ABSTRACT
Elimination of internal eliminated sequences (IES) during macronuclear
development of the hypotrichous ciliate
Stylonychia lemnae
was analyzed in one cluster of macronuclear precursor DNA sequences. The results
indicate that IES elimination is a highly ordered process, it starts very early
during macronuclear development and has only finished immediately before DNA
fragmentation takes place. It occurs in distinct steps and the IES are
eliminated in a specific order, where a defined IES is only removed after
complete elimination of other IES. Transfection experiments clearly demonstrate
that the structure of the IES itself is not sufficient for its correct excision
but other
cis
-acting sequences or additional structural requirements are needed for IES
elimination.
DNA elimination processes are frequently observed in differentiating eukaryotic
cells. They may concern either whole chromosomes such as in differentiating
soma cells of
Sciara coprophila
(
1
) or only parts of chromosomes as observed for example in nematodes or
Cyclops
(
2
,
3
). In addition, excision of defined DNA sequences are described. The best
studied example being the processing of the immunoglobulin and T-cell receptor genes (
4
). The mechanisms of these elimination processes are not completely understood
but it is very likely that recombination processes are involved in most cases (
5
,
6
), similar to the switching of the MAT locus in yeast (
7
) or antigen variation in trypanosomes (
8
).
An extreme form of DNA elimination is observed during macronuclear development
of hypotrichous ciliates where up to >95% of micronuclear DNA sequences are
eliminated. After sexual reproduction of these cells a new macronucleus is
formed from a micronuclear derivative in a complex nuclear differentiation
process. The results of this macronuclear differentiation are the elimination
of micronuclear specific DNA sequences, the specific fragmentation of the
macronuclear genome into short gene-sized DNA molecules and the specific amplification of these molecules (
9
). Macronuclear precursor sequences occur in the micronuclear genome in
clusters, each cluster being separated by long spacer regions (
10
). Very often macronuclear precursor sequences are interrupted by short non-coding DNA sequences (IES, internal eliminated sequences) or long
transposon-like elements (
11
-
14
). Moreover, in some precursor sequences the different exons are scrambled
requiring extensive elimination and recombination events for the creation of a
functional macronuclear gene (
15
-
17
). Both IES and transposon-like elements are excised from the chromosome in the form of DNA circles (
18
). A coordinated en masse excision of transposon-like sequences followed by IES excision prior to polytene chromosome
formation was described in
Euplotes
(
19
,
20
). Models for the elimination of IES and transposon-like elements have been proposed: excision involves staggered cuts in the
chromosome that are subsequently filled in and religated (
21
,
22
). However,
cis
- or
trans
-acting factors involved in these elimination events have not yet been
described.
Recently, we have constructed a DNA vector which carries two macronuclear
precursor sequences flanked by micronuclear specific sequences. The precursor
sequences present on this vector are interrupted by five IES. When injecting
this vector into the developing macronucleus the macronuclear precursor
sequences were correctly fragmented and telomeres were added
de novo
but we did not observe elimination of the IES (
23
). Therefore, we decided to analyze the excision of the IES in this gene cluster
in vivo
in more detail. In this report, we show that elimination of IES starts very
early during macronuclear development and has finished immediately prior to DNA
fragmentation. Moreover, the IES in a given precursor sequence or sequence
cluster are excised in a defined order suggesting a highly ordered elimination
of IES during macronuclear development.
The hypotrichous ciliate
Stylonychia lemnae
was grown in neutral Pringsheim solution and fed daily with the algae
Chlorogonium elongatum
(
24
). To achieve conjugation, cells of two different mating types were mixed and
kept at room temperature; usually 60-90% conjugation was observed the next morning. The different stages of
macronuclear development were determined by phase contrast microscopy and cells
in the same developmental stage selected by the size of the macronuclear
anlagen.
In 6 h intervals after separation of conjugating cells 20 cells of the same
developmental stage were collected under the microscope and the DNA was
isolated as described by Wen
et al.
(
23
) and dissolved in 10 [mu]l TE. Aliquots of 2 [mu]l were used for one PCR reaction. PCR was done as described by Saiki
et al.
(
25
). Primers used for PCR analysis and PCR reaction conditions are indicated in
Figure
1
. For one PCR reaction 250 ng of the corresponding primers were used. Primer P11
is derived from the 3'-region of the 1.1 kb macronuclear precursor sequence, its sequence
is 5'-GCGGGTACCATCAGATAACTAGCAAC, primer P13 is derived from the 3'-end of the 1.3 kb precursor sequence, its sequence is
5'-CAGATACAACGTCCCTCAAC, primer P1 is derived from the 5'-end of the 1.1 kb sequence, its sequence is 5'-GCTATCAATCAAGTGCTGGAGCTT, primer P3 is
derived from the 5'-end of the 1.3 kb precursor gene, its sequence is 5'-GAATATCTGAGAGTAGCAA, primer P2 derives from the 5'-end of the 1.3 kb macronuclear gene, its
sequence is 5'- GGCTCGAGTTGCTACTCTCAGATATTC, primer P5 is derived from IES4, its
sequence is 5'- CTCATTCTTATTATAATCCCATATAAGCAC, primer P9 derives from the 3'-end of the 1.3 kb macronuclear gene, its sequence is 5'-GGCTCGAGTTGCTACTCTCAGATATTC. The
following PCR program was applied: DNA was denatured at 94oC for 3 min, then in 40 cycles denaturation for 1 min at 94oC and extension for 2 min at 72oC, the last extension step was done for 10 min. Following the PCR
reaction samples were separated on 2% agarose gels, blotted onto nylon
membranes (Qiagen) and hybridized with random primed labeled probes (
26
) or probes labeled by tailing with terminal transferase (
27
) using digoxigenin-dUTP (Dig, Boehringer, Mannheim). The vector pCE5 used for injection
experiments and the injection procedure were described earlier (
23
).
The macronuclear precursor sequences analyzed were originally isolated from a
genomic micronuclear DNA library. A homology to two macronuclear DNA molecules
with sizes of 1.1 and 1.3 kb were found on a 3.5 kb
Eco
RI fragment. All sequences required for the formation of the 1.3 kb macronuclear
DNA molecules are present on this 3.5 kb
Eco
RI fragment, but 100 bp of the 1.1 kb macronuclear DNA molecule are lacking;
they must be separated from the rest of the precursor sequence by a very long
IES or transposon (
14
). Two IES (IES 1 and IES 2) with sizes of 41 and 31 bp are found in the 1.1 kb
precursor sequence and three IES (IES 3-5) with sizes of 10, 60 and 70 bp are found in the 1.3 kb macronuclear
precursor sequence. The two precursor sequences are separated from each other by a 11 bp spacer sequence (Fig.
1
a). This 3.5 kb
Eco
RI fragment was used for the construction of a vector used for microinjection
into the developing macronucleus. These injection experiments showed that all
sequences required for correct fragmentation and telomere addition are present
on the vector DNA (
23
).
To study IES elimination from the two macronuclear precursor sequences
in vivo
a PCR analysis of the macronuclear precursor sequences was made at different stages of macronuclear development. In 6 h
intervals after separation of conjugating cells (6-42 h) 20 exconjugants at the same developmental stage, as determined by phase contrast microscopy, were isolated and the DNA prepared. Polytene chromosomes first became visible about 10 h after conjugant cell separation and were fully developed ~35-40 h after cell separation. After that time DNA fragmentation
and elimination took place. This total cell DNA which contained micronuclear DNA and DNA
from the developing macronucleus was used as a template for PCR analysis. The
sequences representing the primers P11 and P13 used for the amplification of
the 1.3 kb precursor sequence are localized downstream of IES 5 (P13) and in
the subtelomeric region of the 1.1 kb precursor sequence (P 11, Fig.
1
a). Using this primer combination only the non-fragmented precursor sequence is amplified. Therefore it can be used both
to determine the timing of IES removal and the time point of DNA fragmentation.
A similar primer combination was used for amplification of the 1.1 kb
macronuclear precursor sequence. Primer P1 binds upstream of IES 1, primer P3
in the 5' subtelomeric region of the 1.3 kb precursor gene. The PCR fragments
synthesized from the DNA isolated at different time intervals after separation
of conjugating cells were separated on an agarose gel. By hybridization with
probes derived from the original 3.5 kb
Eco
RI fragment it could be shown that these PCR fragments were derived from the
precursor sequence (Fig.
1
). Six hours after conjugants separated only the micronuclear version of the 1.3
kb precursor sequence (1.4 kb, see Fig.
1
b) was seen. Twelve hours after cell separation an additional band with a size
of about 1.33 kb was observed. In later stages (24 hours after cell separation)
a third band with a size of about 1.26 kb appeared. Twelve hours later (36
hours after cell separation) only the original 1.4 kb and the 1.26 kb fragments
were observed. Finally 42 hours after separation of conjugants only the 1.4 kb
fragment was synthesized indicating that fragmentation was completed at this
time point (Fig.
1
b). A similar result was obtained with the 1.1 kb macronuclear precursor
sequence. Up to 6 hours after conjugants separated, only the PCR fragment
synthesized from the micronuclear version of the precursor sequence was
observed. At 12 h after cell separation an additional, ~30-40 bp smaller, PCR fragment was synthesized. Twelve hours later (24
hours after conjugant cells separated), a third band appeared which represented the precursor sequence after complete IES elimination but before fragmentation. Finally fragmentation occurred (Fig.
1
c).
The results obtained by PCR analysis show that IES elimination starts very early during macronuclear development but only finishes just before
DNA fragmentation occurs. Moreover, IES removal does not occur simultaneously
but in distinct steps. To further demonstrate that the PCR fragments shown in
Figure
1
b and c are derived from processing intermediates and to determine whether IES
removal occurs randomly or whether there exists an order in IES elimination,
the different PCR products were hybridized with probes derived from the
different IES and the probes used in Figure
1
b and c. Figure
1
d and e show the hybridization to the PCR products synthesized 24 h after cell
separation, i.e. the stage where all the different PCR products were
synthesized. In the case of the 1.3 kb precursor sequence the IES 4 probe
hybridized to two fragments while the IES 5 probe hybridized only to the
original unprocessed precursor sequence. As expected, none of the IES probes
hybridized with the smallest PCR product (Fig.
1
d). This clearly demonstrates that the first IES to be removed from this
precursor sequence was IES 5 followed by IES 4. In no case was a hybridization
of IES 5 observed to the intermediate processing product (1.33 kb) demonstrating that IES 4 was only excised after complete elimination of IES 5 (Fig.
1
d). Due to the small size of IES 3 it is not possible to determine the exact
timing of IES 3 elimination, however according to the estimated sizes of the
PCR fragments it is very likely that it is removed after IES 5 elimination. The
same experiment was done for IES 1 and IES 2 elimination in the 1.1 kb
macronuclear precursor sequence. In this case IES 2 seemed to be removed first
(Fig.
1
e).
We investigated whether transfection experiments could provide further information about the time course and mechanisms of IES elimination.
The vector pCE5 containing the 1.1 kb and 1.3 kb macronuclear precursor
sequence, from which the 1.3 kb macronuclear sequence was modified by the
insertion of a polylinker region (Fig.
2
a, ref.
23
), was injected at the same time intervals after cell separation as in the
studies described above. After injection, cells were allowed to complete
macronuclear development, DNA was then isolated from vegetative cells and used
for PCR analysis of the 1.3 kb gene. Primers used for this analysis are
indicated in Figure
2
a. Using the primer combination P2/P9 two PCR fragments were synthesized from
pCE5 injected cells at any time point during macronuclear development; while
the 1190 bp fragment was synthesized from normal macronuclear DNA the 1830 bp
fragment was synthesized from the modified 1.3 kb macronuclear DNA molecule.
Using the P2/P5 primer combination a 1200 bp fragment, identical to that
synthesized from pCE5, was observed. Sometimes a smaller PCR fragment with a
size of 730 bp appeared which was probably synthesized from micronuclear DNA.
While correct fragmentation and telomere addition occurred (
23
, data not shown), in no case, even when the vector was injected at a stage
before IES excision was observed, were the IES eliminated from the injected
sequence demonstrating that the structure of the IES itself is not sufficient
for correct excision of IES (Fig.
2
b).
Internal eliminated sequences (IES) are frequently found in macronuclear
precursor sequences in hypotrichous ciliates. They are small unique sequences
with sizes between 10 to several hundred base pairs and are very A+T rich. All
IES are bordered by 2-6 bp direct repeats. In
Euplotes crassus
they all include the dinucleotide 5'-TdA-3' (
9
,
11
,
14
,
18
,
28
). The excision of IES has been extensively studied in
Euplotes crassus
. IES elimination is found in a short time range between the end of polytene
chromosome formation and DNA fragmentation (
19
,
28
). A special class of IES, the TelIES (telomeric repeat like IES), are removed
significantly later than the other IES, just prior to DNA fragmentation (
29
). However, the excision time and order of IES within a precursor sequence or
sequence cluster has not yet been analyzed. IES excision seems to be a form of
recombination process in which at least the larger IES are eliminated in a
circular form with a heteroduplex junction (
21
). Transposase-like enzymes may be involved in this elimination process as suggested for
transposon-like elements in hypotrichous ciliates which are also eliminated during
macronuclear development (
9
,
30
). According to this model the sequence organization of the IES itself is sufficient for the precise excision of these sequences. In
Tetrahymena
experimental evidence was found that sequences flanking the IES are required for
correct deletion of sequences (
31
,
32
).
We have analyzed IES elimination from one gene cluster of the hypotrichous ciliate
Stylonychia lemnae
. The analyzed chromosomal region comprises the complete precursor sequence required for the formation of
a 1.3 kb macronuclear DNA molecule and for 1 kb of a 1.1 kb macronuclear DNA
molecule (
14
). Three IES are found in the 1.3 kb and two IES are found in the 1.1 kb
precursor sequence. To study the time course of IES excision a PCR analysis at
different time points of macronuclear development was made from DNA isolated
from 20 cells at approximately the same developmental stage. As in other
hypotrichous ciliates, such as
Euplotes
and
Oxytricha
, IES elimination occurs prior to DNA fragmentation in
Stylonychia
. But it starts very early during macronuclear development at the beginning of
polytene chromosome formation and has only finished just prior to DNA
fragmentation. Moreover, it seems to occur in distinct steps: in early stages
during macronuclear development only one IES per precursor sequence is removed.
In a later stage a mixture of PCR fragments are observed: they are synthesized
from precursor sequences from which one or more IES are eliminated. Using our
approach we cannot decide whether this mixture of differently processed
precursor sequences occurs within one cell or whether it is due to the fact
that the selected 20 cells were not at a completely identical stage of
macronuclear development. Later, as expected only the original micronuclear
version and the precursor sequence after IES-elimination are seen. Hybridization of probes derived from the different
IES to the different PCR products clearly demonstrate that the IES in a given
gene are excised in a defined order. Moreover, these experiments provide strong
evidence that only after complete elimination of one IES the other IES are
removed; either the complete removal of one IES is a necessary prerequisite for
the excision of the other IES or it may be a consequence of the IES excision
machinery. It may well be that such an order exists not only in one gene but
also in a whole gene cluster. When a vector containing the precursor sequences
under investigation was injected at time points before and during IES
elimination, IES excision from the injected DNA was not observed. Therefore,
the structure of the IES is not sufficient for correct excision of these sequences, either integration into the the chromosomal context, assembly into the correct chromatin structure
or other
cis
-acting sequences not present on this vector are necessary for IES
excision.
Sequential excision of DNA sequences is well documented in the formation of
immunoglobulin genes and T-cell receptor genes (
4
-
6
). While the mechanism of DNA sequence excision is basically understood (
33
,
34
), very little is known about the factors determining the timing and order of
sequence deletion. The sequential excision of IES during macronuclear development relies on the
highly ordered DNA deletion processes during immunoglobulin gene formation. To
date there exists no experimental data that the structure of the IES, even the direct repeats bordering the IES,
are important for precise excision. In contrast, in
Tetrahymena
evidence was provided that distinct sequences flanking the IES are required for
correct elimination (
31
,
32
). None of the IES analyzed in this study showed any sequence pecularity which
could explain the time and order of elimination. Speculations could be that the
initation of IES excision starts at a sequence located outside the precursor
sequence followed by branch migration to the IES as described for recombination
processes (
35
); alternatively there are specific interactions between the IES and some other
cis
- or
trans
-acting factors as suggested for
Tetrahymena
(
31
). Finally, the question about the possible biological function of this ordered
excision process arises. It could be assumed that such an order is necessary
for the formation of functional macronuclear genes. This seems obvious in the
case of scrambled genes where rearrangement of the different exons during
macronuclear development must occur in a defined order (
15
-
17
). Sequential deletion of DNA sequences therefore might be an old and
generalised mechanism of processing eukaryotic genes. Since in ciliates a
molecular analysis of single cells in a very defined developmental stage can be
made and a very efficient transfection procedure for single cells exists (
23
), they may provide a useful system for the analysis of programmed DNA
rearrangement processes.
This work was supported by the Deutsche Forschungsgemeinschaft and the Alfred
Krupp von Bohlen und Halbach foundation.
*To whom correspondence should be addressed. Tel: +49 2302 669 144; Fax: +49
2302 669 220; Email: lipps@natwi.natwi.uni-wh.de
REFERENCES
Return