Mechanisms of chromosome banding

Chromosoma (Bed.) 50, 111--145 (][975) 9 by Springer-Verlag 1975 Mechanisms of Chromosome Banding V. Quinacrine Banding D a v i d E. Comings, B r u c e W. Kovacs, E v a n g e l i t a Avelino, a n d D a v i d C. t t a r r i s Department of Medical Genetics, City of Hope ~National Medical Center Duarte, California Abstract. A series of biochemical investigations were undertaken to determine the mechanism of Q-banding. The results were as follows: 1. I n agreement with previous studies, highly AT-rich DNA, such as poly(dA).poly(dT), markedly enhanced quinaerine fluorescence while GC containing DNA quenched fluorescence. These effects persisted at DNA concentrations comparable to those in the metaphase chromosome. 2. Studies of quinacrine-DNA complexes in regard to the hypoehromism of quinaerine, DNA Tm, DNA viscosity, and equilibrium dialysis, indicated the quinacrine was bound by intercalation with relatively little side binding. 3. Single or double stranded nucleotide polymers, in the form of complete or partial helices, were 1000-fold more effective in quenching than solutions of single nucleotides, suggesting that base stacking is required for quenching. 4. Studies of polymers in the A conformation, such as transfer RNA and DNA-I~NA hybrids, indicated that marked base tilting does not affect the ability of nucleic acids to cause quenching or enhancement of quinacrine fluorescence. 5. Salts inhibit the binding of quinacrine to DI~A. 6. Spermine, polylysine and polyarginine, which bind in the small groove of DNA, inhibited quinacrine binding and quenching, while histones, which probably bind in the large groove, had little effect. This correlated with the observation that removal of histones with acid has no effect on Q-banding. 7. Mouse liver chromatin was separated into five fractions. At concentrations of quinacrine from 2 • 10_8 to 2 • 10-5 M all fractions inhibited to varying degrees the ability of the chromatin DNA to bind quinacrine and quench quinacrine fluorescence. At saturating levels of quinacrine two fractions, the 400 g pellet (rich in heteroehromatin) and a dispersed euchromatin supematant fraction, showed a decreased number of binding sites for quinacrine. These two fractions were also the richest in non-histone proteins. 8. DNA isolated from the different fractions all showed identical quenching of quinacrine fluorescence. 9. Mouse GC-rich, mid-band, AT-rich, and satellite DNA, isolated by CsC1 and Cs~SO4-Ag+ centrifugation all showed identical quenching of quinacrine fluorescence, indicating that within a given organism, except for very AT or GC-rich satellites, the variation in base composition is not adequate to explain Q-banding. - - We interpret these results to indicate that: (a) quinacrine binds to chromatin by intercalation of the three planar rings with the large group at position 9 lying in the small groove of DNA, (b) most pale staining regions are due to a decrease binding of quinacrine, and (c) this inhibition of binding is predominately due to non-histone proteins. Introduction Since highly AT-rich D N A enhances a n d D N A c o n t a i n i n g GC bases quenches q u i n a c r i n e fluorescence (Weisblum a n d de H a s e t h , 1972, 1973 ; P a c h m a n n a n d Rigler, 1972 ; MicheIson et al., 1972 ; Selander a n d 8 Chromosoma(Berl.), Bd. 50 112 D.E. Comingsetal. de la Chapelle, 1973) it has been suggested that variations in the base composition of DNA along the chromosome form the primary basis for Q-banding. This seems quite reasonable since independent evidence from DNA ultracentrifugation (Tobia et al., 1971 ; Bostock and Prescott, 1971; Comings, 1972a), staining with fluorescent labeled antinueleoside antibodies (Dev et al., 1972; Schreek et al., 1973), and acridine orange banding (Comings, 1973; Bobrow and Madan, 1973) all agree that Qand G-banded regions, composed of intercalary heterochromatin (Comings, 1972b}, contain relatively AT-rich DNA. Further support for the importance of base composition comes from the observation that chromosomal regions containing AT-rich satellite DNA fluoresce intensely with quinacrine (Ellison and Barr, 1972). Despite this evidence, there are two observations which suggest that protein-DNA interactions may also play a role in Q-banding. First, the centromeric heterochromatin of the mouse, which contains an AT-rich satellite, fluoresces poorly (Rowley and Bodmer, 1971; Dev etal., 1971); the eentromeric heterochromatin of the muntjac, which contains a satellite that has the same base composition as main band DNA, fluoresces poorly (Comings, 1971); and the secondary constrictions on human chromosomes 1, 9 and 16 contain AT-rich (Corneo et al., 1970) satellite II (Jones and Corneo, 1971) but fluoresce poorly (Caspersson et al., 1970). This argument has been countered by the proposal that guanine bases appropriately interspersed in these satellite sequences could quench fluorescence, even though the total DNA was relatively AT-rich (Weisblum and de Haseth, 1973). However, the base sequence of mouse satellite shows that it contains regions of five adjacent AT pairs (Southern, 1970), and thus might enhance fluorescence (Pachmann and Rigler, 1972). Second, Gottesfeld et al. (1974) have separated rat liver and Drosophila chromatin into extended and condensed regions by treatment with DNase-II and precipitation in magnesium chloride. The condensed fraction showed less quenching of quinacrine fluorescence than the extended fraction but the DNA extracted from both showed the same effect on quinacrine fluorescence, indicating the importance of pro~einDNA interactions in Q-banding. They suggest both fractions bound quinacrine but the heterochromatic DNA in a more C-like conformation was less effective in quenching fluorescence than euchromatic DNA in a more B-like conformation. Since these observations strongly suggest that protein-DNA interactions can play a significant role in Q-banding, the effect of various salts, spermidine, histones, polylysine, polyarginine, and chromatin subfractions have been examined to elucidate the factors involved in the enhancement and quenching of quinacrine fluorescence. Mechanisms of Chromosome Banding. V 113 Materials and Methods Reagents. Quinacrine dihydrochloride (Sigma) was used throughout. The calf t h y m u s DNA, nucieotides, and calf t h y m u s histone fractions were obtained from Sigma. The remaining types of DNA, polynucleotides, transfer RNA, a n d amino acid polymers were obtained from Miles Laboratory. The following polyamino acids with their molecular weights and degree of polymerization (D.P.) were used: poly-l-lysine 2200 (D.P. 10), 5900 (D.P. 28), 7200 (D.P. 34), 22700 D.P. (108), a n d poly-l-arginine 13900 (D.P. 72). The complexing of the proteins to DNA was b y simple mixing. I n most cases a n equal volume of a D N A containing solution and a protein containing solution were slowly mixed with constant vortexing. I n various experiments the components were dissolved in 5 m ~ phosphate, 1 mM EDTA, p~I 6.8 ; 5 mM tris-HC1, 1 mM EDTA, p H 7.0; or 1 mM tris, 1 mM EDTA, p i t 7.0. No differences were noted with the use of phosphate versus tris. The concentration of bacterial and m a m m a l i a n DNA was determined optically based on 1.0 O.D. ~ 50 ~g/ml ~ 1.6 • 10-4 M of nucleotides. Instrumentation. Optical density measurements were made with a Beckman double b e a m recording spectrophotometer, Model 25. To obtain the optical density of a quinacrine solution, a tracing from 525 to 400 n m was made a n d the O.D. at 525 n m was subtracted from the pea!( O.D. Fluorometrie measurements were made in a n Aminco-Bowman spectrofluorometer with excitation a t 425 n m a n d emission at 495 nm. I n most quinacrine fluorescence experiments a concentration of 2 X 10-6 M was used. Although quinacrine b o u n d to D N A undergoes a change in the shape of t h e absorbance curve the peak only shifts from 425 to 430 nm. This was n o t corrected for a n d all excitation was a t 425 nm. For DNA b o u n d quinacrine this results in a slight under-measurem e n t of the fluorescence or a slight exaggeration of the degree of D N A induced quenching. Circular dichroism measurements were done with a Jasco ORD-CD spectro0.002x polarimeter, where A e = %-et~ [ D N A ] ' with x = t h e difference in decameters (which are the dimensions on the chart paper) between the baseline with buffer alone a n d the CD reading a t a given wavelength, 0.002 is a sensitivity setting, and [DNA] = the molarity of D N A nucleotides. The light p a t h was 1 cm. Calf DNA with a Ae of 2.5 at 275 n m (Allen et al., 1972) was used to standardize the machine. For relative viscosity a Zimm-Crothers (1962) low shear viscometer (Beckman) was utilized with relative viscosity ~ r p m for D3NA solution/rpm of buffer alone. Isolation o] Chromatin Sub]factions. The method of chromatin subfractionation was similar to t h a t used previously (Mattoecia a n d Comings, 1971) with a few modifications described in detailel sewhere (Comings and Harris, unpubl.). Briefly, nuclei purified b y centrifugation t h r o u g h 2.4 M sucrose, 10-a M iris, p]~[ 7, 10-3 MgC12, were washed three times in 0.14 M NaC1, 0.01 M tris, p H 7, 10-3 M Ha bisulfide, 10-a M cadmium sulfate, 5 • 10-3 M MgC12, 3 • 10-3 M CaCI~, a n d resuspended in 0.34 M sucrose. After sonication the chromatin was centrifuged at 400 g for 5 minutes to produce the 400 g pellet containing nucleoli and heterochromatin. The s u p e r n a t a n t was centrifuged at 4000 g for 20 minutes to produce a heterochromatin pellet, also containing nucleoli and heteroehromatin. The supern a t a n t was centrifuged a t 20000 g for 20 minutes to produce a n intermediate pellet. The s u p e r n a t a n t was t h e n b r o u g h t to 0.14 M with NaC1 a n d the chromatin pelleted at 3000 g for 10 minutes to produce the euchromatin pellet. The supern a t a n t represented "euchromatin s u p e r n a t a n t " . The protein/DNA ratios of these 8* 114 D.E. Comings et al. fractions (mean of six determinations) were as follows: 400 g 2.34, heterochromatin 2.24, intermediate 2.02, euehromatin pellet 2.04, and euchromatin supernatant 4.62. The percent that the DNA of each fraction represented of the total nuclear DNA was 400 g 3.4%, heterochromatin 12.8 %, intermediate 27.7 %, euchromatin pellet 51.5%, and euchromatin supernatant 4.6%. For equilibrium dialysis experiments the chromatin fractions were resuspended in 10-a M tris, 10-a M EDTA, exposed to 2 to 5 9g/ml of l~Nase I (Sigma) and 2-5 U/ml of l~Nase T l (Cal-Biochem) at room temperature for 30 minutes, then dialized against the same buffer at 4 ~ overnight with two changes. This resulted in the complete removal of RNA as determined by the orcinol test. Isolation o] Satellite and Main Band DNA Fractions. Mouse DNA was isolated from fresh liver by the technique of Murmur (1961). Satellite DNA was isolated by the technique of Lieberman (1973) with Ag+-Cs2SO~ at molar ratio of Ag+ to DNA of 0.3. The DNA, at a concentration of 1000 ~g per tube, was first centrifuged at 44000 rpm for 24 hours, then at 33000 rpm for 65 hours. By analytical ultracentrifugation the satellite fraction contained 60% satellite and 40% AT-rich DNA. AT-rich, mid-band, and GC-rich fractions of mouse DNA were prepared by preparative ultracentrifugation in CsC1, with each fraction representing approximately one-third of the main band. Results and Discussion C h a r a c t e r i s t i c s o[ the D y e Quinacrine (Atebrin, Mepacrine) is one of t h e acridine f a m i l y of d y e s possessing a core s t r u c t u r e of t h r e e f i a t a r o m a t i c rings (Fig. 1). Other m e m b e r s of this f a m i l y include ~cridine orange, a n d proflavine. These relationships are e m p h a s i z e d because of t h e well k n o w n a b i l i t y of ~cridine orange a n d p r o f l a v i n e to b i n d to a n d i n t e r c a l a t e between t h e s t a c k e d bases of D N A ( L e r m a n , 1961, 1963). As will be e m p h a s i z e d in this p a p e r , m a n y of t h e o b s e r v a t i o n s on i n t e r a c t i o n b e t w e e n aminoacridines a n d D N A (Peacocke, 1973) also h o l d for quinacrine. DNA Alone Calf t h y m u s D N A , with ~n A T concent of 57 percent, m ~ r k e d l y q u e n c h e d quinacrine fluorescence, while h i g h l y A T - r i c h D N A , such as NH CH~/'CH3 \(CH2)3 N (C2H5)2 CH30~/~ | Quinacrine Fig. 1 (see t e x t ) Mechanisms of Chromosome Banding. V n 2.4 n n t15 i Quinacrine 2 x 10 -s M 2.5 /o 2.2 2i] ~o-o-o-o-o i Poly(dA) 9Poiy(dT) 2,0 1,9 ~,8 1,7 1.6 o 1.5 ].4 1.5 1.2 1,1 9 o 'e .7 .6 .4 ,~ ,2 9 Calf DNA ~176176 ~176176176 ~ .1 10-8 10-7 10-6 ]0-5 10-4 10-~ 10-2 Moles DNA Nucleotides Fig. 2. Effect of varying concentrations of calf thymus DNA and poly(dA) poly(dT) 9 on the fluorescence of quinacrine dihydrochloride at 2 x 10-8 M in 1 mM tris, 1 mqY[ EDTA, pH 7.0. In this and all subsequent figures the DNA concentrations are plotted on a log scale poly(dA), poly(dT), markedly enhanced fluorescence (Fig. 2) (Weisblum and de Haseth, 1972, 1973 ; Pachmann and l~igler, 1972 ; Selander, 1973). Studies of artificial polynucleotides suggest enhancement only occurs when guanine is missing from a sufficiently long stretch of D N A to leave several adjacent AT pairs (Pachmann and Rigler, 1972). The m a x i m u m quenching or enhancement began at D N A concentrations of 1.5 X 10 -s M and persisted up to a concentration of 1.1 • 10 -2 M or 3.4 mg/ml. This approaches the concentration of D N A in the metaphase chromosome of approximately 7.0 X 10 -2 M or 22 mg/ml (Comings et al., 1973), and indicates that quenching or enhancement of quinaerine fluorescence persists at high concentrations of D N A . Conclusion: D N A containing guanine quenches ]luorescence, highly AT-rich D N A enhances ]luorescence, and this e]]ect persists at the concentration o[ D N A in the metaphase chromosome. D. E. Comings et al. 116 [~]/[0~A] , ,,-; ~4 . . . . . ~.0 --c'~aD~oo." C0rf DNA .9 \ % o 6.0 x 10 -5 M Ouinacrine 98 [*, .4 .5 .3 li I %00 ,2 oo ...... o o.l,~e n 10-7 ,9 9 4 .8 cJ (5 ~o~ 6 10-6 a 1.O n o o o oe ~ .6 n 1 0 - 5 10-4 10-5 10-2 Moles DNA Nucleotides 2 g: .5 Fig. 3. Correlation between quenching of quinacrine fluorescence and hypochromism of quinacrine optical density by varying concentrations of DNA. The concentration of quinaerine for both fluorescence and O.D. was 6.0 • 10-6 When planar dyes intercalate (Lerman, 1961) in DNA their extinction coefficient at peak adsorption is decreased (Bradley and Felsenfeld, 1959; Paehmann and Rigler, 1972; Blake and Peaeocke, 1968; Kurnick and Radcliffe, 1962) due to interaction of transition dipoles of stacked molecules (Tinoco, 1960; Van Holde, 1971). This is similar to the hypochromic effect observed when nucleotides are stacked in the DNA helix (Tinoco, 1960; Van Holde, 1971). This effect allows direct monitoring of the binding of quinaerine to D N A or chromatin (Gottesfeld et al., 1974). I n Fig. 3 the relative fluorescence and relative O.D. of 6 • 10-5 M of quinacrine was determined in the presence of varying concentrations of DNA. This indicated there was a close correlation between hypochromism and fluorescence quenching. Both began at a DNA concentration of 3 • 10 -6 M and the effect was m a x i m u m at 5 • 10 -5 M. A similar relationship between fluorescent enhancement and hypochromism was noted with poly(dA)-poly(dT) (not shown). Conclusion: Quenching o/quinacrine fluorescence correlates closely with intercalative binding o/quinacrine to D N A . O t h e r E v i d e n c e ~or Intercalation A further characteristic of intercalation of acridines is t h a t the DNA helix is stabilized to thermal denaturation (Gersch and Jordan, 1965; Kleinwachter and Koudelka, 1964; Kleinwachter etal., 1969). To Mechanisms of Chromosome Banding. V 117 Ouinacrine ~'Hydrophobic Bonds' Fig. 4 (see text) examine this, the effect of 40 ~M of quinacrine on 50 ~g/ml of calf thymus DNA ([dye]/[DNA] of 0.375) was determined. The T m of the calf DNA alone (in 1 mM tris, 1 mM EDTA, p H 7.0) was 57.5 ~ while the Tm of the dye-DNA complex was markedly increased to 98.0 ~ This remarkable stabilization to thermal denaturation was even greater than that reported for other acridines (Gersch and Jordan, 1965; Kleinwachter and Koudelka, 1964; Kleinwachter et al., 1969). This effect is presumably due to an increase in the hydrophobic stacking free energy. To be more specific, much of the stabilization of the DNA helix is due to the strong hydrophobic bonding between the stacked planar bases (Crothers and Zimm, 1964; Yang and Samejima, 1969). Much of the strong type I intercalative binding of acridine dyes to DNA (Peacocke, 1973; Blake and Peacocke, 1968) is due to the formation of hydrophobic bonding between the three planar rings of acridines and the planar rings of the bases. In addition, the three rings serve to form a bridge across the hydrogen bonded base pairs and further strengthen the helix, especially in salt concentrations low enough not to inhibit the binding of the quinacrine to DZX~A(Fig. 4). Intercalation also increases the relative viscosity of DNA due to extension of the helix (Lerman, 1961; Peacocke, 1973; Lloyd et al., 1968 ; Armstrong et al., 1970 ; Drummond et al., 1966 ; Cohen and Eisenberg, 1969). To examine this, the relative viscosity of 75 ~g/ml of calf DNA in 1 mM tris, 1 mM EDTA, p H 7.0, was determined with a ZimmCrothers low shear viscometer (Zimm and Crothers, 1962) and compared to the relative viscosity of the DNA in the presence of 60 ~M quinacrine (r = 0.25). The relative viscosity of the DNA was 1.57 while that of the dye-DNA complex was 2.08. This increase is similar to that observed by Kurnick and Radcliffe (1962). Studies of flow dichroism Lerman (1963) and circular dichroism (Comings and Kovacs, unpublished) also indicate quinacrine is intercalated. Conclusion: The complexing of quinacrine to D N A increases the D N A thermal stability and viscosity. This is consistent with binding by intercalation. D. E. Comings et al. 118 A Free 'lli Deoxynucfeos[des .10 i .9 LL .5 .2 ,1 L 10-6 10-7 B r I 10-5 Conc. (moles) i 1AM i x t u r e s of Free Deoxynucleosides 1.o . ~ ~ = ~ ~: - ~ ~. ~ : _ - ~. . . ~ .~ ~ o .9 - - - a .8 .7 ~ GC ~ ................~ . ~~x'-"~-~'-~-~. i ".'~o-.,-- 0 % ~----~.~.. "~+'- 20 % '~.... f 60% ~ 4o% ,'~*~--,00% \ \ .6 t __ 10 - 4 9 \ 80% 4 ii~m CT DNA .5 .2 .I i 1.7 x 10-z ~ 1ZxfO-6 r i,TxlO"5 Moles DNA r i 1.7x10-4 i r ].?xlO -5 Nucleotides Fig. 5A and B. (A) The effect of increasing concentrations of the free nucleosides deoxycytidine (C), deoxyguanosine (G), deoxyadenosine (A), and thymidine (T) on quinacrine fluorescence. (B) The effect of increasing concentrations of mixtures of free deoxyribonueleotides, made to a GC content of 0 to 100%, on quinacrine fluorescence. :No significant quenching was present until a concentration of 1.7 • 10-3 'M (total nucleosides) was reached. Comparable quenching occurred with calf thymus (CT) DNA at a 1000-fold lower concentration Nucleosides and Nucleotides Alone To d e t e r m i n e whether the stacked helical s t r u c t u r e of D N A was a p r e r e q u i s i t e for quenching or e n h a n c i n g of quinacrine fluorescence, t h e free deoxynucleosides (Fig. 5A) a n d d e o x y n u c l e o t i d e s were e x a m i n e d . N o n e of t h e four nucleosides or nucleotides showed a n y effect on fluorescence e x c e p t when r e l a t i v e l y high c o n c e n t r a t i o n s of 1 0 - 3 M were r e a c h e d w h e r e u p o n t h e y all caused a slight degree of quenching. This was p r o b a b l y due to p a r t i a l base stacking. T h e r e was no difference in t h e degree of quenching when g u a n i n e m o n o p h o s p h a t e was c o m p a r e d t o d e o x y - g u a n i n e - m o n o p h o s p h a t e , or when adenosine monophos- Mechanisms of Chromosome Banding. V 119 phate was compared to deoxyadenosine monophosphate. Thus, the presence of the 5' phosphate or the 2' O H group on the sugar had no effect on fluorescence. Selander and de la Chapelle et al. (1973) have reported t h a t in an artificial mixture of the four bases, quenching increased as the percent GC content increased, except at 100% GC where there was a slight decrease in the degree of quenching. We have repeated this experiment with a similar result (Fig. 5B). The degree of quenching increased up to 80% GC, then decreased slightly at 100% GC. I t is important to point out t h a t none of the individual nucleosides or nucleotides or mixtures of nucleosides caused enhancement of fluorescence, and when the deoxynucleosides are in the form of a D~qA helix comparable quenching occurs at a 1000 fold lower nucleotide concentration. Conclusion: Comparison of the effect of free nucleosides and free nucleotides alone with the effect o / D N A is consistent with the conclusion that base stacking is required/or quenching or enhancement o/quinacrine fluorescence. T h e R o l e 0I B a s e T i l t i n g To evaluate the effect of base tilting, the fluorescence of quinacrine in the presence of transfer I~NA and a DNA-I~NA hybrid was examined. X-ray diffraction studies indicate t h a t the helical structure of both of these compounds is in the A configuration (Arnott, 1970 ; Arnott et al, 1966, 1968 ; Milman et al., 1967). I n studies of DNA films, Tunis-Schneider and Maestre (1970) have shown t h a t the circular dichroism spectra of DNA in the A form is non-conservative with a strong positive Cotton effect and a small negative Cotton effect. The m a x i m u m elliptieity was shifted from 275 nm for D N A in the B form to 262 for D N A in the A form. The transfer I~NA showed an A-type pattern (Fig. 6) (Green and Mahler, 1970). I n A-I~NA the bases are tilted approximately 14 ~ from the perpendicular to the helix axis (Arnott, 1970). The pitch is 28 A with 11 bases per turn for a distance between the bases (along the helix axis) of 2.5 A (Arnott, 1970), compared to 3.4 A for B-DNA, and 3.32 A for C-DNA. The quenching caused b y transfer R N A was almost comparable to t h a t of calf thymus D N A (Fig. 6 inset). This indicates t h a t a nucleic acid can cause quenching of quinacrine fluorescence despite significant tilting of the bases. R~qA-DNA hybrids also have an A conformation (Arnott, 1970) and the circular dichroism spectrum of such a hybrid, poly(rA).poly(dT) was typical of the A conformation (Fig. 7A). Since poly(dA).poly(dT) causes significant enhancement of quinacrine fluorescence, poly(rA), poly(dT) should also cause enhancement and this was observed (Fig. 7B). 120 D.E. Comings et al. ,6 § 1.1 [ ~.a .7 .s L - 2 200 \ x %, __~/ 220 /o, //// x 4-o .~...crD.A \\ • ,\ ~X o \0\ \ _ / I ~x/,o X'• t RNA \ \x , ~ 10_5 10_4 ,~[/X ]0_6 Moles Nucleotides / /o\ ~" \ \ ~/ ~10_7 +] _1 x t RNA ---_.._ ~F Ae +2 ix \ ] / \\ 240"o j / 260 ~nm 280 .500 520 Fig. 6. Circular dichroism of yeast transfer RNA and calf thymus DNA. The spectrum of calf thymus DNA is typical of the B conformation, and the spectrum of the transfer RNA is typical of the A conformation. The inset shows that transfer RNA is effective in quenching quinacrine, similar to DNA, despite its A conformation with tilting of the bases Since tilting of the bases changes the interbase distance when measured along the helix axis, but not when measured on a plane perpendicular to the bases (Crick and Watson, 1953), the interbase distance available for intercalation would not be affected for nucleic acids in the A or C conformation. Conclusion: Tilting el the bases does not a/[ect the ability o/ nucleic acids to quench or enhance quinacrine fluorescence. Single versus Double Stranded DNA Calf thymus DNA was denatured b y boiling for 10 minutes and rapidly cooled. The degree of quenching of quinacrine fluorescence by this single stranded DNA was similar to t h a t of native DNA (not shown). Weisblum and de Haseth (1973), Selander (1973) and Miehelson et al. (1972) obtained similar results with single stranded synthetic polynueleotides. Poly(dT), poly(dA), poly(A) and poly(U) enhanced fluorescence, poly(dG) and poly(G) quenched fluorescence, and poly(dC) and poly(C) had little effect on fluorescence. These results do not invalidate the conclusion t h a t base stacking is 121 Mechanisms of Chromosome Banding. V ~, +lO[ +81 poty(rA). poly(d T ) ! L ~ +6 +4t +2~o AeO CT DNA -2I~ a/( B 20 1.9 1,8 c o/ 1,5 -4 o ~_ 1 , 4 -6 '>~ 1.5 1.2 a~ 1 . 1 / j" 1.0 -]0 .9 i i 1 200 220 240 ~6 0 280 500 320 Xnm I Poly (r A). Poly(dT) .8 10"7 10-6 t0-5 10-4 10-5 Moles DNA Nucleotides Fig. 7. (A) Circular dichroism of poly(rA).poly(dT) shows a spectrum typical of the A conformation. Calf thymus DNA is in the B conformation. (B) Poly(rA). poly(dT) causes marked enhancement of quinacrine fluorescence, similar to that with poly(dA), poly(dT), despite marked tilting of the bases important since the bases remain stacked in most single stranded polynucleotides because of hydrophobie interactions between the planar molecules (Crothers and Zimm, 1964; Holcomb and Tinoco, 1965; u and Samejima, 1969; Green and Mahler, 1970). Strong interealative binding of other acridine dyes to single stranded nucleic acids has been reported (Ichimura et al., 1969; Peacoeke, 1973; Blake and Peaeocke, 1968; Kurnick and Radcliffe, 1962). For a base pair, a variation in the plane of the purine in relation to the pyrimidine is termed twist (Arnott, 1970). Since base pairing and twisting are absent in single stranded DNA, twisting plays no role in the quenching or enhancement of quinaerine fluorescence. Conclusion: Base twisting does not play a role in quinaerine quenching or enhancement. ElIect oI Salts Lithium chloride was used to investigate the effect of salts on the ability of D N A to quench quinacrine fluorescence. This compound was D.E. Comings et al. 122 0 M +5 +z L / =/u--a~ [ ~I ! § [- L / /z \ '~,-?~\ , I \~ / 0 . 5 M Li CI "~,~V, \X z~ -~-.~- h.xx/4,0 ,~,/o_ M _.~.,~ Li C[ 3 l.l x \"~'~X/' -21 .7 \~ ~ / X\ o=.6 ~ 5 \x_/ -5 ~ A ~ . ~ . _ ii+12 M A alone 2 3 -4 200 220 240 Ouinocrine C 260 knm DH 280 500 10-7 10-6 10-5 10-4 10-5 Moles DNA Nucleotides 2xlO-SM Fluorescence .1S ,]6 C.T. DNA Lilhlum Chloride 1.0 ~-o~,-~ ~ 9 o ~-., -o + .SM L~Cl .9 .8 24 9 ~r 0J ~ 4 + .]M L[CI ,10 DNA alone .08 B p_ .5 c/ ,]2 Ci 0 10 20 50 40 50 CaJf Thymus DNA p g / m [ 60 70 80 ~ .5 2 CoIf DNA .1 0 L__r 11 ~ 9 r r r r r 7 5 Moles Li CI ~ 3 i 1 0 Fig. 8. (A) Circular dichroism of calf thymus DNA in the presence of varying concentrations of lithium chloride. At 11 M the spectrum resembles that of DNA in the C conformation. At less than 0.5 M LiCl the CD changes are negligible. (B) Relative quinacrine fluorescence in the presence of calf thymus DNA and varying concentrations of lithium chloride. As little as 0.25 M of lithium chloride inhibits the ability of the D:NA to quench quinacrine fluorescence. (C) The optical density of 2 • 10-5 M of quin~crine in the presence of varying amounts of calf thymus DNA and varying concentrations of lithium chloride. (D) The relative effect of lithium chloride on fluorescence of quinacrine, O.D. of quinacrine at 425 am, and circular dichroism of 75 Fg/ml of calf thymus DNA. To allow the CD to be on the same relative scale the As at 275 nm was multiplied by 0.2 and this subtracted from 1.0 chosen since at higher concentrations it also induces D N A to take on the C conformation (Marvin et al., 1961) and a n y role that conformational changes m i g h t play could also be examined. These changes can be followed b y circular diehroism. As shown in Fig. 8A, at concentrations of LiCI of 4 M or greater the D N A undergoes a B-->C shift in conformation (Adler et al., 1971 ; Studdert et al., 1972; H a n l o n Mechanisms of Chromosome Banding. V 123 Quinacrine binding io Calf DNA 40,000- 0.0 M KCl slope =-2.4 x I0 -6 30,000- r/c I 0.05 M KCI slope : -5,0 • 20,000- 10,000- i ' ~ O . l . M KCI slope=2.7 xlO 5 ;j_ II IIqlp ~~ 1 i II Ii i .2 .5 .22 .27 i i ~ i .4 .5 .6 ,7 -- Fig. 9. Scatchard plot of the binding of quinacrine dihydrochloride to calf DNA. r = [dye]/[DNA]. C concentration of free quinacrine. For dialysis, 0.5 ml of 75 t~g/ ml of calf DNA in 5 m ~ phosphate, 1 mM EDTA, pH 6.8, was placed inside the bags, and 18 ml of the same buffer outside containing from 1 to 100 t~M of quinacrine. The test tubes were gently shaken for 48 hours at 4~ Aliquots were removed, diluted with three parts methanol and the concentration of dye determined utilizing an extinction coefficient of 8000. In subsequent experiments it was found that the extinction coefficient varied somewhat according to the concentration of the dye. When this was taken into consideration the results (Fig. t9) were only slightly different et al., 1972) as i n d i c a t e d b y a progressive decrease in t h e ~ a t 275 n m (Tunis-Schneider a n d Maestre, 1970). T h e r e is little change in conform a t i o n a t LiC1 c o n c e n t r a t i o n s of 0.5 M or less. As shown in Figs. 8 B a n d C, LiC1 c o n c e n t r a t i o n s as little as 0.i M i n h i b i t t h e a b i l i t y of calf D N A to b o t h quench quinacrine fluorescence a n d decrease its optical density. T h e r e l a t i v e effects of LiC1 on fluorescence, O.D. b y circular diehroism are shown in Fig. 8D. This indicates t h a t t h e ionic effects of t h e salt are t o : (a) i n h i b i t q u i n a c r i n e from b i n d i n g to D N A b y intercalation, (b) t h i s p r e v e n t s t h e D N A from quenching quinacrine fluorescence, a n d (c) c o n f o r m a t i o n changes are n o t p l a y i n g a role in t h e salt 124 D. E. Comings et al. A Ethylene Glycol +3 ~o%- +2 _.~-~'~'5~ 40 %, -.?o,~, ~ . / ,~ . J ' -./ ~\\,, 6o O/o----g~., ~ ..... % - +1 ae +0 ~. ~/ -5 -4 200 2~o 2s 2~o 2 4 o ~250 2~o 2r ~ 280 ~o 3oo:=--31o )%nm ~/ .~ \\~ 9 ~ 6 9/ ] \ k \ Relative Hypochromism o.... Re ative "-- e / C i r c u l a r ~ e Dichroism Relotive Fluorescence 100 so 6o 4o 2o & % Ethylene @ycof Fig. 10. (A) The effect of increasing concentrations of ethylene glycol on the circular diehroism of calf DNA. In 90 and 100% ethylene glycol the DNA adopts a C-like conformation. The 0 to 90% solutions contain 0.05 KF and 1 mM EDTA. (B) The effect of increasing concentrations of ethylene glycol on the relative fluorescence and hypochromism of quinacrine and the cireul&r diehroism of calf DNA (75 tzg/ml). The relative fluorescence and relative O.D. = values in solution with DNA/values in comparable solution without DATA. For O.D. quinacrine was at 2 • 10-~ M. To produce a comparable relative circular dichroism the As at 280 nm was multiplied by 0.2 and this subtracted from 1.0 effects. Similar studies w i t h KC1 g a v e t h e s a m e results. T h e salts could also p r e v e n t Q-banding. W h e n methanol-acetic acid f i x e d m o u s e c h r o m o s o m e s were placed in 0.5 M LiC1, t h e n stained in the presence of LiC1 and quinacrine, there was no fluorescence of either the c h r o m o s o m e s or nuclei. Mechanisms of Chromosome Banding. V 125 To further examine the effects of salts, equilibrium dialysis studies were done and presented as Seatchard (1949) plots (Fig. 9). In the absence of KC1 the slope was very steep (--2.4 • l0 s) and showed an intercept at an r value of 0.27, indicating one dye molecule bound for every two base pairs. This binding is typical of Type I intercalativc binding (Peacocke, 1973 ; Blake and Peaeoeke, 1968). At higher r values there was minimal additional binding suggesting that in contrast to dyes like acridine orange (Frederieq and Houssier, 1972) there was little Type I I side binding. In the presence of 0.05 of 0.i M KCI binding was inhibited and the slope decreased by a factor of 10. Salts also inhibit the ability of other aminoacridine dyes to bind to DNA (Peacocke and Skerrett, 1956; LePecq and Paoletti, 1967). Conclusion: Equilibrium dialysis experiments indicate the major mode o[ binding is by intercalation with little external binding. Salts inhibit the quenching o[ quinacrine [luorescenee by preventing it from binding to DNA. Ef[ect of Ethylene Glycol In addition to salts, ethylene glycol can also cause a B - + C shift in the conformation of DNA in solution (Green and Mahler, 1968; Nelson and Johnson, 1970). The reason salts can inhibit the quenching of quinacrine fluorescence before they cause a significant B - ~ C shift in the conformation of DNA, is that their charge is playing an important role in inhibiting quinacrine binding. Perhaps if the C configuration were induced by a non-ionic compound quinaerine would bind well to DNA and there might be a good correlation between the amount of DNA in the C form and the amount of quinaerine fluorescence. To investigate this the effects of ethylene glycol were examined. As shown in Fig. 10A, concentrations of ethylene glycol ranging from 20 to 100% caused a progressive shift from the circular dichroism spectra typical of the B form to that typical of the C form of DNA. Fig. 10B shows the effect of increasing concentrations of ethylene glycol on the relative fluorescence and relative hypochromism of quinacrine and circular diehroism of DNA. At 0 % ethylene glycol the calf thymus DNA quenched the quinacrine fluorescence to 32% of the control value. This quenching progressively decreased as the percent of ethylene glycol increased. The relative hypochromism exactly paralleled the relative fluorescence. At 0% ethylene glycol the peak O.D. of quinacrine in the presence of DNA was 0.66 of the value in the absence o~ DNA. With increasing concentrations of ethylene glycol this relative hypochromism decreased, indicating an inhibition of intercalative binding of quinacrine to DNA. In the range from 0 to 40% ethylene glycol, the effect on circular 126 D.E. Comings et al. dichroism exactly paralleled the effect on relative fluorescence and hypochromism. This suggests one of two possibilities. 1. The primary effect of ethylene glycol is on the conformation of DNA and this inhibits the binding of quinacrine to DNA and this prevents the quenching of quinacrine fluorescence. 2. The ethylene glycol results in dehydration of the DNA and this causes both a change in the conformation of the DNA and inhibits qninacrine binding. The above observations that the tilting of the bases per se, in RNA and DNA-RNA hybrids, had no effect on the ability of nucleic acids to quench quinacrine fluorescence, and the known correlation between dehydration of DNA and its assumption of the C-conformation (Chang et al., 1973), strongly support the second explanation. Since the hypochromism of planar dyes in the presence of DNA is due to an interaction of the transition dipoles of stacked molecules (Tinoco, 1960), i.e. intercalation, the above experiments have indicated that the intercalative mode of binding has been inhibited b y salts and ethylene glycol. I t has provided no information on binding by side stacking. This is important since it would influence whether regions of the chromosome containing DNA in a C-like conformation would result in an increase or decrease in quinacrine fluorescence. For example, a decrease in intercalative binding could result in (a) a failure of the quinacrine to bind to DNA and thus cause a decrease in fluorescence in that region, or (b) a shift from intercalative binding with quenching to side stacking without quenching, thus producing greater fluorescence in that region. To investigate side stacking an equilibrium dialysis experiment was done with 18 ml of 95% ethylene glycol, 5% 0.05 M NaF, 0.001 M EDTA on the outside of the bags, and 0.5 ml of the same plus 35 ~g/ml of calf DNA on the inside of the bags. A series of 13 such tubes containing from 1 to 100 ~M of quinaerine were allowed to equilibrate on a shaker at 4~ for four days. At the end of this time the concentration of qninacrine inside the dialysis bags was no greater than that on the outside, indicating the ethylene glycol inhibited binding by both intercalation and side stacking. Conclusion: Since the tilting o] the bases as seen in the A ]orm o] nucleic acids does not inhibit binding o] quinacrine, it is unlikely ihat the minor degree o] tilting seen in the C-lotto o] D N A would be responsible/or the inhibition o/quinacrine binding. It is more likely that conditions that cause a shift to the C-con]ormation, such as high salt and ethylene glycol and certain proteins, do so by dehydration of the DIYA (Chang et al., 1973) and this same dehydration excludes quinacrine ]rom binding either by intercalation or by side stacking. Thus, i/localized areas of the chromosome exist in a C-like con]ormation the ]actors which cause this con]ormation change also prevent quinacrine ]rom binding. iVIeehanisms of Chromosome Banding. V N N- / \N 127 I H H I H H H H H H H H H H \1 I \/ \! \! \/ \! C ~. N C C C C H ~.C ~ ~C2/~ "~C// ~C / ~N ,-j ~C / ~N/" H H H H H H H H H H H H H I Fig. 11 (see text) gllect o/Spermine Spermine is a polyamine widely distributed in animal tissues and microorganisms (Tabor and Tabor, 1964). I t has the structure shown in Fig. 11 (Liquori et al., 1967). A related molecule, spermidine, is missing the portion of the molecule to the left of the dotted line. The effect of spermine on cellular metabolism has been attributed to its ability to complex with nucleic acids (Oehoa and Weinstein, 1965). This binding results in the neutralization of the phosphate groups (Tabor, 1962; Felsenfeld, 1962), an increase in the DNA T m (Tabor, 1962; Nandel, 1962) and protection against shear (Kaiser et al., 1963). These effects can be explained by the manner in which spermine binds to DNA. X-ray diffraction studies suggest that the two nitrogen groups on one end of the molecule bind to adjacent phosphate groups on one nucleotide chain, the molecule then crosses the small groove where the remaining two nitrogens bind to adjacent phosphates on the other chain (Suwalsky ctal., 1969; Liquori et al., 1967). Because of this detailed information available on the binding of spermine to DNA, it was utilized to examine its effects on the binding of quinacrine to DNA. As shown in Fig. 12A, complexes of spermine with calf DNA at a molar ratio of 1:8 and 1:4 produced no significant changes in the circular dichroism of DNA, while a significant decrease in the positive Cotton effect was seen with ratios of 1 : 2 and 1 : 1. Similarly, ratios of 1 : 8 and 1 : 4 caused little inhibition of quinacrine fluorescence while ratios of 1 : 2 and 1 : 1 decreased quenching by about half (Fig. 12). The interrelationship of all three parameters; fluorescence, hypochromism, and eireular dichroism, are shown in Fig. 12C. The DNA induced hypochromism is moderately inhibited at a spermine to DNA ratio of 1:8 while this produces no effect on the cireular diehroism or fluorescence. At higher spermine to DNA ratios the change in the circular dichroism and fluorescence quenching is relatively greater than the change in hypochromism. Conclusion: Spermine, binding in the small groove o[ DNA, inhibits the binding o] quinacrine to DNA, inhibits the quenching o/ quinaerine, and induces con]ormation changes in the D1VA. 9 Chromosoma (Berl.), Bd. 50 D. E. Comings et al. 128 A +3 f " ONA alone ,__,, 9S--~ "~'t ,~..~o +2 1:8 DNA ~,..Spermlne +1 0 z~e \-\. -1 -2 -3 Coil DNA +$permine -4 2~o 2~o ~o ~6o ~o 2~o ~;o ~o ~Io 320 knm C B 7.0 ,i~--~-o "l z ~ - - z~- - A "~- S per mi n e o n e ~ 1.0 .9 .8 ~"~i~~iL',,~ ~ ~~789 ~.~ ..,F.~__]: I -"h ~.5 ~ ~ i .6 i .~ O.D. "x~mm~i .~.~. tJ .4 a:: .3 n ~ .2 9 ..,e-I 25 Ae ~'\F[uor eseence .]1 , i T i 0 10-6 10-5 ]0 -4 10-3 moles DNA nucleotides ,I, ,12 ~14 Spermine / DNA I I 1:8 0NA alone Fig. 12. (A) The effect of varying molar ratios of spermine, calf DNA on the circular diehroism spectra of DNA (75 ~g/ml). (B) The effect of varying molar ratios of spermine: calf DNA on quenching of quinaerine fluorescence by DNA. (C) Comparison of the effect of varying molar ratios of spermine: DNA on quinacrine hypochromism (O.D.), quinacrine fluorescence, and circular dichroism. To make the circular dichroism comparable to the others the z]e at 275 nm was multiplied by 0.25 and this was subtracted from 1.0 EHect of t g i s t o n e s To e x a m i n e the effect of histones on quinacrine fluorescence, whole histone, lysine-rich and ~rginine-rieh calf t h y m u s histones were complexed to calf D N A , at weight-to-weight ratios of historic to D N A of Mechanisms of Chromosome Banding. V 1.] ]2 E 1.0 ].rh 129 l.l l.O 9 .S I \~ .8 .75 \\o .7 .6 LL .5 ~ ~ .25 -51 ~_m__ I ~ l . ~ . _ ' .4 .3 .2 ' .1 0 -7 C.T, DNA Whole Histone (mixing) i ]0-6 i ]0-5 1 .o .3 ~.DNA Alone .2! 9l J 0 ~ 10-4 10-8 C,T. DNA Arginine- rich Hislone (mixing} 10-7 , ]0-6 one 10'-5 Moles D N A Nucleofides Fig. 13. The effect of whole histone, ]ysine-rich, and arginine-rich calf thymus histone on the ability of calf DNA to quench quinacrine fluorescence. Solutions of histone and DNA in 50 mM phosphate buffer, pH 7.0 were combined by mixing 0.25 to 1, by simple mixing of solutions in 50 ~M phosphate buffer, p i t 7.0. Except for a slight enhancement o/quenching at lower molarities of DNA, this resulted in essentially no change in the quenching of quinacrine by D N A (Fig. 13). This is not entirely surprising since the removal of histones by t r e a t m e n t with 0.2 N HC1 has no effect on Qbanding (Comings, 1971), or on G-banding (Comings and Avelino, 1974). Histones probably bind in the large groove of D N A (Olins and Olins, 1971; Simpson, 1970; Farber et al., 1971). Conclusion: Histones, which probably bind in the large groove o / D N A , have little effect on the ability o[ D N A to quench quinaerinc fluorescence. Effect o[ Polytysine Poly-l-lysine preparations of four different average molecular weights were eomplexed with calf thymus DNA b y simple mixing. The molar ratio (r) of monomer lysine to monomer nucleotide ranged from 0.25 to 2.0. The effect of these complexes on the ability of calf thymus D N A to quench quinacrine fluorescence is shown in Fig. 14. Polylysine in the absence of DNA, at concentrations comparable to the highest molar ratios used, had no effect or only a slight enhancing effect on quinacrine fluorescence. Monomer lysine at an r of 1.0 had no effect on inhibiting quenching. I n this series, polylysine of 7 200 molecular weight was most effective. There was, however, some variation from one set of experiments to another and all molecular weight classes showed some effect in either this or other sets of experiments. For example, poly-19* D. E. Comings et al. 130 1.2 Lysine or Polylysine Alone Polylysine 1.1 A~ A ~ - .~/ Alonel2 ' ' ].25 I 1.0 .9 o L~ if 69 m \: ~ D N A5 AIone ~r DNA-Lysine 1:1 ,3 .2 .1 C.T. DNA Poly-Z-Lys,ine 2,200 00-7 10-6 10-5 . 5 ~ ~ DNA Alone .2 I C7. DNA .1 p Poly-Z-Lysine m . w l . 7,200 ! m.wt. 0 I 10 -7 10-4 t, i i i lO-6 10-5 10-4 70 1.o.v.8~*~i~i'\~, .9 .7 .6 .6 .5 9el .3 .2 O 10- Alone .5 .4 39 f ~ i ~25.\,,A DNA A,oo~ "-"~-,~-':I .2 C:r. DNA Poly-,g-Lysine m. wt. 5,900 i i p 10-s 10-~ 10-4 Moles DNA Nucleotides C,T, D N A I1 0 Poly-Z- Lysine mIwL 22,700 I ]0 -7 10 6 I 10-5 ]O-J{ Moles DNA Nucleotides Fig. 14. The effect of complexing poly-l-lysine to calf DNA on the ability of DNA to quench quinacrine fluorescence. The molar ratios of polylysine (as monomer lysine) to DNA (as nucleotides) range from 0.25 to 2 lysine of molecular weight 2200 was quite effective in inhibiting the ability of poly(dG), poly(dC) to quench quinaerine fluorescence (Fig. 15A). A marked effect of polyJ-lysine 7200 molecular weight was also noted with Micrococcus lysodeikticus DNA (Fig. 15B), and poly-1arginine was also highly effective (Fig. 15C). To ascertain whether the enhancement of quinacrine fluorescence was also affected, the studies were repeated with poly(dA)-poly(dT). Here again the polylysine of M1 molecular weights, and especially 7 200, inhibited the ability of the DNA to enhance quinaerine fluorescence (not shown). X-ray diffraction studies suggest po]ylysine, polyarginine, and protamine probably lie in the small groove of DNA (Wilkins, 1956; Feughe]man et al., 1955; Suwalsky and Traub, 1972). Conclusion: Poly-l.lysine and poly-l-arginine, which can bind in the small groove o / D N A , can markedly inhibit the ability o / D N A to quench or enhance [luorescenee. Presumably a macromolecule binding in the small groove of DNA will inhibit intercalation o/quinacrine. ~ech~nisms of Chromosome Banding. V A 131 B 1.2 poly(dG). 0oty (d C) poly-~--lysine m.wt. 2 , 2 0 0 1.1 1.0 1,2 1.1 1.0 _ _ ~ I1-~.111 I]&4:iL-- ~1.,~II ~ n ' ~ - - 1-0 .9 ~'.8 .8 ~.7 .7 .6 .6 ~.5 .5 ~.-,,t--- DNA Alone A--A-q~-- Lysine (1,25) .4 .5 .2 " ~ o - ~ - - DNA c;ione .3 0 10-7 I 10-6 i 10-5 i 10-4 .2 M. lysod.DNA 9I Poly- ~, - Lysine m.wt. 7,200 f~ 1~)-7 Moles DNA Nucleotides 10-6 10-5 10-4 Moles DNA Nucleotides C 1.5 1.2 1.1 1.0 .9 g.8 Sonicated Complexes t.k W. .5 .4 .5 .2 .1 0 10-7 ~ DNA a.._,__Q "~--Olone C.T. DNA poly-2~- orginine m. wt.15,500 ~ 10-~ i 10- 5 q 10- 4 Moles DNA Nucleotides Fig. 15. (A) The effect of poly-l-lysine, 2200 molecular weight, on the ability of poly(dG).laoly(dC) to quench quinacrine fluorescence. (B) The effect of poly-1lysine, 7200 molecular weight, on the ability of Micrococcus lysodeikticus DNA to quench quinacrine fluorescence. (C) The effect of poly-l-arginine, 13500 molecular weight, on the ability of calf thymus DNA to quench quinacrine fluorescence E[[ect o[ I s o t a t e d C h ~ o m a t i n FT'actions Studies of the effect of salts, spermine0 histones and polylysine help to define the effect of various agents on quinacrine binding but since the ultimate objective is to explain Q-banding, examination of the behavior of chromatin subIractions is most relevant. Isolated and washed mouse liver nuclei were sonicated in 0.34 M sucrose, 0.01 M tris, pH 7.0, subfractionated by differential centrifugation into 400 g, heterochromatin, intermediate fraction, euehromatin pellet, and euchromatin supernatant fractions (see Methods) (Comings and Harris, unpubl.). The 400 g and D. E. Comings et al. 132 Buffer Alone 9,000-1 i/.I 9 9 o~_o 9 o_O~o"-/2f- - _ -~" 7000J em | 6,000 4 | u/o- u -- [] -- o ~u n'~-~'-~ 9 , - - e Eu Sup 8,000 -- ' ~ u~ __i~--+ n e ~- * ~~- - - +6 ,~ z~"//+ ~,~,--6~--6 ~ : / il ~nt % a ' / . a ~ - ~ o ~ r - ~ o / I &'~L%~ 5,0001 ~'~&~ .... . ~ " " E 0 Peli~, %'~" DN A 4'000 1 5000 , , j , , 10 20 50 40 50 60 70 80 90 100 , 120 , 140 160 , 180 200 /J-M Quinacrine Fig. 16. The extinction coefficient of various concentrations of quinaerine in the presence of buffer alone, DNA, and various chromatin fractions. Eu Sup euehromatin supernatant; Het heteroehromatin; Int intermediate fraction. The buffer was 5 mN tris, 1 mE EDTA, pH 7.0 heterochromatin fractions contain C-type heterochromatin, both free and nueleolar associated (Mattoecia and Comings, 1971). The euehromatin supernatant constitutes about 5 % of the total D N A and is rich in non-histones with a p r o t e i n / D N A ratio of 4.6. The other fractions have protein/DNA ratios of 2.0 for the euchromatin pellet to 2.3 for the 400 g pellet, and have about equal amounts of histone and non-histone protein (Comings and Harris, unpubl.). Since some fractions, especially those containing nucleoli and the euehromatin supernatant contained significant amounts of RNA, all chromatin fractions were first treated with RNase (see Methods). The plan was to examine the binding of quinacrine to these subfractions, by equilibrium dialysis. This was complicated b y the fact t h a t the extinction coefficient of quinacrine varied markedly in the presence of buffer, DNA or the various nuclear subfractions. To correct for this the extinction coefficient of 1 to 200 ~zM of quinacrine in buffer alone, 75 ~g/ml of calf DNA, or 75 ~g/ml of ehromatin D N A was determined (Fig. 16). I n buffer alone the extinction coefficient was 8000 at 1 ~xM, increased to 8 750 at 20 ~M and remained relatively constant to 200 FM. I n the presence of calf DNA, at concentrations of quinacrine increasing from 1 to 50 FM, all the quinacrine remained bound to D N A and the extinction coefficient remained at 5000. After 50 FM, or an r of 0.21, the intercalative mode of binding was saturated and additional dye did not show D N A induced hypochromism and the extinction coefficient progressively rose to 7400 at 200 ~xM. The extinction coefficient of the Mechanisms of Chromosome Banding. V A- 133 Equilibrium Dialysis -Quinocrine vs. RNase Treated Chromatin Fractions .3- C.T. DNA ~ ' ~ / / _ r .2- 9 O /e / " ~ 0 1'0 2~0 ~ 3% ~ 40 5'0 60 p.M Free Dye 7'0 8'0 90 100 Fig. 17. Results of equilibrium dialysis, quinaerine against DNA and RNase treated ehromatin fractions plotted as r = ([Q]/[DNA]) versus FM quinacrine chromatin fractions indicated whether quinacrine was being prevented from binding to the D N A or not. Thus the euchromatin supernatant with a high non-histone to D N A ratio, showed the least hypochromism and thus the least intercalative binding of DNA. The heteroehromatin fraction also showed marked inhibition of binding. The euchromatin pellet was most like D N A and showed marked hypochromism at all concentrations of quinacrine. With these extinction coefficients it was possible to perform accurate equilibrium dialysis. I n Fig. 17 the results are plotted on the basis of r (moles dye/moles DXA) v e r s u s FM of free quinaerine. With quinacrine concentrations of I to 40 FM binding was markedly inhibited in chromatin compared to D N A and the inhibition was greatest for the euchromatin supernatant and the 400 g pellet. This is illustrated more dramatically b y the use of Scatchard plots (Fig. 18). With calf thymus D N A alone the slope was very steep and showed an intercept at an r value of 0.226 or one dye molecule bound per every two base pairs. There was a slight amount of side stacking at higher values but this abruptly stopped at an r of 0.28, probably due to precipitation of DNA. The euchromatin pellet and heterochromatin fractions showed significantly less binding with lower slopes, but the intercept at saturation was the same as with D N A alone. This was similar to the finding of Gottesfeld et al. (1974) with their supernatant and pellet fractions. By contrast, the 400 g fraction, which is richest in constitutive heterochromatin and satellite 134 D.E. Comings et al. 12111098- Equilibrium Dialysis Quinacrine vs. DNA and RNase T r e a t e ( e~L./C.T, DNA Chromatin 7r/c x 10 -3 6- / 54- Eu Pellet o __liB\ ',\" 52- 4oog.-q_ po \ iI \_ 1 Eo su0o - \. i ',A .1 t .150 , t .172 r .2 ~' .226 't .5 Fig. 18. Equilibrium dialysis of quinacrine versus DNA and various RNase treated chromatin fractions depicted by Scatchard plots, r [Q]/[DNA], c concentration of free quinacrine DNA, showed both inhibition of quinacrine binding and a decreased r at saturation of 0.172. This was even more striking with the euchromatin supernatant where the r at saturation was 0.130. These results correlated well with the effect of the subfractions on inducing quenching of quinacrLne fluorescence (Fig. 19). Here the euchromatin supernatant showed the least quenching, the 400 g fraction the next least quenching, the D N A the most quenching, and the other fractions were intermediate. Although we cannot necessarily relate the euchromatin supernatant fraction to "active chromatin" it is similar to active chromatin by virtue of its richer load of non-histone proteins (Frenster et al., 1963; Marushige and Bonnet, 1971; Comings, 1972b; Simpson and Recck, 1973). The fractions with the next largest amount of non-histone proteins were the 400 g pellet and heterochromatin, which by SDS gel electrophoresis possess a number of non-histone bands that are not present Mechanisms of Chromosome Banding. V ~,0 o D~D-- D /Eu 135 Sup .9- .8.7- .6Lt_ .5- Xx\ \ ~ 400 g , ~o\ ' \\~ \ \ \ Heterochromc~tin Fu Pehet / .4,3~- ,2- C,T. DNA~ J ~* .110-7 10-6 10-4 , ~ 10'_5 ; Moles DNA Nucleotides Fig. 19. t{elative fluorescence of quinacrine in the presence of varying concentrations of DNA and mouse chromatin subfractions in other fractions (Comings and Harris, unpub].). There thus seems to be a correlation between an increased amount of non-histones and a decreased interealative binding with quinaerine. When the fractions are treated with pronase (25 ~g/ml, 37~ for 1 hour) the differences shown in Fig. 19 disappear. This agrees with the finding of Gottesfeld et al. (1974). In Q-banding of mouse chromosomes the C-band heteroehromatin stains poorly (Rowley and Bodmer, 1971; Dev et al., 1971) and in spectrophotometric titration and equilibrium dialysis studies the 400 g fraction, richest in this type of heterochromatin, also binds quinacrine poorly. The interbands also stain poorly and these regions correlate well with early replicating, genetically active ehromatin (Gannet and Evans, 1971; Zakharov and Egolina, 1972; Comings and Okada, 1973). Our euchromatin supernatant, with some of the characteristics of active chromatin, also binds quinacrine poorly. Conclusion: Spectrophotometric titration and equilibrium dialysis experiments indicate that two mouse chromatin ]factions show a decreased number o] binding sites ]or quinacrine compared to D N A and the other chromatin ]factions. These are also the two ]factions which contain the most nonhistone proteins and may correlate with chromosomal regions that stain lightly during Q-banding. One valid question is, how can fractions which represent less than 10% of the chromatin be responsible for the presence of pale staining regions constituting 20 to 30% of the chromosomes ? The major answer is that the present procedures for fractionation of chromatin sonieation 136 D.E. Comings et al. and centrifugation do not give pure subfractions. There is contamination of the heterochromatin with euehromatin and vice versa. Our approach to this problem has been to separate the nuclear sonicate into a number of fractions and take the extremes at both ends as being most like heterochromatin and euchromatin. Since these could be varied at will according to the amount of sonication and speeds of centrifugation, the absolute amounts in the fractions does not necessarily bear any relationship to the amounts visible cytologically. The important observation is that subfractions do exist that show a decreased number of binding sites for quinacrine compared to whole chromatin and DNA. Comparison o[ Satellite DNA with GC-rich and AT-rich DNA The conclusion that base composition is very important in Qbanding, and that the centromeric heterochromatin in the mouse stains lightly because of the critical placement of guanine bases in the satellite DNA (Weisblum and de Haseth, 1973) can be simply tested by subfractionating the mouse DNA into GC-rich, mid-band, AT-rich, and satellite portions by CsC1 and Cs2SO4 ultracentrifngation (see Methods), and examining their effect on quinacrine fluorescence. When this was done there were no detectible differences in the ability of these DNA's to quench quinacrine fluorescence (Fig. 20). The respective buoyant densities of these fractions were: GC-rich - - 1.703, mid-band - - 1.701, AT-rich - - 1.697, and satellite - - 1.690 gm/cc. On this basis the base composition of the satellite DNA was 31%, the AT-rich DNA was 38%, and GC-rich DNA was 44% GC (Mandel et al., 1968). Although Selander and de la Chapelle (1973) obtained significant differences in the quenching of quinacrine mustard exposed to DNA's of varying base composition, this was based on an artificial mixture of bacterial DNA's where base sequence could be playing a significant role. When they fractionated DNA from the same organism the changes in degree of quenching with base composition were considerably less marked. The DNA isolated from the different chromatin subfractions also gave indistinguishable quenching curves (the same as those shown in Fig. 20) again indicating the differences in the binding of quinacrine to the fractions were due to chromosomal proteins. In agreement with these findings, Bostock and Christie (1974) have recently reported that there was little difference in the degree of quenching of quinaerine by satellite versus main band DNA of Mus musculus or Dipodomys ordii. Conclusion: Since there were insigni[icant di//erences between the ability o/mouse satellite and main band D N A /ractions to quench quinaerine /luorescenee, the decreazed staining o/ mouse C-band region containing satellite D N A must be due to protein-DNA interactions. Mechanisms'of Chromosome Banding. V 1 i Mouse O N A - Q u i n a c r i n e 1.0 - a I 3 - - ~ a - ~ , o 1 2.0 xlO-6M g"Oo 9 ~o .8 \ uj 9 ~.7 ~' hJ .5 .5 9 / * S a t e l l i l e - CsS04 o Main Bond-CsSO,~ .2 137 /o 9 G C - r i c h - CsCI o Mid Band-CsCI .1 9 A T - r i c h - CsCI / I 10 -7 10-6 MOLES I 10 . 5 10 -4 DNA NUCLEOTIDES Fig. 20. Effect of satellite and main band DNA separated by Cs~SO4-Ag+ccntrifugation, and GC-rich, mid-band, and AT-rich DNA separated by CsCI centrifugation on the fluorescence of 2.0 X l0 -6 M quinacrine EI[ect o[ ~uinacrine Concentration I t is important to point out t h a t discussing Q-banding in terms of the in vitro studies of the relative fluorescence of quinacrine + D N A compared to the fluorescence of quinacrine alone, can be misleading. This is because quinacrine has a high affinity for and is markedly concentrated b y the D ~ A , and its fluorescence markedly increases as the concentration of quinacrine increases. I n the equilibrium dialysis experiments in which the bags contained 75 Fg/ml of DNA, the quinacrine was concentrated 50 to 100 fold over the amount of quinacrine outside the bags. The higher concentrations of D N A in the chromosome would concentrate it even further. The relationship between fluorescence and the concentration of quinacrine, in the presence or absence of 1.3 mg/ml of calf DNA, is shown in Fig. 21A. The arrow indicates 2 • 10-6 M, the concentration of quinacrine used in the in vitro experiments. The fluorescence reading at this concentration, in the presence of DNA, was 1.6 relative units. At a concentration of 10-~ M quinacrine D. E. Comings et al. 138 A , ~ i B i 55 ~4o I __ D N A ........~ e / ' " I~ '~ 40 ~- , /o~ ~:35 ~ 3o o ilOL, fL, 25 ~ 2o 0 5 40 [QUfNACRJNE] in pM 010-7 10-6 10 -5 10 -4 10 .3 10 -2 [QUINACRINE] Fig. 21. (A) The relationship of varying concentrations of quinacrine in the presence (q-DNA) and absence (--DNA) of 1.3 mg/ml of calf DNA. (B) The steep portion of the q-DNA curve of A with both relative fluorescence and quinacrine concentration plotted on the same scale the fluorescence was 40 units, a 25 fold increase. The fluorescence then decreased due to concentration quenching (Guilbault, 1973) 1. I n Qbanding the unbound quinacrine is washed away in the rinses. Thus, the base line fluorescence of unbound quinaerine is negligible and all observable fluorescence is due to variable binding or variable fluorescent quantum efficiency of quinacrine. When the fluorescence and the quinaerine concentration for the steep part of the curve is plotted on the same scale (Fig. 21 B), their relationship is approximately linear with a slope of +1.3. Along the metaphase chromosome, where the linear variation in D N A concentration is minimal, differences in the amount of quinacrine binding due to protein-DNA interactions arc adequate to account for the approximately two-fold difference in intensity of Q-bands v e r s u s interbands. B y contrast, the difference in base composition causes only a negligible difference in fluorescence (Fig. 20). I n the interphase nucleus, where the concentration of DNA in heterochromatin is two-fold greater than in euchromatin (Lima-de-Faria, 1959; Berlowitz, 1965), the resulting different concentration of quinacrine could produce significant differences in fluorescent intensity. This is probably the explanation for the bright quinacrine fluorescence of the centromeric heterochromatin in the mouse interphase nucleus (Natarajan and Gropp, 1 The in vitro fluorescence was measured at right angles in i cm cuvettes. At high concentrations of quinacrine an inner-cell effect of adsorption of the exciting light (Guilbault, 1973) contributes to the quenching. This would be largely eliminated on the slide. Mechanisms of Chromosome Banding. V 139 1972) compared to its relatively dull fluorescence in metaphase chromosomes. Although Feulgen staining usually shows no or little variation in the concentration of DNA along the length of the chromosome (Comings et al., 1973; u and Sanchez, 1973), under certain circumstances some investigators have obtained evidence for enough variation to contribute, at least in part, to the banding patterns (Barr et al., 1973; Golomb and Barr, 1974; Rodman and Tahiliani, 1973; McKay, 1973; Heneen and Caspersson, 1973). Because of the ability of DNA to markedly concentrate quinacrine, a portion of the observed Q-banding could be due to minor variations in the DNA content along the chromosome. Conclusion: With the exception o/highly A T-rich satellites, di]/erences in the concentration o] quinacrine along the chromosome are capable o/ producing greater changes in fluorescence intensity than the variations in base composition o] the D N A o / a given organism. Interpretation The similarities and differences between these results and those of Gottesfeld et al. (1974) should be pointed out. Both studies agree there are differences in the ability of chromatin fractions to cause quenching of quinacrine fluorescence and these differences disappear when either the purified DNA is examined or when the proteins arc released by treatment with pronase. Both studies agree that differences in base composition are inadequate to explain Q-banding, except for brightly fluorescing regions containing highly AT-rich satellite DNA. Gottesfeld et al. (1974) found that in the presence of 44 ~M of quinacrine, their heterochromatin fraction caused the least hypochromism of quinacrine O.D. ; whole chromatin was intermediate, and euchromatin caused the most hypochromism (their Fig. 5). This agrees with our Fig. 17 where at 44 ~M of quinacrine the heterochromatin caused less hypochromism than the intermediate fraction, and the euchromatin pellet caused the greatest hypochromism. In addition, their heterochromatin fraction caused the least quenching of quinacrine, whole chromatin was intermediate, and euchromatin caused the most quenching (their Fig. 1). This agrees with our Fig. 19, where at DNA concentrations of 10-5 to 10-6 M the 400 g pellet and hcterochromatin fraction caused less quenching than the euchromatin pellet. Finally, they found that with saturating levels of quinacrine the heterochromatin, euchromatin and DNA all had the same number of quinacrine binding sites. This agrees with our Fig. 18 where the intercept (n) for DNA, heterochromatin and euchromatin pellet were all the same. They also observed differences in the CD spectra of euchromatin versus 140 D.E. Comings et al. heterochromatin, and interpreted their observations to indicate that quinacrine was evenly bound to the different ehromatin fractions but possessed different quantum yields and suggested that DNA in the C-like configuration resulted in a greater quantum yield than DNA in the B configuration. In our studies we observed that further fractionation of the chromatin into five, instead of two fractions, uncovered two fractions that did show a decreased number of binding sites for quinaerine--the 400 g pellet and the euehromatin supernatant. These fractions also showed the least hypochromism and the least quenching of quinacrine, and possessed the most non-histone proteins. A f t e r the chromatin fractions were treated with RNase, to remove associated RNA in the A configuration (Comings, unpubl.), there were no significant differences in the circular dichroism of the different chromatin fractions (except for a slightly more B-like conformation of the euehromatin supernatant). Studies of the effect of salts, ethylene glycol, and spermine indicated that factors which change the hydration of DNA both induce a change in the conformation of the DNA and inhibit quinacrine binding, and the conformation of the DNA p e r 8e is probably not involved. In the i n vitro studies the decreased ability of chromatin to quench quinacrine fluorescence (Fig. 19) was due to decreased binding of quinacrine to DNA (Fig. 18) rather than to decreased quantum efficiency of bound quinatrine. We interpret our results to indicate that negative or pale staining in Q-banding is due to the presence of non-histone proteins some of which bind or cover the small groove of DNA and inhibit the intercalative binding of quinacrine. This is still hypothetical since at present there is little information on the mode of binding of non-histones to DNA. / CH 3 / CH2CH 3 In quinacrine, the large NHCH--CH2CH2N--CH2CH 3 group on the 9 position of acridine probably lies in the small groove of DNA and thus binding of quinacrine to DNA is sensitive to the presence of non-histone proteins in this groove. This is consistent with the observations of Limon (1974, unpublished) that the ability to produce good banding is in part dependent upon the size of this group at the 9 position. Banding was absent when this group contained less than 3 CH 2 groups. In conclusion we feel there are three factors involved in quinaerine banding. 1. Base composition. This plays a major role in the very brightly fluorescing regions of the chromosome that contain extremely AT-rich satellite DNA (Ellison and Barr, 1972) and in the very poorly fluorescing regions of some chromosomes that contain very GC-rich satellite DNA (Holmquist, unpublished observations). 2. Non-histone protein-DNA interactions. These play the major role in producing the Mechanisms of Chromosome Banding. V 141 p o o r l y staining regions in t h e a r m s of t h e c h r o m o s o m e s a n d some C - b a n d h e t e r o e h r o m a t i n . 3. C h r o m a t i n density. T h e r e is a m o d e r a t e degree of c o n d e n s a t i o n of t h e e h r o m a t i n of Q- a n d G - b a n d s in t h e c h r o m o s o m e arms a n d t h i s could c o n t r i b u t e to t h e differential staining. H o w e v e r , since o t h e r acridine dyes (such as d e r i v a t i v e s of quinacrine) also r e s p o n d t o t h e D N A c o n c e n t r a t i o n b u t give p o o r banding, this is p r e s u m a b l y a m i n o r f~ctor. Acknowledgements. This work was supported by NIH Grant GM 15886 and the Norman V. Wechsler Research Fund. References Adler, A. J., Fasman, G. C. : Circular dichroism studies of lysine-rich histone F1deoxyribonucleic acid complexes. Effects of salts and dioxane upon confignration. J. phys. Chem. 75, 1516-1526 (1971) Allen, S. S., Grey, D.M., l~obert, G.P., Tinoco, I.: The ultraviolet circular dichroism of synthetic and natural DNAs and analysis of spectra for sequence information. Biopolymers 11, 853-879 (1972) Armstrong, R. W., Kurucsev, T., Strauss, U. P. : The interaction between acridine dyes and deoxyribonucleic acid. J. Amer. chem. Soc. 92, 3174-3181 (1970) Arnott, S. : The geometry of nucleic acids. Progr. Biophys. molec. Biol. 21, 265-319 (1970) Arnott, S., Fuller, W., Hodgson, A., Prutton, I." Molecular conformations and structure transitions of I~NA complementary helices and their possible biological significance. Nature (Lond.) 220, 561-564 (1968) Arnott, S., Hutchison, F., Spencer, M., Wilkins, M. H. F., Fuller, W., Langridge, R. : X-ray diffraction studies of double helical ribonucleic acid, Nature (Lond.) 211, 227-232 (1966) Barr, G.F., Mikel, U., Engler, W. F. : Correlates of chromosomal banding at the level of ultrastructure. In: Chromosome identification--technique and applications in biology and medicine (T. C~spersson and L. Zech, eds.), p, 280-289. New York, N. Y.: Academic Press, Inc, 1973 Berlowitz, L. : Analysis of histone in situ in developmentally inactivated chroma~in. Proc. nat. Acad. Sei. (Wash.) 54, 476480 (1965) Blake, A., Peacocke, A. lZ. : The interaction of aminoacridines with nucleic acids. Biopolymers 6, 1225-1253 (1968) Bobrow, M., Madan, K.: The effects of various banding procedures on human chromosomes, studies with acridine orange. Cytogenet. Cell Genet. 12, 145-156 (1973) Bostock, C. J., Christie, S. : Quinacrine fluorescence staining of chromosomes and its relationship to DNA base composition. Exp. Cell Res. 86, 157-161 (1974) Bostock, C. J., Prescott, D. M. : Buoyant density of DNA synthesized at different stages of the S phase of mouse L-cells. Exp. Cell Res. 64, 267-274 (1971) Bradley, D.F., Felsenfeld, G.: Aggregation of an acridine dye on native and denatured deoxyribonucleates. Nature (Lond.) 184, 1920-1922 (1959) Caspersson, T., Zech, L., Johaasson, C., Modest, E . J . : ~dentification of human chromosomes by DNA-binding fluorescent agents. Chromosoma (Berl.) 80, 215-227 (1970) Chang, C., Wciskopf, M., Li, H. J. : Conformation studies of nucleoprotein. Circular dichroism of deoxyribonucleic acid base p~irs bound by polylysine. Biochemistry 12, 3028-3032 (1973) 142 D.E. Comings et al. Cohen, G., Eisenberg, H. : Viscosity and sedimentation study of sonicated DNAproflavine complexes. Biopolymers 8, 45--55 (1969) Comings, D.E.: Heterochromatin of the Indian muntjac: Replication, condensation, DNA ultracentrifugation, fluorescent and heterochromatin staining. Exp. Cell Res. 67, 441-460 (1971) Comings, D.E.: Replicative heterogeneity of mammalian DNA. Exp. Cell Res. 71, 106-112 (1972a) Comings, D. E. : The structure and function of chromatin. Advanc. Human Genet. 3, 237-431 (1972b) Comings, D.E.: Biochemical mechanisms of chromosome banding and color banding with acridine-orange. In: Chromosome identification--techniques and applications in biology and medicine (T. Caspersson and L. Zech, eds.), p. 293-299. :New York, N. Y. : Academic Press, Inc. 1973 Comings, D. E., Avelino, E. : Mechanisms of chromosome banding. II. D istones are not involved. Exp. Cell Res. 86, 202-206 (1974) Comings, D. E., Avelino, E., Okada, T. A., Wyandt, H. E. : The mechanism of Cand G-banding of chromosomes. Exp. Cell Res. 77, 469493 (1973) Comings, D.E., Okada, T. A. : DNA replication and the nuclear membrane. J. molee. Biol. 75, 609-618 (1973) Corneo, G., Ginelli, E., Polli, E. : Repeated sequences in human DNA. J. molec. Biol. 48, 319-327 (1970) Crick, F. H. C., Watson, J. D. : The complementary structure of deoxyribonucleic acid. Proc. roy. Soc. Lond. A 223, 80-96 (1954) Crothers, D.M., Zimm, B.H.: Theory of the melting transition of synthetic polynucleotides: Evaluation of the stacking free energy. J. molec. Biol. 9, 1--9 (1964) Dev, V.G., Grewal, M.S., Miller, D.A., Kouri, R.E., Hutton, J . J . , Miller, O. J. : The quinacrine fluorescence karyotype of Mus musculus and demonstration of strain differences in secondary constrictions. Cytogenetics 10, 436-451 (1971) Dev, V. G., Warburten, E., Miller, O. J., Miller, D.A., Erlanger, B.F., Beiser, S. M. : Consistent pattern of binding of anti-adenosine antibodies to human metaphase chromosomes. Exp. Cell Res. 74, 288-293 (1972) Drummond, D.S., Pritchard, N.J., Simpson-Gildemeister, V. F. M., Peacocke, A. R. : Interaction of aminoacridines with deoxyribonucleic acid. Viscosity of the complexes. Biopolymers 4, 971-987 (1966) Ellison, J. t~., Barr, H. J. : Quinacrine fluorescence of specific chromosome regions: late replication and high AT content in Samoaia leonensis. Chromosoma (Berl.) 36, 375-390 (1972) Farber, J., Baserga, R., Gabbay, E. J. : The effect of a reporter molecule on chromatin template activity. Biochim. biophys. Res. Commun. 43, 675-681 (1971) Felsenfeld, G.: Interaction of nucleic acids with cationic substances. In: The molecular basis of neoplasia, p. 104-119. Austin, Texas: University of Texas Press 1962 Feughelman, M., Langridge, R., Seeds, W.E., Steckes, A. t~., Wilson, H.R., Hopper, C. W., Wilkins, M. H. F., Barclay, R. K., Hamilton, L. D. : Molecular structure of deoxyribonucleic acid and nucleoprotein. Nature (Lond.) 175, 834-838 (1955) :Fredericq, E., Houssier, C. : Study of the interaction of DNA and acridine orange by various optical methods. Biopolymers 11, 2281-2308 (1972) Mechanisms of Chromosome Banding. V 143 Frenster, J.H., Allfrey, V. G., Mirsky, A.E.: Repressed and active chromatin isolated from interphase lymphocytes. Proc. nat. Acad. Sci. (Wash.) 50, 1026-1032 (1963) Ganner, E., Evans, H. J. : The relationship between patterns of DNA replication and of quinacrine fluorescence in the human chromosome complement. Chromosoma (Berl.) 35, 326-341 (1971) Gersch, N. F., Jordan, D. 0. : Interaction of DNA with aminoacridines. J. molec. Biol. 13, 138-156 (1965) Golomb, H.M., Bahr, G, F. : Correlation of the fluorescent banding pattern and ultrastructure of a human chromosome. Exp. Cell Res. 84, 121-126 (1974) Gottesfeld, J.M., Bonner, J., Radda, G.K., Walker, I. O. : Biophysical studies on the mechanism of quinacrine staining of chromosomes. Biochemistry 13, 2937-2945 (1974) Green, G., Mahler, H. R.: Comparative study of polynucleotides in aqueous and glycol solutions. Biochemistry 9, 368-387 (1970) Guilbault, G. G.: Practical fluorescence, theory, methods and techniques, p. 26. New York, N. Y. : Marcel Dekker, Inc. 1973 I-Ianlon, S., Johnson, R. S., Wolf, B., Chan, A. : Mixed conformations of deoxyribonucleic acid in chromatin: a preliminary report. Proc. nat. Aead. Sci. (Wash.) 69, 3262-3267 (1972) Heneen, W.K., Caspersson, T.: Identification of the chromosomes of rye by distribution patterns of DNA. Hereditas (Lund) 74, 259-272 (1973) Holcomb, D.N., Tinoco, I.: Conformation of polyriboadenylic acid: pH and temperature dependence. Biopolymers 3, 121-133 (1965) Ichimura, S., Zama, M., Fujita, H., Ito, T. : The nature of strong binding between acridine orange and deoxyribonucleic acid as revealed by equilibrium dialysis and thermal denaturation. Bioehim. biophys. Acta (Amst.) 19O, 116-125 (1969) Jones, K. W., Corneo, G. :Lccation of satellite and homogeneous DNA sequences on human chromosomes. Nature (Lond.) New Biol. 233, 268-271 (1971) Kaiser, D., Tabor, H., Tabor, C.W.: Spermine projection of coliphage A DNA against breakage by hydrodynamic shear. J. molec. Biol. 6, 141-147 (1963) Kleinwachter, V., Balcarova, Z., Bohaeek, J.: Thermal stability of complexes of diaminoaeridines with deoxyribonucleic acids of varying base content. Biochim. biophys. Acta (Amst.) 174, 188-201 (1969) Kleinwachter, V., Koudelka, J. : Thermal denaturation of deoxyribonucleic acidacridine orange complexes. Biochim. biophys. Aeta (Amst.) 91, 539-540 (1964) Kurnick, N. B., Radcliffe, I. E. : Reaction between DNA and quinacrine and other antimalarials. J. Lab. clin. Med. 60, 669-688 (1962) LePecq, J.-B., Paoletti, C. : A fluorescent complex between ethidium bromide and nucleic acids. J. molec. Biol. 27, 87-106 (1967) Lerman, L. S. : Structural considerations in the interaction of DNA and acridines. J. molec. Biol. ], 18-30 (1961) Lerman, L. S. : The structure of the DNA-aeridine complex. Proe. nat. Acad. Sci. (Wash.) 49, 94-102 (1963) Lieberman, M.W.: Fractionation of mouse DNA in preparative Ag+-Cs2SOa gradients. Biochim. biophys. Acta (Amst.) ]24, 309-319 (i973) Lima-de-Faria, A. : Differential uptake of tritiated thymidine into heterochromatin and euehromatin in Melanopus and Secale. J. biophys, biochem. Cytol. 6, 457-466 (1959) Liquori, A.M., Constantino, L., Crescenzi, V., Elia, V., Giglio, E., Puliti, R., de Santis Savio, M., Vitagliano, V.: Complexes between DNA and polyamines: a molecular model. J. molec. Biol. 24, 113-122 (1967) 10 Chromosoma(Berl.). Bd. 50 144 D.E. Comings et al. Lloyd, P. H., Prutton, R. N., Peacocke, A. R. : Sedimentation studies on the interaction of proflavine with deoxyribonucleic acid. Biochem. J. 107, 353-359 (1968) Mandel, M.: The interaction of spermine and native deoxyribonucleic acid. J. melee. Biol. 5, 435-441 (1962) Mandel, M., Schildkraut, C. L., Marmur, J. : Use of density gradient analysis for determining the guanine plus cytosine content for DNA. In: Methods in enzymology, vol. X I I B, p. 184-195 (L. Grossman and K. Moldave, eds.). New York, N. Y. : Academic Press, Inc. 1968 Marmur, J. : A procedure for the isolation of deoxyribonucleic acid from microorganisms. J. molec. Biol. 3, 208-218 (1961) Marushige, K., Bonner, J.: Fractionation of liver chromatin. Prec. nat. Acad. Sci. (Wash.) 68, 2941-2944 (1971) Marvin, D.A., Spencer, M., Wilkins, M. H. F., Hamilton, L.D.: The molecular configuration of deoxyribonucleic acid. III. X-ray diffraction study of the C form of the lithium salt. J. melee. Biol. 3, 547-565 (1961) Mattoccia, E., Comings, D. E. : Buoyant density and satellite composition of DNA of mouse heterochromatin. Nature (Lend.) New Biol. 229, 175-176 (I971) McKay, R. D. G. : The mechanism of G and C banding in mammalian metaphase chromosomes. Chromosoma (Berl.) 44, 1-14 (1973) Michelson, A. M., Monny, C., Kovoor, A. : Action of quinacrine mustard on polynucleotides. Biochimie 54, 1129-1136 (1972) Mihnan, G., Chamberlain, M., Langridge, R. : The structure of a DNA-I~NA hybrid. Prec. nat. Aead. Sci. (Wash.) 57, 1804-1810 (1967) Natarajan, A. T., Groop, A. : A fluorescence study of heterochromatin and nucleolar organization in the laboratory and tobacco mouse. Exp. Cell Res. 74, 245-250 (1972) Nelson, R. G., Johnson, W. C. : Conformation of DNA in ethylene glycol. Biochem. biophys. Res. Commun. 41, 211-216 (1970) Ochoa, M., Jr., Weinstein, I. B. : Spermine inhibition of polypeptide synthesis in a subcellular system derived from the L 1210 mouse ascites leukemia. Biochim. biophys. Acta (Amst.) 95, 176-179 (1965) Olins, D. E., Olins, A. L. : Model nucleohistones: The interaction of F1 and F2al histones and native T7 DNA. J. melee. Biol. 57, 437-455 (1971) Pachmann, U., Rigler, R. : Quantum yield of acridines interacting with DNA of defined base sequence. Exp. Cell Res. 72, 602-608 (1972) Peacocke, A. R. : The interaction of acridines with nucleic acids. In: Acridines, (R., Aeheson, ed.), 2nd ed., p. 723-757. New York, N. Y. : Interscience Publishers, John Wiley & Sons 1973 Peacocke, A.R., Skerrett, J . N . H . : The interaction of aminoaeridines with nucleic acids. Faraday See. (Lend.) Trans. 52, 261-279 (1956) Rodman, T.C., Tahiliani, S.: The Feulgen banded karyotype of the mouse: Analysis of the mechanisms of banding. Chromosoma (Berl.) 42, 37-56 (1973) Rowley, J.D., Bodmer, W. F. : Relationship of centromeric heterochromatin to fluorescent banding patterns of metaphase chromosomes in the mouse. Nature (Lend.) 231, 503-506 (1971) Scatchard, G.: The attractions of proteins for small molecules and ions. Ann. N. Y. Acad. Sci. 51, 660-672 (1949) Schreck, R.R., Warburton, D., Miller, O.J., Beiser, S.M., Erlanger, B . F . : Chromosome structure as revealed by a combined and immunochemical procedure. Prec. nat. Acad. Sci. (Wash.) 70, 804-807 (1973) Selander, R.-K. : Interaction of quinacrine mustard with mononucleotides and polynucleotides. Biochem. J. 131, 749-755 {1973) Selander, R.-K., de la Chapelle, A. : The fluorescence of quinaerine mustard with nucleic acids. Nature (Lend.) New Biol. ~.4~,240-243 (1973) Mechanisms of Chromosome Banding. V 145 Simpson, R.T.: Interaction of a reporter molecule with chromatin. Evidence suggesting that the proteins of chromatin do not occupy the minor groove of deoxyribonucleic acid. Biochemistry 9, 48144819 (1970) Simpson, R.T., Reeck, G. l~.: A comparison of the proteins of condensed and extended chromatin fractions of rabbit liver and calf thymus. Biochemistry 12, 3853-3858 (1973) Southern, D. 5[.: Base sequence and evolution of guinea-pig s-satellite DNA. :Nature (Lond.) 227, 794-798 (1970) Studdcrt, D. S., Patroni, M., Davis, R. C. : Circular dichroism of DNA: Temperature and salt dependence. Biopolymers 11, 761-779 (1972) Suwalsky, M., Traub, W. : A comparative X-ray study of a nucleoprotamine and DNA complexes with polylysinc and polyarginine. Biopolymers 11, 2223-2231 (1972) Suwalsky, M, Traub, W., Subirana, J. A.: An X-ray study of the interaction of DNA with spermine. J. molec. Biol. 42, 363-373 (1969) Tabor, H., Tabor, C. W. : Spermidine, spermine and related amines. Pharmacol. Rec. 16, 245-300 (1964) Tinoco, I.: Hypochromism in polynucleotides. J. Amer. chem. Soc. 82, 47854790 (1960) Tobia, A., Schildkraut,C. L., Maio, J. J. : DNA replication in synchronized cultured mammalian cells. I. Time of synthesis of molecules of different average guanine cytosine content. J. molec. Biol. ~4, 499-515 (1971) Tunis-Schneider, ~ . J. B., Maestre, M. J. : Circular dichroism spectra of oriented and unoriented deoxyribonucleic acid films--A preliminary study. J. molec. Biol. ~2, 521-541 (1970) Van Holde, K. E.: Physical Biochemistry, p. 168-170. Englewood Cliffs, N. J. : Prentice-Hall, Inc. 1971 Weisblum, B., deHaseth, P. L. : Quinacrine, a chromosome stain specific for deoxyadenylate-deoxythymidylate-richregions in DNA. Proc. nat. Acad. Sei. (Wash.) 69, 629-632 (1972) Weisblum, B., deHaseth, P. L. : Nucleotide specificity of the quinacrine staining reaction for chromosomes. Chromosomes today 4, 35-51 (1973) Wilkins, M. H. F. : Physical studies on the molecular structure of deoxyribonucleic acid. Cold Spr. Harb. Symp. quant. Biol. 21, 75-90 (1956) Yang, J.T., Samejima, T.: Optical rotary dispersion and circular dichroism of nucleic acids. Progr. Nucleic Acid Res. and molec. Biol. 9, 223-300 (1969) Yunis, J . J . , Sanchez, O.: G-banding and chromosome structure. Chromosoma (Berl.) 44, 15-23 (1973) Zakharov, A.F., Egolina, N.A.: Differential spiralization along mammalian chromosomes. I. BUd]~ revealed differentiation in Chinese hamster chromosomes. Chromosoma (Berl,) 88, 341-365 (1972) Zimm, B. H., Crothers, D. M. : Simplified rotating cylinder viscometer for DNA. Proc. nat. Acad. Sci. (Wash.) 48, 905-911 (1962) Received July 8-October 15, 1974 / Accepted by J. G. Gall Ready for press November 6, 1974 Dr. David E. Comings Department of Medical Genetics City of Hope National Medical Center Duarte, California 91010 U.S.A. i0"