Leucine Zipper Residues Specify Heterodimerization and Homodimerization with Max
Myc, Mad and Max bHLHZ segments allow the formation of Myc-Max and Mad-Max heterodimers and the Max-Max homodimer. These dimers are stabilized by hydrophobic and polar/charged interactions via helices, H1 and H2, and the leucine zipper region, Z. These regions are mostly hydrophobic and Van der Waals forces play a big part in the dimerization, but the polar/charged interactions dominate dimer specificity. On the C-terminal end of the leucine zipper region, one or two polar/charged residues form hydrogen bonds.
SHOW DIMERS
Two positively charged residues (Arg423-Arg424) in Myc form an Arg423-Arg424-Gln91-Asn92 tetrad with Max. These two pairs of hydrogen bonds formed by two charged residues in Myc (positions 91 and 92 in Max numbering) solely control heterodimerization specificity with Max.
Similarly, the negatively charged Glu125 of Mad forms a hydrogen bond with Asn92 of Max leading to a Glu125-Gln126-Gln91-Asn92 tetrad. The heterodimerization specificity appears to be controlled by the identity of a single charged residue, Gln125 (position 91 in Max numbering).
The Max homodimer yields a Gln91-Asn92-Gln91-Asn92 tetrad. The tetrad that mediates the homodimerization of Max is held together only by polar interactions, as opposed to charged interactions as Myc-Max and Mad-Max.
The Myc-Max and Mad-Mad dimers form more easily than the Max-Max homodimer. The complementary charged residues at the 91 and 92 positions in Myc and Mad have more binding affinity for Max than the complementary polar residues of Max-Max dimer. This difference between charged and polar residues also explains the disfavored Myc-Myc and Mad-Mad homodimers caused by electrostatic repulsions between the complementary residues. The Max protein is present at constant levels in the cell at all times and has a long half life, while Myc and Mad are short lived proteins and their levels fluctuate in the cell. This difference in half life of the three proteins affect the concentration of the dimers in the cell and therefore the regulation of transcription by these dimers.


Structural Basis of E Box Recognition
Each component of the Myc-Max, Mad-Max or Max-Max dimers bind to half of the target palindromic DNA site, the E-box. All three protein dimers use the same protein-DNA contacts in recognition of the 5'-Cyt(1)-Ade(2)-Cyt(3)-Gua(4)-Thy(5)-Gua(6)-3' E-box sequence. The whole basic region of these dimers undergo a conformational change upon DNA binding when the random coil loop becomes helical. This conformational change is an induced fit mechanism. All contacts from the basic region of the proteins are on the major groove side of the DNA. Four contacts between the each basic region and specific bases in the DNA recognition site (eight contacts total for each dimer) promote formation of protein-DNA complexes. The four contacts between the E-box and the Max protein are shown here as an example.
NE2 of His28 in Max (analogous to His359 in Myc and His61 in Mad) bind to N7 of Gua(3) of the opposite DNA strand. OE1 of Glu32 in Max (Myc:Glu363/Mad:Glu65) binds to N4 of Cyt(3). Additionally, OE2 of Glu32 in Max (Myc:Glu363/Mad:Glu65) binds to N6 of Ade(2) and NH1 of Arg36 in Max (Myc:Arg367/Mad:Arg69) binds to N7 of Gua(1) of the opposite DNA strand. These four contacts are repeated in the successive major groove with the Myc protein.
Additional contacts are also made between Myc and the phosphate backbone of the DNA, but the four contacts between the E-box play a bigger role in DNA binding. NZ of Lys355 forms a hydrogen bond with both phosphate oxygens of Gua(4) and Arg356 hydrogen bonds with phosphate oxygen of Thy(5). Two residues in the loop region of Myc also contribute to DNA binding by making hydrogen bonds. NZ of Lys371 contacts O1P of Ada(2) and NZ of Lys392 contacts O1P of Ada(2).
Although Myc-Max, Mad-Max and Max-Max have similar contacts with DNA, they have very different functions. The only big difference in DNA contact between these protein dimers is that the Max-Max homodimer shows the DNA bent 25 degrees towards the major groove in the cocrystal structure. Both the Myc-Max and the Mad-Max heterodimers show no DNA bending in their crystal structures. It is not known if the bending does not occur in a biological environment and is just due to the crystallization method or if the bending does occur in the biological environment and has a function in DNA looping. These dimer-DNA complexes also mediate the binding of other transcription activators or repressors. Despite the similarities, the E-box binding activities of these three similar protein dimers can be quite opposite.


Bivalent Myc-Max Heterotetramer
Interactions between the leucine zippers of two Myc-Max heterodimers lead to the formation of a heterotetramers. Salt bridges and hydrogen bonds mediate a head to tail assembly of the leucine zippers of each heterodimer making an antiparallel four helix bundle.
NZ Lys77 of Max from one heterodimer interacts with OE2 Glu432# of Myc of the other heterodimer, putting it near the backbone carbonyl oxygens of Lys428# and His429#, which also stabilize the interaction (# will indicate the second heterodimer). The positively charged residues NZ Lys77 and NE2 His81 of Max also interacts with OE1 Glu425#, which puts Lys77 and His81 near the backbone carbonyl oxygens of Glu425# and Lys423# of Myc. Polar contacts between Glu417# of Myc with Arg424 and Arg421 of Myc of the other heterodimer are also made. This allows Arg424 to face towards Max Asp84#, which further stabilizes the tetramerization.
Myc targeted genes contain multiple E-box sites with at least 100 nucleotides between them. This allows DNA looping facilitated by the Myc-Max heterotetramer. The bHLHZ region of Myc-Max can bind to other transcription factor proteins to specific sites at the DNA promoter region. The heterotetramer would be able to bind to additional proteins, indicating cooperative regulation at promoters and enhancers containing multiple E-boxes. Some other studies indicate the formation of higher order oligomers of the Max-Max homodimer, but the details are not well known. The polarity of many of the residues that form the salt bridges and hydrogen bonds in the Myc-Max heterotetramer are different in the Mad protein. In Myc, residues Glu425, Arg423 and His429 are Arg127, Glu125 and Arg131 in Mad respectively. This difference explains the reason that Mad-Max heterotetramers are not formed.


Conclusion
The cocrystal structures of Myc-Max, Mad-Max and Max-Max dimers show how the bHLHZ regions can mediate specific and high-affinity binding to DNA. The cocrystals have also lead to a greater understanding of how these transcription factors interact with each other and how that plays a role in the regulation of transcription. In summary, Mad-Max and Max-Max repress transcription by making Max unavailable to Myc, preventing the formation of the Myc-Max dimer, a transcription activator. The binding of the Mad-Max and Max-Max dimers to the common DNA targets also represses transcription by making these binding sites unavailable for Myc-Max to bind once it is formed. So this complex system of regulation by this family of transcription factors depends on the half-life of the individual proteins, the varying dimerization affinities and the competition for common DNA target sites. These insights in the understanding of regulation were concluded from the detailed studies of these complex structures. The discovery of two residues near the C-terminal leucine zipper region that control the affinity of dimerization of Myc, Mad and Max indicates how the cell controls the relative concentrations of the dimers in the cell. Cooperative activation of transcription by the Myc-Max heterotetramer is explained by the complementary polar residues of the leucine zipper of Myc. This can then be contrasted to Mad-Max dimer, which is unable to form tetramers, and the significance of this phenomenon to transcription. Myc-Max, Mad-Max and Mad-Max cell fate determination can be better understood through the study of these cocrystal structures. The particular study of the Myc-Max heterodimer can have profound implications in the medical field because Myc-Max is an oncoprotein and can give some incites in cancer research.


References:
Nair, S.K. and Burley, S.K. (2003) "X-Ray Structures of Myc-Max and Mad-Max Recognizing DNA: Molecular Bases of Regulation by Proto-Oncogenic Transcription Factors" Cell 112, 193-205

Brownlie, P., Ceska,T.A., Lamers, M., Romier, C., Stier, G., Teo, H., Suck, D. (1997) "The Crystal Structure of an Intact Human Max-DNA Complex: New Insights into Mechanisms of Transcriptional Control" Structure 5, 509-520

Voet, Donald and Voet, Judith G.; Biochemistry; 2nd Edition; John Wiley & Sons, Inc.; c1995

PDB files: 1NKP, 1NLW and 1HLO