J. Cell. Sci. 70, 177-188 (1984) \JJPrinted in Great Britain © The Company of Biologists Limited 1984

DISTRIBUTION OF PROTEIN SULPHYDRYLS AND

DISULPHIDES IN FIXED MAMMALIAN

CHROMOSOMES, AND THEIR RELATIONSHIP TO

BANDING

A. T. SUMNERChromosome Structure Group, MRC Clinical and Population Cytogenetics Unit,Western General Hospital, Edinburgh EH4 2XU, U.K.

SUMMARY

Fixed mammalian mitotic chromosomes, when stained with a variety of sensitive, sulphydryl-specific fluorochromes directly or following reduction of disulphide groups, show uniformfluorescence. Thus sulphydryl and disulphide groups are uniformly distributed along the length ofthe chromosomes, and do not show patterns related to chromosome bands. Performance of G- orC-banding procedures oxidizes sulphydryls to disulphides, but does not produce an inhomogeneousdistribution of these groups. Cross-linking sulphydryl groups has no effect on G-, C- or Q-bandingof chromosomes. Thus there appears to be no connection between chromosome bands and thedistribution or state of oxidation of sulphur in chromosomal proteins.

INTRODUCTION

Ten years ago I suggested, on the basis of indirect evidence, that positive G-bandsin chromosomes might be relatively rich in protein disulphide (SS) cross-links, whilenegative G-bands might have a greater proportion of their protein sulphur in the formof sulphydryl (SH) groups (Sumner, 1974). At that time it was not possible to test thishypothesis by direct cytochemical observations. Nevertheless, the hypothesis wasconsistent with the idea that positive G-bands represent regions of chromosomes thatcondense earlier in mitotic prophase, and with evidence that disulphide cross-linksoccur at a higher concentration in condensed chromatin. For a brief review of theseconcepts, see Sumner (1983).

In recent years several very sensitive fluorochromes specific for sulphydryl groupshave become commercially available, so the opportunity was taken to examinecytochemically the distribution of SH and SS groups on mammalian chromosomes.In addition, the effect of routine chromosome banding procedures on these groupswas studied. Because the results obtained did not support the original hypothesis thatpositive G-bands are rich in SS groups, some of the original experiments wererepeated using different reagents. Since one of the important pieces of evidence forthe hypothesis was that cross-linking of chromosomal SH groups prevented G-banding, different reagents that produce cross-linking of SH groups were tested toassess their effect on chromosome banding. A preliminary account of some of theexperiments described in this paper has already been published (Sumner, 1983).

178 A. T. Sumner

MATERIALS AND METHODS

Chromosome preparations

Cells from cultures of either mouse A9 cells or human lymphocytes stimulated with phytohaemag-glutinin were treated with hypotonic (0-075 M) potassium chloride, fixed with methanol/acetic acid(3:1, v/v) and chromosome preparations made by spreading on glass slides according to conven-tional techniques.

Fluorescent staining of SH and SS

The fluorescent sulphydryl reagents used, their sources, the solutions in which they were used,and their reaction mechanisms are listed in Table 1. Chromosome preparations were stained for 1 hwith these reagents, followed by 1 h in the appropriate solvent to wash out non-specifically boundfluorochrome. The only exceptions were DMS andDMSDS (see Table 1), which were used for 24 hsince the shorter staining time produced negligible fluorescence, and monobromobimane (MBB),which was used for 45 min.

iV-ethyl maleimide (NEM; 0-03g/50ml) was used to block staining of SH groups (Sippel,1973), and tri-n-butyl phosphine (TBP; 0-25%) to reduce SS to SH (Sippel, 1978). Bothsubstances were dissolved in 2-propanol/phosphate buffer pH 6 ( 1 : 1 , v/v) and treatment was for1 h, followed by a wash in the same buffer for 5 min. Using these two reagents, total proteinsulphur (SH + SS) could be demonstrated by reducing chromosome preparations with TBP, while•SS alone could be demonstrated using NEM (to block SH groups), followed by TBP (to reduceSS to SH).

Cross-linking of protein sulphydryls

Two reagents were used to produce cross-linking of sulphydryl groups: sodium tetrathionate(Haest, Kamp, Plasa & Deuticke, 1977; Kaufmann, Coffey & Shaper, 1981) and diamide(Kosower, Kosower&Wertheim, 1969; Haeste/a/. 1977). Sodium tetrathionate was obtained fromFluka, and slides were treated in one or other of the following ways: 0-05% in Tris-HCl buffer(pH7-4) for 1, 3 o r24ha t4°C, or 0-05, 0-5 or 5% (saturated) in the same buffer for 24h at roomtemperature (approximately 22 °C), or 0-05 or 0-5% in 70% (v/v) ethanol for 24 h at room tem-perature. Diamide (diazine-dicarboxylic acid bis-(A', TV-dimethylamide)) was obtained from Sigma,and slides were treated with a 0-1 % solution in phosphate buffer (pH7-2) for 1 or 24 h at roomtemperature. Control slides were treated with buffer or ethanol as appropriate.

Two methods were used to assay the efficacy of cross-linking with these two reagents. The firstassay involved staining chromosome preparations with any convenient SH-specific fluorochromeafter treatment with the cross-linking agent, and also after subsequent reduction with TBP. In-hibition of specific fluorescence of SH groups, reversible by TBP reduction, was taken to indicatethe probable formation of SS cross-links. The second, 'cytological' assay measured differences in drymass and area of fixed diploid human lymphocyte nuclei that had been treated with the reagentsunder study, and then either treated with a 1 % solution of sodium dodecyl sulphate (SDS) for 1 h,or left untreated. Measurements of dry mass and area were made with a Vickers M86 IntegratingMicrodensitometer and Interferometer. For measurement of area, nuclei were heavily stained withGiemsa, but were, of course, left unstained for dry mass measurements. The occurrence of cross-linking was deduced from a reduction of swelling of nuclei and a smaller loss of dry mass in the SDSsolution compared to control slides.

Chromosome banding

Standard techniques were used for chromosome banding, as follows:G-banding: the ASG technique (Sumner, Evans & Buckland, 1971).C-banding: the BSG technique (Sumner, 1972).Q-banding: 0-4% quinacrine in acetate buffer (pH 4-5) for 30 min, followed by mounting in the

same buffer.

Tab

le 1

. Flu

ores

cent

sul

phyd

ryl

reag

ents

use

d - -

Com

poun

d

-- - -

Abb

revi

atio

n S

ourc

e S

olut

ion

Rea

ctio

n m

echa

nism

R

efer

ence

N-(

7-di

met

hyla

min

o-4-

m

ethy

1cou

mar

inyl

)-

mal

eim

ide

3-(4-Maleimidylpheny1)-7-

diet

hyla

min

o-4-

m

ethy

lcou

mar

in

Mon

obro

mob

iman

e

Am

mon

ium

4-c

hlor

o-7-

su

lpho

benz

ofur

azan

Did

ansy

lcys

tine

Dif

luor

esce

in d

isul

phid

e

E5I

E5M

F5M

DA

CM

CP

M

MB

B

SB

F-C

I

DD

C

DF

DS

DM

S

Mol

ecul

ar P

robe

s

Mol

ecul

ar P

robe

s

Mol

ecul

ar P

robe

s

Pol

ysci

ence

s

Mol

ecul

ar P

robe

s

Cal

bioc

hem

Pie

rce

Mol

ecul

ar P

robe

s

Pol

ysci

ence

s

Mol

ecul

ar P

robe

s

3.6

mg/

50 m

l P-

P*

(pH

9)

2 m

g/SO

ml

P-P

(pH

6)

2 m

g/50

ml

P-P

(pH

6)

2 m

g/50

ml

P-P

(pH

6)

2 m

g/50

ml

P-P

(pH

6)

0.02

g d

isso

lved

in 1

ml

acet

one

and

adde

d to

10

0 m

l pho

spha

te/s

alin

et

(pH

71

0.01

g/4

0 m

l bor

ate

(pH

8)

+ 1

m~

-Na

z

ED

TA

0.04

g/4

0 m

l pho

spha

te

(pH

7.2)

10

mg/

40 m

l ph

osph

ate

(pH

7.2)

3

mg/

30 m

l ac

eton

e

4- 4

'-Dim

alei

mid

yl-

DM

SD

S

Mol

ecul

ar P

robe

s 4

mg/

40 m

l P-

P (p

H6)

st

ilbe

ne-2

, 2'-d

isul

phon

ic

acid

* P-P

is

2-p

ropa

nol/

phos

phat

e bu

ffer

in

a 1:

1 m

txtu

re a

t th

e pH

sta

ted.

t M

ulle

r &

Cro

ther

s (1

975)

.

Alk

ylat

ion

Alk

ylat

ion

Cur

tis

& C

owde

n (1

980)

Alk

ylat

ion

Cur

tis

& C

owde

n (1

980)

Alk

ylat

ion

Cur

tis

& C

owde

n (1

980)

Alk

ylat

ion

Sip

pel

(198

1)

Alk

ylat

ion

Gai

ner

& K

osow

er (

1980

)

Alk

ylat

ion

And

rew

s, G

hosh

, Ter

nai

&

Whi

teho

use

(198

2)

SS

/SH

exc

hang

e L

azar

ides

& G

rang

er (

1978

)

SS

/SH

exc

hang

e W

inge

nder

& A

rell

ano

(198

2)

Alk

ylat

ion;

fl

uore

scen

t w

hen

cros

s-li

nkin

g S

H in

hy

drop

hobi

c si

tes

As

DM

S, b

ut c

ross

- li

nkin

g hy

drop

hili

c si

tes

180 A. T. Sumner

RESULTS

Distribution of SH and SS groups on chromosomes

All the fluorescent SH reagents tested gave essentially similar results (Fig. 1). BothSH groups (demonstrated by direct staining) and SS groups (shown by the NEM/TBP/staining sequence) were uniformly distributed on both human (including

Fig. 1. The distribution of SH and SS groups on fixed mouse A9 metaphasechromosomes. The fluorochrome used was DACM. A. SH groups only: the chromosomepreparation had no pretreatment before staining: B. SS groups only: the chromosomal SHgroups were blocked with NEM, and the SS groups reduced to SH with TBP beforestaining, c. SH + SS groups: SS groups were reduced with TBP before staining.

SH and SS distribution in chromosomes 181

prophase) and mouse chromosomes, and patterns corresponding to either G- or C-bands were not seen (with one exception mentioned below). Treatment with NEMbefore staining reduced the fluorescence of chromosomes to negligible levels, but didnot abolish fluorescence completely.

Some SH-specific fluorochromes were better than others, and some producedminor variations in staining pattern. The brightest fluorescence was given by F5M,DACM, CPM and MBB, with CPM perhaps the best of these. E5M produced onlymoderate fluorescence, and E5I gave fluorescence too weak for practical use. DDCgave moderate fluorescence, but characteristically the chromosomes appeared asbright outlines with dimmer interiors, which were, however, still uniform. A similarphenomenon was sometimes seen with DFDS, which produced brighterfluorescence, and occasionally revealed centromeric dots. Of the straightforward SHreagents, SBF-CI produced the least satisfactory but most intriguing results. In aninitial experiment, this fluorochrome produced a pattern apparently similar to G-banding, which, however, faded so rapidly that it could not be recorded photographic-ally. This observation could not be repeated. In subsequent experiments, a variety ofmountants was used in an attempt to increase the brightness and stability of thefluorescence; these are listed in Table 2. Only DPX permitted a level of fluorescenceto be produced that was usable, and the chromosomes were quite uniform.

The two fluorochromes that cross-link SH groups, DMS and DMSDS, alsoproduced uniform fluorescence on chromosomes. Both required 24 h staining to givea usable level of fluorescence. DMS, which is expected to cross-link SH groups inhydrophobic sites, gave a low level of fluorescence when slides were mounted in DPXand illuminated with blue light. Ultraviolet illumination, or mounting in phosphatebuffer at pH 7-2, resulted in negligible fluorescence. The absorption spectrum of theDMS staining solution has a narrow peak at 327 nm (Fig. 2A), which would not resultin efficient excitation of fluorescence in a microscope; this may account for the lowlevel of fluorescence found on stained chromosomes. DMSDS, which should cross-link SH groups in hydrophilic sites, gave a higher level of fluorescence than DMS,which was also improved by mounting in DPX and illuminating with blue light. Theabsorption of the DMSDS staining solution extends into the blue region of thespectrum when fresh (Fig. 2B), and this fact alone may account for the higher levelof fluorescence obtained with DMSDS than with DMS. Both DMS and DMSDSappeared to produce a moderate level of cross-linking in nuclei (Fig. 3).

Table 2. Mountants used for SBF-CI fluorescence

Mountant Reference

Phosphate buffer (pH7-2) —Phosphate buffer (pH7-2) containing 1% sodium dithionite Gill (1979)Citifluor/PBS Davidson & Goodwin (1983)Citifluor/glycerol Davidson & Goodwin (1983)Glycerol —Immersion oil (Leitz, fluorescence-free) —DPX(B.D.H.) —

CEL70

182 A. T. Sumner

20

1-8

1-6

1-4

• 1 2

u

S 10

o 0-80-6

0-4

0-2

00

1-8

1-6

1-4

8cCO

_ Q

< 0-6

0-4

0-2

00

275 300 325 350 400 450

Wavelength (nm)

500 550

B

275 300 325 350 400 450

Wavelength (nm)

500 550

Fig. 2. A. Absorption spectrum of DMS (005mg/ml) in acetone, B. Absorptionspectrum of DMSDS (0-05 mg/ml) in 2-propanol/phosphate buffer (pH 6). ( ) Freshsolution; (—) solution 24-h-old. For all spectra the light-path was 5 mm.

Effect of chromosome banding procedures on SH groups in chromosomes

Chromosome preparations that had been treated with G-banding or C-bandingprocedures (except for Giemsa staining) showed greatly reduced levels offluorescence with MBB or DACM. The remaining fluorescence tended to be con-centrated round the outlines of the chromosomes, but there was no sign of either G-or C-banding. Treatment of G- or C-banded chromosomes with TBP to reduce SSgroups to SH restored strong fluorescence; again this tended to be concentrated at theoutlines of the chromosomes, and occasionally centromeric dots could be seen. Nobanding pattern was visible after TBP reduction of banded chromosomes and stainingwith SH-specific fluorochromes.

SH and SS distribution in chromosomes 183

Dry mass Area

Y////////////A

V//////y7777,

Y//////////77777,

V///////////77;

V7S7///////////,

V////////////////A

Untreated

Propanol/phosphate

DMSDS

Acetone

DMS

Glutaraldehyde

1I

Y//X///////////////X1

jY////////////A

i'//AS/////////////////////////A

1

V//(/////////////////////A

i

10 10

Fig. 3. Results of the cytological cross-linking assay for DMSDS and DMS. The lengthsof the bars are proportional to the ratios for dry mass and area of nuclei treated with 1 %SDS to untreated nuclei, for each cross-linking treatment and the corresponding controls.Values nearest to 1 for both dry mass and area indicate least change and therefore thegreatest degree of cross-linking.

Effect of SH cross-linking on chromosome banding

Prior treatment of chromosome preparations with sodium tetrathionate or diamide,under any of the conditions used, had no effect on G-banding (Fig. 4), C-banding orQ-banding of chromosomes. Glutaraldehyde, a known powerful cross-linking agent,prevents G-banding of chromosomes (Fig. 4D).

Treatment of chromosomes with sodium tetrathionate or diamide greatly reducedthe level of fluorescent staining with MBB or DACM; bright fluorescence was res-tored by reduction with TBP after tetrathionate or diamide treatment.

Typical results of the cytological assay for cross-linking are illustrated in Fig. 5. Itmust be pointed out that the degree of loss of dry mass and that of swelling of nucleivaried considerably from one batch of slides of lymphocyte nuclei to another, depend-ing, no doubt, on differences in fixation and on the age of the slides. Nevertheless,comparison with control slides treated only with the buffers used to dissolve thesodium tetrathionate or the diamide showed that these two cross-linking reagentsreduced both the swelling of nuclei and the loss of dry mass. Both these losses werealso reduced in nuclei treated with buffer alone when compared with untreated nuclei.These effects were, however, relatively small compared to the effect of glutaral-dehyde, which largely prevents loss of dry mass and swelling of nuclei. Thus sodiumtetrathionate and diamide appear to introduce a relatively low level of SS cross-linksinto nuclei and chromosomes in preparations fixed in methanol/acetic acid.

DISCUSSION

Three conclusions can be drawn from the results described in this paper. Firstly,the distribution of SH and SS groups on fixed mammalian prophase and metaphase

Iff A. T. Sumner

•> f

4A I B 9 \

\

Fig. 4. Fixed human metaphase chromosome preparations, G-banded using the ASGtechnique, A. ASG with no pretreatment. B. 0-5% Sodium tetrathionate in Tris-HClbuffer (pH7-S), 24h, followed by ASG. c. Tris-HCl (pH7-5), 24h, followed by ASG(control for tetrathionate treatment), D. 2-5% Glutaraldehyde in phosphate buffer(pH7-2), 6h, followed by ASG.

SH and SS distribution in chromosomes 185

Dry mass

W////////

V//////////7,

v/////////////.V///////77777/

W/////////////////S

Untreated

Tris buffer

0-5% Tetrathionate

5% Tetrathionate

Glutaraldehyde

Area

'///{///////M

V/////////////////A

Dry mass

_ 1

V///////////////

V//////////7/&/,

Y///////////////////ZY,

Y/////////////////.

V/////////////////.

Untreated

Phosphate buffer

Diamide

Area

Y/AY////Y//////////AI

Untreated

Phosphate buffer

Diamide

Glutaraldehyde

V/J///A

'////////////////////////////////A

1

Fig. 5. Results of the cytological cross-linking assay for sodium tetrathionate and diamide.A. Sodium tetrathionate. B. Diamide; the upper and lower halves of the diagram representdifferent experiments. The area bar marked* indicates nuclei so extremely solubilized bySDS (84% of mass lost) that the area measurements on the remaining fragments aremeaningless. For further explanation see Materials and Methods and legend to Fig. 3.

chromosomes is uniform. Secondly, chromosome banding procedures oxidize SHgroups to SS cross-links, but do not cause any changes in the distribution of SH andSS, which remains uniform. Thirdly, cross-linking of SH groups produces relativelylittle stabilization of chromosome structure and has no obvious effect on the patternsproduced by standard chromosome banding procedures.

It was the major prediction of the hypothesis put forward by Sumner (1974) thatSH and SS groups would be non-uniformly distributed along chromosomes in apattern corresponding to G-banding. This idea was consistent with a number ofobservations showing that treatment of chromosomes, either in cell cultures or afterfixation, with reagents that attack SS or SH groups can have a marked effect on

186 A. T. Sumner

chromosome banding (Zakharovei al. 1973; Kato& Yosida, 1972; Utakoji, \9TZa,b\R0nne, 1979). It must nevertheless be pointed out that the precise mechanism ofaction of such reagents in affecting banding has not been established, and such ob-servations merely suggest that in some way SH and SS groups are involved in thoseaspects of chromosome structure that can be revealed by banding techniques. Asmentioned in the Introduction, the hypothesis is also consistent with the idea thatpositive G-bands are centres of chromosome condensation, and that condensedchromatin is richer in SS cross-links (Sumner, 1983). In spite of this variety ofindirect evidence, the observations described here on the uniformity of distributionof SH and SS groups on chromosomes make the original hypothesis untenable; thereis no relationship between banding patterns and SH and SS distribution. Thedistributions of these groups remain uniform both in the earlier (prophase) and later(metaphase) stages of chromosome condensation, and they remain uniform duringbanding procedures (although SH groups are oxidized to SS). Using reagents thatonly fluoresce when cross-linking SH groups (DMS and DMSDS), the distributionof cross-linkable sites, whether hydrophobic or hydrophilic, also appeared to beuniform. It should be noted that Buys and his colleagues illustrated uniformdistribution of SH groups along chromosomes, except for brighter fluorescence in thenucleolar organizing regions, although their studies were not specifically intended toshow this point (Buys & Osinga, 1980; Buys & Stienstra, 1980).

In the original experiments described by Sumner (1974), cross-linking of SHgroups was reported to prevent G-banding of chromosomes, the whole chromosomebeing darkly stained with Giemsa. This observation was one of the main pieces ofevidence for the hypothesis put forward in that paper, but was not consistent with auniform distribution of SH and SS groups on chromosomes. The experiments withsodium tetrathionate and diamide reported in the present paper show in fact thatcross-linking of SH has no effect on G-, C- or Q-banding of chromosomes. Further-more, the cross-linking assays carried out suggest that although most available SHgroups are effectively cross-linked, the number of cross-links introduced is relativelysmall and insufficient to stabilize chromosome structure to any great extent. Theseresults agree with the finding by Sumner (1974) that£-phenylene dimaleimide had noeffect on banding. It has been shown that extensive cross-linking, for example withglutaraldehyde, does prevent chromosome banding (Hancock & Sumner, 1982). Theeffect of mercurial reagents in preventing banding (Sumner, 1974) might thereforebe attributable to their more extensive, non-specific fixative action (Pearse, 1980;Horobin, 1982); it should also be noted that mercurial compounds can react withnucleic acids (Horobin & Flemming, 1982; Takeuchi & Maeda, 1976). Reactions ofmercurial compounds with chromosomes are therefore likely to be complex and notattributable solely to reaction with SH groups.

It must be concluded that there is no evidence from the results described in thispaper that the SH and SS groups of chromosomal proteins are in any way involvedin chromosome banding. Although cross-linking of SH to form disulphide cross-linksmay be involved in chromosome condensation, and different types of chromosomebands may be a reflection of condensation processes, the present work provides no

SH and SS distribution in chromosomes 187

evidence for a connection between the two. Similarly, there is no evidence on howreagents that are believed to react with SH and SS groups might produce bandingpatterns on chromosomes.

REFERENCES

ANDREWS, J. L., GHOSH, P., TERNAI, B. & WHITEHOUSE, M. W. (1982). Ammonium 4-chloro-7-sulphobenzofurazan: a new fluorigenic thiolspecific reagent. Archs Biochem. Biophys. 214,386-396.

BUYS, C. H. C. M. & OSINGA, J. (1980). Abundance of protein bound sulphydryl and disulphidegroups at chromosomal nucleolus organizing regions. Chromosoma 77, 1—11.

BUYS, C. H. C. M. & STIENSTRA, S. (1980). Involvement of sulphydryl groups of chromosomalproteins in sister chromatid differentiation. Chromosoma 77, 325-332.

CURTIS, S. K. & COWDEN, R. R. (1980). Demonstration of sulphydryl and disulphide groups bya fluorescent maleimide procedure. Histochemistry 68, 23-28.

DAVIDSON, R. S. & GOODWIN, D. (1983). An effective way of retarding the fading of fluorescenceof labelled material during microscopical examination. Proc. R. microsc. Soc. 18, 151.

GAINER, H. & KOSOWER, N. S. (1980). Histochemical demonstration of thiols and disulphides bythe fluorescent labelling agent, monobromobimane: an application to the hypothalamo-neurohypophysial system. Histochemistry 68, 309-315.

GILL, D. (1979). Inhibition of fading in fluorescence microscopy of fixed cells. Experientia 35,400-401.

HAEST, C. W. M., KAMP. D., PLASA, G. & DEUTICKE, B. (1977). Intra- and intermolecular cross-linking of membrane proteins in intact erythrocytes and ghosts by SH-oxidizing agents. Biochim.biophys. Acta 469, 226-230.

HANCOCK, J. M. & SUMNER, A. T. (1982). The role of proteins in the production of different typesof chromosome bands. Cytobios 35, 37-46.

HOROBIN, R. W. (1982). Histochemistry, pp. 19-55. Stuttgart: Gustav Fischer.HOROBIN, R. W. & FLEMMING, L. (1982). Is mercury orange a selective stain for thiols? Histochem.

y. 14, 1003-1006.KATO, H. & YOSIDA, T. H. (1972). Banding patterns of Chinese hamster chromosomes revealed

by new techniques. Chromosoma 36, 272-280.KAUFMANN, S. H., COFFEY, D. S. & SHAPER, J. H. (1981). Considerations in the isolation of rat

liver nuclear matrix, nuclear envelope, and pore complex lamina. Expl Cell Res. 132, 105-123.KOSOWER, N. S., KOSOWER, E. M. & WERTHEIM, B. (1969). Diamide, a new reagent for the

intracellular oxidation of glutathione to the disulphide. Biochem. biophys. Res. Commun. 37,593-596.

LAZARIDES, E. & GRANGER, B. L. (1978). Fluorescent localization of membrane sites inglycerinated chicken skeletal muscle fibres and the relationship of these sites to the proteincomposition of the Z disc. Proc. natn. Acad. Sci. U.SA. 75, 3683-3687.

MULLER, W. & CROTHERS, C. M. (1975). Interactions of heteroaromatic compounds with nucleicacids. I. The influence of heteroatoms and polarizability on the base specificity of intercalatingligands. Eur.jf. Biochem. 54, 267-277.

PEARSE, A. G. E. (1980). Histochemistry, Theoretical and Applied, 4th edn, vol. 1, pp. 113-114.Edinburgh: Churchill Livingstone.

R0NNE, M. (1979). The effect of in vitro dithioerythritol exposure on human metaphasechromosome structure. Hereditas 91, 105-109.

SIPPEL, T. O. (1973). The histochemistry of thiols and disulphides. I. The use of iV-(4-amino-phenyl)-maleimide for demonstrating thiol groups. Histochem. J. 5, 413-423.

SIPPEL, T. O. (1978). The histochemistry of thiols and disulphides. II. Methodology of differentialstaining. Histochem. J. 10, 585-595.

SIPPEL, T. O. (1981). New fluorochromes for thiols: maleimide and iodoacetamide derivatives ofa 3-phenylcoumarin fluorophore.J. Histochem. Cytochem. 29, 314-316.

SUMNER, A. T. (1972). A simple technique for demonstrating centromeric heterochromatin. ExplCell Res. 75, 304-306.

188 A. T. Sumner

SUMNER, A. T. (1974). Involvement of protein disulphides and sulphydryls in chromosome band-ing. Expl Cell Res. 83, 438-442.

SUMNER, A. T. (1983). The role of protein sulphydryls and disulphides in chromosome structureand condensation. In Kew Chromosome Conference II (ed. P. E. Brandham & M. D. Bennett),pp. 1-9. London: George Allen and Unwin.

SUMNER, A. T., EVANS, H. J. & BUCKLAND, R. A. (1971). New technique for distinguishingbetween human chromosomes. Nature, New Biol. 232, 31-32.

TAKEUCHI, S. & MAEDA, A. (1976). Interaction of a fluorescent reagent, fluorescein mercuricacetate, with nucleic acids. Biochim. biophys. Ada 454, 309-318.

UTAKOJI, T. (1973a). Differential staining of human chromosomes treated with potassium perman-ganate and its blocking by organic mercurials. In Chromosomes Today, vol. 4 (ed. J. Wahrman& K. R. Lewis), pp. 53-59. Jerusalem: Israel Universities Press.

UTAKOJI, T. (19736). Differential staining of hurnan chromosomes after oxidation treatment. InChromosome Identification - Technique and Applications in Biology and Medicine. Nobel Sym-posium 23 (ed. T. Casperson & L. Zech), pp. 346-349. New York: Academic Press.

WINGENDER, E. & ARELLANO, A. (1982). Synthesis and properties of the new thiol-specific reagentdifluorescein disulphide: its application on histone-histone and histone-DNA interactions.Analyt. Biochem. \TJ, 351-360.

ZAKHAROV, A. F., SELEZNEV, J. V., BENJUSCH, V. A., BARANOVSKAYA, L. I. & DEMINTSEVA,

V. S. (1971). Differentiation along human chromosomes in relation to chromosome identifica-tion. Excerpta Med. Int. Congr. Series, vol. 233, p. 193. Amsterdam: Excerpta Medica.

(Received 23 March 1984-Accepted 4 April 1984)