ABSTRACT
We have cloned and characterized the
Dictyostelium discoideum
rep
E gene, a homolog of the human xeroderma pigmentosum (XP) group E gene which
encodes a UV-damaged DNA binding protein. The
rep
E gene maps to chromosome 4 and it is the first gene identified in
Dictyostelium
that is homologous to those involved in nucleotide excision repair and their
related XP diseases in humans. The predicted protein encodes a leucine zipper
motif. The
rep
E gene is not expressed by mitotically dividing cells, and
rep
E mRNA is first detected during the aggregation phase of development when the
cells have ceased dividing and replicating genomic DNA. The mRNA level plateaus
by the time the developing cells have entered multicellular aggregates and remains at the same steady-state level for the remainder of development. In addition, we have
demonstrated that the level of mRNA is very low in developing cells. These
observations suggest that
rep
E may play a regulatory role in development. The data indicate that potential
developmental roles for XP-related genes can be profitably studied in this system.
Xeroderma pigmentosum (XP) is a rare autosomal recessive disease that is marked
by extreme sensitivity to UV light and a high predisposition to skin cancer in
light-exposed areas of the body (
1
). Human XP cells are defective in nucleotide excision repair (NER). They are
divided into eight complementation groups (A-G and a variant group) which differ from each other in the degree of
their repair deficiency. The genes which correspond to the various XP groups
represent different genes whose products are involved in the repair process (
2
). These XP-related genes (named
ERCC
, for excision repair cross-complementing) were identified by their ability to cross-complement a series of UV-sensitive rodent cell lines which were generated by
mutagenesis and have been shown to comprise at least 11 complementation groups
(
2
).
Many XP genes have been shown to share considerable homology with the yeast
RAD
genes and with
Drosophila
homologs of the XP genes (
3
). These homologies with the yeast
RAD
genes provided the initial clues for the repair activity of these genes and
imply that the repair mechanism is highly conserved across these species.
However, it is clear that significant differences in the repair mechanisms
exist between species and each gene has unique additional functions in
different organisms.
Of all the identified XP complementation groups, XPE is the least understood
with respect to its role within the repair pathway. The NER defect in some XPE
cell lines has been correlated with the absence of activity of the UV-DDB (UV-damaged DNA binding) protein. The cognate monkey and human genes
have been cloned and sequenced (
4
-
6
). A yeast homolog has not been identified and no correlation to an existing
ERCC
gene has been made. The XPE protein has not been identified as part of the
repairosome holoenzyme (
7
). However, recent work in which mammalian NER was reconstituted from purified
protein components showed that XPE/UV-DDB plays an accessory role in NER (
8
).
One perplexing problem about XP patients is that even though their genetic
defects are each in single repair genes, they exhibit a wide range of
phenotypes, such as predisposition to cancer, stunted growth and neurological
and developmental abnormalities (
1
). In yeast, some of the XP homologs are actually essential for viability (
9
,
10
). The underlying reason for this pleiotropy is that these repair genes are
involved in multiple processes in addition to NER (
11
). Such processes involve transcription (
7
,
12
-
18
), DNA replication (
19
-
22
) and mitotic (
23
) and meiotic (
24
) recombination.
The involvement of NER gene products in transcription and other general
processes in the cell begins to explain the wide range of phenotypes that are
associated with XP. However, it does not explain why phenotypes which are
associated with defects in such fundamental processes are not more severe and
global. This consideration raises the possibility that these genes may be
affecting specific developmental decisions. Identifying the nature of these
affected processes remains one of the major challenges of this field (
25
). Clearly these problems will benefit from studies on model developmental
systems.
In this paper we report the identification and isolation of
rep
E, the
Dictyostelium
homolog of the primate
UV-DDB
gene.
rep
E is the first XP- or NER-related gene identified in this developing system. We demonstrate
that expression of the
rep
E gene is developmentally regulated. The gene is not expressed in vegetatively
growing cells, but
rep
E mRNA accumulates in the early stages of development. The protein has a leucine
zipper motif, suggesting a role in gene regulation.
A cDNA library in [lambda]Zap and a genomic library in pBLUESCRIPT were gifts from Herbert Ennis (Columbia University). The
rep
E gene was serendipitously identified as part of our studies on the discoidin
genes of
Dictyostelium
. Two positive overlapping cDNA clones (see Results and Fig.
1
B) were identified in an antibody screen of >75 000 phage from the [lambda]Zap expression library while we were attempting to clone the discoidin
II gene, using a specific polyclonal antibody as a probe. The sequence of these
clones did not match any of the peptide sequences we had obtained for the
discoidin II protein, but did reveal their homology to the primate
UV-DDB
/
XPE
gene. We have never explained this aberrant antibody reactivity. The antibody
reacts with a single band on Western blots and it is possible that the cross-reactivity is at the level of secondary structure. The two cloned cDNA
fragments were used to re-screen both the [lambda]Zap and genomic libraries using standard procedures (
26
).
Genomic DNA cut with
Nde
I or
Bcl
I was size fractionated on a sucrose gradient. The DNA fragments from part of
each fraction were separated by electrophoresis and blotted (
27
). The blots were probed with a 5' fragment of the
rep
E cDNA clone, to identify the fractions containing the rest of the gene. The DNA
from the peak fractions was cloned into the compatible vectors [pT7-7 (
28
) for
Nde
I and pUC19 for
Bcl
I] and the clones were screened by hybridization. All DNA manipulations were performed according to standard procedures (
26
). Sequencing was performed using a Sequenase kit from USB and [[alpha]-
32
P]dATP. Routinely, plasmid DNA was purified using a `Wizard prep' kit (Promega).
The
rep
E gene was sequenced in its entirety on both strands.
Strains Ax3 and DH1 were used throughout these studies. Ax3 is a derivative of
the wild-type strain NC4, which is capable of growth on axenic medium. DH1 is a
derivative of Ax3 with a deletion in the
pyr5-6
(UMP synthase) gene, so that it can be used in homologous recombination
experiments with selection for uracil prototropy (a gift from R.Insall). For
this work, the cells were grown on SM agar plates in association with
Klebsiella aerogenes
as a bacterial food source. When the
Dictyostelium
cultures reached mid log phase (2-3 * 10
8
cells/100 mm plate) they were harvested and the remainder of the bacteria were
washed away by differential centrifugation. The cells were plated on buffer-soaked black paper filters for development (10
8
cells/40 mm filter) (
29
). Development under these conditions is completely synchronous. When the cells reached the desired developmental stage, they were
harvested by vortexing the cells off the filter. The cells were then pelleted
and rapidly frozen. Total RNA was isolated at each time point as described (
30
) and used for Northern analyses or reverse transcription-PCR.
First strand cDNA synthesis was performed using a Life Technology (Gibco BRL)
Superscript Preamplification system. For each time point, 2 [mu]g of total RNA sample was treated with DNase I in a final volume of 20 [mu]l 50 mM KCl, 20 mM MgCl
2
. Aliquots of 2.2 [mu]l of the DNase I-treated samples were used for the reverse transcription (RT) reaction,
following the instructions of the manufacturer and using an oligonucleotide
covering position 3473-3488 (3' -> 5', oligo 6) as a primer. Aliquots of 2 [mu]l of each sample were then amplified, using oligo
6 and an oligonucleotide covering position 2711-2726 (5' -> 3', oligo 11) as primers. PCR was performed using a Coy
thermocycler for 25 cycles under the following conditions: 1 min denaturation
at 94oC, 2 min annealing at 40oC and 1 min 45 s elongation at 72oC. Each reaction contained 2 [mu]l 20 [mu]M primers, 10 [mu]l 2.5 mM dNTP and 1 [mu]l
Taq
polymerase (AmpliTaq; Perkin Elmer) in buffer containing 20 mM Tris-HCl, pH 8.4, 50 mM NaCl and 1.5 mM MgCl
2.
Results were analyzed on 1.2% agarose gels in TBE.
The entire cloned genomic sequence was examined for introns using a reverse
transcription-coupled PCR method as described (
31
). Briefly, cDNA was made from mRNA by reverse transcription. The cDNA was then
amplified using overlapping sets of oligonucleotide primers which covered the
entire length of the cloned sequence. A similar series of amplifications was
performed on the genomic clone and the size of the equivalent PCR products were
compared on agarose gels.
Two overlapping cDNA clones (Fig.
1
B, clones pLS1 and pLS2 of 824 and 845 bp, respectively) were identified in an
antibody screen of a [lambda]Zap cDNA library (see Materials and Methods). The sequence of these
clones revealed homology to the primate
UV-DDB/XPE
gene. Keeping with standard
Dictyostelium
gene nomenclature (
32
), we have named the gene
rep
E (
We constructed a restriction map of the region surrounding the
rep
E gene using the pLS3
rep
E clone as a probe (Fig.
1
B). A 2.3 kb
Nde
I fragment which extends 1.6 kb from the 5'-end of the previously cloned genomic sequence seemed to be long
enough to cover the rest of the gene, based on homology to the primate
UV-DDB
gene. A genomic
Nde
I digest was fractionated on a sucrose gradient and the appropriate size
fraction was used to construct a mini-library, from which the 2.3 kb fragment was cloned. Sequence analysis of
this fragment (clone pSKSL2) revealed that it was still missing the 5'-end of the gene due to the presence of an intron. We therefore
repeated this procedure, cloning an overlapping
Bcl
I fragment (clone pSKSL3). This clone contained the 5'-end of the
rep
E coding sequence, 585 bp of 5' upstream region and most of the neighboring upstream gene.
The genomic sequence of the
rep
E gene is 3512 bp long (Fig.
2
). There is a single 92 bp intron at position 2568 which was revealed by
comparing the genomic and cDNA sequences. The absence of additional introns
within the genomic sequence, for which we did not have corresponding cDNA
clones, was established by comparing PCR products of the genomic segment to PCR
products of RT cDNA from the same region (
31
).
The predicted protein sequence of
rep
E shares a 40% identity and over 60% similarity with the primate
UV-DDB
gene (Fig.
3
). A search of the NCBI non-redundant database, using the BLAST program (
39
) has not revealed any homology between the sequence of RepE or the primate UV-DDB protein and the previously identified proteins of the NER complex
(such as the products of the yeast
RAD
genes or members of the
ERCC
family of genes). As such, the
rep
E and
UV-DDB
/
XPE
genes encode a novel, highly conserved family of proteins. A
Drosophila
homolog of primate UV-DDB (dm127) has recently been cloned. It shares 60% homology with the
primate UV-DDB protein and about 40% homology with the RepE protein (M.Takao and
M.Protic, personal communication).
The translated protein sequence of
rep
E contains a region which encodes a leucine zipper motif and an adjacent basic
region, for a putative DNA binding site (double line in Fig.
4
). This region exhibits a very high degree of sequence conservation between the
RepE, the UV-DDB and the
Drosophila
dm127 proteins (64% identity and 82% similarity), as is shown in Figure
4
. Moreover, there is a limited sequence homology to the bacterial UvrB protein,
which is involved in the recognition step of the NER pathway in prokaryotes.
This homology further highlights the functional importance of this sequence in
these proteins.
The RepE protein is expressed in the cells at very low levels. A number of
attempts to identify the mRNA by conventional Northern analysis, as well as by
RNase protection experiments, were unsuccessful. Thus, we prepared
rep
E-specific cDNA from equal amounts of mRNA from time points taken throughout
the entire developmental sequence. The cDNAs were amplified using synthetic
oligonucleotides as primers and analyzed by electrophoresis (Fig.
5
). We were able to show that
rep
E mRNA accumulates in a developmentally controlled manner. It begins to
accumulate at 3 h of development, during the onset of aggregation, peaks at 5-7 h of development, when the cells are forming discrete multicellular
structures, and stays at a relatively constant level throughout the remainder
of development. There is no measurable
rep
E mRNA during vegetative growth on bacteria.
The genes encoding the proteins involved in NER play a pivotal role in repairing
DNA damaged by UV and chemical agents (
3
). Indeed, mutations in these genes result in the pathological conditions of XP,
Cockayne's syndrome (CS) and trichothiodystropy (TTD). Some of the NER proteins play dual roles as part of the basal
transcription complex TFIIH, while others are involved in other fundamental
processes in the cell. These interactions begin to shed some light on why
individuals with these diseases often have associated developmental and
neurological abnormalities. However, it does not explain why these pleiotropic
effects are not more severe. It appears that these genes affect specific
developmental decisions and that model systems are needed for the study of the
role of these gene products in development.
This work presents the identification and characterization of the first XP-related gene in the cellular slime mold
Dictyostelium discoideum
. The organism is ideal for the study of the role of XP- and NER-related genes during development.
Dictyostelium
is a haploid single celled amoebae which grows by mitotic division, feeding on
either bacteria or axenic medium. Following exhaustion or removal of the food
source, cell division and genomic DNA replication cease and starvation signals
the beginning of differentiation. The amoebae aggregate to form cohesive
multicellular assemblies each containing 10
5
cells. These aggregates undergo a complex program of development, resulting in
spatially localized spores resting on top of a cellular stalk (see illustration
in Fig.
5
). The synchronized development allows the correlation of biochemical and
cellular events to the overall program of morphogenesis.
Our findings show that there is no measurable
rep
E mRNA in vegetatively growing cells or at the onset of development. However,
rep
E mRNA begins to be expressed during the aggregation stage of development,
increases to a maximum by the time the cells have entered aggregates and stays
at this steady-state level for the remainder of development. The level of
rep
E mRNA is very low even at its maximum during development. The pattern of
developmental expression of
rep
E is reminiscent of that reported for the
Drosophila
haywire
gene (XPB homolog), where lower levels of mRNA were found in embryos than in
the larva and the adult stage (
40
). Our finding that
rep
E mRNA accumulates during development demonstrates that the expression of the
gene is developmentally regulated.
The
rep
E gene is highly homologous to the
UV-DDB
/
XPE
gene of primates. A
Drosophila
homolog has been found but no yeast equivalent has been reported. The
Dictyostelium
gene encodes a protein with a predicted molecular weight of 127 kDa, which is
identical to that of the primate UV-DDB protein. The primate
UV-DDB
/
XPE
gene encodes a nuclear protein which preferentially binds to UV-damaged DNA (primarily 6-4 pyrimidine dimers) as well as to DNA damaged by chemical agents (
4
,
41
-
43
). There is now considerable evidence indicating that the UV-DDB protein functions as a two subunit complex of p127 and p48 (
5
). The activity of the UV-DDB protein is missing in some XPE patients (
4
,
6
,
44
-
46
). It has been suggested that cells from XPE patients which do not exhibit loss
of DNA binding activity of p127 may lack either the other subunit of the
complex or have a defect in the domain of the XPE protein which is responsible
for the protein-protein interaction, but not the DNA binding domain. Microinjection of
UV-DDB
mRNA (
47
) or purified DDB protein (
48
) have been shown to increase unscheduled DNA synthesis in XPE cells. However,
no sequence analysis of the gene from a human XPE patient has been done in
order to unambiguously correlate an aberrant gene with the XPE phenotype. The
identification of
rep
E in
Dictyostelium
may help to shed light on the role of this protein in repair and/or other
processes, as it will allow us to generate specific mutations in the gene and
analyze the resultant developmental and repair phenotypes.
The predicted
rep
E gene product contains a leucine zipper motif and associated basic region which
suggest that it plays a regulatory role and that it may be localized to the
nucleus (
49
,
50
). There is a 64% identity and 82% similarity around this region between RepE,
the primate and
Drosophila
UV-DDB sequences, indicating the functional importance of this region.
Moreover, there is a 40% homology to the
Escherichia coli
UvrB protein. The latter protein functions in a complex with UvrA (UvrA
2
B) to enable binding to damaged DNA (the recognition step in prokaryotic NER).
The similarity to this protein is very intriguing. Although it has been shown
that eukaryotic NER performs the same functions as the prokaryotic pathway,
there is little homology on the primary sequence level between the protein
components of the two NER pathways. The homology in the region of the leucine
zipper between the RepE, UV-DDB and the prokaryotic UvrB proteins suggests an evolutionary relation
between these proteins. The regulatory role of RepE suggested by the presence
of the leucine zipper is consistent with our data showing that the level of
expression of
rep
E is very low and that its expression is under developmental regulation.
The presence of leucine zippers in proteins
is often an indication that the protein functions as a dimer. A recent study
showed, using the yeast two-hybrid system, that the human UV-DDB protein, which they named X-associated protein (XAP-1), interacts with the hepatitis B virus X protein in
order to initiate viral multiplication
in vivo
(
51
). It was hypothesized that the interaction of the X protein with the DNA repair
protein may modify cellular transcription processes. This protein-protein interaction demonstrates a regulatory function for the UV-DDB/XPE protein separate from its involvement in DNA repair,
similar to many other components of the NER pathway which have dual functions.
We are now trying to identify the protein partner of the
repE
gene product in
Dictyostelium
.
The
rep
E gene product has another striking structural feature, three extended stretches
of poly(Asp). These poly(Asp) regions are absent in the primate homolog. The
function of these regions in the predicted
Dictyostelium
RepE protein is not known and they may be acting merely as spacers between
discrete functional domains of the protein. However, they may have specific
functional roles in this organism. A search of the GenBank for proteins containing poly(Asp) regions resulted in the identification of numerous proteins whose function involves interaction with DNA. Some of
these include: topoisomerase II (
52
), RNA polymerase I (
53
) and the RNA polymerase III large subunit (
54
) from
Plasmodium falciparum
; Suppressor of Zeste from
Drosophila
(
55
); the yeast DNA binding protein BAF1 (
56
) and transcription factor ADR6 (
57
); and the
Dictyostelium
RNA helicase (Mahal,B. and Nellen,W., unpublished, s49262.) and G box binding
protein (
58
). Thus the presence of the poly(Asp) regions in RepE
may further indicate that this protein is involved in interaction with DNA.
Interestingly, the poly(Asp) regions are reminiscent of the poly(Gln) regions
found in the proteins associated with neurodegenerative diseases such as
Huntington's disease. In these diseases the length of these poly(Gln) regions
increases in succeeding generations and the increase in length appears to be
responsible for the phenomenon of genetic anticipation (
59
). Moreover, there is evidence that these long poly(Gln) regions can form polar
zippers with other proteins (
60
). These newly discovered protein-protein interactions may be responsible for the pathology of these
diseases. It is possible that the poly(Asp) regions in the
rep
E gene product may be involved in protein-protein interactions similar to the poly(Gln) polar zippers. The precise
roles of the poly(Asp) regions are directly testable because of the ease with
which homologous gene replacement can be performed in
Dictyostelium
.
The results presented here show that expression of the
rep
E gene is intimately tied to development and encodes motifs that suggest that
the protein has a regulatory role. Efforts are now underway to use this system
to elucidate the precise roles of
rep
E and other XP-related genes in morphogenesis and cell differentiation in this organism.
This work is presented in partial fulfillment of the requirements for the PhD
degree by Sung-Keun Lee and Sung-Lim Yu at the University of Missouri. The work was supported by a
grant from the National Science Foundation (IBN9206891). SA is the recipient of
an American Cancer Society Faculty Research Award (FRA448). S-LY was partially supported by a University of Missouri Graduate
Fellowship. We thank Bill Loomis and Adam Kuspa for chromosome mapping the
rep
E gene. We thank Lisa Sydow and Christine Bonnin for their help in the initial
cloning experiments. We also thank Drs M.Takao and M.Protic for valuable
discussions regarding the primate and
Drosophila
UV-DDB
genes and Vince McGuire for comments on the manuscript.
REFERENCES
Return