Fundamental and Molecular Mechanisms of Mutagenesis

Mutation Research 372 (1996) 17% I X0

PRINS localization of centromeres and telomeres in micronuclei indicates that in mouse splenocytes chromatid non-disjunction is a

major mechanism of aneuploidy

Antonella Russo * , Giovanna Priante, Anna Maria Tommasi Drpnrtment of Ewlog~, Uniwrsity of Padow. Vitr Trivrte 75. l-35123 Padora. Ital!

Abstract

Primed In Situ DNA Synthesis (PRINS) of telomeric and centromeric (minor satellite DNA) sequences has been applied together with the cytokinesis block micronucleus (MN) assay in mouse splenocytes, with the aim of understanding the mechanism of origin of spontaneous and induced MN. Splenocyte cultures were treated in vitro either with the clastogenic agent mitomycin C or with the aneugenic compound colcemid. The relative proportions of MN carrying I to 4 telomeric signals were in agreement with the known mechanism of action of the chemicals tested, i.e., an higher number of MN with less than 4 telomeres were found in MMC - than in colcemid-treated cultures. No MN lacking the telomeric sequences (0 spot) were found, indicating that the observed distributions should not be affected by false-negative data. Furthermore, all MN carrying a single telomere were negative for the centromere, thus indicating that this class represents true chromosome acentric fragments. Finally, MN with 4 telomeric spots always carried the centromeric sequence, as expected on the hypothesis that these MN correspond to whole chromosomes. With respect to centromere-positive MN, more than one half carried 4 telomeric signals (whole chromosomes), and only l/4 or less showed 2 telomeric signals (probably corresponding to a single chromatid). This difference was statistically significant, either in untreated cultures or in cultures exposed to mitomycin C or colcemid. On the whole. these data indicate that non-disjunction followed by whole chromosome loss (with the production of two daughter monosomic nuclei) may be the main mechanism of malsegregation leading to MN formation.

Keywords: PRINS: Telomere; Centromere: Micronucleus assay

1. Introduction

Primed In Situ DNA Synthesis (PRINS) was pro- posed a few years ago (Koch et al.. 1989; Gosden et al., 1991) as a fast approach for the localization of DNA sequences in chromosome preparations. The rationale of PRINS is based on an extremely simpli- fied PCR approach. with a single cycle of anneal-

- Corresponding author

ing/extension. PRINS may be considered alternative to Fluorescence In Situ Hybridization (FISH), the reievant features of the former methodology being: (1) the possibility to label in situ the sequences of interest, using a short oligonucleotide as primer, thus avoiding the problems related to acquisition of probes (in theory, any new published sequence may be used to design the suitable primer); (2) the short time required for completing the whole protocol of se- quence amplification and detection: PRINS may be performed in 2 h or less for the detection of a single

0027.5107/96/$15.00 Copyright G I996 Elsevier Science B.V. All rights reserved PII SOO27-5 107(96)00137-6

sequence. and when a two-colour (double labelling) PRINS is carried out, only 3 h are necessary; there- fore, in contrast with FISH, the results of a PRINS experiment may be collected within a single day.

In spite of these promising features, PRINS has been applied mainly on metaphase spreads and free nuclei, and mostly on cells of the human species. We recently optimized a protocol (Russo et al., 1996) for the simultaneous visualization of centromeric and pericentromeric sequences (minor and major satellite DNA) in cytokinesis-blocked murine splenocytes. In that study PRINS gave excellent results. in terms of quality and resolution of fluorescent spots, and a micronucleus (MN) test carried out after treatment with mitomycin C (MMC) and colcemid confirmed that PRINS may be proposed as a fast alternative to FISH for the characterization of induced chromo- some damage (Russo et al., 1996).

In the present paper, data are shown on the colo- calization of telomeric and centromeric sequences in MN, as a further improvement of our PRINS ap- proach on mouse cells. These results were obtained on cytokinesis-blocked cells prepared from the same splenocyte cultures used to detect minor and major satellite DNA with PRINS tandem labelling (Russo et al., 1996). This allowed us to make an highly accurate comparison between the sensitivities of the two strategies applied for the identification of MN carrying whole chromosomes. From the present study it appears that the localization of telomeric se- quences in MN may represent a powerful tool for understanding the mechanism of origin and the fate of chromosome missegregation.

2. Materials and methods

All the experiments described in this paper were carried out on cell preparations obtained from the same murine splenocyte cultures used in a parallel study in which centromeric and pericentromeric se- quences were detected (Russo et al., 1996). Five male mice (8-14 weeks old) were used; spleens were removed aseptically and squashed to obtained a cell suspension in PBS. Splenocytes were isolated by centrifugation on Ficoll (Hystopaque’“-1077. Sigma, St. Louis, MO, USA), and cultures were started at the initial density of 1 . lOh cells per ml. Treatments

with the clastogenic agent Mitomycin C (MMC, Kyowa Hakko Kogyo, Tokyo, Japan; CAS No. 50- 07-7; final concentration: IO-6 M). and the aneu- genie agent colcemid (COL, Sigma; CAS No. 477- 30-5; 10e7 M). were performed at 22 h from the initiation of the culture. 4 h later Cytochalasin B (Cyt B, Sigma) was added, at the final concentration of 3 kg/ml. Cells were harvested at 44 h after Con A stimulation; colcemid-treated cultures were resus- pended in fresh complete medium (supplemented with Cyt B) in the last 4 h of culture. In general,2 cultures treated with MMC and 2 cultures treated with COL were set up from different mice. Untreated cultures were grown in parallel to treated cultures from each animal used. Slides were prepared by cytocentrifugation (400 rpm for 5 min; Shandon III, UK). air-dried, and then fixed in absolute ethanol at -20°C. Preparations were kept in ethanol until used. More details are given in Russo et al. (1996).

Two-colour PRINS was carried out by using in the first cycle an oligonucleotide which hybridizes with the mammalian telomere (5’-TTAGGG(,,-3’1, and digoxygenin- 11 -dUTP (Boehringer-Mannheim, Germany) as modified nucleotide. In the second cycle. an oligonucleotide corresponding to the minor satellite sequence was used (Mitchell et al., 1993) and the labelled nucleotide was in this case biotin- 16-dUTP (Boehringer). Before the first PRINS cycle the slides were denatured in 0.01 M NaOH/I M NaCl for 30 s, rinsed in 0.01 M Tris-HC1, pH 7.6, for 10 min, and air-dried. Slides were briefly pre- heated (53°C) on the plate of the thermal cycler (OmniGene with In Situ block; Hybaid, UK). 12 ~1 of the reaction mixture (= 300 pmol of the oligo- nucleotide of interest, dATP, dGTP, dCTP 0.2 mM each, dTTP 0.02 mM, the labelled deoxynucleotide, 0.02 mM, the enzyme Taq polymerase (5 U), 7.5 mM MgCl?, in a total volume of 50 ~1 PCR buffer) were placed on each slide. This volume was covered with a 18 X 18 coverslip, which was carefully sealed with nail polish. Annealing and extension were car- ried out, respectively, at 53°C for 10 min, and 72°C for 30 min. At the end of the cycle, coverslips were removed and slides dipped in the blocking solution (50 mM EDTA/SO mM NaCI; 72”C, 2 min), and subsequently in 0.01 M Tris-HCI? pH 7.6, for 5 min. The second PRINS cycle was carried out at the same conditions as above. After a second wash in the

Table

I

Numb

ers

end

types

of

MN

ohwr

ved

in hin

uclea

te \pl

rnoc

ytc~.

II-e

n~cd

wi

th MM

C or

Colce

mid

(COL

), aft

er

vmult

aneo

w de

tectio

n of

centr

ome~

-c\

and

telom

cres

by

PRIN

S

Mous

e No

. ‘rre

almrn

~ Bin

uclca

te/

Binuc

leate

and

age

Binuc

leatr

Tota

l MN

wi

th MN

wi

th 4

MN

with

3 MN

wi

th 2

MN

with

I MN

wi

th

cell\

with

MN

centr

omer

e vi\

iblc

vi\ihl

e vis

lhle

visible

no

vis

ible

I MN

(m

inor

telom

cres

tclom

erer

tel

ornere

\ tel

omcre

tel

omer

es

sr

Mous

e No

. I

- 51

( IO

wee

ks)

MMC.

10

’

M 31

Mous

e No

. 2

- 5x

(14

week

s)

I.137

90s

1679

Mous

e No

. 3

5.5

( I4

week

s) M

M<‘.

IO

’

M 34

Mous

e No

. 4

- so

( I6

weA)

CO

L. IO

’

M 23

Mous

e No

. 5

- 42

(8 we

ekc)

COL,

IO-

M 2S

Mous

e No

. I

5 ~

51

MMC.

IO

F”

M 33

COL.

IO

’ M

24

I645

2037

IS20

7335

_-

.

2361

12x3

x.542

2942

3.51x

.- utc

lllte)

~~

+ cc

ntr.

E to

t lO

I +

centr

. to1

+

ctm

r. to

t +

LYnt

r. 5

14

17

203 9

I42 I ‘1

x3

64

292

22s

5 3

(ho%

)

I I

93

37(40

%)

IS

I5

I4 9(6

4%)

4 4

17

9Kw)

6

6

213

X6

(40%

) 37

37

9 SW

%)

3 3

I so

I I I

(74

%)

ho

60

I ‘1

I3 (74

%)

IO

IO

X6

64

(74%

) 2x

28

64

40

(63%

) 24

24

306

123

(40%

) 52

52

Xi 17

S(74

c/r)

xx

xx

4 3

0

26

IX 2x

5 3

3

x 3

3

54

2x

74

3 2

2

36

2x

39

4 I

4

27

21

27

24

II I2

x0

46

102

63

39

66

2 2

0 0

0 0

0 21

3x

0 0

0 I 0

0

s 23

IS

0 0

2 %

3 I 0

0

IS

3 0

0 2

S 4

0 0

25

72

0 0

3x

I9 0

0

I d

176

-1 MMC

* 3 COLCEMID

totals Mthout with \rtth with WI111 centromere centromere <entronierei centromerci cClltrO,litrei

4 telomeres 3 telomeres 2 telorneres

Fio C’ 1. Frequencies of MN. classified wnh respect to presence/absence of ccntromere (minor jatellite DNA) and number of visible telomeres. observed in control and treated splenocyte cultures. I

blocking solution, slides were washed twice (5 min each) in 4 X SSC, pH 7.0,O. 1% Tween 20, and once in PBST (PBS, O.l%, Tween 201, 0.5% non-fat dry milk. A mixture of fluorescein-antidigoxygenin anti- body (1 : 12, v/v; Boehringer) and Texas Red-Avidin D (1 : 500, v/v; Vector Lab., Burlingame. CA. USA) in PBST was used to detect fluorescent spots (30 ~1 under a 20 X 20 coverslip, 30 min at 37°C in a moist chamber). After the last two washes in PBS, slides were mounted in antifade solution (Vectashield I’, Vector Lab.). The antifade solution contained DAPI (5 p&ml) as a counterstain.

Slides were scored under a Zeiss Axioskop fluo- rescent microscope, equipped with filters for UV, blue and green excitation, and with a triple band pass filter for simultaneous vision of DAPI-Fluorescein- Texas Red. At least 1000 binucleate cells with well preserved cytoplasm were scored, at 1000 x magni- fication, per experimental point. The percentage of binucleate cells with respect to mononucleate ones was calculated as a proliferative index of each cul- ture. MN were identified following the standard criteria (Heddle et al., 1990) and classified on the basis of the number of telomeric sequences and

p < 0.001: ,, -z 0.001.

presence/absence of the minor satellite. Each MN category was calculated as percentage of the total MN number. for evaluation of PRINS reproducibility within experimental points, or reported to 1000 binu- cleate cells for comparison with control data.

Statistical comparisons among MN frequencies observed in treated and untreated cultures were done by applying the G test ( Sokal and Rohlf, 1981). The binucleate cell proportions found in control and treated cultures were compared by applying the Stu- dent f-test on mean values after square root transfor- mation.

3. Results

Table 1 shows the results obtained after culturing and treating splenocytes obtained from 5 different mice. For each culture, the percentage of binucleate to mononucleate cells, the number of cells carrying MN and the total number of MN are given. The number and percentage of MN carrying the cen- tromere (minor satellite DNA) is also shown, irre- spective of the number of telomeres detected, and

Fig. 2. Examples of binucleate splenocytes carrying MN with different numbers of telomeres, and with or without the centromere. (a-c) A splenocyte carrying 2 MN, viewed at the UV filter (a), at the blue light (b), detecting the telomeric sequences (fluorescein, yellow spots), and at the green light cc), which allows the detection of centromeric sequences (Texas-Red; red spots). The MN pointed by the small arrow has 2 telomeres and no centromere, the other one (large arrow) shows 3 telomeres and the centromeric sequence. (d-f) A binucleate cell with 2 MN (d; UV filter): each MN shows 2 telomeres (e; yellow spots), and one of them is centromere-positive (f; red spot). Images from the original pictures were captured by scanner, and the resolution of fluorescent signals was improved by using Photostyler 2.0 software.

finally the distribution of MN with respect of the number of visible telomeres and the presence/ab- sence of the centromeric sequence is given. Due to the evident homogeneity among cultures of the same type, these data were pooled (Table 1). MN frequen- cies were also calculated on 1000 binucleate cells, as shown in Fig. 1. In control cultures, a total of 64 MN and 8542 binucleate cells were scored, corresponding to a MN frequency of 7.49%0 (Fig. l), the proportion of MN carrying the minor satellite sequence was 63% (Table 1). The MN observed in MMC-treated cultures were 306 (104%~ in 2942 binucleate cells; Fig. l), and 40% of them carried the minor satellite (Table 1). The MN found in cultures treated with colcemid were 236, corresponding to 67%~ of the binucleate cells scored (Fig. 1); the percentage of centromere-carrying MN induced by colcemid was 74% (Table 1).

Looking at the distribution of telorneres in the MN, in untreated cultures the largest proportion of MN showed 4 or 3 fluorescent spots (48/64), and the same was found in colcemid-treated cultures (15 l/236), in contrast less than one half of the MN carried 3-4 telomeres after treatment with MMC (132/306). MO re than 600 MN were scored and none of them were found which were negative for the telomeric sequence (Table 1). MN with 4 visible telomeres must correspond to whole undivided chro- mosomes, and accordingly they invariably showed the centromeric signal (Table 1 and Fig. 1). MN with only 1 visible telomere should represent acentric fragment, and in fact they never carried the minor satellite sequence, whatever the treatment considered (Table 1). MN with 2 or 3 spots represented a mixture of centromere-positive and centromere-nega- tive MN, with different proportions according to the treatment carried out (Table 1 and Fig. 1). Examples of MN with different numbers of telomeres and with or without the centromere are shown in Fig. 2.

Treated cultures showed a significant (p < 0.01) decrease in the binucleate/mononucleate ratio with respect to control cultures (Table 1). Statistical com- parisons between MN frequencies detected in control and in each treated culture were performed taking into account the total MN frequencies, the frequen- cies of MN lacking the centromere (with no correc- tion for the visible number of telomeres shown), and the frequencies of centromere-carrying MN, with or

without further subdivision per number of telomeres scored (Fig. 1). Both treatments induced an evident increase in MN, with p -=K 0.001. In particular, a statistical induction was observed for MN either lacking or carrying the centromere (minor satellite) sequence (p -=z 0.001). Among MN carrying the centromere, highly significant differences were found between MMC-treated cultures and the controls con- cerning the frequency of MN showing 4, 3 (both: p +K 0.001) or 2 telomeres (p CK 0.001). The same pattern was observed for colcemid-treated cultures. Finally, it may be noticed that MN with the minor satellite sequence and 4 telomeres were the most common type in each culture. In particular, untreated cells carried 5 times more centromere-carrying MN with 4 than 2 telomeres (p < 0.001). and MMC and colcemid-treated cultures gave a 2 : 1 ratio between the same MN types, with p < 0.01 for MMC and p < 0.001 for colcemid.

4. Discussion

In this paper the possibility of detecting telomeric and centromeric sequences in micronucleated mam- malian cells by a two-colour approach with PRINS was evaluated. It was verified that PRINS allows an effective colocalization of telomeres and centromeres in interphase cells. In particular, the following results are notable.

(1) The proportions of MN carrying the cen- tromeric sequences and therefore classified as con- taining a whole chromosome, appeared highly over- lapping to those identified in a parallel study on the same cell preparations in which PRINS was applied to detect minor and major satellite DNA with the classical tandem labelling approach ( Russo et al.. 1996). As an example, after treatment with MMC 40% of the MN were centromere-positive either from mouse No. 1 and mouse No. 3 (Table 1). With the tandem labelling approach, 37.9 and 40.0 were the percentages of MN carrying the whole cen- tromeric region (C MN) from mouse No. 1 and No. 3, respectively ( Russo et al., 1996). Also in the case of treatment with colcemid, it was found that 74% of the MN were centromere-positive in the present study (both mice No. 4 and No. 5; Table I), whereas C MN detected with the tandem labelling approach

were 67.6 and 60.2 from the same mice ( Russo et al., 1996). These data indicate the reliability of PRINS approach for the simultaneous localization of repetitive sequences in mammalian cells. In particu- lar, it must be noticed that minor satellite DNA was detected either during the first (Russo et al., 1996) or the second PRINS cycle (present study), without any quantitative differences. In other words, no artifac- tual results are produced in the course of a two-col- our PRINS with respect to order of sequence amplification.

(2) Our data indicate that the detection of the actual number of telomeres in MN by PRINS is highly accurate: first, no MN were found lacking the telomeric sequences (0 spot). indicating that the con- tribution of false-negative data is minimal. In addi- tion, all the MN with a single telomere were negative for the centromere. thus indicating that this class represents true acentric chromosome fragments. In agreement with the known mechanisms of action of the chemicals tested, the clastogenic agent MMC induced a majority of MN with 1-2 telomeres. while in untreated cells or in cultures treated with colcemid the most common type of MN carried 4 telomeres. All MN with 4 telomeric spots, corresponding to a whole undivided chromosome, carried as expected the centromere (minor satellite) sequence.

(3) Among the centromere-positive MN, which were statistically increased after treatment with ei- ther the clastogenic agent MMC or the aneugenic compound colcemid, a high prevalence carried 4 telomeric signals. These MN unambiguously corre- spond to whole chromosomes (2 chromatids); their relative proportion in each culture type was com- pared, by following a conservative approach, with that of the MN representing most probably a single chromatid (showing the centromere and 2 telomeric signals), and a significant difference was detected.

A previous study which attempted the simultane- ous detection of telomeric and centromeric (minor satellite) sequences by FISH, in conjunction with the MN assay in mouse bone marrow smears (Schriever-Schwemmer and Adler. 1994) obtained a low resolution of telomeres in MN. and the Authors concluded that the approach was not suitable for aneuploidy assessment. Another study carried out on mouse cultured fibroblasts with telomeric and peri- centromeric (major satellite) DNA probes (Miller

and Niisse, 1993) reported that a correct estimation of the number of telomeres in MN was difficult. In contrast, the data presented here on mouse spleno- cytes cultured and treated in vitro are strictly consis- tent with respect to presence/absence of telomeric and centromeric sequences; in addition, the number of telomeres were estimated with a very good repro- ducibility between replicated experiments, and with patterns in agreement with the treatment protocols considered. Our data therefore indicate that telomeric sequences. together with centromeric sequences, may be considered useful molecular markers to distin- guish MN which arise after chromosome malsegre- gation compared with those derived from chromo- some breaks. It is possible that the different strate- gies of sequence localization implied by the FISH and PRINS methodologies are responsible for the different outcomes of the present and other studies (Miller and Niisse, 1993; Schriever-Schwemmer and Adler, 1994).

Traditionally, MN containing the centromeric re- gion are thought to represent anaphase lagging chro- matids. It is assumed also that when this error oc- curs, one of the daughter cells from a single cell division will be euploid, and the second one mono- somic. Because of the possibility of classifying MN with confidence with respect to the number of telom- eres and the presence or absence of the centromere, the data obtained in this study strongly suggest that non-disjunction followed by whole chromosome 10s~ may be the main mechanism of malsegregation lead- ing to MN formation; in this hypothesis, two daugh- ter monosomic nuclei are produced. It should be noticed that the relative proportions of MN with different numbers of telomeres evaluated in this study did not differ between the observed spontaneous and induced centromere-carrying MN. Furthermore, comparable distributions were found after cells were treated with MMC. a bifunctional alkyating agent known as clastogen, but already reported to have an aneugenic potential and to induce detachment/dis- ruption of kinetochores (Renzi et al., 1996: Russo et al., 1996) or with colcemid, which is known to affect polymerization of microtubules. These data suggest that the region of attachment between kine- tochore and spindle fibers may be the most critical cell target with respect to chromosome missegrega- tion, and that prophase-metaphase. instead of

180 A. Russo PI al. / Mutation Resratrh 3 72 ( 19961 173- 180

anaphase, is the relevant phase of the cell cycle during which this type of error occurs. Further exper- iments are requested to clarify this point, by testing chemicals at different mechanism of action and at different dose levels. Colocalization of centromeric and pericentromeric domains of the mammalian chromosome has contributed in recent years to a fine characterization of chromosome damage by means of the MN assay (Eastmond et al., 1993; Eastmond et al., 1995). In this study we demonstrated that PRINS is a promising new approach for the molecular char- acterization of MN in mammalian cells, allowing in particular to provide new information on the mecha- nism(s) of induction of aneuploidy by simultaneous detection of telomeres and centromeres in MN.

Acknowledgements

We gratefully acknowledge Prof. A.G. Levis (Padova, Italy) for helpful discussion and critical reading of the manuscript. This work was supported by the Italian Association for Cancer Research (AIRC).

References

Eastmond, D.A. D.S. Rupa, H.W. Chen and L.S. Hasegawa (1993) Multicolor fluorescence in situ hybridization with cen- tromeric DNA probes as a new approach to distinguish chro- mosome breakage from aneuploidy in interphase cells and micronuclei, in: B.K. Vig (Ed.), Chromosome Segregation and Aneuploidy. NATO .4SI Series. pp. 377-390.

Eastmond, D.A.. M. Schuler and D.S. Rupa (1995) Advantages and limitations of using fluorescence in situ hybridization for

the detection of aneuploidy in interphase human cells, Muta- tion Res.. 348. 153-162.

Gosden. J., D. Hanratty. J. Starling, J. Fantes. A. Mitchell and D. Poneous (1991) Oligonucleotide-primed in situ DNA synthe- sis (PRINS): a method for chromosome mapping, banding, and investigation of sequence organization, Cytogenet. Cell Genet., 57, 100-104.

Heddle. J.A. A. Bouch, M.A. Khan and J.D. Gingerich (1990) Concurrent detection of gene mutations and chromosomal aberrations induced in viva in somatic cells, Mutagenesis. 5. 179-184.

Koch. J.. S. Kolvraa. K.B. Petersen. N. Gregersen and L. Bolund (1989) Oligonucleotide-priming methods for the chromosome-specific labelling of alpha satellite DNA in situ, Chromoaoma, 98, 259-265.

Miller, B.M. and M. Nusse (1993) Analysis of micronuclei in- duced by 2-chlorobenzylidene malonitrile (CS) using tluores- cence in situ hybridization with telomeric and centromeric DNA probes, and Bow cytometry. Mutagenesis, 8, 35-41.

Mitchell. A.. L. Nicol, P. Mallory and D. Kipling (1993) Novel structural organization of a Mu.7 ~~usc~lus DBA/2 chromo- some shows a fixed position for the centromere. J. Cell Sci.. 106. 79-85.

Renzi, L., F. Pacchierotti and A. Russo (I 996) The centromere as a target for the induction of chromosome damage in resting and proliferating mammalian cells: assessment of mitomy)cin C-induced genetic damage at kinetochores and centromeres by a micronucleus test in mouse splenocytes. Mutagenesis. I I. 133-138.

Russo, A... A.M. Tommasi and L. Renzi (1996) Detection of minor and major satellite DNA in cytokinesis-blocked mouse splenocytes by a PRINS tandem labelling approach, Mutagen- esis, I I, in press.

Schriever-Schwemmer. G. and I.-D. Adler t 1994) Differentiation of micronuclei in mouse bone marrow cells: a comparison betw,een CREST staining and fluorescence in situ hybridiza- tion with centromeric ad telomeric DNA probes, Mutagenesis, 9, 333-330.

Sokal. R.R. and F.J. Rohlf (198 I I Biometry. The Principle and Practice of Statistics in Biological Research, 2nd Edn., Free- man New York.