Functional differences between HOX proteins conferred by two residues in the homeodomain N-terminal arm

Diversification and Functional Evolution of HOX Proteins

Gene duplication and divergence is a major contributor to the generation of morphological diversity and the emergence of novel features in vertebrates during evolution. The availability of sequenced genomes has facilitated our understanding of the evolution of genes and regulatory elements. However, progress in understanding conservation and divergence in the function of proteins has been slow and mainly assessed by comparing protein sequences in combination with in vitro analyses. These approaches help to classify proteins into different families and sub-families, such as distinct types of transcription factors, but how protein function varies within a gene family is less well understood. Some studies have explored the functional evolution of closely related proteins and important insights have begun to emerge. In this review, we will provide a general overview of gene duplication and functional divergence and then focus on the functional evolution of HOX proteins to illustrate evolutionary changes underlying diversification and their role in animal evolution.

Transcription Factor Homeobox D9 Drives the Malignant Phenotype of HPV18-Positive Cervical Cancer Cells via Binding to the Viral Early Promoter

Simple Summary

Transcription factor homeobox D9 (HOXD9) was previously reported to bind to the P97 promoter of HPV16 to induce viral E6/E7 oncogenes. In this article, we investigated whether HOXD9 regulated the P105 promoter of HPV18 and examined the role of HOXD9 in intracellular signaling of cervical cancer (CC). HOXD9 was directly bound to the P105 promoter and regulated the expression of E6/E7 genes of HPV18. The HOXD9 knockdown suppressed the E6/E7 gene expression in HPV18-positive cervical cancer cells. It decreased the expression of E6, activated the p53 pathway, and induced apoptosis. In addition, downregulation of the E7 gene expression activated the Rb pathway, causing G1 arrest in the cell cycle and markedly suppressing cell proliferation. Our results indicate that HOXD9 has pivotal roles in the proliferation and immortalization of HPV18-positive cervical cancer cells through activating the P105 promoter.

Abstract

Persistent infections with two types of human papillomaviruses (HPV), HPV16 and HPV18, are the most common cause of cervical cancer (CC). Two viral early genes, E6 and E7, are associated with tumor development, and expressions of E6 and E7 are primarily regulated by a single viral promoter: P97 in HPV16 and P105 in HPV18. We previously demonstrated that the homeobox D9 (HOXD9) transcription factor is responsible for the malignancy of HPV16-positive CC cell lines via binding to the P97 promoter. Here, we investigated whether HOXD9 is also involved in the regulation of the P105 promoter using two HPV18-positive CC cell lines, SKG-I and HeLa. Following the HOXD9 knockdown, cell viability was significantly reduced, and E6 expression was suppressed and was accompanied by increased protein levels of P53, while mRNA levels of TP53 did not change. E7 expression was also downregulated and, while mRNA levels of RB1 and E2F were unchanged, mRNA levels of E2F-target genes, MCM2 and PCNA, were decreased, which indicates that the HOXD9 knockdown downregulates E7 expression, thus leading to an inactivation of E2F and the cell-cycle arrest. Chromatin immunoprecipitation and promoter reporter assays confirmed that HOXD9 is directly associated with the P105 promoter. Collectively, our results reveal that HOXD9 drives the HPV18 early promoter activity to promote proliferation and immortalization of the CC cells.

Transcription factor homeobox D9 is involved in the malignant phenotype of cervical cancer through direct binding to the human papillomavirus oncogene promoter

OBJECTIVE

To determine the involvement of homeobox D9 (HOXD9) in the survival, proliferation, and metastasis of cervical cancer cells through regulating the expression of human papillomavirus (HPV) 16 E6/E7 genes using the P97 promoter.

METHODS

One hundred cases of cervical cancer (CC), CC cell lines SKG-I, SKG-II, SKG-IIIa, SKG-IIIb, HeLa, and SiHa, and a human tumor xenograft mouse model were used to examine the roles of HOXD9 in CC. Knockdown experiments employed RNA interference of HOXD9. qPCR, functional assays, western blotting, DNA microarray, and luciferase and ChIP assays were applied for assessments.

RESULTS

All CC cell lines expressed HOXD9 mRNA and protein. In uterine CC, HOXD9 gene expression was significantly higher than in normal cervical tissues. A positive correlation of lymphovascular space invasion and lymph node metastasis with high levels of HOXD9 expression was found in patient samples. HOXD9-knockdown cells in the mouse xenograft model only formed small or no tumors. Knockdown of HOXD9 markedly reduced CC cell proliferation, migration and invasion, induced apoptosis, increased P53 protein expression, and suppressed HPV E6/E7 expression by directly binding to the P97 promoter of HPV16 E6/E7 genes. A positive correlation between HOXD9 and HPV16 E6 expression was found in CC patients.

CONCLUSIONS

HOXD9 promotes HPV16 E6 and E7 expression by direct binding to the P97 promoter, which enhances proliferation, migration, and metastasis of CCr cells. Our results suggest that HOXD9 could be a prognostic biomarker and potential therapeutic target in CC.

Activation of Ftz-F1-Responsive Genes through Ftz/Ftz-F1 Dependent Enhancers

The orphan nuclear receptor Ftz-F1 is expressed in all somatic nuclei in Drosophila embryos, but mutations result in a pair-rule phenotype. This was explained by the interaction of Ftz-F1 with the homeodomain protein Ftz that is expressed in stripes in the primordia of segments missing in either ftz-f1 or ftz mutants. Ftz-F1 and Ftz were shown to physically interact and coordinately activate the expression of ftz itself and engrailed by synergistic binding to composite Ftz-F1/Ftz binding sites. However, attempts to identify additional target genes on the basis of Ftz-F1/ Ftz binding alone has met with only limited success. To discern rules for Ftz-F1 target site selection in vivo and to identify additional target genes, a microarray analysis was performed comparing wildtype and ftz-f1 mutant embryos. Ftz-F1-responsive genes most highly regulated included engrailed and nine additional genes expressed in patterns dependent on both ftz and ftz-f1. Candidate enhancers for these genes were identified by combining BDTNP Ftz ChIP-chip data with a computational search for Ftz-F1 binding sites. Of eight enhancer reporter genes tested in transgenic embryos, six generated expression patterns similar to the corresponding endogenous gene and expression was lost in ftz mutants. These studies identified a new set of Ftz-F1 targets, all of which are co-regulated by Ftz. Comparative analysis of enhancers containing Ftz/Ftz-F1 binding sites that were or were not bona fide targets in vivo suggested that GAF negatively regulates enhancers that contain Ftz/Ftz-F1 binding sites but are not actually utilized. These targets include other regulatory factors as well as genes involved directly in morphogenesis, providing insight into how pair-rule genes establish the body pattern.

HOX transcription factors are potential targets and markers in malignant mesothelioma

BACKGROUND

The HOX genes are a family of homeodomain-containing transcription factors that determine cellular identity during development and which are dys-regulated in some cancers. In this study we examined the expression and oncogenic function of HOX genes in mesothelioma, a cancer arising from the pleura or peritoneum which is associated with exposure to asbestos.

METHODS

We tested the sensitivity of the mesothelioma-derived lines MSTO-211H, NCI-H28, NCI-H2052, and NCI-H226 to HXR9, a peptide antagonist of HOX protein binding to its PBX co-factor. Apoptosis was measured using a FACS-based assay with Annexin, and HOX gene expression profiles were established using RT-QPCR on RNA extracted from cell lines and primary mesotheliomas. The in vivo efficacy of HXR9 was tested in a mouse MSTO-211H flank tumor xenograft model.

RESULTS

We show that HOX genes are significantly dysregulated in malignant mesothelioma. Targeting HOX genes with HXR9 caused apoptotic cell death in all of the mesothelioma-derived cell lines, and prevented the growth of mesothelioma tumors in a mouse xenograft model. Furthermore, the sensitivity of these lines to HXR9 correlated with the relative expression of HOX genes that have either an oncogenic or tumor suppressive function in cancer. The analysis of HOX expression in primary mesothelioma tumors indicated that these cells could also be sensitive to the disruption of HOX activity by HXR9, and that the expression of HOXB4 is strongly associated with overall survival.

CONCLUSION

HOX genes are a potential therapeutic target in mesothelioma, and HOXB4 expression correlates with overall survival.

Hox transcription factors influence motoneuron identity through the integrated actions of both homeodomain and non-homeodomain regions

BACKGROUND

Hox transcription factors play a critical role in the specification of motoneuron subtypes within the spinal cord. Our previous work showed that two orthologous members of this family, Hoxd10 and Hoxd11, exert opposing effects on motoneuron development in the lumbosacral (LS) spinal cord of the embryonic chick: Hoxd10 promotes the development of lateral motoneuron subtypes that project to dorsal limb muscles, while Hoxd11 represses the development of lateral subtypes in favor of medial subtypes that innervate ventral limb muscles and axial muscles. The striking degree of homology between the DNA-binding homeodomains of Hoxd10 and Hoxd11 suggested that non-homeodomain regions mediate their divergent effects. In the present study, we investigate the relative contributions of homeodomain and non-homeodomain regions of Hoxd10 and Hoxd11 to motoneuron specification.

RESULTS

Using in ovo electroporation to express chimeric and mutant constructs in LS motoneurons, we find that both the homeodomain and non-homeodomain regions of Hoxd10 are necessary to specify lateral motoneurons. In contrast, non-homeodomain regions of Hoxd11 are sufficient to repress lateral motoneuron fates in favor of medial fates.

CONCLUSIONS

Together, our data demonstrate that even closely related Hox orthologues rely on distinct combinations of homeodomain-dependent and -independent mechanisms to specify motoneuron identity.

Targeting the HOX/PBX dimer in breast cancer

The HOX genes are a family of closely related transcription factors that help to define the identity of cells and tissues during embryonic development and which are also frequently deregulated in a number of malignancies, including breast cancer. While relatively little is known about the roles that individual HOX genes play in cancer, it is however clear that these roles can be both contradictory, with some members acting as oncogenes and some as tumor suppressors, and also redundant, with several genes essentially having the same function. Here, we have attempted to address this complexity using the HXR9 peptide to target the interaction between HOX proteins and PBX, a second transcription factor that serves as a common co-factor for many HOX proteins. We show that HXR9 causes apoptosis in a number of breast cancer-derived cell lines and that sensitivity to HXR9 is directly related to the averaged expression of HOX genes HOXB1 through to HOXB9, providing a potential biomarker to predict the sensitivity of breast tumors to HXR9 or its derivatives. Measuring the expression of HOX genes HOXB1-HOXB9 in primary tumors revealed that a subset of tumors show highly elevated expression indicating that these might be potentially very sensitive to killing by HXR9. Furthermore, we show that while HXR9 blocks the oncogenic activity of HOX genes, it does not affect the known tumor-suppressor properties of a subset of HOX genes in breast cancer.

Classification of sequence signatures: a guide to Hox protein function

Hox proteins are part of the conserved superfamily of homeodomain-containing transcription factors and play fundamental roles in shaping animal body plans in development and evolution. However, molecular mechanisms underlying their diverse and specific biological functions remain largely enigmatic. Here, we have analyzed Hox sequences from the main evolutionary branches of the Bilateria group. We have found that four classes of Hox protein signatures exist, which together provide sufficient support to explain how different Hox proteins differ in their control and function. The homeodomain and its surrounding sequences accumulate nearly all signatures, constituting an extended module where most of the information distinguishing Hox proteins is concentrated. Only a small fraction of these signatures has been investigated at the functional level, but these show that approaches relying on Hox protein alterations still have a large potential for deciphering molecular mechanisms of Hox differential control.

The truncated Hoxa1 protein interacts with Hoxa1 and Pbx1 in stem cells

Hox genes contain a homeobox encoding a 60-amino acid DNA binding sequence. The Hoxa1 gene (Hox1.6, ERA1) encodes two alternatively spliced mRNAs that encode distinct proteins, one with the homeodomain (Hoxa1-993), and another protein lacking this domain (Hoxa1-399). The functions of Hoxa1-399 are unknown. We detected Hoxa1-993 and Hoxa1-399 by immunoprecipitation using Hoxa1 antibodies. To assess whether Hoxa1-399 functions in cellular differentiation, we analyzed Hoxb1, a Hoxa1 target gene. Hoxa1-993 and its cofactor, Pbx1, bind to the Hoxb1 SOct-R3 promoter to transcriptionally activate a luciferase reporter. Results from F9 stem cells that stably express ectopic Hoxa1-399 (the F9-399 line) show that Hoxa1-399 reduces this transcriptional activation. Gel shift assays demonstrate that Hoxa1-399 reduces Hoxa1-993/Pbx1 binding to the Hoxb1 SOct-R3 region. GST pull-down experiments suggest that Hoxa1-399, Hoxa1-993, and Pbx1 form a trimer. However, the F9-399 line exhibits no differences in RA-induced proliferation arrest or endogenous Hoxb1, Pbx1, Hoxa5, Cyp26a1, GATA4, or Meis mRNA levels when compared to F9 wild-type.

HOX transcription factors are potential therapeutic targets in non-small-cell lung cancer (targeting HOX genes in lung cancer)

The HOX genes are a family of homeodomain-containing transcription factors that determine the identity of cells and tissues during embryonic development. They are also known to behave as oncogenes in some haematological malignancies. In this study, we show that the expression of many of the HOX genes is highly elevated in primary non-small-cell lung cancers (NSCLCs) and in the derived cell lines A549 and H23. Furthermore, blocking the activity of HOX proteins by interfering with their binding to the PBX co-factor causes these cells to undergo apoptosis in vitro and reduces the growth of A549 tumours in vivo. These findings suggest that the interaction between HOX and PBX proteins is a potential therapeutic target in NSCLC.

A mutation in HOXA2 is responsible for autosomal-recessive microtia in an Iranian family

Microtia, a congenital deformity manifesting as an abnormally shaped or absent external ear, occurs in one out of 8,000-10,000 births. We ascertained a consanguineous Iranian family segregating with autosomal-recessive bilateral microtia, mixed symmetrical severe to profound hearing impairment, and partial cleft palate. Genome-wide linkage analysis localized the responsible gene to chromosome 7p14.3-p15.3 with a maximum multi-point LOD score of 4.17. In this region, homeobox genes from the HOXA cluster were the most interesting candidates. Subsequent DNA sequence analysis of the HOXA1 and HOXA2 homeobox genes from the candidate region identified an interesting HOXA2 homeodomain variant: a change in a highly conserved amino acid (p.Q186K). The variant was not found in 231 Iranian and 109 Belgian control samples. The critical contribution of HoxA2 for auditory-system development has already been shown in mouse models. We built a homology model to predict the effect of this mutation on the structure and DNA-binding activity of the homeodomain by using the program Modeler 8v2. In the model of the mutant homeodomain, the position of the mutant lysine side chain is consistently farther away from a nearby phosphate group; this altered position results in the loss of a hydrogen bond and affects the DNA-binding activity.

Antagonism of HOX/PBX dimer formation blocks the in vivo proliferation of melanoma

Malignant melanoma is a cancer that arises from melanocyte cells in a complex but well-studied process, and which can only be successfully treated prior to metastasis as it is highly resistant to conventional therapies. A number of recent reports have indicated that members of the HOX family of homeodomain-containing transcription factors are deregulated in melanoma, and may actually be required to maintain proliferation. In this report, we describe the use of a novel, cell-permeable antagonist of the interaction between HOX proteins and PBX, a second homeodomain-containing transcription factor that modifies HOX activity. This antagonist can block the growth of murine B16 cells and trigger apoptosis both in vitro and in vivo when administered to mice with flank tumors.

Elucidation, Quantitative Refinement, and in Vivo Utilization of the HOXA13 DNA Binding Site

Mutations in Hoxa13 cause malformations of the appendicular skeleton and genitourinary tract, including digit loss, syndactyly, and hypospadias. To determine the molecular basis for these defects, the DNA sequences bound by HOXA13 were empirically determined, revealing a novel high affinity binding site. Correlating the utilization of this high affinity binding site with genes exhibiting perturbed expression in Hoxa13 mutant limbs, we identified that HOXA13 suppresses the expression of the BMP antagonist, Sostdc1. In the absence of HOXA13 function, Sostdc1 is ectopically expressed in the distal limb, causing reduced expression of BMP-activated genes and decreased SMAD phosphorylation. Limb chromatin immunoprecipitation revealed HOXA13 binding at its high affinity site in two conserved Sostdc1 regulatory sites in vivo. In vitro, HOXA13 represses gene expression through the Sostdc1 high affinity binding sites in a dosage-dependent manner. Together, these findings confirm that the high affinity HOXA13 binding site deduced by quantitative analyses is used in vivo to facilitate HOXA13 target gene regulation, providing a critical advance toward understanding the molecular basis for defects associated with the loss of HOXA13 function.

Interchange of DNA-binding modes in the deformed and ultrabithorax homeodomains: a structural role for the N-terminal arm

The deformed (Dfd) and ultrabithorax (Ubx) homeoproteins regulate developmental gene expression in Drosophila melanogaster by binding to specific DNA sequences within its genome. DNA binding is largely accomplished via a highly conserved helix-turn-helix DNA-binding domain that is known as a homeodomain (HD). Despite nearly identical DNA recognition helices and similar target DNA sequence preferences, the in vivo functions of the two proteins are quite different. We have previously revealed differences between the two HDs in their interactions with DNA. In an effort to define the individual roles of the HD N-terminal arm and recognition helix in sequence-specific binding, we have characterized the structural details of two Dfd/Ubx chimeric HDs in complex with both the Dfd and Ubx-optimal-binding site sequences. We utilized hydroxyl radical cleavage of DNA to assess the positioning of the proteins on the binding sites. The effects of missing nucleosides and purine methylation on HD binding were also analyzed. Our results show that both the Dfd and Ubx HDs have similar DNA-binding modes when in complex with the Ubx-optimal site. There are subtle but reproducible differences in these modes that are completely interchanged when the Dfd N-terminal arm is replaced with the corresponding region of the Ubx HD. In contrast, we showed previously that the Dfd-optimal site sequence elicits a very different binding mode for the Ubx HD, while the Dfd HD maintains a mode similar to that elicited by the Ubx-optimal site. Our current methylation interference studies suggest that this alternate binding mode involves interaction of the Ubx N-terminal arm with the minor groove on the opposite face of DNA relative to the major groove that is occupied by the recognition helix. As judged by hydroxyl radical footprinting and the missing nucleoside experiment, it appears that interaction of the Ubx recognition helix with the DNA major groove is reduced. Replacing the Dfd N-terminal arm with that of Ubx does not elicit a complete interchange of the DNA-binding mode. Although the position of the chimera relative to DNA, as judged by hydroxyl radical footprinting, is similar to that of the Dfd HD, the missing nucleoside and methylation interference patterns resemble those of the Ubx HD. Repositioning of amino acid side-chains without wholesale structural alteration in the polypeptide appears to occur as a function of N-terminal arm identity and DNA-binding site sequence. Complete interchange of binding modes was achieved only by replacement of the Dfd N-terminal arm and the recognition helix plus 13 carboxyl-terminal residues with the corresponding residues of Ubx. The position of the N-terminal arm in the DNA minor groove appears to differ in a manner that depends on the two base-pair differences between the Dfd and Ubx-optimal-binding sites. Thus, N-terminal arm position dictates the binding mode and the interaction of the recognition helix with nucleosides in the major groove.

Functional comparison of the Hoxa 4, Hoxa 10, and Hoxa 11 homeoboxes

A number of models attempt to explain the functional relationships of Hox genes. The functional equivalence model states that mammalian Hox-encoded proteins are largely functionally equivalent, and that Hox quantity is more important than Hox quality. In this report, we describe the results of two homeobox swaps. In one case, the homeobox of Hoxa 11 was replaced with that of the very closely related Hoxa 10. Developmental function was assayed by analyzing the phenotypes of all possible allele combinations, including the swapped allele, and null alleles for Hoxa 11 and Hoxd 11. This chimeric gene provided wild-type function in the development of the axial skeleton and male reproductive tract, but served as a hypomorph allele in the development of the appendicular skeleton, kidneys, and female reproductive tract. In the other case, the Hoxa 11 homeobox was replaced with that of the divergent Hoxa 4 gene. This chimeric gene provided near recessive null function in all tissues except the axial skeleton, which developed normally. These results demonstrate that even the most conserved regions of Hox genes, the homeoboxes, are not functionally interchangeable in the development of most tissues. In some cases, developmental function tracked with the homeobox, as previously seen in simpler organisms. Homeoboxes with more 5' cluster positions were generally dominant over more 3' homeoboxes, consistent with phenotypic suppression seen in Drosophila. Surprisingly, however, all Hox homeoboxes tested did appear functionally equivalent in the formation of the axial skeleton. The determination of segment identity is one of the most evolutionarily ancient functions of Hox genes. It is interesting that Hox homeoboxes are interchangeable in this process, but are functionally distinct in other aspects of development.

Missense mutations of human homeoboxes: A review

The homeodomain (encoded by the homeobox) is the DNA-binding domain of a large variety of transcriptional regulators involved in controlling cell fate decisions and development. Mutations of homeobox-containing genes cause several diseases in humans. A variety of missense mutations giving rise to human diseases have been described. These mutations are an excellent model to better understand homeodomain molecular functions. To this end, homeobox missense mutations giving rise to human diseases are reviewed. Seventy-four independent homeobox mutations have been observed in 17 different genes. In the same genes, 30 missense mutations outside the homeobox have been observed, indicating that the homeodomain is more easily affected by single amino acids changes than the rest of the protein. Most missense mutations have dominant effects. Several data indicate that dominance is mostly due to haploinsufficiency. Among proteins having the homeodomain as the only DNA-binding domain, three "hot spot" regions can be delineated: 1) at codon encoding for Arg5; 2) at codon encoding for Arg31; and 3) at codons encoding for amino acids of recognition helix. In the latter, mutations at codons encoding for Arg residues at positions 52 and 53 are prevalent. In the recognition helix, Arg residues at positions 52 and 53 establish contacts with phosphates in the DNA backbone. Missense mutations of amino acids that contribute to sequence discrimination (such as those at positions 50 and 54) are present only in a minority of cases. Similar data have been obtained when missense mutations of proteins possessing an additional DNA-binding domain have been analyzed. The only exception is observed in the POU1F1 (PIT1) homeodomain, in which Arg58 is a "hot spot" for mutations, but is not involved in DNA recognition.

A conformational change in PBX1A is necessary for its nuclear localization

The fly homeodomain (HD) protein EXTRADENTICLE (EXD) is dependent on a second HD protein, HOMOTHORAX (HTH), for nuclear localization. We show here that in insect cells the mammalian homolog of EXD, PBX1A, shows a similar dependence on the HTH homologs MEIS1, 2, and 3 and the MEIS-like protein PREP1. Paradoxically, removal of residues N-terminal to the PBX1A HD abolishes interactions with MEIS/PREP but allows nuclear accumulation of PBX1A. We use deletion mapping and fusion to green fluorescent protein to map two cooperative nuclear localization signals (NLSs) in the PBX HD. The results of DNA-binding assays and pull-down experiments are consistent with a model whereby the PBX N-terminus binds to the HD and masks the two NLSs. In support of the model, a mutation in the PBX HD that disrupts contact with the N-terminus leads to constitutive nuclear localization. The HD mutation also increases sensitivity to protease digestion, consistent with a change in conformation. We propose that MEIS family proteins induce a conformational change in PBX that unmasks the NLS, leading to nuclear localization and increased DNA-binding activity. Consistent with this, PBX1 is nuclear only where Meis1 is expressed in the mouse limb bud.

The design and analysis of a homeotic response element

We have shown that the 26 bp bx1 element from the regulatory region of Distal-less is capable of imposing control by the homeotic genes Ultrabithorax and abdominal-A on a general epidermal activator in Drosophila. This provides us with an assay to analyze the sequence requirements for specific repression by these Hox genes. Both the core Hox binding site, 5'-TAAT, and the adjacent EXD 5'-TGAT core site are required for repression by Ultrabithorax and abdominal-A. The Distal-less bx1 site thus fits with the model of Hox protein binding specificity based on the consensus PBX/HOX-family site TGATNNAT[g/t][g/a], where the key elements of binding specificity are proposed to lie in the two base pairs following the TGAT. A single base pair deletion in the bx1 sequence generates a site, bx1:A(-)mut, that on the consensus PBX/HOX model would be expected to be regulated by the Deformed Hox gene. We observed, however, that the bx1:A(-)mut site was regulated predominantly by Sex combs reduced, Ultrabithorax and abdominal-A. The analysis of this site indicates that the specificity of action of Hox proteins may depend not only on selective DNA binding but also on specific post-binding interactions.

Determinants of the DNA-binding specificity of the Avian homeodomain protein, AKR

AKR (Avian Knotted-Related) was the first example of a vertebrate homeodomain protein with a highly divergent Ile residue at position 50 of the DNA-recognition helix. The protein was cloned from a liver cDNA expression library of a day-9 chick embryo by virtue of its ability to bind to the F' site in the proximal promoter of the avian apoVLDLII gene. Expression of the apoVLDLII gene is completely estrogen dependent, and mutation or deletion of the F' site decreases estrogen inducibility 5- to 10-fold. Subsequent data indicated that AKR is capable of repressing the hormone responsiveness of the apoVLDLII promoter, specifically through binding to F'. Involvement of the F' site in the hormone-dependent activation of apoVLDLII gene expression, as well as AKR-mediated repression, strongly suggests that both positive and negative regulatory factors interact with this site. Although several mammalian proteins have now been isolated whose homeodomains share many of the structural features of AKR, including the Ile at position 50, little is known of their functions in vivo or the identities of the genes they regulate. Consequently, the elements through which they exert their effects and the structural determinants of their binding specificities remain largely uncharacterized. In this study, we defined the sequence specificity of binding by AKR using polymerase chain reaction-assisted optimal site selection and determined the affinity with which the protein binds to both the optimized site and the F' site. Additionally, we generated a three-dimensional model of the AKR homeodomain binding to its optimized site and probed the validity of the model by examining the consequences of mutating amino acid residues in recognition helix 3 and the N-terminal arm on the binding specificity of the homeodomain. Finally, we present evidence that the F' site itself may act as an estrogen response element (ERE) when in the vicinity of imperfect or canonical EREs and that AKR can repress hormone inducibility mediated via this site.

Differential regulation of basal and cyclic adenosine 3',5'-monophosphate-induced somatostatin gene transcription in neural cells by DNA control elements that bind homeodomain proteins

A number of genes encoding neuropeptides are expressed in the peripheral and central nervous systems, in different endocrine organs, and in specialized cells distributed along the gastrointestinal tract. Whether expression of the same neuropeptide gene in different tissues is regulated by similar transcriptional mechanisms or by mechanisms that differ in a cell-specific manner remains unclear. We report on promoter studies on the regulation of the somatostatin gene in immortalized neural precursor cells derived from developing rat forebrain. Expression of the somatostatin gene in these cells was determined by RT-PCR/Southern blot analysis, by immunocytochemistry, and by RIA. We show that in cerebrocortical and hippocampal cells, expression of the somatostatin gene is regulated by several negative and positive DNA cis-regulatory elements located throughout the promoter region. The somatostatin cAMP-response element appears to play a prominent role in neural somatostatin gene expression by acting as a strong enhancer even in the absence of cAMP stimulation. Site-directed mutagenesis followed by transient transfection assays indicated that SMS-TAAT1, SMS-TAAT2, and SMS-UE, three previously identified homeodomain protein-binding regulatory elements that enhance transcription in pancreatic cells, act as repressors of transcription in neural cells. Electrophoretic mobility shifts assays indicate that those elements bind protein complexes that differ between neural and pancreatic cells. Our results support the notion that expression of the somatostatin gene in neural cells occurs via transcriptional mechanisms that are different from those regulating expression of the same gene in pancreatic cells.

A conserved C-terminal domain in PBX increases DNA binding by the PBX homeodomain and is not a primary site of contact for the YPWM motif of HOXA1

HOX proteins are dependent upon cofactors of the PBX family for specificity of DNA binding. Two regions that have been implicated in HOX/PBX cooperative interactions are the YPWM motif, found N-terminal to the HOX homeodomain, and the GKFQ domain (also known as the Hox cooperativity motif) immediately C-terminal to the PBX homeodomain. Using derivatives of the E2A-PBX oncoprotein, we find that the GKFQ domain is not essential for cooperative interaction with HOXA1 but contributes to the stability of the complex. By contrast, the YPWM motif is strictly required for cooperative interactions in vitro and in vivo, even with mutants of E2A-PBX lacking the GKFQ domain. Using truncated PBX proteins, we show that the YPWM motif contacts the PBX homeodomain. The presence of the GKFQ domain increases monomer binding by the PBX homeodomain 5-fold, and the stability of the HOXA1.E2A-PBX complex 2-fold. These data suggest that the GKFQ domain acts mainly to increase DNA binding by PBX, rather than providing a primary contact site for the YPWM motif of HOXA1. We have identified 2 residues, Glu-301 and Tyr-305, required for GKFQ function and suggest that this is dependent on alpha-helical character.

Distinct HOX N-terminal arm residues are responsible for specificity of DNA recognition by HOX monomers and HOX.PBX heterodimers

Dimerization with extradenticle or PBX homeoproteins dramatically improves DNA binding by HOX transcription factors, indicating that recognition by such complexes is important for HOX specificity. For HOX monomeric binding, a major determinant of specificity is the flexible N-terminal arm. It makes base-specific contacts via the minor groove, including one to the 1st position of a 5'-TNAT-3' core by a conserved arginine (Arg-5). Here we show that Arg-5 also contributes to the stability of HOX.PBX complexes, apparently by forming the same DNA contact. We further show that heterodimers of PBX with HOXA1 or HOXD4 proteins have different specificities at another position recognized by the N-terminal arm (the 2nd position in the TNAT core). Importantly, N-terminal arm residues 2 and 3, which distinguish the binding of HOXA1 and HOXD4 monomers, play no role in the specificity of their complexes with PBX. In addition, HOXD9 and HOXD10, which are capable of binding both TTAT and TAAT sites as monomers, can cooperate with PBX1A only on a TTAT site. These data suggest that some DNA contacts made by the N-terminal arm are altered by interaction with PBX.

A molecular code dictates sequence-specific DNA recognition by homeodomains

Most homeodomains bind to DNA sequences containing the motif 5'-TAAT-3'. The homeodomain of thyroid transcription factor 1 (TTF-1HD) binds to sequences containing a 5'-CAAG-3' core motif, delineating a new mechanism for differential DNA recognition by homeodomains. We investigated the molecular basis of the DNA binding specificity of TTF-1HD by both structural and functional approaches. As already suggested by the three-dimensional structure of TTF-1HD, the DNA binding specificities of the TTF-1, Antennapedia and Engrailed homeodomains, either wild-type or mutants, indicated that the amino acid residue in position 54 is involved in the recognition of the nucleotide at the 3' end of the core motif 5'-NAAN-3'. The nucleotide at the 5' position of this core sequence is recognized by the amino acids located in position 6, 7 and 8 of the TTF-1 and Antennapedia homeodomains. These data, together with previous suggestions on the role of amino acids in position 50, indicate that the DNA binding specificity of homeodomains can be determined by a combinatorial molecular code. We also show that some specific combinations of the key amino acid residues involved in DNA recognition do not follow a simple, additive rule.

An extradenticle-induced conformational change in a HOX protein overcomes an inhibitory function of the conserved hexapeptide motif

HOX homeoproteins control cell identities during animal development by differentially regulating target genes. The homeoprotein encoded by the extradenticle (exd) gene can selectively modify HOX DNA binding, suggesting that it contributes to HOX specificity in vivo. HOX-EXD interactions are in part mediated by a conserved stretch of amino acids termed the hexapeptide found in many HOX proteins. Here, we demonstrate that a 20 bp oligonucleotide from the 5' region of the mouse Hoxb-1 gene, a homolog of Drosophila labial (lab), is sufficient to direct an expression pattern in Drosophila that is very similar to endogenous lab. In vivo, this expression requires lab and exd and, in vitro, LAB requires EXD to bind this oligonucleotide. In contrast, LAB proteins with mutations in the hexapeptide bind DNA even in the absence of EXD. Moreover, a hexapeptide mutant of LAB has an increased ability to activate transcription in vivo. Partial proteolysis experiments suggest that EXD can induce a conformational change in LAB. These data are consistent with a mechanism whereby the LAB hexapeptide inhibits LAB function by inhibiting DNA binding and that an EXD-induced conformational change in LAB relieves this inhibition, promoting highly specific interactions with biologically relevant binding sites.

The specificity of homeotic gene function

How transcription factors achieve their in vivo specificities is a fundamental question in biology. For the Homeotic Complex (HOM/Hox) family of homeoproteins, specificity in vivo is likely to be in part determined by subtle differences in the DNA binding properties inherent in these proteins. Some of these differences in DNA binding are due to sequence differences in the N-terminal arms of HOM/Hox homeodomains. Evidence also exists to suggest that cofactors can modify HOM/Hox function by cooperative DNA binding interactions. The Drosophila homeoprotein extradenticle (exd) is likely to be one such cofactor. In HOM/Hox proteins, both the conserved 'YPWM' peptide motif and the homeodomain are important for interacting with exd. Although exd provides part of the answer as to how specificity is achieved, there may be additional cofactors and mechanisms that have yet to be identified.

Gsh-1: a novel murine homeobox gene expressed in the central nervous system

We report the characterization of Gsh-1, a novel murine homeobox gene. Northern blot analysis revealed a transcript of approximately 2 kb in size present at embryonic days 10.5, 11.5, and 12.5 of development. The cDNA sequence encoded a proline rich motif, a polyalanine tract, and a homeodomain with strong homology to those encoded by the clustered Hox genes. The Gsh-1 expression pattern was determined for days E8.5 to E13.5 by whole mount and serial section in situ hybridizations. Gsh-1 transcription was restricted to the central nervous system. Expression is present in the neural tube and hindbrain as two continuous, bilaterally symmetrical stripes within neural epithelial tissue. In the mesencephalon, expression is seen as a band across the most anterior portion. There is also diencephalon expression in the anlagen of the thalamus and the hypothalamus as well as in the optic stalk, optic recess, and the ganglionic eminence. Moreover, through the use of fusion proteins containing the Gsh-1 homeodomain, we have determined the consensus DNA binding site of the Gsh-1 homeoprotein to be GCT/CA/CATTAG/A.

Segmental expression of Hoxb-1 is controlled by a highly conserved autoregulatory loop dependent upon exd/pbx

Comparison of Hoxb-1 regulatory regions from different vertebrates identified three related sequence motifs critical for rhombomere 4 (r4) expression in the hindbrain. Functional analysis in transgenic mice and Drosophila embryos demonstrated that the conserved elements are involved in a positive autoregulatory loop dependent on labial (lab) family members. Binding of Hoxb-1 to these elements in vitro requires cofactors, and the motifs closely resemble the consensus binding site for pbx1, a homolog of the Drosophila extradenticle (exd) homoedomain protein. In vitro exd/pbx serves as a Hoxb-1 cofactor in cooperative binding and in Drosophila expression mediated by the r4 enhancer is dependent on both lab and exd. This provides in vivo and in vitro evidence that r4 expression involves direct autoregulation dependent on cooperative interactions of Hoxb-1 with exd/pbx proteins as cofactors.

Gsh-2, a murine homeobox gene expressed in the developing brain

A novel murine dispersed homeobox gene, designated Gsh-2, is described. Analysis of cDNA sequence, including the full open reading frame, reveals an encoded homeodomain that is surprisingly similar to those of the Antennapedia-type clustered Hox genes. In addition, the encoded protein includes polyhistidine and polyalanine tracts, as observed for several other genes of developmental significance. In situ hybridizations showed Gsh-2 expression in the developing central nervous system, including the ganglionic eminences of the forebrain, the diencephalon, which gives rise to the thalamus and hypothalamus, and in the hindbrain. Furthermore, a random oligonucleotide selection and PCR amplification procedure was used to define a target DNA binding sequence, CNAATTAG, as a first step towards the identification of downstream target genes.