Comparing anterior and posterior Hox complex formation reveals guidelines for predicting cis-regulatory elements

Abstract

Hox transcription factors specify numerous cell fates along the anterior-posterior axis by regulating the expression of downstream target genes. While expression analysis has uncovered large numbers of de-regulated genes in cells with altered Hox activity, determining which are direct versus indirect targets has remained a significant challenge. Here, we characterize the DNA binding activity of Hox transcription factor complexes on eight experimentally verified cis-regulatory elements. Hox factors regulate the activity of each element by forming protein complexes with two cofactor proteins, Extradenticle (Exd) and Homothorax (Hth). Using comparative DNA binding assays, we found that a number of flexible arrangements of Hox, Exd, and Hth binding sites mediate cooperative transcription factor complexes. Moreover, analysis of a Distal-less regulatory element (DMXR) that is repressed by abdominal Hox factors revealed that suboptimal binding sites can be combined to form high affinity transcription complexes. Lastly, we determined that the anterior Hox factors are more dependent upon Exd and Hth for complex formation than posterior Hox factors. Based upon these findings, we suggest a general set of guidelines to serve as a basis for designing bioinformatics algorithms aimed at identifying Hox regulatory elements using the wealth of recently sequenced genomes.

Figure 1. Configurations of Hox, Exd, and Hth binding sites in Drosophila and vertebrate cis-regulatory elements

Schematics of eight Hox regulated enhancer elements with the Exd (blue), Hth (yellow) and Hox (pink) sites highlighted. A. Abdominal Hox target elements from Drosophila. The DMXR element contains both Exd/Hox and Hth/Hox sites that are bound by the Abd-A and Ubx (A/U) Hox factors to repress Dll expression in the abdomen (Gebelein et al., 2004). The RhoA element contains consecutive Exd/Hth/Hox binding sites that are bound by Abd-A to activate rhomboid expression in developing abdominal sensory cells (Li-Kroeger et al., 2008). B. The EVIII and Lab 48/95 elements are both activated by the Lab Hox factor in the developing gut endoderm and each contains a Hox, Exd, and Hth binding site (Ebner et al., 2005; Ryoo et al., 1999). C. Four mouse cis-regulatory elements containing Pbx/Hox and distant Meis binding sites. All are activated by the HoxB1 vertebrate homologue of Lab in collaboration with the Pbx and Meis proteins within the developing hindbrain (Ferretti et al., 2005; Ferretti et al., 2000; Manzanares et al., 2001; Popperl et al., 1995; Tumpel et al., 2007).

Figure 2. Dependence of Abd-A tetramer formation on the presence of two Hox sites within DMXR

A. The DMXR probes used in EMSAs with the Hox1, Exd, Hth, and Hox2 sites highlighted. B. The DMXR Hox point mutations used in gel shift analysis. C. Comparative EMSAs of Hox complex formation on the DMXR, DMXR1, and DMXR2 probes. Conditions are as follows: First lane is probe alone, second lane is 75 × 10^-9 M of Exd/Hth heterodimers, third and fourth lane contain the same amount of Exd/Hth with either 17 × 10^-9 M (low amount) or 75 × 10^-9 M (high amount) of Abd-A, respectively and the fifth lane contains 75 × 10^-9 M of Abd-A alone. Schematics at left denote color-coded complexes formed (Exd, blue; Hth, yellow; Abd-A, pink). D and E. Competition DNA binding assays for Abd-A/Exd/Hth complexes on labeled DMXR. Each lane contains 75 × 10^-9 M of Exd/Hth and Abd-A. The first lane contains no competitor DNA whereas subsequent lanes contain either 10×, 50×, or 250× of the indicated cold competitor. Schematics at left denote complexes (Exd, blue; Hth, yellow; Abd-A, pink). Graph depicts the average percent of DMXR probe bound from three different experiments in the presence of different amounts of each cold competitor probe. For this analysis, the amount of probe bound was determined using phosphor-imaging densitometry, and 100% binding was assigned to the amount of probe bound in the absence of competitor. Standard error is noted.

Figure 3. Dependence of Hox complex formation on the Exd, Hth, and Hox sites in DMXR2 and RhoA

A. Sequence comparison of the DMXR2, RhoA, and Meis-Hox site previously identified using site selection assays (SELEX) (Shen et al., 1997). Lower case letters in the Meis-Hox SELEX denote nucleotides under less constraint. The Hth (yellow) and Hox (pink) sites are highlighted and mismatches from SELEX consensus are in red text. The Exd binding sites are highlighted in blue. B. The DMXR2 Exd, Hth, and Hox2 point mutations used in gel shift analysis. C. DNA binding competition assays for Abd-A, Exd and Hth complexes on RhoA, RhoA-T>G, DMXR2, and DMXR2-G>T. Labeled RhoA probe was bound with a constant amount (75 × 10^-9 M) of Exd/Hth and Abd-A. Different amounts of competitor were added as indicated. Schematics at left denote color-coded complexes (Exd, blue; Hth, yellow; Abd-A, pink). The amount of probe bound in absence of competitor was assigned 100% binding and the graph (at right) depicts the average amount of probe bound in presence of competitor from three different experiments with standard error noted. D. Graph of DNA binding competition assays of Exd/Hth (75 × 10^-9 M) binding on RhoA, RhoA-T>G, DMXR2, and DMXR2-G>T in the absence of Hox factors. Three experiments were performed and the average amount of probe bound in absence and presence of competitor was compared with standard error noted. E. Comparative EMSAs using wild type and mutant Exd/Hth heterodimers with Abd-A on DMXR2. Equimolar amounts of Exd/Hth, Exd51A/Hth, and Exd/Hth51A proteins (30 × 10^-9 M) were used with three concentrations of Abd-A (7.5 × 10^-9 M, 22.5 × 10^-9 M, and 70 × 10^-9 M). F. Comparative EMSAs using wild type and mutant Exd/Hth heterodimers with Abd-A on RhoA. Equimolar amounts of Exd/Hth, Exd51A/Hth, and Exd/Hth51A proteins (15 × 10^-9 M) were used with three concentrations of Abd-A (7.5 × 10^-9 M, 22.5 × 10^-9 M, and 70 × 10^-9 M). G. Assessing the dependence of Hox complex formation on Exd and Hth binding to DMXR2 and RhoA. Comparative DNA binding assays were performed in triplicate using equimolar amounts of Exd/Hth, Exd51A/Hth, and Exd/Hth51A proteins and Abd-A. The average amount of probe bound by Abd-A and the wild type Exd/Hth proteins was assigned to 100% for each probe tested (blue bar). The amount of probe bound by Exd51A/Hth (red bar) and Exd/Hth51A (yellow bar) compared to wild type was determined. Standard error bars are noted and * denotes significant difference from wild type binding (p-value < 0.001). H. DNA binding competition assays for Abd-A, Exd and Hth complexes on wild type, Exdm, Hthm, and Hox2m DMXR2 probes. Labeled DMXR2 probe was bound with a constant amount (75 × 10^-9 M) of Exd/Hth and Abd-A. Different amounts of competitor were added as indicated. Schematics at left denote color-coded complexes (Exd, blue; Hth, yellow; Abd-A, pink). The amount of probe bound in absence of competitor was assigned 100% binding and the graph (at right) depicts the average amount of probe bound in presence of competitor in three different experiments. I. DNA binding competition assays for Abd-A, Exd and Hth complexes using wild type, Exdm, Hthm, and Hoxm RhoA probes. Labeled RhoA probe was bound with a constant amount (75 × 10^-9 M) of Exd/Hth and Abd-A. The amount of probe bound in absence of competitor was assigned 100% binding and the graph depicts the average amount of probe bound in the presence of competitor in three different experiments.

Figure 4. Role of Hth binding for Hox complex formation on Exd/Hox sites

A. The DMXR1 Hox1, Exd, and Hth point mutations used in gel shift analysis. B. Sequence comparisons of the DMXR1, DMXR1-Con, DMXR1-ΔA, and DMXR1-HoxC probes used in gel shift assays. C. Comparative EMSAs using wild type and mutant Exd/Hth heterodimers with Abd-A on DMXR1. Equimolar amounts of Exd/Hth, Exd51A/Hth, and Exd/Hth51A proteins (30 × 10^-9 M) were used with three concentrations of Abd-A (7.5 × 10^-9 M, 22.5 × 10^-9 M, and 70 × 10^-9 M). D. Dependence of Abd-A/Exd/Hth binding to DMXR1 on the Hox1, Exd, and Hth sites. DNA competition assays were performed using labeled DMXR1 and DMXR1, Hox1m, Exdm, and Hthm probes as cold competitors. The amount of probe bound in absence of competitor was assigned 100% binding and the graph depicts the average amount of probe bound in presence of each competitor from three different experiments. E. Comparative EMSAs using wild type and mutant Exd/Hth heterodimers with Abd-A on DMXR1-Con. Equimolar amounts of Exd/Hth, Exd51A/Hth, and Exd/Hth51A proteins (30 × 10^-9 M) were used with three amounts of Abd-A (7.5 × 10^-9 M, 22.5 × 10^-9 M, and 70 × 10^-9 M). F. Dependence of Abd-A/Exd/Hth binding to DMXR1-Con on the Hox1, Exd, and Hth sites. DNA binding competition assays were performed using labeled DMXR1-Con and different amounts of DMXR1-Con wild type, Hox1m, Exdm, Hthm, and ExdmHthm probes as cold competitors. The amount of probe bound in absence of competitor was assigned 100% binding and the graph depicts the average amount of probe bound in presence of competitor from three different experiments. G. DNA binding competition assays of Abd-A, Exd, and Hth complex formation on the DMXR1, DMXR1-Con, DMXR1-ΔA, and DMXR1-HoxC probes. Labeled DMXR1 probe was bound with a constant amount of Exd/Hth and Abd-A. Different amounts of competitor were added as indicated. Schematics at left denote color-coded complexes (Exd, blue; Hth, yellow; Abd-A, pink). Data from three independent experiments is graphed at right. Note only the DMXR1-Con is significantly different from the wild type DMXR1 (p-value < 0.01) H. Assessing the dependence of Hox complex formation on Exd and Hth binding to DMXR1, DMXR1-Con, DMXR1-ΔA, and DMXR1-HoxC. Comparative DNA binding assays were performed in triplicate using equimolar amounts of Exd/Hth, Exd51A/Hth, or Exd/Hth51A proteins and Abd-A. The amount of probe bound by Abd-A and wild type Exd/Hth was assigned to 100% for each probe tested (blue bar). The amount of probe bound by Exd51A/Hth (red bar) and Exd/Hth51A (yellow bar) compared to wild type was then determined. * denotes a significant difference from wild type binding (p-value < 0.001).

Figure 5. Comparisons between Lab-Exd-Hth binding sites

A. DNA binding competition assays for Lab, Exd and Hth complexes on Lab48/95, EVIII, Hoxb1 R3-PM2, Hoxa2 PM-PH2, Hoxa3-PHP1, and Hoxb2-PP2 probes (see Figure 1 for sequences). Labeled Lab48/95 probe was bound with a constant amount of Exd/Hth (75 × 10^-9 M) and Lab (110 × 10^-9 M). Different amounts of competitor were added as indicated. Schematics at left denote color-coded complexes (Exd, blue; Hth, yellow; Lab, pink). The amount of probe bound in absence of competitor was assigned 100% binding and the graph (at right) depicts the average amount of probe bound in presence of competitor from three different experiments with standard error noted. B. Comparative EMSAs using wild type and mutant Exd/Hth heterodimers with Lab on Lab48/95, EVIII, Hoxb1 R3-PM2, Hoxa2 PM-PH2, Hoxa3-PHP1, and Hoxb2-PP2 probes. Equimolar amounts of Exd/Hth, Exd51A/Hth, and Exd/Hth51A proteins (30 × 10^-9 M) were used with three different amounts of Lab (12 × 10^-9 M, 36 × 10^-9 M, or 110 × 10^-9 M). C. Assessing the dependence of Lab complex formation on Exd and Hth binding to Lab48/95, EVIII, Hoxb1 R3-PM2, Hoxa2 PM-PH2, Hoxa3-PHP1, and Hoxb2-PP2. Comparative DNA binding assays were performed in triplicate using equimolar amounts of Exd/Hth, Exd51A/Hth, and Exd/Hth51A proteins (30 × 10^-9 M) and Lab (110 × 10^-9 M). The amount of probe bound by Abd-A and wild type Exd/Hth was assigned to 100% for each probe tested (blue bar). The amount of probe bound by Exd51A/Hth (red bar) and Exd/Hth51A (yellow bar) compared to wild type was then determined. * denotes a significant difference from wild type binding (p-value < 0.001).

Figure 6. Differences in complex formation between anterior and posterior Hox factors

A. The DNA probes tested for Hox complex formation in gel shift assays using Lab, Scr and Abd-A. Comparisons between the Exd/Hox sites, Hth sites and orientations between sites are highlighted. The percent of probe bound by Lab (110 × 10^-9 M), Scr (100 × 10^-9 M) and Abd-A (75 × 10^-9 M) with a constant amount of Exd/Hth (75 × 10^-9 M) in triplicate is noted. B. Comparative EMSAs on the DMXR1-Con probe using wild type and mutant Exd/Hth heterodimers (30 × 10^-9 M) with Lab (110 × 10^-9 M), Scr (100 × 10^-9 M) or Abd-A (75 × 10^-9 M) as indicated. C-E. Assessing the dependence of Hox complex formation on Exd and Hth binding to DMXR1-Con (C), Lab48/95 (D), and, Hoxa3-PHP1 (E). Comparative DNA binding assays were performed in triplicate using equimolar amounts of Exd/Hth, Exd51A/Hth, and Exd/Hth51A proteins (30 × 10^-9 M) and Lab (110 × 10^-9 M), Scr (100 × 10^-9 M) or Abd-A (75 × 10^-9 M). The amount of probe bound by each Hox factor with wild type Exd/Hth was assigned to 100% (blue bar). The amount of probe bound by Exd51A/Hth (red bar) and Exd/Hth51A (yellow bar) compared to wild type was then determined. * denotes a significant difference from wild type binding (p-value < 0.001).