Methyl Green Staining and DNA Strands in Vitro: High Affinity of Methyl Green Dye to Cytosine and Guanine

Acta Histochem. Cytochem. 36 (4): 361–366, 2003 Prof. K Watanabe Memorial Article

Methyl Green Staining and DNA Strands In Vitro: High Affinity of Methyl Green Dye to Cytosine and Guanine

Shinobu Umemura1, Johbu Itoh2, Susumu Takekoshi1, Hideaki Hasegawa2, Masanori Yasuda1, R. Yoshiyuki Osamura1 and Keiichi Watanabe1

1Department of Pathology, 2Laboratory of Structure and Function Research, Tokai University School of Medicine, Bohseidai, Isehara, Kanagawa 259–1193, Japan

Received May 23, 2003; accepted June 25, 2003

We occasionally encounter decreased more strongly than low concentrated staining of methyl green (MG) in heated SSC or distilled water. Denatured and/or sections for antigen retrieval. To exam- DNase treated DNA showed decreased ine the relation between MG staining staining. It was especially noteworthy and conditions of DNA strands, DNA in that oligonucleotide DNA strands of poly various conditions were blotted on the C and poly G strands were positively transfer membrane and stained. It was stained, but very faint MG staining was found that staining intensity of MG was also observed for poly A while no stain- changed due to structural alterations of ing was noted for poly T. In conclusion, DNA strands. Double stranded DNA in we presented that MG staining intensity various solutions, denatured single- was altered by the conditions of DNA in stranded DNA, DNase treated samples, vitro. Intense MG staining for basic acids and short length single-stranded oligo- with three hydrogen bonds, cytosine and nucleotides were examined. DNA diluted guanine, suggest a possible mechanism in high concentrated SSC was stained for MG staining.

Key words: methyl green, DNA strand, cytosine and guanine, hydrogen bond, dot blotting

I. Introduction ever, stained as strongly as RNA by pyronin [5, 6]. Sub- sequent studies showed that the affinity of MG and pyronin Methyl green (MG) staining is widely utilized for to DNA or RNA were not determined by chemical differ- nuclear staining, ever since Brachet reported that Pappen- ences, but depended on the conditions of gene strands [11] heim’s (1899) mixture of methyl green was useful for stain- or, more simply, by the staining speeds by both dyes [9]. We ing chromatin DNA [2, 3]. MG is especially useful for therefore had a hypothesis that the decreased MG staining nuclear counterstaining in immunohistochemistry, because is caused by conformational changes of DNA caused by the color of MG itself is of low contrast compared to heating. Many histochemical stainings are of practical use, Hematoxyrin, and can highlight positive nuclear staining, but the principles and mechanisms of staining have not especially in black-and-white photographs. However, we oc- always been well documented. Especially when using paraf- casionally encounter decreased or no MG staining in heat- fin sections taken from in vivo tissue, we cannot observe induced antigen retrieved sections [7], the reason for which how MG dye binds to some of the specific molecules such is not fully understood. as DNA strands, or how it is altered by heating or other The specificity of MG and pyronin for staining DNA forms of pretreatment. To examine the relationship between and RNA, respectively, was argued in the earlier half of the MG staining and DNA strands, we studied the changes in 20th century. Kurnick showed that highly “polymerised” MG staining for DNA by using dot blotting on membrane DNA was stained weakly by pyronin at one-fifth intensity of in vitro. We presented here the finding that the staining RNA, and that the DNA-histone was also stained weakly as intensity of MG changes in relation to the conditions of the one-sixth of intensity. “Depolymerised” DNA was, how- DNA strands.

Correspondence to: Shinobu Umemura, Department of Pathology, Tokai University School of Medicine, Bohseidai, Isehara, Kanagawa 259–1193, Japan.

361 362 Umemura et al.

II. Materials and Methods DNA treated by DNase Two types of DNA samples were treated by DNase Methods either mildly or strongly as follows: A sample was prepared The principle and strategy of the method is based on by commercially available “activated DNA” (Worthington those of dot blot hybridization. The protocol was briefly Biochemical Corp., Freehold, NJ), by exposure of 250 g described as follows: DNA samples were applied on the DNA to 5P104 g DNase [1]) at 37LC for 15 min. This filter with the aid of a vacuum manifold with dot-like spaces DNase treatment improves the DNA as a primer for E. coli for each sample. The transfer membrane (Immobilon-N, polymerase, but is not enough to produce acid-soluble Millipore, Bedford, MA) was prepared by immersing in nucleotides. This sample was used as DNA mildly treated by 100% methanol for 20 sec and washed in distilled water DNase. The other sample was calf thymus DNA digested (DW). The membrane was attached to dot blot apparatus by DNase I (Stratagene, LaJolla, CA) under the condition (Bio-Rad Lab., Hercules, CA) with manufactured papers. for complete digestion of DNA strands, that is, an exposure DNA samples were prepared as described below, and 50– of 250 g DNA to 250 U DNase I at 37LC for 10 min. This 100 l was applied for each dot. The samples were incubated sample was applied as DNA treated by DNase strongly. The at room temperature (RT) for 5 min, vacuumed by running DNA samples treated by DNase were blotted on the mem- water, and then washed. The detached membrane was im- brane for MG staining, and examined by loading on the 2% mersed in DW, air-dried, and baked at 80LC for 2 hr to fix agarose gel with a marker of /Hind III. the DNA samples on the membrane. The membrane was stained by Veronal-Acetate buffered 1% MG (pH 4.0) for Oligonucleotide DNA strands 10 min, and then washed well in DW and air-dried. The inten- Oligonucleotides in 30 mer length were synthesized. sity of staining and changes of color tones were examined. To avoid dimer or loop formation between each DNA strand, oligonucleotides were made for poly T, poly A, poly DNA samples C and poly G separately. These oligonucleotides were Double-stranded DNA applied separately, or mixtures of poly A and poly T, and Calf thymus DNA (Worthington Biochemical Corp., poly C and poly G diluted in DW or 2P SSC buffers were Freehold, NJ) was diluted at the concentrations of 2.0, 1.0, hybridized in a tube at melting temperature, 37.0LC and 0.5, 0.25, and 0.125 mg/ml by different kinds of solutions. 80LC, respectively, and applied on to the filter. The DNA was diluted in distilled water (DW), sodium chlo- P ride-sodium citrate (SSC) buffer, pH 7.0, 10 SSC contains III. Results 3 M sodium chloride and 0.3 M sodium citrate, or 0.01 M phosphate buffered saline (PBS), and used as double-strand- Conditions of DNA strands and MG staining ed DNA. Double-stranded DNA in different buffers DNA diluted in highly concentrated SSC (10P, 5P) was Denatured single-stranded DNA and re-hybridized double- stranded DNA Calf thymus DNA diluted by DW or 2P SSC at the concentration of 2.0 mg/ml was heated at 100LC for 5 min, and then immediately cooled in an ice water bath, and applied as denatured single-stranded DNA. The heated DNA samples were placed at room temperature (RT), and applied as re- hybridized DNA for the following time courses (30 min, 90 min and 4 hr).

Scanning electron microscopic study for DNA strands DNA samples right after heating and 60 min after heating were examined by scanning electron microscope. DNA samples were diluted in a solution containing 100 mg/ml cytochrome C at a concentration of 2 g/ml. DNA molecules with negatively charged phosphate base were spread to the positively charged network of mono-layered protein, cytochrome C, at the water surface. DNA samples diluted in DW were used for this experiment, and no salt-containing Fig. 1. Methyl green (MG) staining for double stranded DNA in solutions of PBS or SCC were applied. DNA samples different solutions. Calf thymus DNA was diluted by different solutions from the concentration of 2.0 mg/ml to 0.125 mg/ml, and were transferred onto the grid by a spreading method, and blotted on transfer membrane, and then stained by MG. The solu- observed after shadowing for enhancement. tions used for dilutions were 10P, 5P, 2P SSC, 0.01 M PBS and DW (see Materials and Methods). The solutions without DNA were also blotted and stained. Methyl Green Staining In Vitro 363

Fig. 2. MG staining for denatured single-stranded DNA and re-hybridized double-stranded DNA. DNA samples diluted in 2P SSC (a) and in DW (b) were denatured by heating, then allowed to sit at room temperature to re-hybridize. Samples were blotted on the membrane, before heating, right after heating, 30 min, 1 hr 30 min and 4 hr later heating.

Fig. 3. Scanning microscopic observation of DNA strands. Denatured DNA strands right after heating (A) are thinner than re-hybridized DNA at 60 min after heating (B). Bar200 nm. 364 Umemura et al.

intensity, which was detected more clearly in the samples diluted by 2P SSC (Fig. 2). Attempts to observe the constructive changes of DNA strands after heating were carried out using a scanning electron microscope. DNA samples right after heating showed sparse distribution of DNA strings (Fig. 3A). In contrast, re- hybridized DNA strands 60 min after heating formed tangles with increased thickness (Fig. 3B).

DNA treated with DNase Staining intensities of DNA samples treated by DNase were compared. Mild treatment by DNase showed remarkably decreased staining intensity of MG (Fig. 4A). Strong DNase treatment completely negated MG staining. The digestive effects of DNase were confirmed by loading samples on 2% agarose gel (Fig. 4B). The DNA was frag- mented into small pieces loading out of the gel by strong DNase treatment.

Oligonucleotide DNA strands Oligo DNA strands of single strands of poly A, poly T, poly C and poly G in 30 mer length showed different staining intensities depending on base type (Fig. 5). Poly A and poly T showed no or very weak staining for MG independent of the sample buffer (DW or 2P SSC). On the other hand, poly C and poly G were weakly positive especially in 2P SSC buffer. In DW solution, poly C oligonucleotide strands were weakly positive for MG, and poly CG strands detected were of a brownish color for an unknown reason.

IV. Discussion We presented evidence that the staining intensities of Fig. 4. MG staining for DNase treated DNA. DNase treated and MG were changed due to the structural alterations of DNA untreated samples were blotted on to the membrane, and stained by strands, for instance, denatured single-stranded, DNase MG (A). Two types of DNase treated samples were prepared. One treated samples, or short length single-stranded oligo DNA. is a commercially available activated DNA that was mildly treated It was especially noteworthy that poly C and poly G strands by DNase (see Materials and Methods). The other is a sample were positively stained by MG. DNA strands in solution are strongly treated by DNase. DNA samples without DNase treatment present in hybridized (double stranded) or separated (dena- and those treated in strong condition were loaded on to 2% agarose tured) conditions. The homeostatic ratio between hybridized gel (B). DNA and denatured DNA changed in relation to solution temperature (designated as melting curve) [12]. The right- shift of the melting curve in high salt solution may be a strongly stained with occasional aggregation in speckled possible explanation for the results of the present study, i.e. pattern (Fig. 1). DNA diluted in 2P SSC and 0.01 M PBS double-stranded DNA in higher SSC buffer increased stain- showed similar staining intensity, while DNA diluted in DW ing intensity (Fig. 1) and DNA samples in SSC buffers after was lightly stained. All of the solutions without DNA were heating showed increased intensity compared to those in negative. DW (Fig. 2). Staining intensity of MG for DNA samples in SSC buffer right after heating was not constantly weaker Denatured single-stranded DNA and re-hybridized double- than DNA samples before heating but occasionally stronger stranded DNA as shown in Fig. 2. It was speculated that DNA strands in DNA samples were denatured to single-stranded DNA SSC buffer were quickly re-hybridized before blotting onto by heating at 100LC for 5 min. Samples were then left in RT the membrane. Thus, even if the concentration of DNA is the and allowed to be re-natured to double stranded DNA. Stain- same, in the salt-rich solution the DNA strands tend to be ing intensity of denatured single-stranded DNA right after compactly hybridized, or quickly re-hybridized after dena- heating was decreased remarkably in DW diluted DNA. turing, and these conditions lead to the strong MG staining. Re-hybridized DNA strands showed an increased staining The present study showed that decreased staining MG Methyl Green Staining In Vitro 365

Fig. 5. MG staining for oligonucleotides. Oligonucleotide strands of each base, poly A, T, C and G, were blotted on to the membrane separately or in hybridized condition.

is speculated that not all three of the benzene rings are inter- calated between the double stranded DNA. Interestingly, the present study showed that MG was not completely negative for single stranded DNA. Particularly, poly A/poly T oligonucleotides were not stained, but poly C/poly G was con- siderably stained, which cannot be attributable only to intercalation. The differences between poly C/G and poly A/T may provide a clue to the staining mechanism of MG, because the C/G and A/T bases have different numbers of hydrogen bonds. C/G has three hydrogen bonds, while A/T has two [10]. The positive charge of the benzene rings may provide MG with the ability to bind basic acids. Vercanteren proposed the hypothesis that the stereochemical factor in Fig. 6. The chemical structure of MG. MG dye has three benzene the nucleic acid molecule was the presence of a negatively rings, two of which are positively charged. charged phosphate residue [13]. The present study is the first to show that the basic acids cytosine and guanine are related to MG staining. intensity was caused by either denaturing DNA to single In conclusion, we presented evidence that MG staining strands (Fig. 3), or by fragmentation of DNA strands into intensity is altered by the conditions of DNA in vitro, and is small pieces (Fig. 4), or by oligonucleotides (short and sin- related to certain kinds of basic acid. gle-stranded DNA) (Fig. 5). The mechanisms of MG staining has not been well understood, but the decreased staining V. Acknowledgments intensity for single stranded DNA suggests a mechanism of staining by intercalation of dye or nonintercalating groove- The present study was first carried out with the late binding, similar to that for Ethidium Bromide or DAPI. MG Prof. K. Watanabe in 1997–1998 and was interrupted for dye has three benzene rings, two of which are positively four years after his retiring from Tokai University School of charged (Fig. 6). The formulae of Malachite Green and Medicine. On 2 October 2002, Prof. Watanabe called on me Crystal Violet are similar to MG, but they have only one to discuss a plan to complete this work. We had just restarted positive charged site. Malachite Green and Crystal Violet do the plan when he tragically died. We dedicate this work to not stain DNA, therefore the two positive charged sites of his memory. The authors thank Saburo Aita, JEOL Datum MG are assumed to be the binding sites to the negatively Ltd., Tachikawa, Tokyo, for special support to the electron charged phosphate residues of DNA strands. Unlike DAPI microscopic study portion. [4] or Hoechst 33258 [8], MG dye is a non-planar dye, and it 366 Umemura et al.

VI. References 8. Loontiens, F. G., McLaughlin, L. W., Diekmann, S. and Clegg, R. M. (1991) Binding of Hoechst 33258 and 4', 6-diamidino-2- 1. Apochian, H. V. and Kornberg, A. (1962) Enzymatic synthesis of phenylindole to self-complementary decadeoxynucleotides with deoxyribonucleic acid. J. Biol. Chem. 237; 519–525. modified exocyclic base substituents. Biochemistry 30; 182–189. 2. Brachet, J. (1940) La détection histochimique des acides pen- 9. Marshall, P. N. and Horobin, R. W. (1973) Measurements of the tosencléiques. C. R. Soc. Biol. 133; 88–90. affinities of basic and “mordant” dyes for various tissue substates. 3. Brachet, J. (1942) La localization des acides pentosenucléiques Histochemie 36; 303–312. dans les tissues animaux et les oufs d’Amphibiens en voie de 10. Pearse, A. G. E. (1985) Staining of DNA and RNA with basic développement. Archs. Biol. 53; 207–257. dyes. In “Histochemistry. Theoretical and Applied”, vol. 2, 4. Kubista, M., Akerman, B. and Norden, B. (1987) Characteriza- Analytical Technology, 4th ed., Churchill Livingstone, London, tion of interaction between DNA and 4',6-diamidino-2-phenylin- pp. 632–641. dole by optical spectroscopy. Biochemistry 26; 4545–4553. 11. Rosenkranz, H. S. and Bendich, A. (1958) On the nature of the 5. Kurnick, N. B. (1950) The quantitative estimation of desoxy- deoxyribonucleic acid-methyl green reaction. J. Biochem. Cyto- ribosenucleic acid based on methyl green staining. Exp. Cell Res. chem. 4; 663–665. 1; 151–158. 12. Tuan, D., Chetsanga, C. and Doty, P. M. (1977) Sequence specif- 6. Kurnick, N. B. (1950) Methyl green-pyronin. I. Basis of selective ic interaction of the chromosomal proteins with DNA. Nucleic staining of nucleic acids. J. Gen. Physiol. 33; 243–265. Acids Res. 4; 3415–3439. 7. Leong, A. S.-Y. and Leong, F. J. W.-M. (2002) Microwave-stim- 13. Vercanteren, R. (1950) The structure of desoxyribose nucleic acid ulated antigen retrieval. An update. Acta Histochem. Cytochem. in relation to the cytochemical significance of the methylgreen- 35; 367–374. pyronin staining. Enzymologia 14; 134–140.