number 9. Nat New Biol 238: 122–124 39. Gagné R, Laberge C (1972) Specific cytological recognition of the heterochromatic segment of number 9 chromosome in man. Exp Cell Res 73:239–242 40. Alves P, Jonasson J (1978) New staining method for the detection of sister chromatid exchanges in BrdU-labelled chromosomes. J Cell Sci 32:185–195 41. Takayama S, Sakanishi S (1977) Differential Giemsa staining of sister chromatids after extraction with acids. Chromosoma 64: 109–115 42. Jan KY, Su PF, Lee TC (1985) Reversal differential staining of sister chromatids induced by Hoechst plus black light and endonucleases. Expl Cell Res 157:307–314 43. Lisanti JA, Stockert JC (1973) Observations on the staining of centromeric heterochromatin with Giemsa. Experientia 29:887–888 44. Sumner AT (1990) Chromosome banding. Unwin Hyman, London
38
Juan C. Stockert et al.
45. Stockert JC, Pinna-Senn E, Bella JL, Lisanti JA (2005) DNA-binding fluorochromes: correlation between C-banding of mouse metaphase chromosomes and hydrogen bonding to adenine-thymine base pairs. Acta Histochem 106:413–420 46. Ribas M, Korenberg JR, Peretti D et al (1994) Sister chromatid differentiation in 5-bromo-2′deoxyuridine-substituted chromosomes: a study with DNA-specific ligands and monoclonal antibody to histone H2B. Chromosome Res 2:428–438 47. Perry P, Wolff S (1974) New Giemsa method for the differential staining of sister chromatids. Nature 251:156–158 48. Paulson JR, Laemmli UK (1977) The structure of histone-depleted metaphase chromosomes. Cell 12:817–828 49. Verma RS, Babu A (1995) Human chromosome. Principles and techniques, 2nd edn. McGraw-Hill, New York 50. Czepulkowski B (2001) Analyzing chromosomes. Springer, New York
51. Kimura E, Hoshi O, Ushiki T (2004) Atomic force microscopy of human metaphase chromosomes after differential staining of sister chromatids. Arch Histol Cytol 67:171–177 52. Rooney DE (2001) Human cytogenetics, constitutional analysis, 3rd edn. Oxford University Press, Oxford, UK 53. Hazen MJ, Villanueva A, Stockert JC (1987) Induction of sister chromatid exchanges in Allium cepa meristematic cells exposed to meso-tetra (4-pyridyl) porphine and hematoporphyrin photo-radiation. Photochem Photobiol 46:463–467 54. Clark G (1981) Staining procedures used by the Biological Stain Commission, 4th edn. Williams and Wilkins, Baltimore 55. Kiernan JA (1990) Histological and histochemical methods: theory and practice, 2nd edn. Pergamon Press, Oxford, New York 56. Espada J, Valverde P, Stockert JC (1993) Selective fluorescence of eosinophilic structures in grasshopper and mammalian testis stained with haematoxylin-eosin. Histochemistry 99:385–390
1
Introduction Dyes are widely used in biochemical and biological investigations, as well as in medical diagnostic procedures. In particular they are applied as analytical tools for the localization and identification of cell and tissue structures [1, 2]. Molecular approaches to the design of dyes [3, 4] and staining procedures [5] have a considerable history. Efforts to achieve a molecular understanding of the mechanisms of such biological staining are indeed of even longer standing [6–9]. Various aspects of these approaches have been recently reviewed [10–12]. Given this considerable literature, it is both unfortunate and puzzling that biomedical professionals often do not understand, or are not interested in, the physical and chemical aspects of biological staining methods. The contents of this chapter, and others in this volume, attempt to help improve this situation.
Juan C. Stockert et al. (eds.), Functional Analysis of DNA and Chromatin, Methods in Molecular Biology, vol. 1094, DOI 10.1007/978-1-62703-706-8_3, © Springer Science+Business Media New York 2014
25
26
Juan C. Stockert et al.
Fig. 1 Chemical structures of azure B and eosin Y. In the latter dye the planes of the xanthene chromophore and that of the pendant phenyl group are nearly perpendicular to each other
It is known that certain staining procedures used for microscopy give differential coloration to chromatin. Such selective reactions depend on a variety of mechanisms. These include ortho- and metachromatic colors due to binding of dye monomers (intercalation) or dye aggregates (stacking) to DNA, physical phenomena which arise both with acridines [13] and thiazines [14, 15]. Other processes, such as nuclear staining by cuprolinic blue [16] and ruthenium red [17], involve chemical redox reactions. Finally, multi-dye systems such as Giemsa staining [18, 19] are also used. Although the mechanisms of differential staining by these dyes are not yet precisely understood at the molecular level, probably involving both physical and chemical parameters, it is likely that differential chromatin reactivity—“stainability”—also results from specific prestaining treatments. Multi-dye staining mixtures involving azure B plus eosin Y, the topic of this chapter, have been described by a diversity of names. For an amusing summary of this, see the title of a paper by Krafts [20]. The present paper uses the most common generic term for such stains, namely, “Giemsa.” Note however that the purple staining of chromatin characteristic of this stain has been termed the “Romanowsky-Giemsa effect.” 1.1 Giemsa Staining Mechanism
Giemsa stains usually contain polychrome methylene blue plus eosin Y. The “polychrome” epithet describes the presence of methylene blue plus its demethylated homologues, in particular azure B; see Fig. 1 for key dye structures. A recent review on the history and application of these dye mixtures has been published [21].
Differential Giemsa Staining of Chromatin
27
Giemsa stains are currently applied in bacteriology, cytogenetics, cytology, hematology, histopathology, and parasitology [12, 22, 23] to reveal basophilic cytoplasm, chromatin, ortho- and metachromatic cytoplasmic granules, and nucleoli in different colors. Typically, condensed chromatin DNA (in nuclei and chromosomes) gives rise to a purple (magenta) color, whereas RNA-containing basophilic cytoplasm and nucleoli appear in blue [24, 25]. The Giemsa stain is an important polychromatic staining method for peripheral blood, bone marrow, cervical smears, Plasmodium, Leishmania, and Trypanosoma parasites, as well as for banded and harlequin metaphase chromosomes. Some of the diverse end users routinely use the terms Jenner, Leishman, MayGrünwald, or Wright stains rather than Giemsa. Properties and applications of such dye mixtures, as well as their individual thiazine and eosin components, have been described [1, 26, 27]. The principles and methodology of Giemsa staining for cytogenetics, cytopathology, and hematology—with particular reference to equipment, reagents, and troubleshooting—have been recently reviewed [28–30]. In the present chapter, only the differential Giemsa staining of chromatin will be considered. Differential coloration of chromatin by thiazine dyes and Giemsa stain is influenced by dye concentration, staining time, local availability of binding sites, and different diffusion rates of dyes [12, 25, 31, 32]. The purple color arising during Giemsa chromatin staining, with an absorption peak around 550 nm, originates from a complex between cationic thiazines and eosin Y, which is formed preferentially on certain polyanions such as the DNA present in chromatin [12, 25]. Wittekind et al. [22, 33] demonstrated directly that a Giemsa-type stain could be prepared by mixing two pure dyes, namely, azure B (AB) and eosin Y (EY). It is known that the purple color of Giemsa-stained nuclei and chromosomes (whether banded or not) arises from the precipitation of a hydrophobic thiazine eosinate complex with a molar ratio 2:1 [31, 34]. This complex would be the same compound that slowly precipitates from aqueous Giemsa solutions, in which strong hydrophobic interactions occur between stacked thiazine and eosin dyes. Interestingly, after extraction from chromatin and DNA, the ratio thiazine/eosin Y is 2.3:1 (chromatin) and 2.8:1 (DNA) [31]. The thiazine dye present in excess of the 2:1 ratio probably corresponds to additional DNA intercalated dye (which would be 3:1 at saturation). The precise structure of the DNA-AB-EY complex remains unknown, but possible structures can be considered. Early views, with eosin Y attached directly to two intercalated thiazines [31], are now disregarded [35]. In any event, the sequence of the process is as shown in Fig. 2. First, DNA intercalation (a, b) and external (c) dye binding occur, followed by formation of a 2:1 AB-EY complex on DNA (d) with an orthogonal relation between
28
Juan C. Stockert et al.
Fig. 2 Schematic representation of DNA, DNA-AB, and AB2-EY complexes (AB: azure B, EY: eosin Y). Thick vertical lines are negatively charged deoxyribosephosphate chains. (a) DNA (bp: base pairs). (b) DNA with intercalated thiazines (iT). (c) Intercalated DNA with additional external thiazines (eT). (d) DNA-AB2-EY complex based on that described by Friedrich et al. [35], with the suggested orthogonal position of xanthene (X) and thiazine rings; ph: EY phenyl group. (e) AB2-EY complex after dissociation (arrow) from DNA, showing stacked coplanar xanthene and thiazine chromophores. In (c–e) only one side of external DNAdye interactions is shown for clarity
chromophores [35]. It must be noted that in this case, a DNA-AB complex with 1:1 phosphate-AB stoichiometry was first obtained and subsequently stained by EY. However the purple compound formed by one eosin Y and two thiazine dye molecules does not necessarily remain attached to DNA, and purple AB-EY that precipitates on its own might cause the color reaction of chromatin [25]. If this were so, two thiazine dyes should first bind to DNA phosphates at the correct distance apart, in agreement with early studies on thiazine binding to phosphate groups; see Bergeron and Singer [36]. Subsequently they would combine with one EY to form the purple AB-EY complex that precipitates, thereby freeing DNA phosphates to bind more thiazines and EY again [25, 34] (Fig. 2e). This cyclic self-catalytic process (DNA → DNA-AB2 → DNAAB2-EY → DNA + AB2-EY) would be responsible for the intensification of the purple staining of chromatin DNA (and other polyanions) by a “template effect” [12]. The continuous accumulation of precipitate by self-assembly of the AB2-EY
Differential Giemsa Staining of Chromatin
29
Fig. 3 Frontal (a) and lateral view (b) of a molecular model of two units of a proposed azure B-eosin Y (AB2-EY) complex
complex on a nanostructured polyanionic scaffold can explain the color enhancement effect which facilitates the visualization of certain small purple stained structures such as centromeres and chromosome bands. To illustrate the possible molecular structure of the AB2-EY complex, a simple and plausible model was constructed (Fig. 3). Geometry optimization was performed with the molecular mechanics force field (MM+) method using HyperChem 8 software and Polak-Ribière algorithm converged at a conjugate gradient of 0.1 kcal/(Ǻ mol). Fig. 3a, b shows two face-to-face related AB2-EY complexes in which each unit is a sandwich of two AB dyes (with S atoms in cis orientation) and one EY (with a heterocyclic electronegative O atom stacked between the electropositive S atoms). In this model strong electrostatic, hydrophobic, and π-electron interactions occur between the planar tricycle rings of AB and EY chromophores, which show identical aromatic size as assessed by the largest conjugated fragment values of 18 (see Horobin, ref. 8) and opposite partial charges. Although this is the optimal geometry, it may be noted that other AB-EY arrangements are possible. One wonders if other alternative complexes might be responsible for the historically well-known [7] “carmine” or “red” (i.e., not purple) nuclear coloration of Giemsa-stained malarial parasites. As to why this might occur, the Plasmodium falciparum genome is known to contain the unusual feature of “super AT-rich regions” [37]. 1.2 Differential Giemsa Staining
Native DNA is the substrate required to induce purple Giemsa staining in normal chromatin. However following various specific treatments some regions of chromatin DNA become modified (mainly disorganized), resulting in differential staining. Purple and
30
Juan C. Stockert et al.
blue chromatin colors were first reported after staining with Giemsa at high pH [38, 39] or following specific pretreatments [18, 19]. Although a differential color reaction in chromosomes seems to be a simple model for Giemsa staining, no clear relation to the previous treatment and staining conditions is apparent. Note, for instance, the contrasting results in the case of differentially bromosubstituted chromatids when subjected to other procedures such as direct staining with alkaline Giemsa solution [40], treatment with hot perchloric or hydrochloric acid [41], or DNase digestion after Hoechst 33258 plus UV treatment [42], which give a reverse staining pattern in comparison with that produced by the conventional fluorescence-plus-Giemsa (FPG) method. 1.2.1 G- and C-bands
It is known that G- and C-banding are related in a sequential manner to the progressive disorganization (e.g., decondensation, dispersion, degradation) of chromatin [43, 44]. Formation of the purple complex was originally postulated to explain the staining of chromosome G-bands [31, 34]. However, G-bands can be seen in the absence of eosin Y, and most thiazine dyes when applied alone give a metachromatic G-banding pattern [32]. In these cases, G-bands appear highly stained without formation of a purple eosin Y containing compound. After appropriate pretreatments and Giemsa staining, C- and G-bands appear in purple, whereas the remaining (non-banded) chromosome arms are blue. However, there is no significant difference in DNA content between both regions [31, 34]. After a staining time long enough to produce purple bands, other regions are still blue [18]. This suggests that changes in the chromatin compaction and DNA strandedness have occurred, with an overall decrease in chromatin packing. Changes in color and staining intensity of C-bands from mouse chromosomes are shown in Fig. 4a, b. It is known that treatments that produce C-bands cause considerable swelling of chromatin, with concomitant protein extraction and DNA denaturation [18, 31]. Highly repetitive adenine-thymine-rich DNA located in the centromeric heterochromatin of mouse chromosomes (see ref. 45) reassociates very rapidly [13], and then the occurrence of purple and blue staining at a specific time would indicate differential DNA strandedness [18]. Interestingly, the centromeric heterochromatin of non-banded and C-banded chromosomes shows an opposite Giemsa staining reaction. It is pale violet-blue or blue in nonbanded chromosomes, but becomes purple in C-bands. The contrary occurs in the chromosome arms. The highly compact centromeric heterochromatin would not allow the rapid formation of the purple compound, and at short staining times it stains blue mainly by thiazine monomers. After procedures involving protein extraction, DNA denaturation, and loss of chromatin compaction, C-bands stain purple. In this context, it is known that sperm nuclei
Differential Giemsa Staining of Chromatin
31
Fig. 4 (a, b) Giemsa-stained mouse metaphase chromosomes before (a) and after (b) C-banding. Arrows indicate weakly stained centromeric heterochromatin (a) and purple C-bands (b). Giemsa stain in (a) was removed with 5 % acetic acid in ethanol before production of C-bands by heating (b). (c) Harlequin banding in Allium cepa roots showing sister chromatids exchanges (SCE). Procedure involved BrdU labeling of chromosomes, followed by photosensitization (10−7 M meso-tetra-(4-pyridyl)-porphine, red light irradiation), and FPG staining [53]. Note the purple and blue TT and BT chromatids, respectively; the large arrow shows a “dot” SCE. Scale bars: 5 μm (b) and 2 μm (c)
are stained blue by toluidine blue but they become swollen and violet after reduction of disulfide bridges [15]. Thus not only a differential compaction but a differential reactivity would occur between C-bands and chromosome arms, with a conspicuous inversion of colors and staining intensity (Fig. 4a, b) (see spectral profiles in Fig. 5a). The compact chromatin DNA in the centromeric heterochromatin would limit the external thiazine binding to DNA phosphates, and therefore the purple complex cannot be formed if no external thiazines are available to complex with eosin Y. 1.2.2 Sister Chromatid Differentiation
When using selective Giemsa staining to differentiate between sister chromatids, purple and blue colors are again produced. The method involves treatment of cells with the thymidine analogue 5-bromo2′-deoxyuridine (BrdU) and incorporation of bromouracil (instead thymine) during DNA replication, followed by treatment with a photosensitizing agent and then UV irradiation [19, 46]. Using this FPG staining technique [47], normal chromatids (thymine– thymine (TT) in each DNA chain) and bromo-substituted chromatids (BT, BB) can be easily distinguished. Purple and blue colors are seen in TT/BT chromatids, respectively, and the same occurs in the case of BT/BB chromatids, but here the staining time must be somewhat prolonged to produce the purple color. When sister chromatid exchanges (SCE) occur, Giemsa-stained “harlequin”
32
Juan C. Stockert et al.
Fig. 5 Scheme (left side) of differential Giemsa-stained C-band (a), harlequin chromosome with three SCE (b), and nuclear halo (c), with their respective spectral profiles (right side). C C-band, E euchromatin, TT and BT thymine/thymine- and bromo/thymine-containing chromatids, respectively, CC compact chromatin, H chromatin halo, NM nuclear membrane remnant. Note the contrasting absorption curves from purple- (P) and blue (B)-stained structures
chromosomes show TT and BT chromatids with purple and blue colors, respectively (Fig. 4c). The corresponding spectral profiles are seen in Fig. 5b. It is significant that single-stranded DNA and undercondensed chromatin with loss of histones (H2B) occur in bromo-substituted chromatids [46].
Differential Giemsa Staining of Chromatin 1.2.3 Chromatin Halos
2
33
Analogous selective color reactions occur in other chromatin model systems (Fig. 5c). After depletion of histones by polyanions such as heparin and dextran sulfate, chromatin expands to form a halo composed of naked DNA fibers [48]. To further confirm that purple and blue colors arise from compact and disperse chromatin, respectively, nuclear halos were induced in nucleated erythrocytes from unfixed chicken blood smears by treatment with 2 mg/ml dextran sufate and 0.1 % Nonidet P-40 in PBS for 30 min [48]. Compact nuclear remnants formed by DNA-nonhistone proteins and surrounding DNA halos show different colors and spectral profiles (Fig. 5c).
Materials Suitable metaphase chromosome preparations can be obtained using any of the current cytogenetic methods (short-term blood cultures, bone marrow, cultured cell lines, lymphoid tissues, or solid animal tumors), using metaphase arrest, adequate hypotonic treatment, fixation in freshly made methanol-glacial acetic acid (3:1, v/v), and the conventional air- or flame-drying techniques. These preparative procedures are used routinely, and have been described in detail [44, 49–51].
3
Methods In this section, only the Giemsa staining method for chromatin will be described.
3.1 C-bands, G-bands, and Harlequin Chromosomes
Banding is a well-known method in cytogenetics. C- and G-bands can be induced by a variety of procedures, such as those using alkali, hot saline solutions, proteolytic enzymes (e.g., trypsin), reduction of disulfide bonds, urea, or combined treatments [18, 43, 44, 49, 50, 52]. In the case of harlequins, DNA labeling with BrdU and the precise FPG procedure for plant and animal cells are well documented and described elsewhere [19, 46, 51, 53]. In this volume, chapters entitled “FISH Methods in Cytogenetic Studies,” “Atomic Force Microscopy for Analyzing Metaphase Chromosomes. Comparison of AFM Images with Fluorescence Labeling Images of Banding Pattern,” and “Ultrastructural and Immunofluorescent Methods for the Study of the XY Body as a Biomarker” also describe the preparative methods for microscopical studies on chromosomes. Regarding certain chromosome preparative steps before Giemsa staining, see Note 1.
3.2
Practical procedures for Giemsa staining have been described many times, and well-known variants are described by Clark [54] and
Giemsa Staining
34
Juan C. Stockert et al.
Kiernan [55]. The staining method is relatively easy and rapid, although a few important points must be taken into account to achieve successful results. Prepare the staining solution as follows: 1. Filter the stock solution of Giemsa stain prior to use (see Note 2). 2. Dilute with 0.03 M Sørensen phosphate buffer pH 6.4–6.8 or 0.03 M HEPES buffer pH 6.5 (Giemsa/buffer 1:10, v/v) (see Note 3). A crucial factor is the freshness of the aqueous Giemsa solution. 3. Place the slide with chromosomes under study on a horizontal staining rack over a sink and flood the slide with the diluted Giemsa solution. Leave for 10–15 min at room temperature (see Note 4). 4. Wash preparations twice with tap water for 0.5–1 min, air-dry, and mount directly in DePeX (Serva, Heidelberg, Germany) (see Note 5). 3.3
4
Results
After Giemsa staining, chromosome regions that remain relatively compact following specific banding and harlequin procedures are purple. In contrast, disorganized or disperse chromatin should show a pale blue or blue–gray color (see Note 6). Purple chromosome regions correspond to C- and G-bands (see Fig. 4b). Purple chromatin corresponds either to TT chromatids (as opposed to BT ones, see Fig. 4c) or to BT chromatids (as opposed to BB ones).
Notes 1. Chromosome preparation procedures. (a) Aldehyde fixation of chromosome preparations (e.g., using formaldehyde or glutaraldehyde) must be avoided because it prevents banding [13, 43]. (b) Chromosome pretreatments (banding, harlequin). Differential Giemsa staining is strongly dependent on the chromatin disorganization produced by specific pretreatments. Therefore, variations in the staining time are required, depending on the fixing and spreading conditions, as well as the specific preparative methods applied for chromosome analysis. If banding and harlequin pretreatment methods are more vigorous than necessary, more chromatin disorganization occurs and more prolonged Giemsa staining is needed to achieve good color contrast. 2. Generally, adequate Giemsa stain batches can be purchased from any of the current suppliers (e.g., Sigma-Aldrich,
Differential Giemsa Staining of Chromatin
35
St. Louis, MO, USA; Merck, Darmstadt, Germany; Fluka, Buchs, Switzerland; BDH, Poole, UK; Polysciences, Warrington, PA, USA). Using stains which are Biological Stain Commission certified is an additional and simple way to obtain effective batches. Commercial Giemsa solutions contain thiazine dyes and eosin Y dissolved in nonaqueous solvents, generally methanol and also glycerol, which increase the stability of the diluted Giemsa solution. The stock Giemsa solution should not be used if it contains a visible precipitate. 3. The purple coloration is sensitive to the pH and ionic strength of the aqueous Giemsa solution. A buffer solution (pH 6.4– 6.8) is currently used, but tap water, if not too alkaline, will serve quite adequately. Distilled water is usually too acid (pH 5.5). At pH 6.5–8 the color is intense, but it decreases at pH 5–6 and is abolished at pH 3.5. Giemsa in concentrated salt solutions gives no purple reaction, because inorganic cations prevent thiazine binding to phosphates. To achieve correct purple (or blue) staining, pretreatment solutions and Giemsa stain must not contain di- or trivalent cations (e.g., Mg2+, Ca2+, Al3+), as these block phosphate groups. However, washing the stained material with salt solutions gives better purple staining and banding patterns because stacked thiazines are removed. 4. Staining time with Giemsa must be assessed for each specimen depending on pretreatments, fixation, and thickness of the sample. This final factor is not usually important since cell monolayers and metaphase chromosome smears are very thin microscopic samples. 5. Preparations should be not dehydrated because ethanol removes xanthene and thiazine dyes and abolishes the differential staining. 6. Staining colors and intensity. (a) True Romanowsky-Giemsa effect in untreated cells gives purple nuclei (and chromosomes) and blue cytoplasms. Formation of the purple complex is a gradual process. Depending on the type of material, Giemsa staining for 1 min gives blue chromatin, after 2–3 min color is blue– violet, and purple coloration usually occurs only after 5–7 min. Therefore, if nuclei stain blue, the staining time usually needs to be extended. (b) Purple vs. blue chromatin is not easily observed, and may simply be overlooked. It is often assumed that successful differential Giemsa staining is achieved when banded (C, G) vs. non-banded chromosome arms are well contrasted, e.g., purple++ vs. purple + or blue++ vs. blue+. Thus banded and non-banded regions are simply described as “dark-” and “light-stained” chromosome segments.
36
Juan C. Stockert et al.
The same occurs in the case of unsubstituted vs. differentially bromo-substituted chromatids, TT and BT being described as dark- and light-stained chromatids, respectively. BT and BB are also dark- and light-stained chromatids. When no correct purple/blue ratio is obtained, the best strategy is to modify the staining time. Obviously, in photomicrographs obtained through a filter to improve contrast (e.g., green), color differences are abolished. (c) EY is a fluorescent dye. EY emission is easily observed in hematoxylin-eosin-stained preparations [56]. However, no EY fluorescence occurs in the purple complex on account of the strong quenching effect of thiazines.
Acknowledgements We thank M. Cañete, J. Espada, and A. Villanueva for valuable collaboration. This work was supported by a grant (CTQ201020870-C03-03) from the Ministerio de Ciencia e Innovación, Spain. References 1. Horobin RW, Kiernan JA (2002) Conn’s biological stains. A handbook of dyes, stains and fluorochromes for use in biology and medicine, 10th edn. Bios Scientific Publishers, Oxford 2. Zollinger H (2003) Color chemistry. Synthesis, properties, and applications of organic dyes and pigments, 3rd edn. VHCA, Zürich, WileyVCH, Weinheim 3. Scott JE (1973) Affinity, competition and specific interactions in the biochemistry and histochemistry of polyelectrolytes. Biochem Soc Trans 1:787–806 4. Scott JE (1980) The molecular biology of histochemical staining by cationic phthalocyanine dyes: the design of replacements for Alcian blue. J Microsc 119:373–381 5. Horobin RW (1988) Understanding histochemistry: selection, evaluation and design of biological stains. Horwood, Chichester 6. Mann G (1902) Physiological histology. Clarendon, Oxford 7. Baker JR (1958) Principles of biological microtechnique: a study of fixation and dyeing. Methuen, London 8. Horobin RW (1982) Histochemistry. An explanatory outline of histochemistry and biophysical staining. Gustav Fischer, Butterworths, Stuttgart, London
9. Stockert JC (1985) Cytochemistry of nucleic acids: binding mechanisms of dyes and fluorochromes. Microsc Electron Biol Celular 9:89–131 10. Dapson RW (2005) Dye-tissue interactions: mechanisms, quantification and bonding parameters for dyes used in biological staining. Biotech Histochem 80:49–72 11. Stockert JC, Abasolo MI (2011) Inaccurate chemical structure of dyes and fluorochromes found in the literature can be problematic for teaching and research. Biotech Histochem 86:52–60 12. Horobin RW (2011) How Romanowsky stains work and why they remain valuable ̶ including a proposed universal Romanowsky staining mechanism and a rational troubleshooting scheme. Biotech Histochem 86:36–51 13. Stockert JC, Lisanti JA (1972) Acridine-orange differential fluorescence of fast- and slowreassociating DNA after in situ DNA denaturation and reassociation. Chromosoma 37:117–130 14. Armas-Portela R, Cañete M, Stockert JC (1981) Orthochromatic and metachromatic staining reactions of chromatin by thiazine dyes. Biol Zentralbl 100:673–678 15. Barrera C, Mazzolli AB, Pelling C, Stockert JC (1993) Metachromatic staining of human
Differential Giemsa Staining of Chromatin
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26. 27.
28. 29.
30.
sperm nuclei after reduction of disulphide bonds. Acta Histochem 94:141–149 Juarranz A, Cañete M, Stockert JC (1987) Colour differences in the chromatin staining by Cuprolinic blue. Z Mikrosk Anat Forsch 101(532–536):1987 Gutiérrez-Gonzálvez MG, Armas-Portela R, Stockert JC (1987) Differential staining of biological structures by ruthenium red. J Microsc 145:333–340 Lisanti JA, Stockert JC (1974) Color differences in heated and Giemsa stained mouse chromosomes. Stain Technol 49:253–256 Hazen MJ, Villanueva A, Juarranz A et al (1985) Photosensitizing dyes and fluorochromes as substitutes for 33258 Hoechst in the fluorescence–plus−Giemsa (FPG) chromosome technique. Histochemistry 83:241–244 Krafts K (1993) The Ehrlich-ChenzinskyPlehn-Malachowski-Romanowsky-NochtJenner-May-Grünwald-Leishman-ReuterWright-Giemsa-Lillie-Roe-Wilcox stain. The mystery unfolds. Clin Lab Med 13:759–771 Krafts K, Hempelmann E, Oleksyn BJ (2011) The color purple: from royalty to laboratory, with apologies to Malachowski. Biotech Histochem 86:7–35 Wittekind DH (1983) On the nature of Romanowsky-Giemsa staining and its significance for cytochemistry and histochemistry: an overall view. Histochem J 15:1029–1047 Barcia JJ (2007) The Giemsa stain: its history and applications. Int J Surg Pathol 15: 292–296 Wittekind DH (1979) On the nature of Romanowsky dyes and the RomanowskyGiemsa effect. Clin Lab Haematol 1:247–262 Horobin RW, Walter KJ (1987) Understanding Romanowsky staining. I: The RomanowskyGiemsa effect in blood smears. Histochemistry 86:331–336 Power KT (1982) The Romanowsky stains: a review. Am J Med Technol 48:519–523 Green FJ (1990) The Sigma-Aldrich handbook of stains, dyes and indicators. Aldrich Chemicals Co., Milwaukee, Wisconsin Dolan M (2011) The role of the Giemsa stain in cytogenetics. Biotech Histochem 86:94–97 Dunning K, Safo AO (2011) The ultimate Wright-Giemsa stain: 60 years in the making. Biotech Histochem 86:69–75 Krafts KP, Pambuccian SE (2011) Romanowsky staining in cytopathology: history, advantages and limitations. Biotech Histochem 86:82–93
37
31. Sumner AT, Evans HJ (1973) Mechanisms involved in the banding of chromosomes with quinacrine and Giemsa. II. The interaction of the dyes with the chromosomal components. Exp Cell Res 81:223–236 32. Comings DE (1975) Mechanisms of chromosome banding. IV. Optical properties of the Giemsa dyes. Chromosoma 50:89–110 33. Wittekind DH, Kretschmer V, Löhr W (1976) Kann Azur B-Eosin die May-GrünwaldGiemsa-Färbung ersetzen? Blut 32:71–78 34. Sumner AT (1980) Dye binding mechanisms in G-banding of chromosomes. J Microsc 119: 397–406 35. Friedrich K, Seiffert W, Zimmermann HW (1990) Romanowsky dyes and RomanowskyGiemsa effect. 5. Structural investigations of the purple DNA-AB-EY dye complexes of Romanowsky-Giemsa staining. Histochemistry 93:247–256 36. Bergeron JA, Singer M (1958) Metachromasy: an experimental and theoretical reevaluation. J Biophys Biochem Cytol 4:433–457 37. Woynarowski JM, Krugliak M, Ginsburg H (2007) Pharmacogenomic analyses of targeting the AT-rich malaria parasite genome with AT-specific alkylating drugs. Mol Biochem Parasitol 154:70–81 38. Bobrow M, Madan K, Pearson PL (1972) Staining of some specific regions of human chromosomes, particularly the secondary constriction of number 9. Nat New Biol 238: 122–124 39. Gagné R, Laberge C (1972) Specific cytological recognition of the heterochromatic segment of number 9 chromosome in man. Exp Cell Res 73:239–242 40. Alves P, Jonasson J (1978) New staining method for the detection of sister chromatid exchanges in BrdU-labelled chromosomes. J Cell Sci 32:185–195 41. Takayama S, Sakanishi S (1977) Differential Giemsa staining of sister chromatids after extraction with acids. Chromosoma 64: 109–115 42. Jan KY, Su PF, Lee TC (1985) Reversal differential staining of sister chromatids induced by Hoechst plus black light and endonucleases. Expl Cell Res 157:307–314 43. Lisanti JA, Stockert JC (1973) Observations on the staining of centromeric heterochromatin with Giemsa. Experientia 29:887–888 44. Sumner AT (1990) Chromosome banding. Unwin Hyman, London
38
Juan C. Stockert et al.
45. Stockert JC, Pinna-Senn E, Bella JL, Lisanti JA (2005) DNA-binding fluorochromes: correlation between C-banding of mouse metaphase chromosomes and hydrogen bonding to adenine-thymine base pairs. Acta Histochem 106:413–420 46. Ribas M, Korenberg JR, Peretti D et al (1994) Sister chromatid differentiation in 5-bromo-2′deoxyuridine-substituted chromosomes: a study with DNA-specific ligands and monoclonal antibody to histone H2B. Chromosome Res 2:428–438 47. Perry P, Wolff S (1974) New Giemsa method for the differential staining of sister chromatids. Nature 251:156–158 48. Paulson JR, Laemmli UK (1977) The structure of histone-depleted metaphase chromosomes. Cell 12:817–828 49. Verma RS, Babu A (1995) Human chromosome. Principles and techniques, 2nd edn. McGraw-Hill, New York 50. Czepulkowski B (2001) Analyzing chromosomes. Springer, New York
51. Kimura E, Hoshi O, Ushiki T (2004) Atomic force microscopy of human metaphase chromosomes after differential staining of sister chromatids. Arch Histol Cytol 67:171–177 52. Rooney DE (2001) Human cytogenetics, constitutional analysis, 3rd edn. Oxford University Press, Oxford, UK 53. Hazen MJ, Villanueva A, Stockert JC (1987) Induction of sister chromatid exchanges in Allium cepa meristematic cells exposed to meso-tetra (4-pyridyl) porphine and hematoporphyrin photo-radiation. Photochem Photobiol 46:463–467 54. Clark G (1981) Staining procedures used by the Biological Stain Commission, 4th edn. Williams and Wilkins, Baltimore 55. Kiernan JA (1990) Histological and histochemical methods: theory and practice, 2nd edn. Pergamon Press, Oxford, New York 56. Espada J, Valverde P, Stockert JC (1993) Selective fluorescence of eosinophilic structures in grasshopper and mammalian testis stained with haematoxylin-eosin. Histochemistry 99:385–390
