A Robust Procedure for Distinctively Visualizing Zebrafish Retinal Cell Nuclei Under Bright Field Light Microscopy

Abstract

To simultaneously visualize individual cell nuclei and tissue morphologies of the zebrafish retina under bright field light microscopy, it is necessary to establish a procedure that specifically and sensitively stains the cell nuclei in thin tissue sections. This necessity arises from the high nuclear density of the retina and the highly decondensed chromatin of the cone photoreceptors, which significantly reduces their nuclear signals and makes nuclei difficult to distinguish from possible high cytoplasmic background staining. Here we optimized a procedure that integrates JB4 plastic embedding and Feulgen reaction for visualizing zebrafish retinal cell nuclei under bright field light microscopy. This method produced highly specific nuclear staining with minimal cytoplasmic background, allowing us to distinguish individual retinal nuclei despite their tight packaging. The nuclear staining is also sensitive enough to distinguish the euchromatin from heterochromatin in the zebrafish cone nuclei. In addition, this method could be combined with in situ hybridization to simultaneously visualize the cell nuclei and mRNA expression patterns. With its superb specificity and sensitivity, this method may be extended to quantify cell density and analyze global chromatin organization throughout the retina or other tissues.

Keywords

Feulgen nuclei staining zebrafish retina photoreceptor aging degeneration

The zebrafish is a good model to study the etiology of age-related macular degeneration in humans because similar to the human macula, the zebrafish retina is also enriched with cones (Østerberg 1935; Clark 1981; Branchek and Bremiller 1984; Curcio et al. 1990; Bilotta et al. 2001; Doerre and Malicki 2002). To study retinal degeneration, it is desirable to quantify precisely how the retina loses cells both spatially and temporally. A direct way to quantify cell loss is to count the number of cell nuclei under light microscopy because each retinal cell has one cell nucleus.

Direct nuclear counting can, however, be difficult because of the high nuclear density in the retina. This difficulty arises from the lower vertical resolution of light microscopy compared to its planar resolution (Bertero and De Mol 1996; Lauer 2002). The vertical resolution can be greatly influenced by the thickness of tissue sections. For thick sections, when the boundaries of the cell nuclei are oriented obliquely to the axis of optical imaging, the images of two closely juxtaposed cell nuclei may merge together from a top view (Fig. 1A; Fig. 2). In thin sections, this problem may be less severe (Fig. 1B). However, thin tissue sectioning provides weaker nuclear signals, which could be problematic when nonspecific cytoplasmic staining is too strong (Fig. 1C). Therefore, to reliably quantify the loss of retinal cells in particular retinal regions under bright field light microscopy, it is necessary to specifically and sensitively stain the cell nuclei in thin tissue sections.

Figure 1.

Schematic drawings illustrating how the thickness of tissue sections and cytoplasmic background signals may affect the imaging effects of two closely juxtaposed cell nuclei under bright field light microscopy. (A) In thick sections, the images of two closely juxtaposed cell nuclei may merge together from the top view. (B) In thin sections, the images of two closely juxtaposed cell nuclei will appear separated from each other from the top view. (C) Weak nuclear signals and high cytoplasmic signals may make the images of the two nuclei appear merged from the top view.

Figure 2.

Effects of section thickness on nuclear distinguishability. (A–D) Individual nuclei show more clear boundaries in 2-μm sections (A) than in 3-μm (B), 4-μm (C), and 6-μm (D) sections. Arrows indicate the image overlapping of neighboring nuclei.

The above need can be met by integrating JB-4 plastic resin’s superb capability of preserving tissue morphology and Feulgen reaction’s highly specific nuclear staining. JB-4 embedding is based on glycol methacrylate polymerization (Higuchi et al. 1979; Cole 1982; Helander 1983; Ladekarl 1994; Miller and Meyer 1990). Because the polymerization occurs after the monomeric reagents infiltrate the biological samples, JB-4 embedding faithfully preserves sample morphology. In addition, the hard JB-4 resin is suitable for sectioning at thicknesses as low as 0.5 µm (Bancroft and Gamble 2008).

To stain the cell nuclei, four general types of dyes or reagents are frequently used. These are based on electrostatic binding, hydrophobic interactions, covalent modifications, or immuno-reactivities (Kiernan 2008). Feulgen nuclear staining, invented by Feulgen and Rossenbeck (1924), is based on a covalent reaction between Schiff’s reagents and the aldehyde groups generated by HCl hydrolysis of the bond between the deoxyribose residues and the bases of DNA molecules. This reaction generates pinkish products and consequently stains the cell nuclei that store the majority of the cellular DNA (Overend and Stacey 1949; Lessler 1953; Van Duijn 1956; Chieco et al. 1994). Thus, this chemical reaction underlies the superb specificity of Feulgen nuclear staining (Pearse 1953; Kurnick 1955). Procedures that combined JB4 embedding and Feulgen staining have been reported before (Schulte and Wittekind 1989). However, this study was performed on 4-µm-thick sections of tissues with sparsely distributed nuclei. Because thick sectioning is not suitable for counting densely packed zebrafish retinal cells, a specialized protocol for visualizing the cell nuclei in thin sections of the zebrafish retina was needed.

Here, we report a protocol that combines JB4 plastic embedding and Feulgen staining to visualize zebrafish retinal cell nuclei. We analyzed the effects of JB4 section thickness, extent of HCl hydrolysis, and Schiff reaction conditions on nuclear staining and visualization. As a result, we optimized a protocol that produces sensitive and specific nuclear staining of thin sections of the zebrafish retina. In conjunction with three-dimensional reconstruction algorithms, this method has the potential to be utilized for reliable nuclear quantification.

Materials and Methods

Zebrafish Care

AB and TU wild-type zebrafish fish were maintained in a 14-hr light and 10-hr dark cycle. All experimental procedures conformed to University of Pittsburgh standards for use and care of animals in research.

Fixation and JB-4 Embedding and Sectioning

Wild-type adult zebrafish eyes were fixed in 4% paraformaldehyde in 1x phosphate buffered saline (PBS) at room temperature overnight. The fixed tissues were then dehydrated in alcohol and embedded in JB-4 resin (Polysciences, Inc. JB-4 embedding kit, Catalog #: 00226) following the manufacturer’s procedure. To determine the effects of section thickness on nuclear visualization, the JB-4-embedded tissues were sectioned at 2, 3, 4, or 6 µm with a disposable tungsten carbide blade (Delaware Diamond Knives, Inc.) on a Shandon Finesse microtome (Thermo Electron Co.). Sections were collected on glass slides.

Feulgen Staining

To find out the optimal conditions for HCl hydrolysis of tissue DNA molecules, JB-4 sections of zebrafish eyes were incubated with 1 N, 2 N, 3 N, 4 N, or 5 N HCl in coplin jars at room temperature for periods from 30 minutes up to overnight. The slides were then washed with distilled water for 10 min to remove excess HCl and then incubated with Schiff’s reagents (Santa Cruz Biotechnology, Inc.; Catalog #: sc-301793) for 2 hrs at room temperature. To determine a sufficient time for Schiff reaction, sections that were treated with 3 N HCl overnight were incubated with Schiff’s reagents at room temperature for 30 min to 5 hrs. To determine the durability of the Feulgen staining, the stained sections (by 3 N HCl overnight hydrolysis and 2 hrs of incubation with the Schiff’s reagents) were rinsed for 30 seconds with tap water and then washed with distilled water for 5 min, 30 min or overnight at room temperature. The washed sections were air dried and mounted with Permount (Fisher Scientific Inc.) under coverslips.

Combination of Whole-Mount In Situ Hybridization and Feulgen Staining

To determine if Feulgen staining can be combined with the whole-mount in situ hybridization technique, 5-dpf (days postfertilization) GFP-expressing transgenic zebrafish eyes were subjected to in situ hybridization to visualize the GFP mRNA expression pattern, embedded in JB-4, sectioned at 2 μm for better resolution, and then further processed for Feulgen staining. In situ hybridization analyses were performed according to a previously published method (Zou et al. 2006). As a control, 5-dpf wild-type zebrafish eyes without in situ hybridization were fixed, embedded in JB-4, sectioned, and then Feulgen-stained.

Methylene Blue-Azure II Staining and Nuclear Fast Red Staining

To compare Feulgen staining with methylene blue-Azure II and nuclear fast red staining, 2-μm JB-4 sections were stained directly with methylene blue-Azure II (0.13% methylene blue [Sigma-Aldrich Co.; Catalog #: 7220-79-3], 0.02% azure II [Icn Biomedicals, Inc.; Catalog #: 150419], 10% glycerol, 10% methanol, 45 mM phosphate buffer Ph 6.9) or nuclear fast red (Trevigen Inc.; Catalog #: 4800-30-17) for 30 min at room temperature. To remove excess dyes, the stained sections were then rinsed with tap water for 30 sec and further destained in distilled water for 5 min, 30 min or overnight at room temperature. The washed sections were then mounted in Permount under coverslips.

Bright Field Light Microscopy

The samples were observed under an Olympus BX60 microscope using an Olympus PlanApo 60×/1,40 oil or a UplanApo 20×/0,80 oil objective and photographed with a SPOT RT camera and Spot software version 4.6 (Diagnostic Instruments, Inc.). Images were taken under an identical exposure time. In general, over 10 images of different sections were taken for each staining procedure. These images were compared visually to select representative images for that particular staining procedure.

Results and Discussion

Feulgen staining is based on two basic chemical reactions (Overend and Stacey 1949; Lessler 1953; Fig. 3N). First, HCl hydrolysis breaks up the bond between a base and a deoxyribose residue of a DNA molecule, resulting in an aldehyde group on the dexoyribose residue of the DNA backbone (Fig. 3N). Second, the resulting aldehyde group condenses with Schiff’s reagents to produce a pink-colored quinonoid compound (Fig. 3N). Thus, the amount of the pink product correlates directly with the amount of DNA material (Böhm and Sprenger 1968; Duijndam and Van Duijn 1975; Böcking et al. 1995). When biological samples are embedded in JB-4 resin, the JB-4 matrix will hinder the penetration of the chemical reagents and slow down the above mentioned chemical reactions. Thus, compared with thin sections, although thick JB-4 sections contain more DNA material for Feulgen staining, the reduction in the chemical reactions implies that staining intensity may not increase proportionally with the increase in thickness. In addition, thick sections could jeopardize the image quality because of the limitation of the vertical resolution of light microscopy (Bertero and De Mol 1996). Thus, to best visualize zebrafish retinal nuclei, we need to seek a balance between signal intensity and image resolution. We therefore systematically analyzed, as presented below, how HCl hydrolysis and Schiff reaction conditions, as well as the thickness of JB4 sections, affect Feulgen staining of zebrafish retinal nuclei.

Figure 3.

Effects of HCl hydrolysis conditions and Schiff reaction durations on Feulgen staining. (A–J) More concentrated HCl or longer incubation gave stronger Feulgen staining than less concentrated HCl or shorter incubation. (K–M) Feulgen staining increased with prolonged Schiff reaction, reaching a maximal intensity after 2 hrs incubation (data not shown). (N) The chemical reactions of HCl hydrolysis of the DNA molecules and Schiff reaction (Pearse 1968, 1972; Kiernan 2008). RPE, retinal pigment epithelium; ONL, outer nuclear layer; INL, inner nuclear layer; RGC, retinal ganglion cell layer.

HCl Hydrolysis Conditions for Feulgen Staining

HCl hydrolysis for Feulgen staining can be carried out under many different conditions. To determine the optimal HCl concentrations for Feulgen staining of zebrafish retina embedded in JB-4, we treated 4 µm JB-4 sections in 1 N, 2 N, 3 N, 4 N, or 5 N HCl for various durations before treating them with Schiff’s reagents. We then judged the extent of Feulgen staining by visual inspection of the intensity of the pink staining. We found that higher HCl concentrations and longer incubations produced stronger staining than lower HCl concentrations and shorter incubation times (Fig. 3A–J). Furthermore, lower HCl concentrations were able to be combined with longer incubation times to increase the staining intensity. Thus, for quick and reliable hydrolysis, we recommend incubating JB-4 embedded tissue sections with 5N HCl for 45 min. Alternatively, a 3 N HCl overnight treatment is sufficient to achieve a similar extent of hydrolysis.

Duration of Schiff Reaction on Feulgen Staining

Because the Schiff reaction is a covalent condensation reaction between the aldehyde groups and Schiff’s reagents, the duration of the Schiff reaction will also affect the ultimate amount of pink staining that is produced. We thus next compared the effects of various Schiff incubation durations on the staining intensity, using 4-µm-thick JB4 sections that were treated with 3N HCl overnight. We found that staining became stronger with longer Schiff incubation but reached the maximum intensity after 2 hrs (Fig. 3K–M). Thus, we conclude that a 2-hr incubation with Schiff’s reagent is sufficient for Feulgen staining of JB-4 sections to visualize the cell nuclei.

Section Thickness on Nuclear Distinguishability and Chromatin Morphology

We next compared how the section thickness affects Feulgen staining and the distinguishability of nuclear boundaries. We found that the intensities of staining in 2-, 3-, 4-, and 6-µm sections were all strong enough to visualize the cell nuclei of all types of retinal cells. However, we found that the thicker the section, the harder it was to discern individual nuclei because of the increased image overlap of neighboring nuclei (Fig. 2). Considering that sections can frequently be torn apart when sectioned at 2-µm thickness, we recommend sectioning at 3 µm to reveal discrete nuclei and overall tissue morphology.

Furthermore, we were surprised to notice that the Feulgen staining of 2-µm-thick sections was also strong and sensitive enough to distinguish the euchromatic from heterochromatic regions in cone nuclei, though not in rod photoreceptors; these two types of chromatin display morphological differences that greatly resemble those observed under transmission electron microscopy (TEM; Fig. 4A, B). This capability of Feulgen staining to distinguish euchromatin from heterochromatin in cones may have very useful applications. Because the regulation of global chromatin organization in the cell nuclei is an important epigenetic mechanism to regulate gene expression, it makes sense to examine a causal relationship between chromatin organization and various cone dystrophies. Considering how time-consuming the conventional TEM procedure is, our Feulgen staining procedure offers an easy and robust alternative for examining the global chromatin organization in cone photoreceptors in zebrafish.

Figure 4.

Feulgen staining is sensitive enough to distinguish the euchromatic and heterochromatic regions in the cone nuclei in the zebrafish retina. (A) The euchromatic regions in cones, but not rods, are clearly distinguishable from heterochromatic regions in 2-μm JB-4 sections after Feulgen staining. (B) A transmission electron micrograph of the outer nuclear layer of the adult retina shows the heterochromatin and euchromatin in both rods and cones. UV, UV cone nuclei; long cone, either a green, red, or blue cone nucleus.

Comparisons Among Feulgen, Methylene Blue-Azure II, and Nuclear Fast Red Staining

Methylene blue-azure II and nuclear fast red are cationic dyes that stain tissues through electrostatic binding to negatively charged molecules. Thus, they are frequently used to stain the cell nuclei because DNA is negatively charged. However, they also stain other negatively charged cytoplamic molecules, which results in cytoplamic staining that can be either useful or problematic, depending on the staining purposes.

We thus next compared Feulgen staining with methylene blue-azure II and nuclear fast red staining for nuclear observation. For this comparison, we used 2 µm JB-4 sections rather than thicker sections for better resolution. We found that both methylene blue-azure II and nuclear fast red produced strong nonspecific cytoplasmic staining, making it difficult to distinguish individual retinal cell nuclei. It was particularly difficult to distinguish cone nuclear staining from cytoplasmic staining because of the weaker cone nuclei staining. We then tried to reduce the cytoplasmic background staining by increasing the destaining time. But extensive destaining was not helpful because it nonspecifically reduced both the nuclear and cytoplasmic signals. For example, after a 30 min wash, both the nuclear and cytoplasmic staining had become very faint; after an overnight wash, staining was almost completely removed (Fig. 5A,B). In contrast, covalent Feulgen nuclear staining endured extensive destaining (Fig. 5C). Thus, although Feulgen staining is more time- consuming, it gave much more specific and stable staining than methylene blue and nuclear fast red.

Figure 5.

Feulgen staining provides a more specific and stable staining than methylene blue-azure II and nuclear fast red staining. (A–B) Methylene blue-azure II and nuclear fast red staining showed strong cytoplasmic background. Destaining washes nonspecifically removed both nuclear and cytoplasmic signals. After an overnight (o/n) wash, methylene blue-azure II staining was completely removed. (C) Covalent Feulgen nuclear staining remained largely unaffected after an overnight wash. RPE, retinal pigment epithelium; ONL, outer nuclear layer; INL, inner nuclear layer; RGC, retinal ganglion cell layer.

Combination of Whole-Mount In Situ Hybridization With Feulgen Staining

We next performed Feulgen staining on samples that were already probed by whole-mount in situ hybridization. We found that HCl hydrolysis and Schiff reaction did not affect the blue precipitation produced by alkaline phosphatase from the substrates of nitroblue tetrazolium (NBT) and 5-bromo-4-chloro-3-indolyl phosphate (BCIP; Fig. 6A,B, arrows). However, compared with Feulgen staining of regular 5-dpf fish eyes (Fig. 6C,D), the in situ procedure rendered the cell nuclear boundary fuzzy, which might have been caused by the dispersion of the chromatin material from the nuclei. In addition, it appeared to cause tighter packaging of the cell bodies in the retina. Despite this drawback, this combinatory procedure will still have very useful applications when individual nuclei are not required to be distinguished or when other tissue types with sparsely distributed nuclei are to be examined. Thus, Feulgen staining is compatible with alkaline phosphatase-based colorimetric detection that is frequently used in whole-mount in situ hybridization; the two procedures can be combined to visualize the nuclei and mRNA expression patterns simultaneously.

Figure 6.

Feulgen staining is compatible with alkaline phosphatase-based colorimetric detection that is frequently used in conventional whole-mount in situ hybridization. (A–B) Both the blue in situ hybridization signals (arrows) and Feulgen staining were visible. A peripheral retinal region is magnified 3 times in B. (C–D) Feulgen staining of regular 5-dpf fish that was not pretreated with the in situ hybridization showed more distinguishable nuclear boundaries. A peripheral retinal region is magnified 3 times in D.

Conclusion

In summary, we optimized the following protocol to perform Feulgen staining of JB-4-embedded zebrafish retina to visualize cell nuclei and tissue morphology:

Fix the zebrafish eyes in 4% PFA overnight at room temperature.

Embed the eyes in JB-4 resin according to manufacturer’s protocol (Polysciences).

Section the JB-4-embedded zebrafish eyes at 3 µm for general visualization of the cell nuclei or at 2 µm to visualize the chromatin distribution patterns in the cone photoreceptors.

Collect the sections on glass slides.

Treat the JB-4 sections in 3N HCl overnight or 5N HCl for 45 min at room temperature to hydrolyze DNA molecules.

Wash the slides with distilled water for 10 min to remove excess HCl.

Incubate HCl-hydrolyzed JB-4 sections with Schiff’s reagent (Santa Cruz Biotechnology) for 2 hrs at room temperature.

Wash the slides with distilled water for 5 min to remove excess Schiff’s reagent.

Mount the sections in Permount and seal under a coverslip for microscopy.

This simple and robust method is sensitive and specific for visualizing cell nuclei with various degrees of chromatin condensation. The nuclear visualization is sensitive enough to study the global distribution patterns of the heterochromatin and euchromatin in the zebrafish cone photoreceptors. This allows a convenient comparison of the variations in chromatin organization between different retinal regions. In principle, this method may also be utilized for cell quantification if combined with proper algorithms for three-dimensional reconstruction from consecutive JB-4 sections.

Footnotes

Acknowledgements

We thank Mrs. Lynne Sunderman for proofreading the manuscript.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by an NIH Core Grant (P30EY008098), Eye and Ear Foundation of Pittsburgh, an unrestricted grant from Research to Prevent Blindness, and a NIH R01 grant (EY016099) and a RPB Wasserman Merit Award to XW. Jinling Fu was supported by a State Scholarship Fund awarded by China Scholarship Council and Jilin University.

References

Bancroft

Gamble

. 2008. Theory and practice of histological Techniques. 6th ed. London: Churchill Livingstone.

Bertero

De Mol

. 1996. Super-resolution by data inversion. Progress in Optics. 36:129–178.

Bilotta

Saszik

Sutherland

. 2001. Rod contributions to the electroretinogram of the dark-adapted developing zebrafish. Dev Dyn. 222:564–570.

Böcking

Giroud

Reith

. 1995. Consensus report of the ESACP task force on standardization of diagnostic DNA image cytometry. Anal Cell Pathol. 8:67–74.

Böhm

Sprenger

. 1968. Fluorescence cytophotometry: a valuable method for the quantitative determination of nuclear Feulgen-DNA. Histochemie. 16:100–118.

Branchek

Bremiller

. 1984. The development of photoreceptors in the zebrafish, Brachydanio rerio. I. Structure. J Comp Neurol. 224:107–115.

Chieco

Jonker

Melchiorri

Vanni

van Noorden

CJF

. 1994. A user’s guide for avoiding errors in absorbance image cytometry: a review with original experimental observations. Histochem J. 26:1–19.

Clark

. 1981. Visual responses in the developing zebrafish (Brachydanio rerio). PhD dissertation, University of Oregon, Eugene.

Cole

. 1982. Glycol methacrylate embedding of bone and cartilage for light microscopic staining. J Microsc. 127:139–148.

10.

Curcio

Sloan

Kalina

Hendrickson

. 1990. Human photoreceptor topography. J Comp Neurol. 292:497–523.

11.

Doerre

Malicki

. 2002. Genetic analysis of photoreceptor cell development in the zebrafish retina. Mech Dev. 110:125–138.

12.

Duijndam

WAL

Van Duijn

. 1975. The influence of chromatin compactness on the stoichiometry of the Feulgen-Schiff procedure studied in model films. II. Investigations on films containing condensed or swollen chicken erythrocyte nuclei. J Histochem Cytochem. 23:891–900.

13.

Feulgen

Rossenbeck

. 1924. Mikroskopisch-chemischer Nachweis einer Nukleinsäure von Typus der Thymonukleinsäure und die darauf beruhende selective Färbung von Zellkernen in mikroskopischen Präparaten. Hoppe-Seylers Z Physiol Chem. 135:203–248.

14.

Helander

. 1983. Thickness variations within individual paraffin and glycol methacrylate sections. J Microsc. 132:223–227.

15.

Higuchi

Suga

Dannenberg

Schofield

. 1979. Histochemical demonstration of enzyme activities in plastic and paraffin embedding tissue sections. Stain Tech. 54:5–12.

16.

Kiernan

. 2008. Histological and histochemical methods: theory and practice. 4th ed. Oxford, UK: Scion.

17.

Kurnick

. 1955. Histochemistry of nucleic acids. Int Rev Cytol. 4:221.

18.

Ladekarl

. 1994. The influence of tissue processing on quantitative histopathology in breast cancer. J Microsc. 174:93–100.

19.

Lauer

. 2002. New approach to optical diffraction tomography yielding a vector equation of diffraction tomography and a novel tomographic microscope. J Microsc. 205:165–176.

20.

Lessler

. 1953. The nature and specificity of the Feulgen reaction. Int Rev Cytol. 2:231–247.

21.

Miller

Meyer

. 1990. Methods in laboratory investigation. Effects of tissue preparation on glomerular volume and capillary structure in the rat. Lab Invest. 63:862–866.

22.

Østerberg

. 1935. Topography of the layer of rods and cones in the human retina. Acta Ophthalmol. 13(Suppl 6):l-103.

23.

Overend

Stacey

. 1949. Mechanism of the Feulgen nuclear reaction. Nature. 163:538–540.

24.

Pearse

AGE

. 1953. Histochemistry. London: Churchill Livingstone.

25.

Pearse

AGE

. 1968, 1972. Histochemistry: Theoretical and Applied. 3rd ed. London: Churchill Livingstone.

26.

Schulte

Wittekind

. 1989. Standardization of the Feulgen-Schiff technique. Histochemistry. 91:321–331.

27.

Van Duijn

. 1956. A histochemical specific thionine-SO2 reagent and its use in a bi-color method for deoxyribonucleic acid and periodic acid Schiff positive substances. J Histochem Cytochem. 4:55–63.

28.

Zou

Beermann

Wang

Kawakami

Wei

. 2006. The Fugu tyrp1 promoter directs specific GFP expression in zebrafish: tools to study the RPE and the neural crest-derived melanophores. Pigment Cell Res. 19:615–627.