Polyphenol Compound as a Transcription Factor Inhibitor

A target-based approach has been used to develop novel drugs in many therapeutic fields. In the final stage of intracellular signaling, transcription factor–DNA interactions are central to most biological processes and therefore represent a large and important class of targets for human therapeutics. Thus, we focused on the idea that the disruption of protein dimers and cognate DNA complexes could impair the transcriptional activation and cell transformation regulated by these proteins. Historically, natural products have been regarded as providing the primary leading compounds capable of modulating protein–protein or protein-DNA interactions. Although their mechanism of action is not fully defined, polyphenols including flavonoids were found to act mostly as site-directed small molecule inhibitors on signaling. There are many reports in the literature of screening initiatives suggesting improved drugs that can modulate the transcription factor interactions responsible for disease. In this review, we focus on polyphenol compound inhibitors against dimeric forms of transcription factor components of intracellular signaling pathways (for instance, c-jun/c-fos (Activator Protein-1; AP-1), c-myc/max, Nuclear factor kappa-light-chain-enhancer of activated B cells (NF-κB) and β-catenin/T cell factor (Tcf)).


Introduction
Polyphenol compounds have been widely studied for their bioactive properties as antioxidants and for their broad range of roles in the prevention of cancers, cardiovascular diseases and neurodegenerative diseases. Polyphenol compounds are abundant in numerous groups of natural plant products (fruit, vegetables, grains and nuts) and can be a dietary food constituent. Polyphenols are divided into several classes, i.e., phenolic acids (hydroxybenzoic acids and hydroxycinnamic acids), flavonoids (flavonols, flavones, flavanols, flavanones, isoflavones, proanthocyanidins), stilbenes, and lignans [1][2][3].
The mechanism underlying how polyphenol compounds play roles in the prevention of cancers, cardiovascular diseases and neurodegenerative diseases may be related to their modulation of cell signaling pathways. Therefore, to explain the mechanisms behind the actions of polyphenols, a number of signaling targets have been studied.
In this review, we discuss in particular the disruption of multimeric forms of transcription factors such as c-jun/c-fos (Activator Protein-1; AP-1), c-myc/max, Nuclear factor kappa-light-chain-enhancer of activated B cells (NF-κB) and β-catenin/T cell factor (Tcf) with small polyphenol compounds that block their protein-protein interactions or their interactions with DNA.
Cell proliferation, differentiation and transformation are regulated in response to cell surface stimuli in accordance with intracellular signals. In the final stage of regulation, transcription factors play an important role in the expression of genetic information. AP-1 is a heterodimeric transcription factor formed by the products of the fos and jun proto-oncogene families. The fos and the jun proteins nordihydroguaiaretic acid (NDGA) blocked transcription factor and DNA binding (Table 1). Each bZIP domain of the jun and fos proteins used in electrophoresis mobility shift assay (EMSA) contain only a basic region and a leucine zipper region, not including dephosphorylation and phosphorylation sites activated by protein kinase C (PKC) and Jun N-terminal kinase (JNK) or Fyn-related kinase (FRK), respectively [45,46]. Thus, the possible inhibition sites of these inhibitors are limited to both basic regions and the leucine zippers of c-jun and c-fos proteins. To investigate whether functional group modification of the benzaldehyde moiety in the curcuminoid structure affects the inhibition of fos-jun and DNA complex formation, our group synthesized a variety of curcuminoids as shown in Table 2. Curcumin and some synthetic curcumin derivatives were found to be transcription factor inhibitors following EMSA experiments using in vitro expressed c-fos and c-jun proteins. The inhibitory actions of curcumin, curcuminoids, DHGA and NDGA on jun/AP-1 and DNA binding in cell nuclear extracts may be not via blocking the regulatory domain at c-jun sites that negatively regulate its DNA binding activity, but via direct interference with dimer binding to DNA. The inhibitors showed an inhibitory effect against the formation of the fos-jun-DNA complex in an in vitro assay with the IC 50 values shown in Tables 1 and 2. Curcumin and NDGA were revealed to have an IC 50 value of around 0.3 mM in an in vitro assay, showing a more potent inhibitory effect on the formation of the fos-jun-DNA complex than DHGA. The change of the methoxy functional group of DHGA into a hydroxy group of NDGA caused a marked increase in the inhibitory activity. Remarkably, the synthetic curcumin derivative BJC005 showed the highest inhibitory activity out of all the synthetic curcuminoids tested. BJC005, with an IC 50 of 5.4 µM, shows a much greater inhibitory effect than momordin, reported hitherto as the best fos-jun-DNA complex inhibitor [44,47]. However, some curcuminoids, such as CHC006 and BJC004, showed no inhibitory activity against fos-jun-DNA complex formation [44].
The myc and max proteins, containing basic regions and helix-loop-helix-zipper regions, were over-expressed in E. coli and used for the investigation into the inhibitors of myc-max-DNA complex formation. Curcuminoid 004, 005, and 008 demonstrated relatively selective inhibition of myc-max-DNA binding versus jun-fos-DNA complex formation. When assayed using the myc-max dimer protein expressed in E. coli, NDGA showed a weak inhibitory effect on the dimer and cognate DNA complex formation (~8 mM of IC 50 ). This value represents about 50-fold specificity of NDGA to AP-1. However, the inhibitory effect of NDGA is more specific to β-catenin/Tcf than AP-1 or myc-max transcription factor [48,49]. This suggests that polyphenol compounds can be selective inhibitors against each transcription activator, therefore may be effective modulators of cell signaling and regulation.
Altogether, these results suggest that the substituted position of the benzaldehyde moiety and the polarity property of the substituted group play an important role in the inhibition. Regarding the inhibitory activity of curcuminoids on AP-1, p-position-and polar group-substituted curcuminoids have a tendency to inhibit the fos-jun-DNA complex formation more potently [44]. Considering that the insertion of a nitro group into the benzene ring at CHC 007, 008, 009, 011 and BJC005 contribute to efficient inhibitory activity, the substitution of polar groups in benzene rings is likely to confer a richly negative charge to the molecule itself, making them potent competitors to DNA for interaction with the protein.
The low energy conformations of curcumin and DHGA inhibitors have "V-shaped" two-fold symmetry as shown in Figure 1 [49]. Conformational studies of these inhibitors revealed that they share a common conformational feature. The two-fold symmetry of molecular conformation is also a characteristic of fos-jun or myc-max transcription activators that function as dimer proteins. As a result, a stable binding mode was proposed due to two potential interactions between the protein and the inhibitor. One is a hydrogen bond and/or an electrostatic interaction between the acidic hydroxyl (OH)-group in the phenol ring of the inhibitor and the amino-group of the basic amino acid, which is abundant in the DNA binding region of the dimeric forms of the transcription factor. The other is a hydrophobic interaction between the hydrocarbon chain of the inhibitor and the hydrocarbon chain of the amino acid in the fos-jun protein [49]. The inhibitors have differences in their molecular structures. DHGA and NDGA have the same hydrophobic carbon chain. DHGA has one phenolic OH and one methoxy group on each side while NDGA has two phenolic OH groups on each side. The methoxy group of DHGA cannot form hydrogen bonds. This might make DHGA a less potent inhibitor than NDGA. Although curcumin has one phenolic OH on each side, the conjugated double bond of the hydrophobic carbon chain may cause curcumin to be more hydrophobic than the other two inhibitors. It was also suggested that the rigid conformation of curcumin provides a better acidic OH and hydrocarbon orientation for interactions with the fos-jun protein [49]. Actually, the lowest interaction energy for each inhibitor, which was obtained using Monte Carlo docking simulations, showed a correlation with the IC 50 value of each inhibitor, which was experimentally determined. conjugated double bond of the hydrophobic carbon chain may cause curcumin to be more hydrophobic than the other two inhibitors. It was also suggested that the rigid conformation of curcumin provides a better acidic OH and hydrocarbon orientation for interactions with the fos-jun protein [49]. Actually, the lowest interaction energy for each inhibitor, which was obtained using Monte Carlo docking simulations, showed a correlation with the IC50 value of each inhibitor, which was experimentally determined. NDGA 0.34 [45] 8 [49] 0.01 [48] Curcumin 0.28 [45] 0.6 [49] 0.012 [46] Asarinin 0.3 [49] Resveratrol 0.03 [50,51] T-5224 2 [52] Anthraquinone 0.1 [53] DHGA: dihydroguaiaretic acid; NDGA: nordihydroguaiaretic acid; AP-1: Activator Protein-1.
Resveratrol and anthraquinone derivatives were also shown to inhibit the binding of AP-1 to the AP-1 binding site [50,51]. A new benzophenone derivative T-5224 was found to inhibit the DNAbinding activity of c-fos/c-jun, with IC50 values around 2 mM [52,53]. Although IC50 values around 2 mM is not very efficient, this compound exhibited a good specificity profile since the small-molecule did not affect the activities of other transcription factors including CCAAT-enhancer-binding proteins (C/EBP) and Activating Transcription Factor 2 (ATF-2) (bZIP domain) as well as MyoD (bHLH domain).
Resveratrol and anthraquinone derivatives were also shown to inhibit the binding of AP-1 to the AP-1 binding site [50,51]. A new benzophenone derivative T-5224 was found to inhibit the DNAbinding activity of c-fos/c-jun, with IC50 values around 2 mM [52,53]. Although IC50 values around 2 mM is not very efficient, this compound exhibited a good specificity profile since the small-molecule did not affect the activities of other transcription factors including CCAAT-enhancer-binding proteins (C/EBP) and Activating Transcription Factor 2 (ATF-2) (bZIP domain) as well as MyoD 0.3 [49] Resveratrol Nutrients 2015, 7, page-page conjugated double bond of the hydrophobic carbon chain may cause curcumin to be more hydrophobic than the other two inhibitors. It was also suggested that the rigid conformation of curcumin provides a better acidic OH and hydrocarbon orientation for interactions with the fos-jun protein [49]. Actually, the lowest interaction energy for each inhibitor, which was obtained using Monte Carlo docking simulations, showed a correlation with the IC50 value of each inhibitor, which was experimentally determined. NDGA 0.34 [45] 8 [49] 0.01 [48] Curcumin 0.28 [45] 0.6 [49] 0.012 [46] Asarinin 0.3 [49] Resveratrol 0.03 [50,51] T-5224 2 [52] Anthraquinone 0.1 [53] DHGA: dihydroguaiaretic acid; NDGA: nordihydroguaiaretic acid; AP-1: Activator Protein-1.
Resveratrol and anthraquinone derivatives were also shown to inhibit the binding of AP-1 to the AP-1 binding site [50,51]. A new benzophenone derivative T-5224 was found to inhibit the DNAbinding activity of c-fos/c-jun, with IC50 values around 2 mM [52,53]. Although IC50 values around 2 mM is not very efficient, this compound exhibited a good specificity profile since the small-molecule did not affect the activities of other transcription factors including CCAAT-enhancer-binding proteins (C/EBP) and Activating Transcription Factor 2 (ATF-2) (bZIP domain) as well as MyoD 0.03 [50,51]
Resveratrol and anthraquinone derivatives were also shown to inhibit the binding of AP-1 to the AP-1 binding site [50,51]. A new benzophenone derivative T-5224 was found to inhibit the DNAbinding activity of c-fos/c-jun, with IC50 values around 2 mM [52,53]. Although IC50 values around 2 mM is not very efficient, this compound exhibited a good specificity profile since the small-molecule did not affect the activities of other transcription factors including CCAAT-enhancer-binding proteins (C/EBP) and Activating Transcription Factor 2 (ATF-2) (bZIP domain) as well as MyoD
Resveratrol and anthraquinone derivatives were also shown to inhibit the binding of AP-1 to the AP-1 binding site [50,51]. A new benzophenone derivative T-5224 was found to inhibit the DNAbinding activity of c-fos/c-jun, with IC50 values around 2 mM [52,53]. Although IC50 values around 2 mM is not very efficient, this compound exhibited a good specificity profile since the small-molecule did not affect the activities of other transcription factors including CCAAT-enhancer-binding proteins (C/EBP) and Activating Transcription Factor 2 (ATF-2) (bZIP domain) as well as MyoD 0.1 [53] DHGA: dihydroguaiaretic acid; NDGA: nordihydroguaiaretic acid; AP-1: Activator Protein-1.
Resveratrol and anthraquinone derivatives were also shown to inhibit the binding of AP-1 to the AP-1 binding site [50,51]. A new benzophenone derivative T-5224 was found to inhibit the DNA-binding activity of c-fos/c-jun, with IC 50 values around 2 mM [52,53]. Although IC 50 values around 2 mM is not very efficient, this compound exhibited a good specificity profile since the small-molecule did not affect the activities of other transcription factors including CCAAT-enhancer-binding proteins (C/EBP) and Activating Transcription Factor 2 (ATF-2) (bZIP domain) as well as MyoD (bHLH domain).

Flavonoids
Many flavonoid target pathways related to intracellular signaling have been suggested and reviewed [57]. Although flavonoids were suggested as potential wnt/beta-catenin signaling modulators in cancer, all of the flavonoids did not show the same mechanism. Also, some compounds showed different mechanisms in different component targets. For instance, flavonoids like genistein, tangeretin, epigallocathechin gallate (EGCG), quercetin, apigenin and isoquercitrin were reported to modulate wnt/β-catenin signaling [57][58][59][60][61][62][63][64]. But these flavonoids do not all affect the same targets. Genistein seems to inhibit not only wnt/receptor binding but also Akt action, leading to β-catenin activation [54,65]. Moreover, genistein can inhibit the binding of β-catenin/Tcf complexes with consensus DNA [66]. Quercetin not only targets Akt and MAPK but also interferes with the binding of Tcf complexes to DNA [57,65,67,68]. Flavonoid compounds that disrupt the transcription factor complex are listed in Table 3.
Common flavonols including quercetin, kaempferol, and isorhamnetin as well as flavones such as baicalein and one of several known isoflavones, genistein, showed inhibitory effects in β-catenin activated cells [66,67]. The mechanism underlying the reduction of β-catenin/Tcf transcriptional activity by quercetin, kaempferol, isorhamnentin, baicalein and genistein was due to the decreased binding of β-catenin/Tcf complexes with consensus DNA. EMSA data showed that the binding of Tcf complexes to DNA was impaired by inhibitors. Conforming to these results, coimmunoprecipitation of β-catenin and Tcf-4 using Tcf-4 antibodies showed that the association of β-catenin with Tcf-4 was blocked by inhibitors. With genistein, the decreased distribution of β-catenin in the nucleus caused by decreased phosphorylation of Akt and GSK3b contributed as a partial mechanism for reduced β-catenin/Tcf transcriptional activity [66]. Thus, these results agree with the earlier reports by Jaiswal et al. and Sarkar et al. [54,65].

Flavonoids
Many flavonoid target pathways related to intracellular signaling have been suggested and reviewed [57]. Although flavonoids were suggested as potential wnt/beta-catenin signaling modulators in cancer, all of the flavonoids did not show the same mechanism. Also, some compounds showed different mechanisms in different component targets. For instance, flavonoids like genistein, tangeretin, epigallocathechin gallate (EGCG), quercetin, apigenin and isoquercitrin were reported to modulate wnt/β-catenin signaling [57][58][59][60][61][62][63][64]. But these flavonoids do not all affect the same targets. Genistein seems to inhibit not only wnt/receptor binding but also Akt action, leading to β-catenin activation [54,65]. Moreover, genistein can inhibit the binding of β-catenin/Tcf complexes with consensus DNA [66]. Quercetin not only targets Akt and MAPK but also interferes with the binding of Tcf complexes to DNA [57,65,67,68]. Flavonoid compounds that disrupt the transcription factor complex are listed in Table 3.
On the other hand, flavanone and flavonoids that include the same skeleton such as naringenin did not show inhibitory activity against β-catenin/Tcf and DNA binding, in discord with other types of flavonoids [68,71]. From the three dimensional structure, the 2-substituted B ring of flavanone seems to be puckered from the plane, with a different shape from the other flavone skeletons containing conjugated double bonds of C3 rings as shown in Figure 2. This puckered structure may be unfit to block β-catenin/Tcf and DNA complex formation. Although flavonoids such as 7,8-dihydroxyflavanone showed inhibitory activity against junfos/DNA binding, the inhibitory action was not as effective as on β-catenin/Tcf, with a 1.8 mM IC50 value obtained from EMSA [49]. Although there are few reports on the inhibitory effect of flavonoids against AP-1 or other transcription factors, our unpublished data showed that kaempferol inhibited fos-jun-DNA binding and myc-max-DNA binding with IC50 values of 2.9 mM and 2 mM, respectively [49]. Also, tanshinone I and 7,8 dihydroxyflavanone showed inhibitory activities against fos-jun-DNA binding, with IC50 values of 2.6 and 1.8 mM, respectively [49]. All these data demonstrate that flavonoids show selective inhibitory activity against β-catenin/Tcf rather than against the AP-1 transcription factor. However, quercetin, kaempferol, apigenin, luteolin and chrysin were reported to inhibit the NF-kB transcription factor by interfering with complex formation [69,70]. Interestingly, flavanone and flavonoids that include a flavanone skeleton such as naringenin and EGCG showed no inhibitory activity on the direct binding of β-catenin/Tcf with DNA ( [49,68,71]. Although genistein, kaempferol, isorhamnentin and baicalein possess a common phenylbenzopyrone structure (C6-C3-C6), they are mainly categorized into flavanols (kaempferol and isorhamnentin), flavone (baicalein), and isoflavone (genistein) according to the substitution state and direction of the phenyl ring. Genistein showed the least potent blocking activity in EMSA while baicalein showed the most [66]. Considering the relationship between the structure and β-catenin/Tcf inhibitory activity, the direction of the B phenyl ring at the 2-subtituted site (flavanol and flavone) appears to be more effective than at the 3-substituted (isoflavone) site. It also seems that hydroxyl group substitution at the 3-site (flavanol) reduces the potency of inhibition compared to without substitution (flavone).
On the other hand, flavanone and flavonoids that include the same skeleton such as naringenin did not show inhibitory activity against β-catenin/Tcf and DNA binding, in discord with other types of flavonoids [68,71]. From the three dimensional structure, the 2-substituted B ring of flavanone seems to be puckered from the plane, with a different shape from the other flavone skeletons containing conjugated double bonds of C3 rings as shown in Figure 2. This puckered structure may be unfit to block β-catenin/Tcf and DNA complex formation.

Non Flavonoid Polyphenol β-catenin Inhibitors
Because wnt/β-catenin signaling is involved in a broad range of biological systems, small molecular inhibitors of wnt/β-catenin signaling have received attention as targets for therapeutic interventions. A variety of inhibitors against other components in wnt/β-catenin signaling were recently reviewed [73]. Out of several small molecules, polyphenol compounds that disrupt the β-catenin/Tcf-4 complex are listed in Table 4.

Non Flavonoid Polyphenol β-catenin Inhibitors
Because wnt/β-catenin signaling is involved in a broad range of biological systems, small molecular inhibitors of wnt/β-catenin signaling have received attention as targets for therapeutic interventions. A variety of inhibitors against other components in wnt/β-catenin signaling were recently reviewed [73]. Out of several small molecules, polyphenol compounds that disrupt the β-catenin/Tcf-4 complex are listed in Table 4.

Conclusions
A useful starting point for investigations into target-specific drug design is to identify effective targets that elicit responses in physiological states and to compile an abundant candidate library of effective leading compounds. Polyphenol compounds are abundant and ideally include more comprehensive structural variance, with effects on intracellular signaling induced by the altered responsiveness (for instance, inhibition against tumor necrosis factors (TNF)-induced, transforming growth factor (TGF)-induced, α-melanocyte-stimulating hormone (α-MSH)-induced, oxidative stress-induced and UV-induced signaling).
A general concern is that many of the targets for intracellular signaling may play vital roles in other signaling cascades. Moreover, a critical issue that should be addressed is whether targeting of the dimeric forms of transcription factors is sufficiently specific. For instance, AP-1, myc-max, NF-κB and β-catenin/Tcf transcription factors play specific roles in a variety of signaling pathways. Therefore, broad inhibition of the transcription factors above is predicted to cause toxicity; this is why specificity will be an important determinant of the success of transcription factor inhibiting drugs. Thus, whether the inhibition of AP-1, myc-max, NF-κB and β-catenin/Tcf can be discriminated by chemical inhibitors remains to be explored. Theoretically, DNA can form very large interfaces with proteins and the buried surfaces of the protein-nucleic acid interfaces have a broad range of sizes [80]. Usually, protein-protein recognition sites are formed by protein surfaces with good shape and electrostatic complementarity [81]. Protein-protein interfaces are frequently hydrophobic and bury a large portion of the nonpolar surface area [82]. However, in protein-DNA interaction, the DNA binding surface of the protein harbors basic amino acids because of the negative charge of DNA. It has been reported that standard sizes for the interface range from 1200-2000 A˚ [83]. The interface ~5 [73,79] PKF115-584

Conclusions
A useful starting point for investigations into target-specific drug design is to identify effective targets that elicit responses in physiological states and to compile an abundant candidate library of effective leading compounds. Polyphenol compounds are abundant and ideally include more comprehensive structural variance, with effects on intracellular signaling induced by the altered responsiveness (for instance, inhibition against tumor necrosis factors (TNF)-induced, transforming growth factor (TGF)-induced, α-melanocyte-stimulating hormone (α-MSH)-induced, oxidative stress-induced and UV-induced signaling).
A general concern is that many of the targets for intracellular signaling may play vital roles in other signaling cascades. Moreover, a critical issue that should be addressed is whether targeting of the dimeric forms of transcription factors is sufficiently specific. For instance, AP-1, myc-max, NF-κB and β-catenin/Tcf transcription factors play specific roles in a variety of signaling pathways. Therefore, broad inhibition of the transcription factors above is predicted to cause toxicity; this is why specificity will be an important determinant of the success of transcription factor inhibiting drugs. Thus, whether the inhibition of AP-1, myc-max, NF-κB and β-catenin/Tcf can be discriminated by chemical inhibitors remains to be explored. Theoretically, DNA can form very large interfaces with proteins and the buried surfaces of the protein-nucleic acid interfaces have a broad range of sizes [80]. Usually, protein-protein recognition sites are formed by protein surfaces with good shape and electrostatic complementarity [81]. Protein-protein interfaces are frequently hydrophobic and bury a large portion of the nonpolar surface area [82]. However, in protein-DNA interaction, the DNA binding surface of the protein harbors basic amino acids because of the negative charge of DNA. It has been reported that standard sizes for the interface range from 1200-2000 A˚ [

Conclusions
A useful starting point for investigations into target-specific drug design is to identify effective targets that elicit responses in physiological states and to compile an abundant candidate library of effective leading compounds. Polyphenol compounds are abundant and ideally include more comprehensive structural variance, with effects on intracellular signaling induced by the altered responsiveness (for instance, inhibition against tumor necrosis factors (TNF)-induced, transforming growth factor (TGF)-induced, α-melanocyte-stimulating hormone (α-MSH)-induced, oxidative stress-induced and UV-induced signaling).
A general concern is that many of the targets for intracellular signaling may play vital roles in other signaling cascades. Moreover, a critical issue that should be addressed is whether targeting of the dimeric forms of transcription factors is sufficiently specific. For instance, AP-1, myc-max, NF-κB and β-catenin/Tcf transcription factors play specific roles in a variety of signaling pathways. Therefore, broad inhibition of the transcription factors above is predicted to cause toxicity; this is why specificity will be an important determinant of the success of transcription factor inhibiting drugs. Thus, whether the inhibition of AP-1, myc-max, NF-κB and β-catenin/Tcf can be discriminated by chemical inhibitors remains to be explored. Theoretically, DNA can form very large interfaces with proteins and the buried surfaces of the protein-nucleic acid interfaces have a broad range of sizes [80]. Usually, protein-protein recognition sites are formed by protein surfaces with good shape and electrostatic complementarity [81]. Protein-protein interfaces are frequently hydrophobic and bury a large portion of the nonpolar surface area [82]. However, in protein-DNA interaction, the DNA binding surface of the protein harbors basic amino acids because of the negative charge of DNA. It has been reported that standard sizes for the interface range from 1200-2000 A˚ [83]. The interface Not Determined [73][74][75][76]

Conclusions
A useful starting point for investigations into target-specific drug design is to identify effective targets that elicit responses in physiological states and to compile an abundant candidate library of effective leading compounds. Polyphenol compounds are abundant and ideally include more comprehensive structural variance, with effects on intracellular signaling induced by the altered responsiveness (for instance, inhibition against tumor necrosis factors (TNF)-induced, transforming growth factor (TGF)-induced, α-melanocyte-stimulating hormone (α-MSH)-induced, oxidative stress-induced and UV-induced signaling).
A general concern is that many of the targets for intracellular signaling may play vital roles in other signaling cascades. Moreover, a critical issue that should be addressed is whether targeting of the dimeric forms of transcription factors is sufficiently specific. For instance, AP-1, myc-max, NF-κB and β-catenin/Tcf transcription factors play specific roles in a variety of signaling pathways. Therefore, broad inhibition of the transcription factors above is predicted to cause toxicity; this is why specificity will be an important determinant of the success of transcription factor inhibiting drugs. Thus, whether the inhibition of AP-1, myc-max, NF-κB and β-catenin/Tcf can be discriminated by chemical inhibitors remains to be explored. Theoretically, DNA can form very large interfaces with proteins and the buried surfaces of the protein-nucleic acid interfaces have a broad range of sizes [80]. Usually, protein-protein recognition sites are formed by protein surfaces with good shape and electrostatic complementarity [81]. Protein-protein interfaces are frequently hydrophobic and bury a large portion of the nonpolar surface area [82]. However, in protein-DNA interaction, the DNA binding surface of the protein harbors basic amino acids because of the negative charge of DNA. It has been reported that standard sizes for the interface range from 1200-2000 A 0 [83]. The interface between transcription factor and DNA is enough for multi-facial binding of polyphenol compounds showing approximate sizes from 7 to 10 A 0 .
This review discusses polyphenol compounds with roles in direct inhibition against transcription factor complexes such as AP-1, myc-max, NF-κB and β-catenin/Tcf and suggests structure-activity relationships for the development of small-molecule drug candidates that modulate this clinically relevant signaling pathway. The structural diversity of curcumin derivative compounds or flavonoids may provide specificity for recognizing transcription factors. Polyphenol groups have structural variability and abundance in nature and show distinctive inhibitory activities against various transcription factors. Altogether, polyphenol transcription factor inhibitors have significant value as clinical drug candidates. However, another aspect that has to be considered is that some polyphenol compounds that reach cells and tissues are chemically different from the original dietary form because of structural modifications due to the conjugation process that takes place in the small intestine and, mostly, in the liver [84]. As animal trials are performed in future studies, they will provide extremely valuable information and overcome some of the difficulties associated with bioavailability in humans as clinical drug candidates.