Regulation of a remote Shh forebrain enhancer by the Six3 homeoprotein

Abstract

In humans, SHH haploinsufficiency results in holoprosencephaly (HPE), a defect in anterior midline formation. Despite the importance of maintaining SHH transcript levels above a critical threshold, we know little about the upstream regulators of SHH expression in the forebrain. Here we describe a rare nucleotide variant located 460 kb upstream of SHH in an individual with HPE that resulted in the loss of Shh brain enhancer-2 (SBE2) activity in the hypothalamus of transgenic mouse embryos. Using a DNA affinity-capture assay, we screened the SBE2 sequence for DNA-binding proteins and identified members of the Six3 and Six6 homeodomain family as candidate regulators of Shh transcription. Six3 showed reduced binding affinity for the mutant compared to the wild-type SBE2 sequence. Moreover, Six3 with HPE-causing alterations failed to bind and activate SBE2. These data suggest a direct link between Six3 and Shh regulation during normal forebrain development and in the pathogenesis of HPE.

Figure 1. SBE2 activity in the rostral hypothalamus is compromised by a sequence variant found in an individual with HPE

(Top) Physical map displaying the distribution of genes (black rectangles) and regulatory sequences (colored rectangles) spanning 1 Mb upstream of SHH on human chromosome 7 (from ref. 13). The sequence tracing to the right is from an individual with lobar HPE who is heterozygous for a C/T transition mutation in SBE2. The mutation resides within a 10 bp-block of SBE2 sequence that was 100% conserved in human, mouse, chicken and frog (red base in boxed sequence alignment). (a-l) X-gal staining of representative embryos carrying wild type SBE2(C) (a, d, g, j), mutant SBE2(T) (b, e, h, k) or a 10 bp deletion SBE2(Δ10 bp) (c, f, i, l) at E10.5. Dashed lines in (a–c) indicate the planes of section shown in panels (d-l). The number of embryos showing representative reporter activity over the total number of transgenic embryos is indicated for each construct (a–c). Abbreviations: SBE, Shh brain enhancer; SFPE, Shh floor plate enhancer; ZRS, zone of polarizing activity regulatory sequence.

Figure 2. Six3/6 proteins bind directly to SBE2

EMSAs performed with Cos-1 cell extracts transfected with flag-tagged Six3 (lanes 3–5,10,11) or Six6 (lanes 6–8,12,13) expression vectors and incubated with SBE2(C) (lanes 1–8), SBE2(T) (9–13) or WEE (14–17) radiolabeled probes. Specific protein/DNA complexes were supershifted in the presence of an αFlag antibody (lanes 4,7,15,17), but not a nonspecific antibody (lanes 5,8). Note that in addition to the supershift, incubation with the αFlag antibody also disrupted Six3/6-SBE2 and Six3/6-WEE complex formation (lanes 4,7,11,13,15,17). The asterisk indicates the formation of a nonspecific complex that is more effectively competed away in the presence of Six3/6 and high affinity probes.

Figure 3. Overlap of Shh and Six3/Six6 expression in the ventral diencephalon

Whole mount (a–e) and transverse sections (f–j) showing the colocalization of Shh (salmon-gal), Six3 and Six6 (alkaline phosphatase) expression in the mouse embryonic forebrain at E10.5. The expression of Shh was monitored using a lacZ reporter line (447L17βlacZ) that recapitulates endogenous Shh expression in the hypothalamus in an SBE2 dependent manner. (k-l) Shh expression (alkaline phosphatase) in wild type and Six6^−/− embryos at E10.5.

Figure 4. Six3/Six6 proteins bind SBE2(C) with higher affinity than SBE2(T)

(a) EMSAs performed with increasing amounts of radiolabeled SBE2(C) (lanes 1–5) and SBE2(T) (lanes 6–10) probes incubated with Cos-1 cell lysates transfected with pCDNA3-Six3-Flag. (b) Dose responsive curves for data shown in (a). Each point along the curve is the average band intensity from three independent experiments (*p<0.05, Student’s t-test). (c) Competitive EMSAs showing the binding of Six3 to a radiolabeled SBE2(C) probe. Cos-1 cell lysates transfected with pCDNA3-Flag (lane 1) or pCDNA3-Six3-Flag (lanes 2–8) were analyzed for binding to a 35 bp probe overlapping wild type SBE2(C). Increasing concentrations of cold wild type SBE2(C) competitor (lanes 3–5), was more efficient at displacing radiolabeled SBE2(C) probe from Six3, compared to increasing concentrations of cold SBE2(T) competitor (lanes 6–8). A nonspecific probe (lane 9) did not significantly alter the shifted complex. (d) Graphical representation of the data in (a). The relative intensities of the retarded bands were quantified and plotted against competitor concentration. Each data point on the curve is an average of five independent experiments. At sub-saturating concentrations of competitor (50x, 100x), SBE2(C) (blue line) was significantly more effective at interfering with complex formation than SBE2(T) (red line) (*p<0.05, Student’s t-test). (e) ChIP from embryos using anti-Six3 or anti-IgG antibodies. QPCR results from three independent experiments reveal a significant enrichment of SBE2 DNA in Six3 versus IgG bound chromatin isolated from forebrain but not posterior trunk regions of E8.75 mouse embryos(*p<0.01, Student’s t-test). A negative control sequence, 6.5 kb downstream of SBE2, was not enriched in Six3 bound chromatin, whereas a positive control sequence, Pax6 SIMO, was enriched to a similar degree as SBE2.

Figure 5. HPE causing mutations in Six3 alter binding and activation of SBE2

(a) Top: Schematic of Six3 protein showing the location of amino acid substitutions resulting from five different HPE causing point mutations (mt1-5) in either the six domain or homeodomain. An additional mutation (mt6) in the groucho interaction domain interferes with Six3 repressor activity but was not identified in HPE patients. Bottom: Cell lysates transfected with pCDNA3-Flag (Lane1), pCDNA3 (wt)Six3-Flag (Lane 2), or pCDNA3 (mt1-6)Six3-Flag (Lanes 3–8) were analyzed for binding to a 35 bp probe overlapping wild type SBE2(C). The Six3/SBE2 complex is indicated with a bracket. Weak or no complex formation was observed for mt1,2 (lanes 3–4) and mt3–5 (lanes 5–7), respectively. Whereas, DNA binding activity was retained by mt6 (lanes 8). αFlag immunoblot demonstrates that wild type and mutant Six3 proteins were expressed at equivalent levels. (b) Six3 regulates SBE2 activity in Cos-1 cells. pCMV-Six3 stimulated wild type SBE2(C)-lacZ expression (black bars), compared to the empty expression plasmid. This transcriptional activation by Six3 was attenuated in cells expressing SBE2(T)-lacZ (red bars). Each bar represents an average of six replicates. (c) SBE2(C)-lacZ activation is compromised by mutations in Six3. Wild type and F88E (mt6) forms of Six3 activated reporter expression while the other Six3 mutants (mt1-5) showed reduced capacity to stimulate SBE2 (black bars). Each bar represents an average of three replicates. Asterisk indicates significant difference from wild type (p<0.001).