ABSTRACT
By designing recombinant genes containing tandem copies of the coding region of the BHLH domain of MASH-1 (MASH-BHLH) with intervening DNA sequences encoding linker sequences of 8 or 17 amino acids, the two subunits of the MASH dimer have been connected to form the single chain dimers MM8 and MM17. Despite the long and flexible linkers which connect the C-terminus of the first BHLH subunit to the N-terminus of the second, a distance of ~55 Å, the single chain dimers could be produced in Escherichia coli at high levels. MM8 and MM17 were monomeric and no `cross-folding' of the subunits was observed. CD spectroscopy revealed that, like wild-type MASH-BHLH, MM8 and MM17 adopt only partly folded structures in the absence of DNA, but undergo a folding transition to a mainly [alpha]-helical conformation on DNA binding. Titrations by electrophoretic mobility shift assays revealed that the affinity of the single chain dimers for E box-containing DNA sequences was increased ~10-fold when compared with wild-type MASH-BHLH. On the other hand, the affinity for heterologous DNA sequences was increased only 5-fold. Therefore, the introduction of the peptide linker led to a 4-fold increase in DNA binding specificity from -0.14 to -0.57 kcal/mol.
The basic helix-loop-helix family of eukaryotic transcription factors relies on a simple structural motif for sequence-specific DNA recognition. The DNA binding activity of these proteins is confined to ~60 amino acids, named the basic helix-loop-helix (BHLH) domain (1-3; Fig. 1A). The BHLH domain comprises two regions of distinct function in DNA recognition, the helix-loop-helix domain, which mediates dimerization, and the basic region, which contacts the DNA through direct interactions with the phosphate backbone and the nucleobases (Fig. 1B; 4,5). Results from circular dichroism (CD) spectroscopy showed that in the absence of DNA BHLH proteins can form stable dimers, which are found in a concentration-dependent equilibrium with the monomer (6,7). Dimerization is accompanied by a folding transition from the largely unfolded monomer to a mainly [alpha]-helical dimer, in which helices 1 and 2 are separated through a loop of ~8 amino acids. The same transition can be induced by addition of DNA, even at concentrations where the BHLH domain alone is mainly unfolded (7-9). NMR spectroscopy and ITC experiments have shown that in the absence of DNA the basic region remains unfolded, even at concentrations where the dimer is the predominant species (10,11). However, upon DNA binding the basic region also adopts an [alpha]-helical conformation. The crystal structure analyses of the DNA complexes of the BHLH proteins E47 and MyoD revealed that the basic region is simply the N-terminal end of helix 1 and that helices 1 and 2 form the tightly packed core of the dimers (Fig. 1B; 4,5).
The gene encoding MM8 was constructed in three steps. Plasmid pJGetMASH-BHLH, which contains a fragment of the MASH-1 cDNA coding for the BHLH domain from G(106) to D(172) (8), was digested with restriction enzymes PstI and BamHI. The resulting vector fragment was ligated with a cassette with sequence
To construct the expression plasmid for MM17, pJGetMM8 digested with AgeI and KpnI and the DNA sequence coding for the additional amino acids of the linker were inserted through ligation with the following double-stranded oligonucleotide
The DNA sequence of all constructs was verified using the dideoxy sequencing method (27).
BL21(DE3)pLysS cells containing the MM8 or MM17 expression plasmids were grown at 37°C on LB medium with 100 mg/l ampicillin and 50 mg/l chloramphenicol until the OD600 reached 0.4. Then IPTG was added to a final concentration of 1 mM. Cells were harvested 3 h after induction by centrifugation and pellets were frozen at -20°C.
MM8 and MM17 were purified essentially as described for the BHLH region of MASH-1 and for the MASH mutant MASH-GGC (7,8). The purified proteins were homogeneous as judged by SDS-PAGE and cation exchange chromatography on a Resource-S (Pharmacia) HPLC column. MALDI-TOF mass spectrometry showed molecular masses of 16 001 and 16 581 for MM8 and MM17 respectively, which corresponded well with the calculated masses of 15 977 and 16 580 for the single chain dimers without their N-terminal methionines. Sequencing by Edman degradation gave the correct N-terminal sequences and confirmed that the N-terminal methionine had been removed proteolytically. Protein concentrations were determined by measuring the UV absorption at 215 and 220 nm (28). The yields for the preparations were ~4 mg purified protein/l culture.
CD spectra were measured using a Jasco J600 spectropolarimeter. The buffer was 1 mM Tris-HCl, pH 7. Spectra were measured for a concentration range of 100 nM-5 µM. For DNA binding experiments the protein concentrations were 0.5 µM for MM8 and MM17; for MASH-BHLH a concentration of 1 µM was used.
Oligonucleotides were purchased from Microsynth, desalted on Sephadex and precipitated with ethanol. Double-stranded MCK-S oligonucleotide, containing a central E box sequence and SP-1 oligonucleotide, were used as specific and heterologous DNA probes respectively (Fig. 1D). Single-stranded oligonucleotides were labelled with [[gamma]-32P]ATP (Amersham) in the presence of T4 polynucleotide kinase (NEB) and complementary strands (10% excess) were annealed by heat denaturation followed by slow cooling to room temperature.
Electrophoretic mobility shift assays (EMSA) were performed as previously described (7,8). Bacterially expressed proteins were serially diluted into EMSA buffer (50 mM Tris-HCl, pH 7.9, 6 mM MgCl2, 40 mM ammonium sulphate, 0.2 mM EDTA, 1 mM DTT, 5% glycerol). This solution was incubated in the presence of 10 nM labelled oligonucleotide for 10 min at room temperature. Samples were applied to 4% polyacrylamide gels in 0.9× TAE, pH 7.9. After electrophoresis the gels were dried and exposed to Kodak X-OMAT-S film at -70°C. Quantitative data were obtained with a Packard Instantimager using system software. The fraction [Phi] of DNA bound was determined as the activity of the retarded band (corresponding to the protein-DNA complex) divided by the sum of the activities of the retarded and unretarded (corresponding to the free DNA) bands. Plotting [Phi] against the concentration of unbound protein allowed determination of the concentration [P]1/2 at which half of the protein binding sites were filled (8). The best fit for DNA binding of `single chain dimers' to the binding isotherm (1)
The association reaction between BHLH proteins and DNA is characterized through the energetic coupling of protein folding, dimerization and DNA binding (7,8,29,30). Data from CD and nuclear magnetic resonance spectroscopy revealed that in the absence of DNA the helix-loop-helix domain can form a stably folded dimer which is found in a concentration-dependent equilibrium with the unstructured monomer with dimerization constants between 1 and 50 µM (6-8,10). However, at the concentrations where half maximal DNA binding occurs (10-500 nM) BHLH proteins are largely unfolded monomers in solution. Folding and dimerization are induced upon DNA binding. Therefore, the favourable free energy of the association reaction is reduced, because some energy must be spent on dimerization and folding at concentrations where dimerization is unfavourable. We have shown that linking the subunits of MASH-BHLH through a disulfide bond not only obviated the requirement for dimerization, but also induced the protein to adopt the folded conformation even in the absence of DNA (7). Here we tested the hypothesis that linking the C-terminus of the first BHLH subunit to the N-terminus of the second through a peptide linker should result in increased DNA binding activity without significantly altering the conformational properties of the protein.
According to crystal structure analyses of the DNA complexes of MyoD and E47 the shortest path between the C- and N-termini of the two protein subunits is ~55 Å (Fig. 1B; 4,5). Therefore, a linker of 17 amino acid residues seemed sufficient to connect the two monomers, resulting in the `single chain dimer' MM17 (Fig. 1B). Since the primary sequence of MASH-BHLH suggested that the nine N-terminal amino acids might not adopt an [alpha]-helical conformation, we also constructed MM8, in which two MASH-BHLH domains are connected through an eight residue linker. Successful construction of an active `single chain dimer' depends on a linker that neither interferes with folding and association of the two BHLH domains nor reduces stability and recognition properties of MASH-BHLH. Many surface loops in natural proteins consist of glycine, threonine and serine residues and we chose these residues for our linkers in order to maximize both flexibility and solubility (Fig. 1B).
CD spectroscopy was used to obtain structural information about MM8 and MM17. The CD spectrum of a 1 µM solution of wild-type MASH-BHLH revealed that ~25% of the amino acids were in an [alpha]-helical conformation (Fig. 2A; 31,32). Even though the amount of [alpha]-helical structure was higher in the single chain dimers (~38%), a significant portion of the peptides remained unstructured (Fig. 2A). This is in sharp contrast to the behaviour of the MASH mutant MASH-GGC, in which under oxidizing conditions the BHLH subunits are held together through a disulfide bond at the C-terminal end of helix 2 (7). Oxidized MASH-GGC was stably folded and mainly [alpha]-helical. The disulfide linkage keeps two segments of the BHLH domain in close proximity, which in the folded `dimer' are in direct contact. On the other hand, in the `single chain dimers' two parts of the peptides are held together that are remote from each other even in the folded conformation (Fig. 1B).
MASH-BHLH undergoes a concentration-dependent transition from a mainly unfolded monomer to a stably folded dimeric form with a dimerization constant of ~2 µM (7). On the other hand, the CD spectra of MM8 and MM17 were essentially unchanged over the concentration range 0.1-5 µM (corresponding to 0.2-10 µM monomer equivalents), as expected for a unimolecular folding reaction (Fig. 2B and data not shown). The predominant species of MM8 and MM17 are, therefore, monomers and no evidence for significant `cross-folding' of the BHLH subunits to form dimeric species or higher aggregates or linear polymers was observed.
The sizes of the DNA complexes of MM8 and MM17 were compared with wild-type MASH-BHLH complexes in electrophoretic mobility shift assays. MCK-S, a 17 bp DNA fragment from the IgH enhancer-like element of the muscle creatine kinase gene, was used as a probe (33; Fig. 1D). Incubation of this oligonuclotide with MM8 and MM17 respectively produced mobility shifts of approximately the same magnitude as binding to dimeric wild-type MASH-BHLH (Fig. 3A), suggesting that the structures of the complexes were similar. If a single DNA binding domain had formed by cross-folding of BHLH domains from different single chain dimers retardation of the mobility of the complexes would have been significantly greater.
In order to obtain structural information the DNA complexes of MM8 and MM17 were studied by CD spectroscopy. Upon addition of 1 equiv. double-stranded oligonucleotide containing an E box sequence to a solution of MM8 or MM17 a folding transition from a largely unfolded to a mainly [alpha]-helical conformation was observed (Fig. 2A and C). A similar change in the CD spectrum occurred when MCK-S was added to wild-type MASH-BHLH (Fig. 2A; 7,8). Interestingly, the amount of helicity observed in the different complexes varied. In the DNA complex of MM17 90% of all residues were in an [alpha]-helical conformation, an increase of 5% when compared with the wild-type complex (Fig. 2A and data not shown). On the other hand, the percentage of [alpha]-helicity was ~75% in the MM8 complex (Fig. 2A). This might be a consequence of the shorter length of the linker used in MM8. Either the N-terminal end of the basic region or the C-terminal part of helix 2 might have to unfold partly to allow proper folding of MM8 on the DNA. However, if so, this local unfolding did not diminish the DNA binding affinity of MM8 (Table 1, vide infra).
The structural changes upon DNA binding observed in both wild-type MASH-BHLH and the `single chain dimers' were in sharp contrast to the behaviour of disulfide-linked MASH-BHLH, which was fully folded even in the absence of DNA. No conformational change could be observed when DNA was added (7), indicating that the processes of dimerization, folding and DNA binding were uncoupled. MM8 and MM17, on the other hand, behave similarly to wild-type MASH-BHLH, in that folding and DNA binding remain coupled processes. Since the two subunits are covalently linked in the single chain dimers, no dimerization occurs on DNA binding. However, the subunits of MM8 and MM17 still undergo a conformational rearrangement which brings the two subunits into the intimate contact needed for formation of the proper complex.
Table 1.
As had previously been observed with MASH-BHLH and other BHLH proteins, the coil to [alpha]-helix transition was not only induced through addition of E box-containing DNA, but also by completely unrelated DNA (Fig. 3C; 7-9,11). Interestingly, the complex of MM8 with MCK-S contained slightly more [alpha]-helical residues than the complex with heterologous DNA. The same observation was made for the DNA complexes of MM17 (data not shown). While these observations were difficult to interpret, they nevertheless suggested a small difference in the geometry of the specific and the non-specific complexes of MM8 and MM17. It is noteworthy that no difference in the CD spectra of the specific and non-specific complexes of wild-type (8) and disulfide-linked MASH-BHLH had been observed (7).
Earlier work had shown that MASH-BHLH binds to DNA with moderate affinity and low DNA sequence specificity (Table 1; 7,8,11). In EMSA titration experiments, the apparent dissociation constants were measured for complexes of the `single chain dimers' with oligonucleotides containing an E box and with completely heterologous DNA (Fig. 1D). Increasing amounts of the proteins were added to a constant amount of DNA and the extent of complex formation was measured (Fig. 3B). The protein concentration at which half of the DNA binding sites are occupied, [P]1/2, was determined from the graphs describing the dependence of [Phi], the fraction of DNA bound, on the concentration of the unbound protein (Fig. 3C).
Further evidence that the linker might restrict the conformational mobility of the adjacent basic region was provided by the observation that not only the DNA binding affinity but also the DNA binding specificity was increased in the single chain dimer when compared with MASH-BHLH. While the affinity for E box-containing DNA was increased in MM8 and MM17 by 10- to 14-fold, the affinity for heterologous DNA was only 4- to 6-fold higher (Table 1). As a consequence, the free energy of transferring a protein molecule from the heterologous SP-1 DNA to an oligonucleotide containing an E box was decreased from -0.14 kcal/mol for wild-type MASH-BHLH to -0.59 kcal/mol for MM8 and to -0.57 kcal/mol for MM17. Limiting the number of accessible conformations of the basic region through introduction of the linker could stabilize the complex with specific DNA to a greater extent than the complex with heterologous DNA. Interestingly, while the association reaction between the single chain dimers and MCK-S was more exergonic by ~1.2 kcal/mol than the binding reaction of disulfide-linked MASH-BHLH (Table 1; 7), the specificity increase was slightly smaller. [Delta][Delta]Gobs for MM8 and MM17 were -0.59 and -0.57 kcal/mol respectively, while for disulfide-linked MASH-BHLH [Delta][Delta]Gobs was -0.71 kcal/mol (7).
In summary, the single chain dimers MM8 and MM17 are stable, soluble, cooperatively folded proteins which bind to DNA with enhanced affinity and specificity. Unlike disulfide-linked MASH-BHLH (7), MM8 and MM17 preserve most of the characteristic DNA binding properties of wild-type MASH-BHLH. While MM8 and MM17 do not rely on dimerization for binding, they undergo substantial conformational rearrangement for DNA binding, indicating that conformational rigidity is not a requirement for enhanced DNA binding specificity of BHLH proteins.
To the best of our knowledge the linker in MM17 is the longest linker which has been used to successfully connect two protein domains (with the exception of the linkers used to create single chain antibodies). It shows that protein subunits can be successfully connected even when the appropriate C- and N-termini are remote from each other. Despite the fact that the MM17 linker must transverse >55 Å from one side of the BHLH dimer to the other, it is resistant to protease digestion in E.coli and does not interfere with either protein folding or DNA binding.
The single chain dimers of MASH-BHLH provide the opportunity to address several questions concerning molecular recognition. Since amino acids in the two domains can be varied independently, it should be possible through mutagenesis to direct the single chain dimers to asymmetric DNA target sequences. In addition, single chain dimers can be displayed on the surface of filamentous phage particles and new DNA binding properties can be selected for through random mutagenesis.
We thank the members of our laboratory for support and discussions, Dr Lesley A.Tannahill for critical reading of the manuscript and Dr P.L.Luisi for the use of his CD spectrometer.
Nucleic Acids Research
Pages
Introduction
Materials And Methods
Construction of expression plasmids for MM8 and MM17
Purification of MM8 and MM17
CD spectroscopy
Oligonucleotides
Electrophoretic mobility shift assays
Results And Discussion
Design, expression and purification of `single chain dimers' of MASH-BHLH
CD spectroscopy of the single chain dimers MM8 and MM17
Structural characterization of the DNA complexes of MM8 and MM17
DNA binding affinity of MM8 and MM17
Specificity of DNA binding
Acknowledgements
References
resulting in plasmid pJgetMABlink1 (lower case letters indicate bases from the coding region of the MASH gene). In a second step a cassette (coding for the second half of the linker) with sequence
5'-g cag ctg ctg ACC GGT GGT ACC GGg
ac gtc gtc gac gac TGG CCA CCA TGG CCc cta g-5'
was inserted into the NdeI site of pJGetMASH-BHLH to give plasmid pJGetMABlink2. In the final step the KpnI-BamHI fragment of the insert in pJGetMABlink2 was inserted between the KpnI and BamH sites of pJGetMABlink1 to yield pJGetMM8.
5'-T ATG GGT ACC GGG GGT GGA AGT AT
AC CCA TGG CCC CCA CCT TCA TAA t-5'
5'-CC GGT GGA GGT AGT GGT GGC GGG TCA GGT GGA GGT AC
A CCT CCA TCA CCA CCG CCC AGT CCA CCT C-5'
was obtained for n = 1.
[Phi] = 1/(1 + [P]1/2/[P]n)
(1)
[P1/2]a (nM)
Kdb (1015)
[Delta]Gobsc (kcal/mol)
[Delta][Delta]Gobsd
MCK-Se
SP-1e
MCK-Se
SP-1e
MCK-Se
SP-1e
MASH-BHLH
458.0 (± 91)
520.0 (± 129)
209.8
270.4
-16.98
-16.84
-0.14
MM8
16.2 (± 5.5)
44.5 (± 1.1)
1.1
7.9
-10.44
-9.85
-0.59
MM17
22.3 (± 6.6)
59.3 (± 1.7)
2.0
14.1
-10.25
-9.68
-0.57
REFERENCES
This page is run by Oxford University Press, Great Clarendon Street, Oxford OX2 6DP, as part of the OUP Journals
Comments and feedback: www-admin{at}oup.co.uk
Last modification: 27 Feb 1998
Copyright© Oxford University Press, 1998.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  C. Krueger, C. Berens, A. Schmidt, D. Schnappinger, and W. Hillen Single-chain Tet transregulators Nucleic Acids Res., June 15, 2003; 31(12): 3050 - 3056. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Liang, J. Chen, M.-L. Tjornhammar, S. Pongor, and A. Simoncsits Modular construction of extended DNA recognition surfaces: mutant DNA-binding domains of the 434 repressor as building blocks Protein Eng. Des. Sel., August 1, 2001; 14(8): 591 - 599. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Sieber and R. K. Allemann Thermodynamics of DNA binding of MM17, a single chain dimer' of transcription factor MASH-1 Nucleic Acids Res., May 15, 2000; 28(10): 2122 - 2127. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||