Histone deacetylases: transcriptional repression with SINers and NuRDs.

The DNA in eukaryotic cells is packaged into chromatin, which functions as a boundary to the transcriptional activation process. The nucleosome is the basic repeating unit of chromatin. The purification and characterization of several chromatin-remodelling complexes and the demonstration that histone acetyltransferases and histone deacetylases are regulatory components of coactivator and corepressor complexes, respectively, demonstrates that the nucleosome is not simply a static architectural feature of chromatin but, rather, plays a dynamic and integral role in the regulation of gene expression. This review focuses primarily on histone deacetylases and deacetylase-containing complexes and their role in mediating transcriptional repression.

The Mlx network: evidence for a parallel Max-like transcriptional network that regulates energy metabolism.

Recent experiments suggest the existence of a transcriptional network that functions in parallel to the canonical Myc/Max/Mad transcriptional network. Unlike the Myc/Max/Mad network, our understanding of this network is still in its infancy. At the center of this network is a Max-like protein called Mlx; hence we have called this network the Mlx network. Like Max, Mix interacts with transcriptional repressors and transcriptional activators, namely the Mad family and the Mondo family, respectively. Similar to Max-containing heterodimers, Mlx-containing heterodimers recognize CACGTG E-box elements, suggesting that the transcriptional targets of these two networks may overlap. Supporting this hypothesis, we have observed genetic interactions between the Drosophila melanogaster orthologs of Myc and Mondo. In higher eukaryotes, two proteins, MondoA and MondoB/CHREBP/WBSCR14, constitute the Mondo family. At present little is known about the transcriptional targets of MondoA; however, pyruvate kinase is a putative target of MondoB/CHREBP/WBSCR14, suggesting a function for the Mondo family in glucose and/or lipid metabolism. Finally, unlike the predominant nuclear localization of Myc family proteins, both Mondo family members localize to the cytoplasm. Therefore, while the Myc and Mondo families may share some biological functions, it is likely each family is under distinct regulatory control.

Simian virus 40 major late promoter: a novel tripartite structure that includes intragenic sequences.

Unlike most genes transcribed by RNA polymerase II, the simian virus 40 late transcription unit does not have a TATA box. To determine what sequences are required for initiation at the major late mRNA cap site of simian virus 40, clustered point mutations were constructed and tested for transcriptional activity in vitro and in vivo. Three promoter elements were defined. The first is centered 31 base pairs upstream of the cap site in a position normally reserved for a TATA box. The second is at the cap site. The third occupies a novel position centered 28 base pairs downstream of the cap site within a protein-coding sequence. The ability of RNA polymerase II to recognize this promoter suggests that there is greater variation in promoter architecture than had been believed previously.

A downstream-element-binding factor facilitates assembly of a functional preinitiation complex at the simian virus 40 major late promoter.

Recent work has shown that many promoters recognized by eucaryotic RNA polymerase II contain essential sequences located downstream of the transcriptional initiation site. We show here that the activity of a promoter element centered 28 base pairs downstream of the simian virus 40 major late initiation site appears to be mediated by a DNA-binding protein, which was isolated by affinity chromatography from HeLa cell nuclear extracts. In the absence of the other components of the transcriptional machinery, the protein bound specifically but weakly to its recognition sequence, with a Kd of approximately 10(-8) M. Analysis of kinetic data showed that mutation of the downstream element decreased the number of functional preinitiation complexes assembled at the promoter without significantly altering the time required for half the complexes to assemble. This suggests that in the absence of the downstream activating protein, preinitiation complexes are at least partially assembled but are not transcriptionally competent.

Pf1, a novel PHD zinc finger protein that links the TLE corepressor to the mSin3A-histone deacetylase complex.

The mSin3A-histone deacetylase corepressor is a multiprotein complex that is recruited by DNA binding transcriptional repressors. Sin3 has four paired amphipathic alpha helices (PAH1 to -4) that are protein-protein interaction motifs and is the scaffold upon which the complex assembles. We identified a novel mSin3A-interacting protein that has two plant homeodomain (PHD) zinc fingers we term Pf1, for PHD factor one. Pf1 associates with mSin3A in vivo and recruits the mSin3A complex to repress transcription when fused to the DNA binding domain of Gal4. Pf1 interacts with Sin3 through two independent Sin3 interaction domains (SIDs), Pf1SID1 and Pf1SID2. Pf1SID1 binds PAH2, while Pf1SID2 binds PAH1. Pf1SID1 has sequence and structural similarity to the well-characterized 13-amino-acid SID of the Mad bHLHZip repressor. Pf1SID2 does not have sequence similarity with either Mad SID or Pf1SID1 and therefore represents a novel Sin3 binding domain. Mutations in a minimal fragment of Pf1 that encompasses Pf1SID1 inhibited mSin3A binding yet only slightly impaired repression when targeted to DNA, implying that Pf1 might interact with other corepressors. We show that Pf1 interacts with a mammalian homolog of the Drosophila Groucho corepressor, transducin-like enhancer (TLE). Pf1 binds TLE in an mSin3A-independent manner and recruits functional TLE complexes to repress transcription. These findings suggest that Pf1 may serve to bridge two global transcription networks, mSin3A and TLE.

Beta-catenin-histone deacetylase interactions regulate the transition of LEF1 from a transcriptional repressor to an activator.

Recent evidence suggests that certain LEF/TCF family members act as repressors in the absence of Wnt signaling. We show here that repression by LEF1 requires histone deacetylase (HDAC) activity. Further, LEF1 associates in vivo with HDAC1, and transcription of a model LEF1-dependent target gene is modulated by the ratio of HDAC1 to beta-catenin, implying that repression by LEF1 is mediated by promoter-targeted HDAC. Consistent with this hypothesis, under repression conditions the promoter region of a LEF1 target gene is hypoacetylated. By contrast, when the reporter is activated, its promoter becomes hyperacetylated. Coexpression of beta-catenin with LEF1 and HDAC1 results in the formation of a beta-catenin/HDAC1 complex. Surprisingly, the enzymatic activity of HDAC1 associated with beta-catenin is attenuated. Together, these findings imply that activation of LEF1-dependent genes by beta-catenin involves a two-step mechanism. First, HDAC1 is dissociated from LEF1 and its enzymatic activity is attenuated. This first step yields a promoter that is inactive but poised for activation. Second, once HDAC1-dependent repression has been overridden, beta-catenin binds LEF1 and the beta-catenin-LEF1 complex is competent to activate the expression of downstream target genes.

SIN3-dependent transcriptional repression by interaction with the Mad1 DNA-binding protein.

The SIN3 gene in Saccharomyces cerevisiae encodes a negative regulator of transcription of a large number of genes. Mouse homologs of SIN3 have been identified through screens for proteins interacting with the mammalian Mad1 protein, a transcriptional repressor. We find that yeast Sin3 (ySin3) interacts with Madl and that, as for mouse Sin3, the N terminus of Mad1 interacts with the PAH2 domain of ySin3. Although Mad1 (a basic helix-loop-helix leucine zipper [bHLH-Zip) protein) forms a heterodimer with the Max bHLH-Zip protein, LexA-Mad1 and VP16-Max do not activate transcription of a reporter gene in a two-hybrid assay. This failure in activation is due to direct repression by ySin3, as LexA-Mad1 and VP16-Max are able to activate the two-hybrid reporter in a sin3 mutant. This inhibition of activation by LexA-Mad1 and VP16-Max requires the PAH2 domain of ySin3 and the N-terminal interaction region of Mad1. These data demonstrate that ySin3 functions as a transcriptional repressor by being brought to promoters by interacting with proteins bound to DNA.

MondoA, a novel basic helix-loop-helix-leucine zipper transcriptional activator that constitutes a positive branch of a max-like network.

Max is a common dimerization partner for a family of transcription factors (Myc, Mad [or Mxi]), and Mnt [or Rox] proteins) that regulate cell growth, proliferation, and apoptosis. We recently characterized a novel Max-like protein, Mlx, which interacts with Mad1 and Mad4. Here we describe the cloning and functional characterization of a new family of basic helix-loop-helix-leucine zipper heterodimeric partners for Mlx termed the Mondo family. MondoA forms homodimers weakly and does not interact with Max or members of the Myc or Mad families. MondoA and Mlx associate in vivo, and surprisingly, they are localized primarily to the cytoplasm of cultured mammalian cells. Treatment of cells with the nuclear export inhibitor leptomycin B results in the nuclear accumulation of MondoA and Mlx, demonstrating that they shuttle between the cytoplasmic and nuclear compartments rather than having exclusively cytoplasmic localization. MondoA preferentially forms heterodimers with Mlx, and this heterocomplex can bind to, and activate transcription from, CACGTG E-boxes when targeted to the nucleus via a heterologous nuclear localization signal. The amino termini of the Mondo proteins are highly conserved among family members and contain separable and autonomous cytoplasmic localization and transcription activation domains. Therefore, Mlx can mediate transcriptional repression in conjunction with the Mad family and can mediate transcriptional activation via the Mondo family. We propose that Mlx, like Max, functions as the center of a transcription factor network.

Mad proteins contain a dominant transcription repression domain.

Transcription repression by the basic region-helix-loop-helix-zipper (bHLHZip) protein Mad1 requires DNA binding as a ternary complex with Max and mSin3A or mSin3B, the mammalian orthologs of the Saccharomyces cerevisiae transcriptional corepressor SIN3. The interaction between Mad1 and mSin3 is mediated by three potential amphipathic alpha-helices: one in the N terminus of Mad (mSin interaction domain, or SID) and two within the second paired amphipathic helix domain (PAH2) of mSin3A. Mutations that alter the structure of the SID inhibit in vitro interaction between Mad and mSin3 and inactivate Mad's transcriptional repression activity. Here we show that a 35-residue region containing the SID represents a dominant repression domain whose activity can be transferred to a heterologous DNA binding region. A fusion protein comprising the Mad1 SID linked to a Ga14 DNA binding domain mediates repression of minimal as well as complex promoters dependent on Ga14 DNA binding sites. In addition, the SID represses the transcriptional activity of linked VP16 and c-Myc transactivation domains. When fused to a full-length c-Myc protein, the Mad1 SID specifically represses both c-Myc's transcriptional and transforming activities. Fusions between the GAL DNA binding domain and full-length mSin3 were also capable of repression. We show that the association between Mad1 and mSin3 is not only dependent on the helical SID but is also dependent on both putative helices of the mSin3 PAH2 region, suggesting that stable interaction requires all three helices. Our results indicate that the SID is necessary and sufficient for transcriptional repression mediated by the Mad protein family and that SID repression is dominant over several distinct transcriptional activators.

Inhibition of cell proliferation by the Mad1 transcriptional repressor.

Mad1 is a basic helix-loop-helix-leucine zipper protein that is induced upon differentiation of a number of distinct cell types. Mad1 dimerizes with Max and recognizes the same DNA sequences as do Myc:Max dimers. However, Mad1 and Myc appear to have opposing functions. Myc:Max heterodimers activate transcription while Mad:Max heterodimers repress transcription from the same promoter. In addition Mad1 has been shown to block the oncogenic activity of Myc. Here we show that ectopic expression of Mad1 inhibits the proliferative response of 3T3 cells to signaling through the colony-stimulating factor-1 (CSF-1) receptor. The ability of over-expressed Myc and cyclin D1 to complement the mutant CSF-1 receptor Y809F (containing a Y-to-F mutation at position 809) is also inhibited by Mad1. Cell cycle analysis of proliferating 3T3 cells transfected with Mad1 demonstrates a significant decrease in the fraction of cells in the S and G2/M phases and a concomitant increase in the fraction of G1 phase cells, indicating that Mad1 negatively influences cell cycle progression from the G1 to the S phase. Mutations in Mad1 which inhibit its activity as a transcription repressor also result in loss of Mad1 cell cycle inhibitory activity. Thus, the ability of Mad1 to inhibit cell cycle progression is tightly coupled to its function as a transcriptional repressor.

Regulation of Myc and Mad during epidermal differentiation and HPV-associated tumorigenesis.

c-Myc and Mad each form heterodimers with Max that bind the same E-box related DNA sequences. Whereas Myc:Max complexes activate transcription and promote cell proliferation and transformation, Mad:Max complexes repress transcription and block c-Myc-mediated cell transformation. Here we examine these antagonistic transcription factors during epithelial differentiation and neoplastic progression. During differentiation of primary human keratinocytes, Mad is rapidly induced and c-Myc is downregulated, resulting in a switch from c-Myc:Max to Mad:Max heterodimers. In normal epidermis and colonic mucosa c-myc expression is restricted to proliferating cell layers, while mad expression is restricted to differentiating cell layers. Using HPV18 transformed keratinocytes that vary in their ability to differentiate in organotypic cultures, we find that Mad induction occurs only in those cells that retain a differentiation response. In the epidermis of transgenic mice in which expression of the HPV16 E6 and E7 oncogenes are targeted to basal keratinocytes, neoplastic progression occurs and is marked by an expansion of c-myc expressing basal-like cells. Expression of mad is found only in growth-arrested differentiating cells on the outer edges of preneoplastic lesions. The squamous cell carcinomas that arise evidence a variable number of sites within the tumor masses where mad expression and morphological differentiation coincide; increasing malignancy correlates with loss of both mad and capability to differentiate. These results indicate that c-Myc and Mad expression are tightly coupled to the transition from proliferation to differentiation of epithelial cells and that restriction of Mad expression may be associated with loss of normal differentiation capability and with tumorigenesis.

Mapping of two genes encoding members of a distinct subfamily of MAX interacting proteins: MAD to human chromosome 2 and mouse chromosome 6, and MXI1 to human chromosome 10 and mouse chromosome 19.

Both the MAD and the MXI1 genes encode basic-helix-loop-helix-leucine zipper (bHLH-Zip) transcription factors which bind Max in vitro, forming a sequence-specific DNA-binding complex similar to the Myc-Max heterodimer. Mad and Myc compete for binding to Max. In addition, Mad has been shown to act as a transcriptional repressor while Myc appears to function as an activator. Mxi1 also appears to lack a transcriptional activation domain. Therefore, Mxi1 and Mad might antagonize Myc function and are candidate tumor suppressor genes. We report here the mapping of the MAD and MXI1 genes in human and mouse by fluorescence in situ hybridization (FISH) and by recombination mapping. The MAD gene was mapped to human chromosome 2 at band p13 by FISH and to mouse chromosome 6 by meiotic mapping. The MXI1 gene was mapped to human chromosome 10 at band q25 and on mouse chromosome 19 at region D by FISH. There was a second site of hybridization on mouse chromosome 2 at region C, which may represent a pseudogene or a related sequence. The mapping results confirm regions of conservation between human chromosome 2p13 and mouse chromosome 6 and between chromosome 10q25 and mouse chromosome 19D. Human chromosomes 2p13 and 10q25 have been involved in specific tumors where the role of Mad and Mxi1 can now be investigated.

Acetyl glycerylphosphorylcholine inhibition of prostaglandin I2-stimulated adenosine 3',5'-cyclic monophosphate levels in human platelets. Evidence for thromboxane A2 dependence.

Previous studies with AGEPC (1-O-hexadecyl/octadecyl-2-acetyl-sn-glyceryl-3-phosphorylcholine) stress the independence of the proaggregatory activity of AGEPC from the platelet cyclooxygenase. However, our dose response analyses in human platelet-rich plasma show distinct primary and secondary waves of aggregation in response to AGEPC. Second wave aggregation is inhibited completely by either 10 micro M indomethacin, a cyclooxygenase inhibitor, or 5.6 micro M 9,11-azoprosta-5,13-dienoic acid, a thromboxane A2 synthetase inhibitor. Simultaneous addition of AGEPC and prostaglandin I2 to platelet-rich plasma results in a marked increase in platelet cyclic AMP, which is not different from the prostaglandin I2 response alone. However, if prostaglandin I2 is added to AGEPC-stimulated platelets at a point where secondary aggregation is just beginning, AGEPC can attenuate prostaglandin I2-stimulated cyclic AMP accumulation. The inhibition by AGEPC is blocked by either cyclooxygenase or thromboxane A2 synthetase inhibitors, and radioimmunoassay of thromboxane B2 confirmed that the inhibition of prostaglandin I2-stimulated cyclic AMP accumulation is due to thromboxane A2 synthesis, and that AGEPC-stimulated secondary aggregation does not start until thromboxane A2 is synthesized. These data suggest that much of the bioactivity of AGEPC is attributable to thromboxane A2.

2-(Aminomethyl)chromans that inhibit iron-dependent lipid peroxidation and protect against central nervous system trauma and ischemia.

A series of 2-(aminomethyl)chromans was developed as potent inhibitors of iron-dependent lipid peroxidation. Compounds within this class are extremely effective at inhibiting lipid peroxidation with IC50's as low as 0.2 microM. Selected members were found to enhance early neurological recovery and survival in a mouse head injury model. In this assay, improvement in the 1-h post-head-injury neurological status (grip test score) by as much as 230% of control was observed. One of the most efficacious compounds (35) was evaluated in two models of cerebral ischemia where significant neuroprotection was observed. These results provide further support for the importance of cerebroprotective antioxidants for the treatment of traumatic and ischemic injury as well as additional evidence for the role of oxygen radicals in postischemic brain damage.

Novel 21-aminosteroids that inhibit iron-dependent lipid peroxidation and protect against central nervous system trauma.

A novel class of 21-aminosteroids has been developed. Compounds within this series are potent inhibitors of iron-dependent lipid peroxidation in rat brain homogenates with IC50's as low as 3 microM. Furthermore, selected members enhance early neurological recovery and survival in a mouse head injury model. Significant improvement in the 1 h post-head-injury neurological status (grip test score) by as much as 168.6% of the control has been observed. The most efficacious compound in this assay (30) showed an increase in the 1-week survival of 78.6% as compared to 27.3% for the vehicle-treated mice in the head-injury model. Based on its biological profile, 21-[4-(2,6-di-1-pyrrolidinyl-4-pyrimidinyl)-1-piperazinyl]-16 alpha- methylpregna-1,4,9(11)-triene-3,20-dione monomethanesulfonate (30) was selected for further evaluation and is currently entering phase I clinical trials for the treatment of head and spinal trauma.

Novel membrane localized iron chelators as inhibitors of iron-dependent lipid peroxidation.

Attachment of various iron chelating moieties to hydrophobic steroids greatly enhanced their abilities to inhibit iron-dependent lipid peroxidation. Using whole rat brain homogenates, lipid peroxidation initiated by the addition of 200 microM Fe2+ was assessed by the formation of thiobarbituric acid reactive products (TBAR). Under these conditions, 50% inhibitory concentrations of Fe3+ chelators such as desferrioxamine or N1,N8-bis(2,3-dihydroxybenzoyl) spermidine hydrobromide (compound II) were around 170 and 50 microM respectively. Coupling desferrioxamine or compound II to a steroid at the D ring increased their potency in lipid peroxidation assays by 5- to 10-fold. Evidence that inhibition of lipid peroxidation by the steroid-chelator adducts was due to iron chelation was suggested by the fact that methylation of the catechol oxygens of compound II, which are essential for chelation, completely eliminated activity of the steroid adduct. A series of 21-aminosteroids which complex Fe2+ iron and potently inhibit iron-dependent lipid peroxidation has also been synthesized. Coupling Fe2+ chelators to hydrophobic steroids increased their inhibitory potencies by as much as 10- to 100-fold. Some steroid-based Fe2+ chelators stimulated lipid peroxidation at low concentrations in the presence of Fe3+. The degree of stimulation was related to the affinity of a compound for Fe2+ with the stronger chelators causing greater stimulation. The most potent inhibitors of lipid peroxidation in the 21-aminosteroid series were found to be those compounds forming the weakest Fe2+ complexes. The findings suggest that it is iron at or near the membrane that is responsible for the catalysis of lipid peroxidation. The compounds described should provide useful tools for studies of the involvement of iron in the lipid peroxidation process.

Neuroprotective efficacy of microvascularly-localized versus brain-penetrating antioxidants.

The 21-aminosteroid (lazaroid) tirilazad mesylate has been demonstrated to be a potent inhibitor of lipid peroxidation and to reduce traumatic and ischemic damage in a number of experimental models. Currently, tirilazad is being actively investigated in phase III clinical trials in head and spinal cord injury, ischemic stroke and subarachnoid hemorrhage. This compound acts in large part to protect the microvascular endothelium and consequently to maintain normal blood-brain barrier (BBB) permeability and cerebral blood flow autoregulatory mechanisms. However, due to its limited penetration into brain parenchyma, tirilazad has generally failed to affect delayed neuronal damage to the selectively vulnerable hippocampal CA1 and striatal regions. Recently, we have discovered a new group of antioxidant compounds, the pyrrolopyrimidines, which possess significantly improved ability to penetrate the BBB and gain direct access to neural tissue. Several compounds in the series, such as U-101033E, have demonstrated greater ability to protect the CA1 region in the gerbil transient forebrain ischemia model with a post-ischemic therapeutic window of at least four hours. In addition, U-101033E has been found to reduce infarct size in the mouse permanent middle cerebral artery occlusion model in contrast to tirilazad which is minimally effective. These results suggest that antioxidant compounds with improved brain parenchymal penetration are better able to limit certain types of ischemic brain damage compared to those which are localized in the cerebral microvasculature. On the other hand, microvascularly-localized agents like tirilazad appear to have better ability to limit BBB damage.

A switch from Myc:Max to Mad:Max heterocomplexes accompanies monocyte/macrophage differentiation.

Mad is a basic-helix-loop-helix-zipper protein that heterodimerizes with Max in vitro. Mad:Max heterodimers recognize the same E-box-related DNA-binding sites as Myc:Max heterodimers. However, in transient transfection assays Myc and Mad influence transcription in opposite ways through interaction with Max; Myc activates while Mad represses transcription. Here, we demonstrate that Mad protein is induced rapidly upon differentiation of cells of the myeloid lineage. The Mad protein is synthesized in human cells as a 35-kD nuclear phosphoprotein with an extremely short half-life (t1/2 = 15-30 min) and can be detected in vivo in a complex with Max. In the undifferentiated U937 monocyte cell line Max was found complexed with Myc but not Mad. However, Mad:Max complexes began to accumulate as early as 2 hr after induction of macrophage differentiation with TPA. By 48 hr following TPA treatment only Mad:Max complexes were detectable. These data show that differentiation is accompanied by a change in the composition of Max heterocomplexes. We speculate that this switch in heterocomplexes results in a change in the transcriptional regulation of Myc:Max target genes required for cell proliferation.

Mad-Max transcriptional repression is mediated by ternary complex formation with mammalian homologs of yeast repressor Sin3.

The bHLH-ZIP protein Mad heterodimerizes with Max as a sequence-specific transcriptional repressor. Mad is rapidly induced upon differentiation, and the associated switch from Myc-Max to Mad-Max heterocomplexes seem to repress genes normally activated by Myc-Max. We have identified two related mammalian cDNAs that encode Mad-binding proteins. Both possess sequence homology with the yeast transcription repressor Sin3, including four conserved paired amphipathic helix (PAH) domains. mSin3A and mSin3B bind specifically to Mad and the related protein Mxi1. Mad-Max and mSin3 form ternary complexes in solution that specifically recognize the Mad-Max E box-binding site. Mad-mSin3 association requires PAH2 of mSin3A/mSin3B and the first 25 residues of Mad, which contains a putative amphipathic alpha-helical region. Point mutations in this region eliminate interaction with mSin3 proteins and block Mad transcriptional repression. We suggest that Mad-Max represses transcription by tethering mSin3 to DNA as corepressors and that a transcriptional repression mechanism is conserved from yeast to mammals.