Nucleic Acids Research, 1993, Vol. 21, No. 5
1133-1139
Repetitive DNA sequences located in the terminal portion
of the Caenorhabditis elegans chromosomes
Giuseppina Cangiano and Adriana La Volpe*
CNR International Institute of Genetics and Biophysics, 10 Via Marconi, 80125 Naples, Italy
Received December 14, 1992; Revised and Accepted January 28, 1993
EMBL accession nos4
ABSTRACT
We describe the distribution along the chromosomes
of Caenorhabditis elegans of two repetitive DNA
families, RcS5 and Cerep3 and interstitial telomerlc
sequences. Both families show, among other
interesting features, a preferential location In the
terminal 30% of the chromosomes. It Is known that in
these regions of the genome the frequency of
recombination is much higher than In the central
portion, genes are rarer and sequences Important for
chromosome disjunction may lie.
INTRODUCTION
As a result of the intensive and multifaceted analysis of genomes,
it is becoming increasingly manifest that most chromosomal DNA
performs functions other than those involved in pure coding
activity. Chromosomes need to attach to the scaffold and to the
nuclear matrix, to duplicate with fidelity and to segregate
efficiently. Homologous and non homologous recombination is
active in somatic tissues and germ line cells. Specific sequences
along the chromosome guarantee the binding of specific matrix
proteins, proper DNA duplication and segregation during the cell
cycle, and probably the enhancement or the repression of
recombination. C.elegans is a well known model system for
developmental genetics and genomic studies. The physical map
is close to completion and a pilot project to sequence the
C.elegans third chromosome has been undertaken (1).
Comparison between the physical map and the genetic map (2)
has confirmed a non uniform distribution of crossovers along the
chromosomes. Crossovers are relatively rare in the central part
of the chromosomes, in the so called 'gene clusters' (where most
mutants map), but their frequency is highly enhanced in the proterminal region of the chromosomes, where genetic loci defined
by mutations are rarer (fig 1). A survey of the distribution of
the coding regions in the genome (by cDNA analysis) has shown
that although variation in the frequency of recombination accounts
for the clustering of most genes in the centre of the chromosomes
on the genetic map, nevertheless coding sequences are physically
somewhat rarer in the so called 'chromosomal arms' (3).
Cytological and genetic studies indicate that C. elegans
centromeres are diffuse along the chromosomes (holocentric
• To whom correspondence should be addressed
+X69085, X69082, X68553 and X68498
chromosomes, 4, 5, 6). However, when the chromosomes
segregate in meiosis they align parallel to the spindle axis such
that one end or the other points towards the spindle pole (Donna
Albertson personal communication).
In an effort to resolve the chromosome structural and functional
organization in C. elegans, we have recently described the
distribution of clusters of satellite-like DNA sequences and short
interspersed elements (7). We observed a uniform distribution
of five repetitive DNA families along the chromosome. Here we
identify and describe a family of repetitive DNA elements, RcS5,
which in contrast is preferentially located at the pro-terminal
portion of the chromosomes.
We further show that another repetitive DNA family previously
described, Cerep3 (8) is similarly distributed. Associated with
these two repetitive DNA families we find a high concentration
of interstitial telomeric repeats.
MATERIALS AND METHODS
N2-JW-H, W-PA-1H, W-PA-2H, DH424, RW7000, RC301,
TR403, CB4507, CB4555, CB4853, CB4854, CB4855, CB4856,
CB4857 are wild type, independently isolated C.elegans strains,
CB1392 is a nucOOl mutant, they were all obtained from the
Caenorhabditis Genetics Centre. DNA from C.remanei, PG-3,
P.redivivus, C.briggsae, PG-2, C.marpasi , all closely related
nematode species, was a gift from L.W.Nawrocki. All the strains
were raised at 18°C. Routine strain maintenance was as described
by Sulston and Hodgkin 1988 (9).
C.elegans DNA purification from liquid cultures was
performed according to Sulston and Hodgkin (9).
The cosmid library, from which the initial representative of
the RcS5 family (RcS5 # 1) was obtained, was a gift from Silvana
Gargano. It was made by inserting genomic DNA, partially
digested with SauETJA, into the BamHI site of tetra-cos vector.
A subset of clones (250), corresponding to about 10% of the
C.elegans genome, was screened by hybridisation with labelled
total genomic DNA as probe. We obtained 12 positive clones,
two of which contain ribosomal genes, three contain the
transposable element Tel, two contain elements belonging to the
same repetitive DNA family, the others all represent individual
elements belonging to independent repetitive DNA families. All
1134 Nucleic Acids Research, 1993, Vol. 21, No. 5
the other cosmid clones derive from a library also constructed
from a partial SauIIIA digest, using the vector Lorist6 (10).
The genomic DNA library in yeast artificial chromosomes was
that made for the C.elegans mapping project and has been
previously described (11, 12, 13, 7). The vector was pYAC4
and the Saccharomyces cerevisiae host strain was AB138O
(MATa, y+- ura-3, trp-1, ade2-l, can-1 100, tys 2-1, his-5).
Southern transfer was performed as described by Southern (14)
with the following modifications: the DNA (1 /tg of total genomic
DNA per lane) was transferred to Hybond-N membrane and
immobilized by UV treatment. Hybridisations of libraries and
Southern nylon filters were performed in 6xSPC (20xSPC =
2M NaCl, 0.6M NajHPO,, 0.02M EDTA pH 6.2), 0.9%
Sarkosyl, 10% dextran sulphate and 100/ig/ml single strand DNA
carrier at 65°C for 6—16 hours. We used approximately 107
cpm of oligolabelled DNA probes prepared as described by
Feinberg and Vogelstein (15, 16) with a specific activity of greater
than 109 cpm//tg per hybridisation bag. The telomeric probe was
a single strand 48mer containing eight copies the TTAGC-C repeat
identified as the Ascaris telomeric repeat (17). It was purchased
from Eurogentech (Belgium) and end-labelled with T4
polynucleotides kinase and [7 32P]-ATP. Hybridisation was
performed in the mixture described above at 60°C.
Following the hybridisation, filters were washed four times
either in l x S P C , 1%SDS at 50°C or in 2xSPC, 1%SDS at
42°C (relaxed conditions) and air dried.
Removal of probe: In order to reuse the filters, after each
independent hybridisation, probes were removed from filters by
repeating the following steps twice: 10 min room temp, in 0.2
M NaOH, 0.5 M NaCl, 10 min room temp, in 0.5 M Tris pH7.4,
0.5M NaCl and 10 min room temp, in lxSSPE(20xSSPE=3M
NaCl, 0.2M NaH2PO4.H20, 0.02M EDTA, pH 7.4). Efficiency
of removal was monitored by re-exposing used filters before
reuse.
Cosmids and plasmids used as probes were maintained in E. coli
rec A strain HB101 (F~, hsd S20, (^m8), recA13, ara-14,
proA2, lacYl, galK2, rpsL20 (Snf), xyl5, mtl-1, supE-44, X")
selecting for kanamycin or ampicillin resistance. The strain was
checked periodically for the recA phenotype.
All RcS5 subclones were constructed in pGEM3/4 or
pUC18/19. Cerep3 plasmid was a gift from Scott Emmons.
Sequencing was performed according to Sanger and Coulson
(18). The complete sequences obtained during this work have
been sent to the EMBL Data Library, and only relevant portions
are shown. Accession numbers are indicated in the legend of
fig. 5.
Computer programs used in the physical mapping analysis are
described in Sulston et al. (19). Sequence data were analysed
with the GCG (20) software package.
RESULTS
Physical distribution of RcS5 and Cerep3 elements in the
genome
The first element belonging to the RcS5 family, called RcS5 # 1,
was isolated from a genomic DNA library using as probe total
genomic C.elegans DNA; clones containing DNA from the
genomic repetitive fraction give much stronger hybridisation
signals (7). The repetitive portion of the original cosmid was then
subdoned and used as probe on two types of indexed library grids
on nylon membrane (12, 7). The first kind of grid represents
a subset of cosmids already positioned on the map, the 1934
I 111 M i l l
Figure 1. Comparison between the genetic map and the physical map of
chromosome 2. On the genetic map, top, all the genetic loci (unspecified) are
shown as vertical bars. The bolded box represents the region named 'gene duster'
and the bare terminal portions of the chromosome are evident. Slanted lines point
at corresponding loci in the genetic map and physical map underneath. Variations
in the frequency of recombination account for the discrepancies between genetic
and physical distances. In order to allow the comparison, the scale used for both
maps is expressed in percent of the chromosome length. Genetic data were derived
from Edgtey et al. (2). Physical data were derived, with permission, from physical
map version 920628 (physical map data may be subjected to changes, Coulson
and Sulston personal communication).
cosmids cover about 80% of the genome in a non-overlapping
manner. The second kind of grid is a set of overlapping YAC
clones representing almost 95 % of the genome. We demonstrated
elsewhere that all the cloned members of a repetitive family can
be positioned in this way on the already mapped portion of the
genome (7). The use of two indexed libraries carried by a
prokaryotic and a eukaryotic host may overcome problems of
instability (in one host or the other) of some DNA sequences.
hi figure 2 we report the physical distribution of RcS5 elements
on the physical map of the five autosomes and on the X
chromosome. RcS5 elements appear to be preferentially located
outside the 'gene clusters'. Another family with a similar
distribution is the Cerep3 family. This latter family has been
identified by Felsentein and Emmons in 1988 and comprises
elements about one kb long, some of which have been shown
to affect replication and segregation of plasmids inSaccharomyces
cerevisiae . The distribution of Cerep3 elements along the
chromosomes is also shown in figure 2.
Strains and species comparison
We have compared the patterns obtained by digesting the genomic
DNA of several independently isolated C. elegans wild type strains
with the restriction endonuclease Eco RI and hybridising the
Southern transfer with RcS5 or Cerep3 probes (figure 3). The
RcS5 probe is an Eco Rl-Sac I fragment from the
RcS5(ZK1091) element containing only RcS5 sequences (see
figure 5 for details), the Cerep3 probe is an EcoRI-BamHI from
pCel7(Cerep3.1) containing a Cerep3 entire element (Felsenstein
and Emmons 1988). Some of the C.elegans strains used in this
experiment have a high copy number of the transposable element
Tel and among these some contain a mutator activity that results
in an increase in the rate of transposition ('hopper' genetic
background, 21). We observe that the RcS5 probe generally
recognises the same number of bands of similar size, suggesting
Nucleic Acids Research, 1993, Vol. 21, No. 5 1135
10
30
) •
40
SO
40
70
10
tO
l»0 0
TCU
DC-41
Ptt
Ft!
CM-3
• Ml
• M4
•Ml
• Ml
daf-7
«P70, ftTC
•OP32
•0P33
rpl-37
fP34 • • * - !
OP>) LNM/1
c
E
«P31 c*b-S
• •1-1
baa- C**-4
Rao/
kla-: J
• MI- i 3 1 d4f.
St P t ll c«b 1
L1S
•*J>-5
mnc-M
•nc-li
y nc-lll
4 •1311a- 1 3
ih-:
?y1
<
ooc-SI
uc-ll
1 la-1)
IDC
fit
SIP/1
I as
hT
ii|ii
\
t
TUVS7
tii-1
i I rtw-3
L " • '
•II cypil«P7l,7»
H4«op-34
C
ll
0
i i
1 i
•>p-72pa
t-123
CPHOT/J
1
Up-]) a
*it-J , 4
«HS-3.
irt-l
• CP4
•«P«
c*b
tctnc toc
|
I bap-1
I «P71
W«P7« • ' • H i « f7 J
f
KP10I
CP10WtP13>
i
e
P4 1
•on
1
TCHKC:
•OP/ 1
hia4 ,••€
C«*-14
py -7
•c-ll
•It-1
d«f-l
• tMOl
MC-7
iri/u
K«l -1
rt»-J
kia-f
TCLIII14A
POP/1
ia-14
itHl
• tP71
•Vb-J
giPl
1
1
Ik
IP*
ht
1 '
El
)4TI III
m
31
1
1
01
f
P
\
J.P1
• •P>
1In •*"
m
A/1
HJ«PI TCWC
• •1'
T
ITTV nJ
•1",
Figure 2. Physical distribution of RcS5 and Cerep3 elements along the physical map of the five autosomes and of the chromosome X. The distribution has been
obtained by hybridising two types of indexed library grids on nylon membrane: a subset of 1934 cosmids already positioned on the map covering about 80% of
the genome in a non-overlapping manner and a set of overlapping YAC clones representing almost 95% of the genome (12, 7). The probes used in the screening
are: a Hindm fragment containing the entire RcS5 # 1 element and a 728bp flanking DNA region (on the left side of the element illustrated in figure 5, ac. no.:
X68553), an EcoRI-BamM fragment from pCel7(Cerep3.1) containing a Cerep3 entire element and about 200bp of flanking sequences on both sides (8), and a
synthetic oligonucleotide 48bp in length containing 8 repeats of the putative telomeric bexamer (TTAGGC). Hybridisations of libraries with RcS5 and Cerep3 probes
were performed in 6XSPC, 0.9% Sarkosyl, 10% dextran sulphate, 100/xg/ml single strand DNA carrier at 65°C, hybridization with the single stranded telomeric
probe was performed in the same mix at 60°C. In all cases, filters were washed four times in 1 x S P C , 1 %SDS at 50°C. The RcS5 elements and the Cerep3 elements
are reported as small vertical rods in the second and diird row respectively. Interstitial telomeric repeats arerepresentedby round bold dots under the physical map.
(Physical map version 920628)
an average constant copy number, although RFLPs are detectable
especially in CB4854 and CB4856 strains. These are not 'hopper'
strains, but show a low Tel copy number and mobility. Those
RFLPs are therefore not likely to be due to accidental
transposition of an active Tel next to a RcS5 element. The
patterns obtained on the same strains with the Cerep3 probe
appear more variable, though not so well resolved due to the
high number of Cerep3 loci in genome.
We have also tested for possible cross-hybridisations between
our repetitive probes and DNAs from other free living nematode
species (figure 4, upper panels). We also have on the same gel
three C.elegans N2 derived lines. No detectable RFLPs were
observed among N2 derived lines. Using an internal fragment
of the RcS5(ZK1091) element, common to ah1 the members of
the family, we did not detect any cross-species hybridisation under
relaxed stringency conditions (hybridisation at 65°C in 6xSPC,
1136 Nucleic Acids Research, 1993, Vol. 21, No. 5
I I I I I I I I I I
• • • •
[ J|
• • •
£Q
£Q
£Jj
£Q
^Q
QQ
Figure 3. Southern transfer and hybridisation with RcS5 and Cerep3 specific
probes of independently isolated C. elegans wild type strains. DNA was digested
with EcoRI and run on a 1 % agarose gel in Tris-Acetate buffer, following the
hybridisation in 6XSPC, 0.9% Sarkosyl, 10% dextran sulphate at 65°C, filters
were washed four times in 1 xSPC, 1%SDS at 50°C. From left to right WPA-lH(Ca.USA), W-PA-2H(Ca,USA), DH424 (Ca.USA), RW7000* (France),
RC301* (Germany), TR403* (Wi, USA), CB4507 (Ca.USA), CB4555*
(Ca.USA), CB4853 (Ca,USA), CB4854 (Ca,USA), CB4855 (Ca,USA), CB4856
(Haway), CB4857 (Ca.USA). Strains with an asterisks have a high copy number
of Tel dements. Origin of the strains is indicated in parenthesis. Upper panel:
the filter was hybridised with an RcS5 specific probe. Lower pand: a second
copy of the same gel transferred on filter was hybridised with a Cerep3 specific
probe.
0.9% Sarkosyl and washes at 42°C in 2xSPC, 1%SDS), even
after a week exposure (not shown). Similarly, we confirmed the
observation of Felsenstein and Emmons 1988 that there is no
cross-species hybridisation with the Cerep3 probe. However a
fragment spanning the RcS5(ZK1091) element and its flanking
sequence gave multiple bands with all nematode species tested,
but not with the yeast DNA used as control, raising the question
of what kind of repetitive sequence, so widespread among other
nematodes, lies next to this RcS5 element (figure 4, lower panels).
Comparison within RcS5 elements and within their flanking
sequences
We have determined the DNA sequence of four independent
members of the RcS5 famfly.Two, RcS5(T25H8) on chromosome
I and RcS5(ZK1091) on chromosome V, represent complete
Figure 4. Southern transfer and hybridisation with RcS5 and Cerep3 specific
probes of different closely related nematode species. DNA in all cases was digested
with EcoRI and run on a 1 % agarose gel in Tris-Acetate buffer, following the
hybridisation in 6xSPC, 0.9% Sarkosyl, 10% dextran sulphate at 65°C, filters
were washed four times in 2 xSPC, 1% SDS at 42°C.The gel in the upper pand
contains, left to right, Cremona, PG-3,C.briggsae, PG-2, Cmarpasi, followed
by three N2 derived CeUgans strains A, B, C, an empty lane and P.redivivus
DNA. The gd was hybridised with an RcS5 specific probe, (Eco Rl-Sac I
fragment from the RcS5(ZK1091) dement containing only RcS5 sequences, see
figure 5 for details) left, and a Cerep3 specific probe (EcoRI-BamHI from
pCel7(Cerep3.1) containing a Cerep3 entire dement, 8) right In the lower panel,
we have a different gel hybridised with two subclones of an RcS5 element. The
gel contains, left to right, Cremona, PG-3, P.redivivus, C.briggsae, PG-2,
Cmarpasi Jaccharomyces cerevisiae DNA as negative control, two empty lanes
and CB1392. The filter on the left was hybridised with the RcS5 internal fragment
described above, the one on the right with a SacI—Hindm fragment spanning
the junction between the RcS5(ZK1091) dement and its rigth flanking sequence
(accession number: X69082).
elements, 1.3kb in length and composed of two terminal inverted
repeats of about 250bp separated by an intervening sequence
around 8O0bp in length. Within the intervening sequence of the
two complete elements we find a duplication of over 150 bp.
RcS5(ZK1091) has also a duplication of 69bp within the inverted
repeat at the 3' end. The other two, RcS5 # 1, unmapped, and
RcS5(T17H5) on chromosome n, are truncated elements missing
the 3' inverted repeat and part of the intervening sequence (see
figure 5) . Sequence identity between elements ranges between
86% and 93%. We did not find any open reading frame inside
the elements making it unlikely that any of these could be an
active member of an unhindered transposon family such as Tc 1.
An interesting feature of die flanking regions of the RcS5
Nucleic Acids Research, 1993, Vol. 21, No. 5 1137
Chromotom* 1
Si
ReS5(T25H8)
u 1 'IOO
a
Chremosom* V
RcS5(ZK1091)
-f
kt
6D V) W \fl
50
^n
lcD
unasslgrwd
RcS5#1
Chromosome
\
HcS5(T17H5)
30^p
^y
Figure 5. A graphic representation of four repetitive elemeius of the RcS5 family. The complete sequence data obtained (including flanking sequences) have been
submitted to EMBL Data Library (accession numbers are: X69O85, X69082, X68553, X68498 respectively). Arrows indicate duplications, white boxes the inverted
terminal repeats, the bold boxes tandem repetitions of telomeric hexamers. On the right, physical locations of the sequeoced elements are shown when known.
elements is the presence of telomeric repeats. Hie inverted repeats
of each element ends with the sequence AGGGTT (in opposite
orientations), and adjacent to each element either at the 5' end
(in RcS5(T17H5)) or at the 3' end (in RcS5(T25H8),
RcS5(ZK1091) and RcS5 # 1) we find a number of repetitions
of the putative telomeric hexamer of C. elegans TTAGGC (as
defined by C. Wicky by means of Bal31 experiments, personal
communication). This finding accounts for the results obtained
in the Southern experiments with the panel of the free living
nematode species: in fact, the probe that only recognises
C. elegans DNA was an Eco RI—Sac I fragment from the
RcS5(ZK1091) element containing only RcS5 sequence, while
the other probe, detecting sequences in all the nematode species,
was a Sacl-Hindm fragment encompassing the 3' end of
RcS5(ZK1091) and entering the flanking region for 98 bp. We
also hybridised the filter shown in figure 4, bottom pannel, with
a oligomer containing 8 tandem repetitions of the TTAGGC
hexamer, and we detected the same pattern obtained with the
Sacl-Hindin probe including the flanking sequence of
RcS5(ZK1091) (data not shown). Evidence for the occurrence
of intrachromosomal telomeric repeats in C. elegans has been
obtained in other laboratories (Donna Albertson, Yuko Kozono,
and Chantal Wicky, personal communication). Sequence
comparison between this 98bp fragment flanking RcS5(ZK1091)
and the sequence of the cloned Ascaris telomere (17) is shown
infig.6 (top).
Distribution of interstitial telomeric repeats in the C. elegans
genome
Transposable elementsflankedby telomere-like repeats have been
isolated in several protozoa (22, 23). Could the occurrence of
intrachromosomal telomeric hexamers be closely linked to RcS5
elements? We mapped the interstitial telomeric repeats making
use of the same indexed filters already used for the physical
mapping of the RcS5 and Cerep3 families. The filters were
19
29
39
49
59
*e«Tr«t«CTT»«C*T»Ata«e«CTatCTTCA6tCnT«CTTAA
i ii III i n n ii i
3'RcS5
AicTel
6*
6*
7»
89
99
3'RcSS TCnA«CTTM(TTTT»C<CrTUCTTT»etCTr»-«CTTU6CTT
«TnAt«TrAS«Tr»CtCn»<«CTTX«CCTT»66CTT»MCn»6<CTT»«CnA«CTT»CCCnX«CTT
St
99
199
119
129
139
149
159
11
3'KcS5
Loor€p
2S
39
49
AGCCTTAKCTTAUCATAAGCTUM
CTCTGCTTCACCCTTTG
II III Illlll I II III III
Ml Illl I Illl I
TT»TCCTT»CtCTTAetCTTA6tTTCTTC(«CTnAT6CTTAMCTT»MCrrACCTTTaCC6CTTT»T6aTA6
2»
231
244
25»
2»
279
2M
it
(4
7»
M
99
3'RcS5 CCTTAATCTO«CTTUeTTOC«TTAACTTrAStCTTA-ACTTAAeCrT
I I I
1 1 I I I 11 11 I 1 1 1 I I I I I
I Illllllll
Illl Illl
Loortp CTTCATSCnAS6Cn»etCTTAttCTT«6Cn»6tCTTA«CTTA«CTTA«i 11IIIIWCTTTATCCTTACS
29)
3*8
31t
32»
339
3*»
35»
36»
Figure 6. Sequence similarity between RcS5 flanking regkm and nematode
telomeres. The figure shows the alignment between the 98bp laying next to
RcS5(ZKl(#l)eleniertandtheA$cam/um&rice«ii3teta
flanking
this and the other RcS5 dements we sequenced we found a number of perfect
telomeric hexamers interspersed with degenerated copies. In the bottom part of
the figure sequence alignment between the same 98bp sequence and a Loa Loa
repetitive DNA sequence is also shown (from 30).
hybridised with a terminal labelled 48mer containing eight tandem
repeats of the TTAGGC hexamer. Interstitial telomeric repeats
once again turned out to be clustered in the terminal 30% of the
chromosomes (see bold dots in figure 2). They appear to be more
abundant than the repetitive elements belonging to RcS5 and
Cerep3 families, although they are often found associated either
with Cerep3 or with RcS5 or with both, in feet cosntids containing
RcS5 and/or Cerep3 often hybridise also with the telomeric probe.
There is a positive correlation between enhancement of crossing
overs in this part of the genome and frequency of interstitial
telomeric repeats (figure 7).
1138 Nucleic Acids Research, 1993, Vol. 21, No. 5
11
1
(IV)
TCWO)
dpy-U
• •1
1 mm mill it
r
I
/1 /
/ / »i t-1
/ / •OTI2
// // • •p-7
II
Mp-7
;i Tt
col-/
ij/l
coff-1
/tar/2
1 KMI
l|
i
ll *
-I
c
/
PI
11 (
HI |
TS4
/$?
l l 1 VU
flit 11 rpf-1
1
rti-
d (
;
;C
m
Kit
[-
SO
•t
II »P3*
H
1BC
I I I TCC
13
1 *.Vr
I'."
ID* •'
cfib •
JHn c
Inll
lift HI1
|
o
ll '•'
IH ctd
1 •
1* '<f'"f
flni
(BKSLI
J B «P
•I
S|
Illrsn-l(
SPA AC / I
1 >H*1 7
ll
CAM/A
I
1
• tktO
tpyf 10
5
0
«
• 1
7J
• 1
m
P7I.7, 1 htp-1
«p-34
• P7 1, 73
• P7 J
Figure 7. Recombination frequencies and repetitive sequences distribution along the chromosomes. In this figure we compare the genetic map, the physical map
and the RcS5 and Cerep3 element distribution of chromosomes n and chromosome IV. As in figure 1, the scale used for both maps is expressed in percent of the
chromosome length. On the genetic map, top, all the genetic loci (unspecified) are shown as vertical bars. The bolded box represents the region named 'gene cluster'
and the bare terminal portions of the chromosomes are evident. Slanted lines point at corresponding loci in the genetic map and physical map underneath. Variations
in the frequency of recombination account for the discrepancies between genetic and physical distances. The RcS5 elements and the Cerep3 elements are reported
as small vertical rods in the second and third row respectively, pseudo-telomeric repeats (TTAGGC) are represented by round bold dots under the physical map.
(Physical map version 920628.)
DISCUSSION
The work reported in this paper is directed at defining the
distribution of repetitive elements in the terminal portions of the
chromosomes of C.elegans. We were particularly interested in
these regions because of their involvement in recombination and
chromosome stability. In fact, recombination frequencies vary
along these chromosomes being higher in the proterminal
portions. C.elegans chromosomes are holocentric (also found in
other nematodes, in some insects and plants). Kinetocores are
not visible during meiosis I, the two main functions of the
centromeres (attachment to the spindle and holding the bivalent
together) being adopted by the chromosome termini (Albertson
personal communication). One end of the chromosome is oriented
towards the spindle pole while the other guarantees the
maintenance of the bivalent (probably by terminalization of the
chiasmata). Notably, it is not always the same end to be used
for each function (Albertson personal communication). Some
mutations, that affect meiotic disjunction, have been shown to
have recombination abnormalities suggesting that exchange may
be necessary for proper disjunction in C.elegans as well as in
other organisms (24, 25). We observe a positive correlation
between abundance of interstitial telomeric repeats, and
enhancement of recombination along the chromosomes (see figure
7). Telomeric repeats have been shown to be highly
recombinogenic in other organisms (26, 27, 28, 29). In
Saccharomyces cerevisiae , for example, telomeric repeats moved
inside the chromosome create a hot-spot of recombination (Petes,
personal communication). This observation suggests that the
occurrence of stretches of degenerate telomeric hexamers,
invading the last 30% of the chromosomes in C. elegans , might
account for the increase in the recombination frequency in the
'chromosome arms'. In this perspective these DNA elements
might play an important functional role in chromosome
disjunction favouring the formation of sub-terminal chiasmata
during meiosis. Middle repetitive DNA families, containing
several TTAGGC repeating units, have been detected in Ascaris
lumbricoides, where these DNA elements are completely
eliminated from the somatic genome during the process of
chromatin elimination (17), and in the filarial species Loa Loa
(30). Klion et al. (30) claim that such repeated DNA is a speciesspecific diagnostic probe. Under our stringency conditions as little
as 8 tandem duplications of the TTAGGC repeat cross-hybridise
with all nematode species analysed. Sequence similarity between
C.elegans interstitial telomeric repeats and Loa Loa repetitive
DNA sequence is shown in figure 6 (bottom).
We also observed that interstitial telomeric repeats are
associated in C. elegans with bonafide repetitive DNA elements.
The RcS5 and Cerep3 families are also preferencially located
in the proterminal last 30% of the chromosomes. Elements
belonging to the Cerep3 family have been shown to direct
replication and correct segregation in Saccharomyces cerevisiae
, stabilising yeast plasmids during mitosis and meiosis and also
lowering the plasmid copy number (8). Taking into account the
non-uniform distribution of Cerep3 elements in C.elegans , it
is suggestive to imagine that they might play a role in segregation
in this organism.
Internally located telomeric sequences in the germ-line
chromosomes of Tetrahymena mark the ends of transposon-like
elements (31). The authors suggest that transposition of these
elements involves a linear form and that telomeric repeats are
added to the ends of these free linear elements. Transposons
bounded by C4A4 telomeric repeats were also found in
Oxytricha jallax (22) and shown to excise precisely during
macronuclear development (23). RcS5 might be a vestigial mobile
element of the kind observed in protozoa and the cause of
interspersion of telomeric repeats in the C.elegans 'chromosome
arms'. Once moved inside the chromosome, telomeric repeats
might have been furtherly amplified by gene conversion, unequalcrossing over or DNA polymerase slippage (32, 33, 34).
Alternatively, the amplification of telomeric repeats in the
chromosome arms has independently occurred and the putative
transposon has acquired the hexamer TTAGGC as preferential
target site of integration.
Nucleic Acids Research, 1993, Vol. 21, No. 5 1139
ACKNOWLEDGEMENTS
We are grateful to Clara Frontali, Alan Coulson and John Pulitzer
for stimulating discussions and helpful suggestions concerning
the manuscript. We thank Donna Albertson, and Chantal Wicky
for suggestions and for sharing informations prior publication.
We thank Scott Emmons for the Cerep3 plasmid, and
L.W.Nawrocki for non-C.elegans nematode DNAs. Some
nematode strains used in this work were provided by the
Caenorhabditis Genetics Center, which is funded by the NIH
National Center for Research Resources. This work was
supported in part by Commission of European Communities
Contract SC1*-CT91-0633, CNR Progetto Speciale
Bioinformatica and CNR Target Project Biotechnology and
Bioinstrumentation.
REFERENCES
1. Sulston, J.E., Du, L., Thomas, K., Wilson, R., Hillier, L., Staden R.,
Halloran, N., Green, P., Thierry-Mieg, Qiu, L., Dear, S., Coulson, A.,
Craxton, M., Durbin, R., Berks, M., Metzstein, M., Hawkins, T.,
Ainscough, R., Waterston, R. (1992) Nature, 356,37-41.
2. Edgley, M.L. and Riddle, D.L. (1990) in S.J. O' Brien (ed.), Genetic Maps
Cold Spring Harbor Laboratory.NY, vol. 5, pp. 3.111-3.133.
3. Waterston, R., Martin, C , Craxton, M., et al., (1992) Nature Genetics,
1, 114-123.
4. Albertson, D.G.and Thomson, J. N. (1982) Chromosoma, 86, 409-428.
5. Herman, R.K., Madl, J.E., Kari, C.K. (1979) Genetics, 92, 419-435.
6. McKim, K.S. and Rose, A.M. (1990) Genetics, 124, 115-132.
7. Naclerio, G., Cangiano, G., Coulson, A., Levitt, A., Ruvolo, V., and La
Volpe, A. (1992) J. Mol. Biol., 226, 159-168.
8. Felsenstein, K. M., and Emmons, S. W., (1988) Mol. Cell. Biol., 8,
875-883.
9. Sulston J.S., and Hodgkin J. (1988) in W.B. Wood (ed.) The Nematode
Caawrhabdah elegans Cold Spring Harbor Laboratory. NY, pp 587-606.
10. Gibson, T.J., Rosenthal, A., Waterson, R.H., (1987) Gene, S3; 283-286.
11. Coulson, A., Waterston, R. Kiff, J., Sulston, J., Kohara, Y. (1988) Nature,
335 ,184-186.
12. Coulson, A., Kozono, Y., Lutterbach, B., Shownkeen, R., Sulston, J.,
Waterston, R. (1991) Bioassay, 13, 413-417.
13. Cangiano G., Ameer H., Waterston R. and La Volpe A. (1990) Nucl. Acids
Res., 18, 5077-5081.
14. Southern, E. M. (1975) J. Mol. Biol., 98, 503-517.
15. Feinberg, A. P. and Vogelstein, B., (1983) Anal Biochem., 132, 6 - 1 3 .
16. Feinberg, A. P. and Vogelstein, B., (1984) Anal. Biochem., 137, 266.
17. MOlkr, F., Wicky C , Spicher A., Tobler H. (1991) Cell, 67 815-822.
18. Sanger, F. and Coulson, A.R. (1975) J. Mol. Biol., 94, 441-448.
19. Sulston, J., Mallett, F., Staden, R., Durbin, R., Horsnell, T., Coulson, A.
(1988) CABIOS, 4, 125-132.
20. Devereux, J., Haeberli, P., and Smithies, O. (1984) Nucl. Acids Res., 12,
387-395.
21. Collins, J., Saari B., and Anderson P. (1987) Nature, 328, 726-728.
22. Herrick G., Cartinhour S., Dawson D., Ang D., Sheets R., Lee A. and
Williams K. (1985) Cell, 43, 759-768.
23. Hunter D. J., Williams K., Cartinhour S. and Herrick G.,(1989) Genes and
Dev. 3,2101-2112.
24. Hodgkin, J., Horvitz, H.R., and Brenner, S. (1979) Genetics, 91, 67-94.
25. Murray, A.W. and Szostak, J.W. (1985) Ann. Rev. CeU. Biol. 1, 289-315
26. Horowitz, H., Thorburn, P., and Haber, J. E. (1984) Mol. CeU. Biol., 4,
2509-2517.
27. Pays, E., and Steinert M. (1988) Ann. Rev. Genec, 22, 107-126.
28. Corcoran, L.M., Thompson, J.K., Walliker, D., and Kemp, D J . (1988)
CeU, 53, 807-813.
29. Puita, A.F., and Zakian, V.A. (1989) Nature, 337, 429-433.
30. Klion, A.D., Raghavan, N., Brindley, P J . and T.B. Nutman (1991) Mol.
Biochem. Parasitol., 45, 297-307.
31. Cherry. J. M. and Blackburn E. H. (1985) CeU, 43, 747-758.
32. Dover, G. (1982) Nature, 299, 111-117.
33. Dover, G. (1987) J. Mol. Evol. 26, 4 7 - 5 8 .
34. Henderson, S.T., and Petes, T.D. (1992) Mol. CeU. Biol. 12, 2749-2757.