Proc. Nati. Acad. Sci. USAVol. 88, pp. 10916-10920, December 1991Genetics

Simultaneous visualization of chromosome bands and hybridizationsignal using colloidal-gold labeling in electron microscopy

(in situ hybridization for electron microscopy/chromosome banding for electron microscopy/double-labeling/biotinylatedprobe/gold-tagged chromosome bands)

RAOUF FETNI*t, RtGEN DROUIN*t, NICOLE LEMIEUXt, PAUL-EMIL MESSIER*, AND CLAUDE-LISE RICHER*t§*N6partements d'Anatomie et de tPathologie, Facultd de MWdecine, Universitd de Montreal, C.P. 6128, Succ. A, Montreal, PQ H3C 3J7, Canada

Communicated by Susumu Ohno, September 6, 1991

ABSTRACT Electron microscopy (EM) is seldom usedwith in situ hybridization to localize DNA sequences becausebanding methods for chromosome identification could not becoupled to EM techniques. We have applied an immunochem-ical replication-banding method specific for EM to solve thisproblem. A thymidine synchronization/BrdUrd release pro-tocol allows BrdUr4 Jcorporation only into late replicatingbands. A biot d DNA probe is hybridized in situ to itscomplementary sequence. The biotinylated probe and theBrdUrd-substituted DNA are simultaneously localized by dif-ferent reporter/detection systems using different-sized colloi-dal gold particles as electron-dense tags. We demonstrate thehigh precision of this mapping procedure by localizing on longprophase chromosomes (>1000 bands per haploid set) thepXBR-1 sequence to a small subregion of the centromericsubband Xpll.1-Xqll.1. This localization to a part of anindividual prophase subband is the most precise localizationever reported on human banded mitotic chromosomes.

Localization ofgene sequences to chromosomes using in situhybridization is usually done with light microscopy (LM).Attempts with electron microscopy (EM) are scarce andlimited to detecting repeated DNA sequences along un-banded chromosomes (1-5). The major limitation to greateruse of EM for DNA sequence mapping was the inability toapply banding techniques along with EM localization tech-niques. A few years ago, Li et al. (6) reported hybridizationof repeated oramplified genes. The radiolabeled probe wasin situ hybridized, and chromosomes were treated withtrypsin to produce a G-band pattern and photographed onglass slides before transfer onto EM grids. Faint G-bandstructures were observed in EM while simultaneously visu-alizing the hybridization sites. There has been no continua-tion to this method because autoradiographic studies in EMare cumbersome, the background is high, two screenings ofslides and grids are necessary, and the contrast between thebands is too weak to allow precise assignment of the hybrid-ization sites to the chromosome bands.Compared to autoradiographic techniques, nonisotopic in

situ hybridization procedures, as developed for LM, shownumerous advantages (7)-greater speed and precision,lower background, longer probe stability, and the availabilityof different reporter/detection systems to simultaneouslylocalize several probes (8-15). These LM techniques, byallowing one-step disclosure of the hybridization sites andchromosome bands, significantly improve the rapidity andprecision of mapping (16-19). Nonisotopic methods are alsosuitable for immunogold labeling and gold particles are thetag of choice for EM detection (20).

The need to localize numerous closely linked genes re-quires sequence mapping at higher resolution. MappingDNAfragments on elongated chromosomes is one way of ensuringa better resolution. This specifies the localization at thesubband level. Another way is to make use of the highresolving power available by EM. Therefore, the highestresolution can be achieved by mapping DNA sequences onlong prophase chromosomes by EM. Thus, the purpose ofthis study was to develop for EM a nonisotopic immuno-chemical procedure that gives, in one step, a specific hybrid-ization signal on banded chromosomes at any mitotic stagesfrom prophase to metaphase. We adapted our gold-labelingmethod, which produces chromosome banding specific forEM (21-23), to produce the immunochemical banding ofchromosomes on which a DNA sequence had just beenhybridized. The biotinylated probe and the bands weredisclosed with gold particles of two different sizes. Wedemonstrated the higher level of resolution afforded by thisprocedure by localizing DNA sequences to a subregion ofsmall prophase subbands with chromosomes showing >1000bands per haploid set.

MATERIALS AND METHODSLymphocyte Culture and Chromosomal Preparations. Hu-

man peripheral lymphocytes were grown for 54 hr andblocked in S phase by thymidine (300 jg/ml) for 17 hr. Theblocking agent was removed and the cells were grown foranother 4.5 hr in the presence ofBrdUrd (30 ,ug/ml) (24). Thisthymidine synchronization/BrdUrd release protocol allowsBrdUrd incorporation only into late replicating bands, the Gand C bands.The best mitoses were selected and located (22, 25). In situ

hybridization, probe detection, and chromosome bandingwere performed on glass slides and the selected mitoses weretransferred to the EM grids at the end of the protocol.

Probes. The DNA probe pXBR-1, chromosome X-specificDNA a satellite (26), and the probe for human ZFY (zincfinger protein, Y linked) gene [p06185, 1.3 kilobases (kb),from American Type Culture Collection] (27) were biotin-labeled by nick-translation.In Situ Hybridization. Slide preparations were dehydrated

in an ethanol series. To denature the double-stranded DNA,the slides were placed for 2 min at 70'C in a 70% solution offormamide in 2x SSC at pH 7. Following denaturation, theywere immediately immersed in 10% formalin for 2-3 min atroom temperature and then quickly dehydrated through 70%,80%, 90%, and 100% ethanol for 2 min each at -200C.Manipulations were resumed only after the alcohol hadevaporated.

Abbreviations: EM, electron microscopy; LM, light microscopy.tPresent address: Beckman Research Institute of the City of Hope,Division of Biology, Duarte, CA 91010-0269.§To whom reprint requests should be addressed.

10916

The publication costs of this article were defrayed in part by page chargepayment. This article must therefore be hereby marked "advertisement"in accordance with 18 U.S.C. §1734 solely to indicate this fact.

Dow

nloa

ded

by g

uest

on

Feb

ruar

y 1,

202

1

Proc. Natl. Acad. Sci. USA 88 (1991) 10917

The probe was denatured at 900C for 10 min and then quicklycooled on ice. Each slide received 20 ng of the biotinylatedprobe in 20 A.l of hybridization mixture and was covered witha plastic film (11). The slides were then placed in a humidifiedplastic box, at 370C, overnight. The slides were washed onceat 420C in 2x SSC containing 50% formamide, at pH 7, for 2min and twice in 2x SSC at 420C for 2 min each. From this stepon, the slides had to be kept humid.Probe Detection. After 2x SSC baths, slides were rinsed for

5 min in a modified isotonic phosphate-buffered saline (PBT-:PBS, pH 7.4/0.1% Tween 20/0.15% bovine serum albumin)and incubated with 50 A.l of rabbit anti-biotin antibody (EnzoBiochem) diluted 1:100 in PBT for 45 min at 370C in a humidchamber. Slides were subjected to two 5-min rinses in PBT.For gold labeling, two methods were used. In the strepta-

vidine method, slides were incubated with 50 A.l ofbiotinylatedgoat anti-rabbit IgG antibody (BRL) diluted 1:100 in PBT for45 min at 370C. Preparations were incubated with streptavidinecomplexed with gold particles [diameter, 30 nm; made in ourlaboratory according to Bendayan (20)] for 45 min at 37TC.Nonspecifically bound gold particles were removed by three5-min rinses in PBT at room temperature. In the protein Amethod, a protein A-gold complex [3-6 nm; prepared accord-ing to Bendayan (20)] diluted in PBS was added and thespecimens were incubated for 30 min at room temperature.The material was then washed three times in PBT.Chromosome Banding and Mitosis Transfer. The banding

technique, as described in detail elsewhere (22, 23), went asfollows. To detect the BrdUrd, a mouse monoclonal anti-BrdUrd antibody (Partec, Switzerland) diluted 1:10 waslayered over the slides, which were placed for 45 min at 37°C.After rinsing, the slides were then incubated for 45 min at37°C with a rabbit anti-mouse antiserum (DAKO, Carpinte-ria, CA), diluted 1:25, and 30 min at room temperature witha protein A-gold complex (6 or 15-17 nm). After three 5-minwashes, the material was fixed in 3.0% phosphate-bufferedglutaraldehyde for 30 min at room temperature (23).

Previously selected, well-spread mitoses were transferredfrom the glass slides to the EM grids (25) and examinedunstained in a Siemens Elmiskop 1A electron microscopeoperated at 60 kV.

RESULTS AND DISCUSSIONThe hybridization sites of the biotinylated probe were spe-cifically revealed with either small (6 nm, Fig. 1 A and B) orlarger (30 nm, Fig. 1C) gold particles. Gold particles werecoupled to either protein A or streptavidine. Fig. 1 shows thehybridization of pXBR-1, with an accumulation of goldparticles on the centromeric region of the X chromosomes.Proteinase K treatment does not improve hybridization orbanding nor does it affect the background. Although thebackground over chromosomes is always very low, thenumber of gold particles is higher outside the chromosomesthan on the chromosomes themselves (Fig. 1B), allowing foran unambiguous hybridization signal along the chromosomalaxis even with small amounts of particles at the hybrid site.It seems that free and complexed gold particles do not bindnonspecifically to chromatin, presumably due to an electro-static repulsion since both possess a net negative charge (4).At high magnification, it is very easy, even with the

smallest particles, to discriminate the gold particles at thehybridization sites or on the bands (Figs. 1 B, C and 2B). Theparticles do not aggregate. The fact that the particles arehomogeneous in size (Figs. 1 B and C and 2B) is an asset fordouble labeling. We repeatedly observed that more compactlabeling is related to the use of the smaller gold particles. Theless compact labeling obtained with larger particles possiblyresults from repulsion caused by strong negative charges onthe colloidal gold particle surfaces and in the chrornatin. Itshould be noted that there is no need for silver enhancementof the smaller particles since this procedure produces parti-cles of heterogeneous sizes and a coalescence of several goldparticles (28).

In our reporter/detection system (7), the reporter is eitherbiotin (DNA sequence labeling) or BrdUrd (chromosome

. ,^. w

*w__.'

..

0

I A B

IL I l J..I

v _|.

_sS

.s;E

.

t, ^

. ,.# ¢

'iiiiaiSt.4_D' n*.i_. )I _.,::_F:

>n

.. ^

C $

FIG. 1. Electron micrographs of human unbanded prometaphase X chromosome hybridized with biotinylated probe pXBR-1. Thehybridization signal is localized at the centromeric region. (A) Detection with protein A-gold complex, 6 nm. (Bar = 1.25 /,m; x7290.) (B) Highermagnification of the hybridization site seen in A. (Bar = 0.22 ,um; x40,860.) (C) Detection with streptavidine coupled with 30-nm colloidal goldparticles. (Bar = 1.25,um; x7290.)

Genetics: Fetni et al.

Dow

nloa

ded

by g

uest

on

Feb

ruar

y 1,

202

1

Proc. Natl. Acad. Sci. USA 88 (1991)

band labeling). We found it essential to complete the report-er/detection system for the biotin before initiating the re-porter/detection system for the BrdUrd because the BrdUrdreporter is in great excess over the biotin reporter. Hence, topermit a good access to the probe, the hybridization signal isrevealed and then, the chromosome bands.The BrdUrd incorporated in the late replicating bands is

revealed with the anti-BrdUrd antibody and gold particles ofa size different than that used for probe detection. The latereplicating G and C bands are revealed on the chromosomesalong with the hybridization signal (Figs. 2, 3 A and B, and 4).We found that using 6-nm gold particles for chromosomebanding produced bands that show a much more compactlabeling than that observed with the 15-nm or larger goldparticles (Figs. 2, 3A, and. 4B). At low magnification, thesmaller particles are not individualized and thus provide aclearer band pattern (Figs. 2A and 3A).A formaldehyde treatment was introduced to obtain the

highest quality of chromosome banding after the in situhybridization. Hot formamide was used to denature thechromosomal DNA for subsequent probe hybridization. Thistreatment is also necessary to optimally expose the BrdUrdepitope (29), which is inaccessible when BrdUrd-substitutedDNA is in the double-stranded configuration (30, 31). Anovernight incubation with the hybridization mixture follow-ing formamide denaturation is associated with a considerablerenaturation of in situ DNA and the banding quality obtainedis consequently decreased (Fig. 3C). Since the duration ofthehybridization cannot be shortened without affecting the ef-ficiency of the hybridization, formaldehyde was used toimpede the in situ DNA renaturation.

Immediately following formamide denaturation, treatmentwith formaldehyde, an amino group reagent, allows high-quality banding even after 18 hr of incubation with thehybridization mixture (Fig. 5, plot A). Reassociation of thetwo strands is prevented by the hydroxymethylation of theamino-groups within guanine and adenine in single-strandedDNA (34). On the other hand, without formaldehyde, thereis a progressive decrease in the banding quality with theincubation time (Fig. 5, plot B); the banding never completely

disappeared, indicating an incomplete in situ renaturation.Indeed, after in situ DNA denaturation, during hybridization,renaturation occurs in two phases (32). In the first phase,when denatured chromosomes and nuclei are subjected toDNA annealing conditions, about 35% of the single-strandedDNA renatures rapidly (half-life, 10-15 sec). In the secondphase, the remaining single-stranded DNA renatures veryslowly and then the BrdUrd becomes unaccessible to theantibody. We did not observe this biphasic reaction. Theslow reannealing phase is consistent with our observation ofa gradual decrease in banding quality over 18 hr whenformaldehyde was not used to prevent renaturation.EM not only allows precise localization of the DNA

fragments within a subband but may be used almost quanti-tatively to reveal the amount of complementary sequences inthe subband. The major site ofhybridization ofpXBR-1 is thecentromeric region of the X chromosome (subband Xpl1.1-Xqll.1) and the probe shows preferential central localizationon the centromeric subband, leaving peripheral portions ofthe subband free of hybridization signal (Fig. 2B). As Yanget al. (26) -reported, other chromosomes show more or lessconsistent but significantly lower levels of hybridization.However, the probe does not hybridize in the very center ofthe centromeric subband in any chromosomes. Indeed, Fig.3B clearly illustrates that for chromosomes 2 and 13, the weakhybridization signal is only found on the distal third of thesubband toward the long arm. The extremely high resolutionafforded by the procedure, which combines long prophasechromosomes (>1000 bands per haploid set) and EM, allowsthe localization of a DNA sequence to a portion of anindividual prophase subband. This resolution shows that thealphoid satellite DNA and related sequences are unevenlydistributed within the centromeric subband. Applying thesame procedure, we were able to clearly visualize, within asubband, single-copy DNA sequences. Fig. 4 shows the1.3-kb probe for ZFY (27) hybridized on the short arm of theY chromosome.The protocol used to incorporate BrdUrd also allows us to

determine the replication time of bands. Indeed, high thymi-dine concentration blocks S phase progression near the R

-o

S!.,

-. t. mw-

,... ;... '.1,;.t

. - -k

.

." -F Or. t-..-.t -. :%twc

I - .,iL :nl4..X

FIG. 2. Electron micrographs of human banded prometaphase X chromosome hybridized with biotinylated probe pXBR-1. The pXBR-1probe was detected with streptavidine coupled to 30-nm colloidal gold particles and is localized at the centromeric region, more precisely atthe subband Xpll.1-Xqll.1. Bands were revealed with anti-BrdUrd antibody and protein A-gold complex, 6 nm. The BrdUrd substitutionoccurred during late S phase, in the G and C bands. The typical band pattern, at a band level of about 800 bands per haploid set, makes thechromosome easy to identify. (A) Low magnification to identify the X chromosome. (Bar = 1.33 g&m; x6750.) (B) Higher magnification of thecentromeric region. The occurrence of large gold particles among the smaller ones indicates that the probe hybridized to a late replicatingsubband. Note that the hybridization signal is localized in the very center of the centromeric subband, whereas the flanking thirds are free oflarge gold particles. (Bar = 0.22 ,um; x40,860.)

10918 Genetics: Fetni et al.

.A

* - Ap,.II a 1.

:.. .. .,P." I -&., ", " .. li

*: a AI%.

i. I

M" I JIF,.0

A

Dow

nloa

ded

by g

uest

on

Feb

ruar

y 1,

202

1

Proc. Nati. Acad. Sci. USA 88 (1991) 10919

,d~~~~~~~~~A

m..^.~~~~~~~~ -."4 . je ' ~~ ..'a . :

CI-S

L- J

FIG. 3. Electron micrographs of chromosomes from partial human mitoses. These chromosomes incorporated BrdUrd during the late Sphase. (A) The typicalOG and C banding, at a band level of about 1250 bands per haploid set, makes chromosomes 2 and 13 easily recognizable.The bands were revealed with anti-BrdUrd antibody and protein A-gold complex, 6 nm. The biotinylated probe pXBR-1 was hybridized anddetected with streptavidine coupled with 30-nm colloidal gold particles. A very weak hybridization signal can be observed on the centromericsubband of each chromosome (subbands 2p11.1-2q11.1 and 13pll.11-13qll). The arrow and arrowhead. indicate the hybridization site ofchromosomes 2 and 13, respectively. (Bar = 1.82 Anm; x4950.) (B) Higher magnification of the centromeric region of chromosomes 2 (top) and13 (bottom) showing weak hybridization signals mainly localized on the distal third of the centromeric subband toward the long arm. The arrowand arrowhead indicate the hybridization site of chromosomes 2 and 13, respectively. (Bar = 0.33 Anm; X27,270.) (C) The bands were revealedafter 16 hr of incubation with the hybridization mixture (minus the probe) but without formaldehyde pretreatment. The decrease in bandingquality is ascribed to a partial DNA renaturation. (Bar = 3.85 ,um; x2340.)

band to G band replication transition in the middle of the Sphase, and the block release into a low concentration of

0 009 a

0'4 0 0It. 't0 0!. .2-MAda- *0

A

ray

e- Il

.:BZ'FIG. 4. Electron micrographs of a human G-banded (late BrdUrd

incorporation) Y chromosome hybridized with biotinylated ZFYgene probe (pO6185, 1.3 kb) (27). (A) High magnification to show thehybridized probe. The hybridization signal, indicated by the arrows,is visualized with small gold particles (3-4 nm), whereas the chro-mosome bands are revealed with larger gold particles (15-17 nm).The background associated with small particles is extremely low,making the hybridization signal very clear. Note the symmetry ofthetwo hybridization sites on the external part of each chromatid. (Bar= 0.16 ,um; x54,000.) (B) Lower magnification to show the easilyrecognizable Y chromosome. (Bar = 1.05 um; x8550.)

BrdUrd substitutes only the DNA replicated after the block(i.e., G and C bands) (24, 35). The occurrence of large goldparticles among the smaller ones indicates that the probehybridized to a late replicating subband (Figs. 2B and 3B).As an alternative to colloidal gold for the electron-dense

marker, the horseradish peroxidase with diaminobenzidine(DAB) staining is commonly used in EM to detect the probe(2) or reveal the bands (28, 36). However, we found that thislabeling shows less contrast, is not quantifiable, and is muchless precise because ofa halo effect (data not shown). Indeed,during the DAB-H202 reaction, the dye deposits becomeconfluent and it is very difficult to avoid exceeding thehybridization sites or the BrdUrd-substituted bands.Any sequence mapping implies accurate localization of

sites on specific chromosomes and therefore requires the useof chromosomal reference points. Localization can be rela-tive to adjacent sequences (16), relative to centromeres ortelomeres (14), or onto an idiogram. Adjacent sequencelocalization provides little information and is limited tofragments 0.5-1.0 megabases apart (16). Localization rela-tively to centromeres or telomeres requires the evaluation ofmany mitoses, and some discrepancy in localization betweenbanding versus measurement has been observed (16). Theinaccuracy of measurements occurs because of the differen-tial condensation of chromosome regions in correlation withmitotic stages (37). Localization to a particular band, whichcan be immediately related to an idiogram, is definitely themost reliable method, particularly when the bands and hy-bridization signal can be simultaneously visualized. Thebands allow very accurate probe localization, are indepen-dent of chromosome condensation, and provide much infor-mation. Indeed, the cytological characteristics of bands areassociated with important functional features such as repli-

e4*

Genetics: Fetni et al.

T.P%,.i.40...# .1

I AEi..

0

,p.., A.

16

A

.*eager""qv4 'N

4",

-z

*,J-"41r'.' '-

*.-j

vZ.".I4.

I.&-

tr

O'V*k-T#,

"-"0&It ..p

i AW

I

I.. .11

Dow

nloa

ded

by g

uest

on

Feb

ruar

y 1,

202

1

Proc. Natl. Acad. Sci. USA 88 (1991)

A

co-0 0 a0

o~~~

We sincerely acknowledge Dr. Gerald Holmquist for reviewing themanuscript, Dr. Moise Bendayan for his help in the preparation ofprotein A and streptavidine gold complexes, and Dr. AntonellaMaresca for the gift of probe pXBR-1. We greatly appreciate theassistance of Emilienne Lambert, Louise Laquerre, and ChristianeRondeau and the photographic work of Jean Lveill6. This work wassupported in part by the Fonds internes de recherche de l'Universitdde Montr6al et de la Facult6 de mddecine. R.F. holds a studentshipfrom the Ministtre de l'enseignement supdrieur d'Algdrie and R.D.holds a research fellowship from the Medical Research Council ofCanada.

1. Hutchison, N. J., Langer-Safer, P. R., Ward, D. C. & Hamkalo,B. A. (1982) J. Cell Biol. 95, 609-618.

2. Manuelidis, L., Langer-Safer, P. R. & Ward, D. C. (1982) J. Cell-B Cd Biol. 95, 619-625.

3. Radic, M. Z., Lundgren, K. & Hamkalo, B. A. (1987) Cell 50,0 1101-1108Nrynsai S HmaoB A 190 yomty 1 441202 4 6 8 10 12 14 16 18 hr 4. Hamkalo, B. A., Narayanswami, S.&Lundgren, K. (1989)Am.J.

INCUBATION TIME Anat. 185, 197-204.5. Narayanswami, S. & Hamkalo, B. A. (1990) Cytometry 11, 144-152.

5. Relationship between banding quality and incubation 6. Li, C.-B., Wu, M., Margitich, I. S. & Davidson, N. (1986) Chro-the hybridization mixture. Plot A, formamide and formalde- mosoma 93, 305-312.eatments; plot B, formamide treatment but no formaldehyde; 7. Lichter, P. & Ward, D. C. (1990) Nature (London) 345, 93-95.,tno formamide and no formaldehyde. (i) The slides were 8. Pinkel, D., Straume, T. & Gray, J. W. (1986) Proc. Natl. Acad. Sci.l with one of the above combinations of formamide and USA 83, 2934-2938.dehyde, dehydrated in ethanol at -200C, and air-dried. (ii) 9. Bhatt, B., Burns, J., Flannery, D. & McGee, J. 0. (1988) Nucleicvere incubated with 50 pAl of the hybridization mixture (minus Acids Res. 16, 3951-3961.be) at 370C in a moist chamber for different periods of time (0, 10. Lawrence, J. B., Vilinave, C. A. & Singer, R. H. (1988) Cell 52,6, 8, 10, 12, 14, 16, and 18 hr). (iii) After incubation, the 11. Viegas-Pdquignot, E., Dutrillaux, B., Magdelenat, H. & Coppey-

osome bands were revealed with the anti-BrdUrd antibody Moisan, M. (1989) Proc. NatI. Acad. Sci. USA 86, 582-586.ated with a fluorescent (LM) or colloidal gold (EM) reporter/ 12. Cherif, D., Bernard, 0. & Berger, R. (1989) Hum. Genet. 81,ion system (29). All specimens were studied in LM and several 358-362.;tudied in EM. The banding quality (0, no banding; 1+, poor 13. Moens, P. B. & Pearlman, R. E. (1989) Chromosoma 98, 287-294.g; 2+, fair banding; 3+, good banding; 4+, best banding) was 14. Lichter, P., Tang, C. C., Call, K., Hermanson, G., Evans, G. A.,ted by two independent observers as described and illustrated Housman, D. & Ward, D. C. (1990) Science 247, 64-69.iere (29). These experiments were repeated three times and the 15. Tkachuk, D. C., Westbrook, C. A., Andreeff, M., Donlon, T. A.,'ere summarized to determine the plots. Without formamide Cleary, M. L., Suryanarayan, K., Homge, M., Redner, A., Gray, J.ent, only poor to fair banding (1+-2+), with variability be- & Pinkel, D. (1990) Science 250, 559-562.mitoses, is obtained. This 1+-2+ banding decreases to 0-1+ 16. Lawrence, J. B., Singer, R. H. & McNeil, J. A. (1990) Science 249,ig due to a slow renaturation process (plot C). About 20% of 928-932.4A is single-stranded in methanol/acetic acid-fixed chromo- 17. Fan, Y.-S., Davis, L. M. & Shows, T. B. (1990) Proc. Natl. Acad.* this in situ denaturation is presumably caused by the fixative Sci. USA 87, 6223-6227.ent (32, 33). 18. Cherif, D., Julier, C., Delattre, O., Derr6, J., Lathrop, G. M. &

Berger, R. (1990) Proc. Natl. Acad. Sci. USA 87, 6639-6643.19. Zhang, F. R., Heilig, R., Thomas, G. & Aurias, A. (1990) Chro-

time, condensation rate, gene density, base composi- mosoma 99, 436-439.rsochore types, chiasmata frequency, and chromosome 20. Bendayan, M. (1984) J. Electron Microsc. Tech. 1, 243-270.sites (Gerald Holmquist, personal communication). 21. Drouin, R., Messier, P.-E. & Richer, C.-L. (1988) Cytogenet. Cellsuccessfully mapped DNA sequences with an ex- Genet. 47, 117-120.sucesy ul.h p i s edwe cocmtnyse propha 22. Drouin, R., Messier, P.-E. & Richer, C.-L. (1989) Chromosoma 98,4y high precision since we concomitantly used prophase * 40-48.

4d chromosomes [at the highest band level, 1250 bands 23. Messier, P.-E., Drouin, R. & Richer, C.-L. (1989) J. Histochem.tologically distinguishable replication units (35)], the Cytochem. 37, 1443-1447.article reporter/detection system, and the high resolv- 24. Drouin, R., Lemieux, N. & Richer, C.-L. (1988) Cytobios 56,)wer afforded by EM. Gold particles can be prepared in 107-125.

wifferenaffordedabyEM .Gol patlesu can be peared.i 25. Messier, P.-E., Jean, P. & Richer, C.-L. (1986) Cytogenet. Celldifferent sizes and are hghly suitable for quantitative Genet. 43, 207-210.ation and multiple labeling (20). The procedure de- 26. Yang, T. P., Hansen, S. K., Oishi, K. K., Ryder, 0. A. &d is simple, reproducible, and reliable since all of the Hamkalo, B. A. (1982) Proc. Natl. Acad. Sci. USA 79, 6593-6597.for probe hybridization, probe detection, and chromo- 27. Page, D. C., Mosher, R., Simpson, E. M., Fisher, E. M. C., Mar-bandingarecarriedoutonusualglassslidesandaredon, G., Pollack, J., McGillivray, B., de la Chapelle, A. & Brown,

banding are carried out on usual glass slides and are L. G. (1987) Cell 51, 1091-1104.gous to those employed for LM. The transfer of selected 28. Vogel, W., Mehnert, K. & Pentz, S. (1990) Hum. Genet. 84,es to a grid for EM occurs at the end of the protocol and 237-240.s analysis of complete mitoses and of chromosomes at 29. Drouin, R., Messier, P.-E. & Richer, C.-L. (1989) Chromosoma 98,aitotic stages from prophase to metaphase. 3.174-180.30. Gratzner, H. G. (1982) Science 218, 474-475.I will certainly be useful in facing the challenge of 31. Moran, R., Darzynkiewicz, Z., Staiano-Coico, L. & Melamed,ing the human genome. Its usefulness becomes partic- M. R. (1985) J. Histochem. Cytochem. 33, 821-827.evident if its high resolving power is coupled with 32. Raap, A. K., Marijnen, J. G. J., Vrolijk, J. & van der Ploeg, M.

reizing DNA sequences to banded chromosomes that (1986) Cytometry 7, 235-242.dizingNAseqencest bande chromoomes tat 33.Kurnit, D. M. (1974) Cytogenet. Cell Genet. 13, 313-329.it a distinct band pattern at the level of 1000-1250 bands 34. Grossman, L., Levine, S. S. & Allison, W. S. (1961) J. Mol. Biol.aploid set. Furthermore, long chromosomes with a 3,47-60.ct band pattern are certainly essential for the precise 35. Drouin, R., Lemieux, N. & Richer, C.-L. (1990) Chromosoma 99,,zatintosubbndo singe-cop seqencesinLM but273-280.zation to asubband of single-copy sequences in LM, but 36. Manuelidis, L. & Chen, T. L. (1990) Cytometry 11, 8-25.xtends the precision beyond this by permitting local- 37. Drouin, R., Lemieux, N. & Richer, C.-L. (1991) Cytogenet. Celln within subbands. Genet. 57, 91-99.

>- 3+>.50

Lo 2+z

z4

14

FIG.time inhyde trplot CtreatedformalThey vthe pro1, 2, 4chrom4associadetectiwere sbandinevalualelsewhdata wtreatmtweenbandinthe DI,somes;treatmo

cationtion, ibreakWe

tremebandeor cylgold ping pomanyevaluEscribestepssomeanalolmitosallow,any nrEM

mappularlyhybri4exhibper hdistinlocaliEM eizatio

10920 Genetics: Fetni et al.

Dow

nloa

ded

by g

uest

on

Feb

ruar

y 1,

202

1