In vitro
and
in vivo
function of the C-terminus of
Escherichia coli
single-stranded DNA binding protein
In vitro and in vivo function of the C-terminus of Escherichia coli single-stranded DNA binding protein
Ute
Curth
,
Jochen
Genschel
,
Claus
Urbanke
and
Joachim
Greipel*
Received April 1, 1996
;
Revised and Accepted May 17, 1996
ABSTRACT
We constructed several deletion mutants of
Escherichia coli
single-stranded DNA binding protein (
Eco
SSB) lacking different parts of the C-terminal region. This region of
Eco
SSB is composed of two parts: a glycine- and proline-rich sequence of
~
60 amino acids followed by an acidic region of the last 10 amino acids which is
highly conserved among the bacterial SSB proteins. The single-stranded DNA binding protein of human mitochondria (
Hs
mtSSB) lacks a region homologous to the C-terminal third of
Eco
SSB. Therefore, we also investigated a chimeric protein consisting of the
complete sequence of the human mitochondrial single-stranded DNA binding protein (
Hs
mtSSB) and the C-terminal third of
Eco
SSB. Fluorescence titrations and DNA-melting curves showed that the C-terminal third of
Eco
SSB is not essential for DNA-binding
in vitro
. The affinity for single-stranded DNA and RNA is even increased by the removal of the last 10 amino
acids. Consequently, the nucleic acid binding affinity of
Hs
mtSSB is reduced by the addition of the C-terminus of
Eco
SSB. All mutant proteins lacking the last 10 amino acids are unable to
substitute wild-type
Eco
SSB
in vivo
. Thus, while the nucleic acid binding properties do not depend on an intact C-terminus, this region is essential for
in vivo
function. Although the DNA binding properties of
Hs
mtSSB and
Eco
SSB are quite similar,
Hs
mtSSB does not function in
E.coli
. This failure cannot be overcome by fusing the C-terminal third of
Eco
SSB to
Hs
mtSSB. Thus differences in the N-terminal parts of both proteins must be responsible for this
incompatibility. None of the mutants was defective in tetramerization. However,
mixed tetramers could only be formed by proteins containing the same N-terminal part. This reflects structural differences between the N-terminal parts of
Hs
mtSSB and
Eco
SSB. These results indicate that the region of the last 10 amino acids, which is
highly conserved among bacterial SSB proteins, is involved in essential protein-protein interactions in the
E.coli
cell.
INTRODUCTION
Due to its role in DNA replication, repair and recombination, single-stranded DNA binding protein of
Escherichia coli
(
Eco
SSB) is essential for survival of the cell (
1
-
3
).
Eco
SSB consists of 177 amino acids and forms a very stable homotetramer (
4
). Secondary structure prediction indicates that the sequence of
Eco
SSB can be divided into two parts, an N-terminal domain (~120 amino acids) rich in [alpha] helices and [beta] sheets and a more or less unstructured C-terminus (
5
). The N-terminal two thirds contain the DNA-binding domain (
6
). The function of the C-terminal third is not yet characterized in detail. The importance of the C-terminus is illustrated by the
ssb113
mutation (
7
), in which the penultimate residue of the
Eco
SSB protein (proline 176) is replaced by serine (
8
). This mutation results in a UV- and temperature-sensitive phenotype. The DNA-binding affinity of the mutated protein is only slightly
affected (
8
). Williams
et al
. (
6
) showed that the C-terminal third of
Eco
SSB becomes more accessible to proteolysis after DNA-binding. Removing the C-terminal fragment resulted in a protein with higher affinity towards single-stranded DNA.
The C-terminal third of
Eco
SSB consists of two regions: a sequence of ~50 amino acids rich in proline and glycine residues is followed by an
acidic region of 10 amino acids containing four aspartate residues. In contrast
with the proline- and glycine-rich sequence, the region of the last 10 amino acids is highly conserved among procaryotic SSB proteins (
1
).
To characterize further the functions of the C-terminal third of
Eco
SSB, we constructed several mutant
Eco
SSB proteins in which different parts of this region were deleted.
Information about the influence of the N-terminal DNA binding domain on the functions of the C-terminal third can be obtained by replacing this domain with the
human mitochondrial single-stranded DNA binding protein (
Hs
mtSSB).
Hs
mtSSB, although sharing only 36% sequence homology with the N-terminus of
Eco
SSB, is a homotetrameric single-stranded DNA binding protein (SSB) with
in vitro
DNA binding properties almost indistinguishable from those of
Eco
SSB (
9
). The size of
Hs
mtSSB (133 amino acids) corresponds to the size of the N-terminal domain of
Eco
SSB and the protein is completely lacking any sequences homologous to the C-terminal third of
Eco
SSB.
MATERIALS AND METHODS
Poly(dT), poly(dA-dT) and poly(rU) were purchased from Pharmacia (Freiburg). Experiments
were performed in a standard buffer containing 0.1 mM EDTA, 20 mM potassium
phosphate pH 7.4 and sodium chloride as indicated. For fluorescence titrations 100 p.p.m. Tween 20 (Serva) was added.
Media
LB medium (
10
) was used for all microbiological experiments. The concentration of antibiotics
used for selection purposes was 100 [mu]g/ml ampicillin and 30 [mu]g/ml chloramphenicol respectively.
Site-directed mutagenesis
Site-directed mutagenesis was performed by the gapped-duplex method using the pBR322 derived plasmids pSF1 (
11
) and pSF1(
Hs
mtSSB) (
9
) both conferring ampicillin resistance. The truncated
Eco
SSB mutants
Eco
SSB Q152* and
Eco
SSB G117* were constructed by changing the codons for Q152 or G117 into the stop
codon TAG and TGA respectively. For construction of the deletion mutant
Eco
SSB [Delta]116-167 and
HsEc
SSB, we first introduced cleavage sites for blunt end restriction endonucleases
at the respective positions in the genes. After cleavage the desired gene
fragments were purified using preparative agarose gel electrophoresis and
ligated.
Eco
SSB [Delta]116-167 is a fusion protein where the amino acids 1-115 of
Eco
SSB are followed by the last 10 amino acids (168-177) of
Eco
SSB. In
HsEc
SSB, the complete sequence of
Hs
mtSSB (1-133) is followed by amino acids 113-177 of
Eco
SSB.
Protein preparation
The plasmids used for overproduction of SSB protein carry the respective gene
under control of the [lambda]P
L
-promotor. After transformation into
E.coli
TGE900 (
12
) carrying the thermosensitive [lambda]cI857 repressor protein, production was induced by a temperature shift
from 30 to 42oC and cells were harvested 3 h after induction. The sequence of the
ssb
gene was confirmed from plasmid prepared from an aliquot of the cells taken
before induction.
Wild-type
Eco
SSB and
HsEc
SSB were prepared using poly(ethyleneimine) precipitation of the protein and
subsequent extraction with 0.4 M NaCl as described by Lohman et al. (
13
). The deletion mutant
Eco
SSB [Delta]116-167 was prepared like wild-type
Eco
SSB but had to be extracted from the poly(ethyleneimine) precipitation with 0.8 M NaCl.
Hs
mtSSB was prepared as described earlier (
9
). All
Eco
SSB mutants lacking the acidic region of the last 10 amino acids were purified
like
Hs
mtSSB.
Determination of concentrations
Protein concentrations are given in units of tetramers throughout the text. They
were determined spectrophotometrically using the following absorption coefficients at 280 nm: [epsilon] = 113 000 M
-1
cm
-1
for wild-type
Eco
SSB (
14
) and
Eco
SSB Q152*, [epsilon] = 88 800 M
-1
cm
-1
for
Eco
SSB G117* and
Eco
SSB [Delta]116-167, [epsilon] = 76 240 M
-1
cm
-1
for
Hs
mtSSB and [epsilon] = 99 000 M
-1
cm
-1
for
HsEc
SSB. Unless stated otherwise, the absorption coefficients were calculated from
the known amino acid composition and the absorption coefficients of the
aromatic amino acids (
15
).
Nucleic acid concentrations are given in units of monomers and determined using the following absorption coefficients: [epsilon]
max
= 8600 M
-1
cm
-1
for poly(dT) (
16
), [epsilon]
260 nm
= 9200 M
-1
cm
-1
for poly(rU) (
17
) and [epsilon]
260 nm
= 6700 M
-1
cm
-1
for poly(dA-dT).
Physicochemical experiments
Fluorescence titrations were carried out in a Schoeffel RRS 1000 spectrofluorimeter as described earlier (
18
). Excitation wavelength was 295 nm and emission was observed at 350 nm. Theoretical binding isotherms
were calculated using the model of Schwarz and Watanabe (
19
) with the binding site size
n
, cooperative binding affinity (
K
@[omega]), and the fluorescence quench (
Q
f
) as parameters. The fluorescence quench is the difference between the normalized fluorescences of the free protein and the protein-DNA complex.
DNA melting curves were measured in a DMR10 (Zeiss) spectrophotometer as described previously (
20
) using a heating rate of 20 K/h. No significant differences between heating and
cooling curve could be observed confirming the reversibility of melting.
Analytical ultracentrifugation was carried out in a Spinco/Beckman model E centrifuge and evaluated as described earlier (
21
).
Subunit exchange experiments
For subunit exchange experiments, two different SSB proteins were mixed at a
concentration of 50 [mu]M each in a buffer containing 0.5 M NaCl, 1 mM EDTA, 20% glycerol and 20 mM
potassium phosphate, pH 7.5 in a volume of 10 [mu]l. The mixture was allowed to react for 3 days at room temperature for those
exchange experiments containing only
Eco
SSB derivatives and for 2 weeks at 37oC for exchanges with
Hs
mtSSB and
HsEc
SSB. The products were diluted by addition of 40 [mu]l 1 M triethanolamine pH 8.5, 20 [mu]l dimethylsuberimidate (30 mg/ml in water) and 130 [mu]l water. Crosslinking then was performed by a 3 h incubation at room temperature and the products were analyzed by SDS-PAGE (12%) (
22
).
Complementation of
ssb
defective strains
For complementation analysis two
E.coli
strains were used. In RDP268 (
23
) the chromosomal
ssb
gene is deleted and SSB function is restored by pACYC
ssb
plasmid conferring chloramphenicol resistance. CS149 carries the
ssb
-3 mutation (G15D) (
24
) on the chromosome rendering the cells unable to survive in presence of 0.9 [mu]g/ml mitomycin C. The strains were transformed with plasmids containing the
respective
ssb
genes. For high copy number plasmids the respective overproduction vectors
(pSF1 derivatives v.s.) were used. In the defective strains the P
L
promotor is not repressed leading to high intracellular SSB concentrations.
Monocopy plasmids were constructed as follows: pSBL5 (
25
) is a derivative of the mini-F plasmid pRE432 (
26
) conferring ampicillin resistance and contains the wild-type
Eco
SSB gene under control of its natural promotor. In this plasmid the nucleic acid
sequence coding for 160 C-terminal amino acids of wild-type
Eco
SSB was replaced by the respective mutant gene fragments. Unfortunately the
absence of suitable restriction sites for exchanging the complete
Eco
SSB gene precluded the construction of
Hs
mtSSB and
HsEc
SSB derivatives of pSBL5. Since pSBL5 is a monocopy vector and the
ssb
gene is under the control of the natural
ssb
promotor, intracellular SSB concentrations in this case correspond to the SSB
level in a normal
E.coli
cell.
For
ssb-3
complementation 100-200 transformed cells were grown on plates containing ampicillin. Colonies were replica plated on LB containing 0.9 [mu]g/ml mitomycin C and incubated at 37oC overnight.
For [Delta]
ssb
complementation transformed cells were grown in the presence of ampicillin but
omitting chloramphenicol. After four subsequent inoculations of 4 ml medium
allowing the cells to grow for ~50 generations, 100-200 cells were plated. Clones which lost the helper plasmid pACYC
ssb
were identified by replica plating on LB plates containing both chloramphenicol
and ampicillin.
If the plasmid tested could complement [Delta]
ssb
~25% of the colonies had lost the helper plasmid indicated by chloramphenicol sensitivity. In this case the absence of wild-type
Eco
SSB was confirmed by western blot analysis.
RESULTS
A schematic depiction of the different deletion mutants of
Eco
SSB, wild-type
Hs
mtSSB and the chimeric
HsEc
SSB constructed for this work is shown in Figure
1
.
CONCLUSIONS
The C-terminal third of
Eco
SSB is neither essential for the binding of nucleic acids nor for tetramer
formation. The negative charges of the last 10 amino acids weaken the binding
of the protein to nucleic acids. The electrostatic repulsion depends on the
distance of these charges from the N-terminal binding domain in the amino acid sequence.
However, the last 10 amino acids of the C-terminal third are essential for
in vivo
function. Mutant SSB proteins missing this acidic region are not functional in
the
E.coli
cell. The vital role of these last 10 amino acids must be some function
different from DNA binding, probably in protein-protein interactions. The sequence between the DNA binding domain and the
last 10 amino acids serves only as a spacer keeping the negative charges away
from the bound DNA.
DNA binding and tetramer formation are both localized in the N-terminal domain. Structural differences in this domain between eucaryotic
mitochondrial and procaryotic SSB proteins prohibit interactions of the
heterologous subunits. The phenotypical effect of mutations localized in
different regions of
Eco
SSB can be overcome by the formation of mixed tetramers.
Thus a number of defined functional properties of these SSB proteins can be
assigned to different structural regions.
ACKNOWLEDGEMENTS
We thank R. D. Porter (Pennsylvania State University) for generously supplying the
E.coli
strain RDP268, J. de Vries (University Oldenburg, Germany) for a gift of the
E.coli
strain CS149 and the plasmids pRE432 and pSBL5, and L. Litz for expert
technical assistance. This work was supported by a grant from the Deutsche
Forschungsgemeinschaft Az. Gr 1396/1-1.
REFERENCES
1 Lohman, T. and Ferrari, M. (1994) Annu. Rev. Biochem., 63, 527-570.MEDLINE Abstract
2 Meyer, R. and Laine, P. (1990) Microbiol. Rev., 54, 342-380.MEDLINE Abstract
3 Greipel, J., Urbanke, C. and Maass, G. (1989) In Saenger, W. and Heinemann, U. (eds), Protein-Nucleic Acid Interaction, Macmillan, London, pp. 61-86.
4 Weiner, J., Bertsch, L. and Kornberg, A. (1975) J. Biol. Chem., 250, 1972-1980.
5 Sancar, A., Williams, K., Chase, J. and Rupp, W. (1981) Proc. Natl Acad. Sci. USA, 78, 4274-4278.MEDLINE Abstract
6 Williams, K., Spicer, E., Lopresti, M., Gugenheimer, R. and Chase, J. (1983) J. Biol. Chem., 258, 3346-3355.MEDLINE Abstract
8 Chase, J., L'Italien, J., Murphy, J., Spicer, E. and Williams, K. (1984) J. Biol. Chem., 259, 805-814.MEDLINE Abstract
9 Curth, U., Urbanke, C., Greipel, J., Gerberding, H., Tiranti, V. and Zeviani, M. (1994) Eur. J. Biochem., 221, 435-443.MEDLINE Abstract
10 Maniatis, T., Fritsch, E. and Sambrook, J. (1982) Molecular Cloning: A Laboratory Manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, New York.
11 Bayer, I., Fliess, A., Greipel, J., Urbanke, C. and Maass, G. (1989) Eur. J. Biochem., 179, 399-405.MEDLINE Abstract
12 Courtney, M., Buchwalder, A., Tessier, L., Jaye, M., Benavente, A., Balland, A., Kohli, V., Lathe, R., Tolstoshev, P. and LeCocq, J. (1984) Proc. Natl Acad. Sci. USA, 81, 669-673.MEDLINE Abstract
13 Lohman, T., Green, J. and Beyer, R. (1986) Biochemistry, 25, 21-25.MEDLINE Abstract
14 Lohman, T. and Overman, L. (1985) J. Biol. Chem., 260, 3594-3603.MEDLINE Abstract
