Hox specificity

Shown is the structure of the Scr-Exd-fkh250 ternary complex. In this image, Scr is in blue, Exd in grey, and DNA in magenta. Two Scr residues are highlighted in green that insert into a particularly narrow region of the DNA minor groove, but only when Scr is bound to Exd and in the presence of the correct DNA binding site. See Joshi et. al., (2007) for further details.

Back to Projects page

The control of Hox specificity and activity in Drosophila

The Hox genes encode homeodomain transcription factors that are important for specifying tissue and cell-type identities, usually in discreet regions along the anterior-posterior (A-P) axis, of many animals. These genes are well conserved throughout metazoans. Moreover, changes in the way Hox proteins regulate their target genes has likely played an important role in the diversification of animal morphologies during animal evolution. Therefore, knowledge of how these proteins function at a mechanistic level is critical for understanding normal animal development, many human developmental disorders, and animal evolution.

The homeodomain is a DNA binding motif that, on its own, has loose DNA recognition properties. This question is especially relevant to the Hox proteins, because their homeodomains are highly similar, especially in the residues known to contact DNA. One question, then, that the lab has been interested in for several years is how these proteins achieve specificity in vivo. The problem can be broken down into at least two steps: 1. How do Hox proteins select and bind to the correct regulatory sequences in vivo? and 2. Once bound, how do these proteins and the complexes they are part of regulate transcription?

Part of the answer to the first question is that the Hox proteins bind to DNA with additional DNA binding proteins, some of which are also homeodomain proteins. In flies, two of these cofactors that we study are called Extradenticle (Exd) and Homothorax (Hth). In vertebrates, the orthologous proteins are called Pbx and Meis, respectively. There are now several examples in the literature of how Hox/Exd/Hth trimers, or their vertebrate equivalents, bind to the regulatory sequences of Hox target genes. Further, there are examples in which the Exd and Hth cofactors increase the specificity of the Hox regulation. For example, an element from the forkhead (fkh) gene called fkh250 is specifically activated by the Hox protein Sex combs reduced (Scr) in conjunction with Exd and Hth and, in vitro, this binding site shows a preference for binding Scr/Exd dimers over other Hox/Exd dimers. Changing the Scr/Exd site to a 'consensus' Hox/Exd site, however, results in a relaxed specificity both in vitro and in vivo. Thus, this 'change of specificity' experiment suggests that Hox/Exd sites come in different flavors; some show specificity for some Hox/Exd dimers, while others do not.

Recently, we have provided atomic resolution support for the idea that Exd enhances allows Hox homeodomains to achieve specificity. In particular, in collaboration with A. Aggarwal's lab at Mt. Sinai, we have solved the structure of Scr-Exd bound to both fkh250 and to a consensus DNA binding site. Strikingly, we found that additional protein-DNA contacts are made in the specific (fkh250) structure that are not observed in the 'consensus' structure. These findings strongly argue that Exd (and perhaps other cofactors) reveal a latent specificity that is built into the Hox homeodomain. Moreover, in collaboration with B. Honig and colleagues at Columbia, we found that, at least for Scr-Exd-fkh250, these specificity-determining residues are reading a DNA structure (a narrow minor groove), instead of making specific hydrogen bonds. Interestingly, these specificity-determining 'signature' residues are conserved in a paralog-specific manner (see YPWM alignment figure and Joshi et. al., (2007) for a complete description of these results). More recently, we have extended these findings to show that the recognition of minor groove shape is a general mechanism used by many DNA binding proteins (see Rohs et al., 2009, Nature, in press).

Hth and Exd cannot, however, be sufficient to explain Hox specificity. Although they seem to play an important role at many target genes, there are some tissues in both the fly and in vertebrates where these proteins are not available to be Hox cofactors. In addition, the in vitro binding specificity observed in the presence of these factors is not sufficient to account for all of Hox targeting specificity. Therefore, an underlying premise we have is that additional factors will also be Hox cofactors and contribute to specificity in vivo.

The second question, about how Hox proteins activate versus repress transcription, stems from the observation that Hox/Exd/Hth complexes are involved in both gene activation as well as gene repression. Therefore, other proteins must be used to determine the sign of the regulation. This underscores our hypothesis that Hox/Exd/Hth complexes are used primarily for targeting specificity, not for determining whether a gene will be activated or repressed.

The lab continues to try to find answers to these and related questions.

Related articles:

Gebelein B, Culi J, Ryoo HD, Zhang W, Mann RS. (2002) Specificity of Distalless repression and limb primordia development by abdominal Hox proteins. Dev Cell. 3:487-98.

Gebelein B, McKay DJ, and Mann RS (2004) Direct integration of Hox and segmentation gene inputs during Drosophila development. Nature (Article) Oct 7;431(7009):653-9.

Noro, B., Culi, J., McKay, DJ, Zhang, W., and Mann RS (2006) Distinct functions of homeodomain-containing and homeodomain-less isoforms encoded by homothorax. Genes Dev. 2006 Jun 15;20(12):1636-50

Joshi, R. Passner, J., Rohs, R., Jain, R. Sosinsky, A., Crickmore, M.A., Jacob, V., Aggarwal, A.K., Honig, B. and Mann, RS. Functional specificity of a Hox protein mediated by the recognition of minor groove structure. Cell (2007) 131(3):530-43.

Rohs, R., West, S., Sosinsky, A., Liu, P. Mann, RS, and Honig, B. The role of DNA shape in protein-DNA recognition. Nature (article) (2009) in press.

Back to Projects page

Hox gene silencing

In new work, recently published in PLoS Genetics, we describe a novel mode of Hox gene regulation. We found that the Hox gene Ubx can be silenced when Ubx protein is ectopically and transiently expressed during development. Ubx silencing can be induced by adding copies of the wild type Ubx gene, suggesting that simply doubling the amount of Ubx expression is sufficient to trigger silencing. Upon thinking about when such a system might be used, we reasoned that fluctuations in Ubx levels in wild populations of D. melanogaster might trigger silencing to keep Ubx protein levels (and haltere sizes) constant. Indeed, we provide evidence that such a silencing system exists in the wild: by crossing Ubx enhancer traps to wild populations of D. melanogaster we find that silencing can be induced to varying frequencies, depending on which wild population is being monitored. These findings suggest that enhancer activities are not 'hard-wired', but are instead sensitive to genetic variations that exist in natural populations. Future work will focus on identifying theloci that vary in natural populations that can induce silencing in the wild, and on the molecular mechanisms that mediate silencing.

Shown is a summary of the Hox complexes from the fruit fly (top row), basal chordate (amphioxus; second row), and mouse (bottom four rows).

Types of Selector Genes. The Hox genes, which specify identities along the A-P axis, belong to a class of genes known as selector genes. Other categories of selector genes are illustrated here and include compartment-specific, field-specific, cell type-specific, and tissue-specific. See the following reference:

Mann, R.S. and S.B. Carroll (2002) Molecular mechanisms of selector gene function and evolution. Curr. Op. Genet. Dev. 12:592-600.

Mann, R.S. and G. Morata (2000). The developmental and molecular biology of genes that subdivide the body of Drosophila. Annual Review of Cell and Developmental Biology (invited review), 16:243-271.

Back to Projects page