Autoregulatory binding sites in the zebrafish six3a promoter region define a new recognition sequence for Six3 proteins

  • Published on

  • View

  • Download

Embed Size (px)


  • Autoregulatory binding sites in the zebrafish six3apromoter region define a new recognition sequence forSix3 proteinsClotilde S. Suh, Staale Ellingsen*, Lars Austb, Xiao-Feng Zhao, Hee-Chan Seo and Anders Fjose

    Department of Molecular Biology, University of Bergen, Norway


    Vertebrate Six3 proteins have important roles during

    the development of eyes and forebrain, and belong to

    the Six Sine oculis family. This family represents adivergent group of the homeodomain (HD) superfam-

    ily of transcription factors [1,2]. The 60 amino acid

    HD, which is a DNA-binding domain, has a conserved

    global fold consisting of three a-helices and a exibleN-terminal arm that becomes more ordered upon

    DNA binding [36]. Binding to specic DNA

    sequences is mediated by interactions between particu-

    lar amino acids in the recognition helix and bases in

    the major groove, and specic contacts between the

    N-terminal arm and the minor groove [4,7,8].

    Specic base contacts in the minor groove involve

    the rst two nucleotides in the TAAT core, and are

    achieved through interactions with residues at posi-


    chromatin; eye development; homeobox;

    transcription factor; transgenic


    A. Fjose, Department of Molecular Biology,

    University of Bergen, PO Box 7803, N-5020

    Bergen, Norway

    Fax: +47 555 89683

    Tel: +47 555 84331


    *Present address

    National Institute of Nutrition and Seafood

    Research, NIFES, PO Box 2029 Nordnes,

    N-5817 Bergen, Norway

    (Received 11 September 2009, revised

    22 December 2009, accepted 29 January



    The homeodomain (HD) transcription factor Six3, which is a member of

    the Six Sine oculis family, is essential for development of the eyes and fore-brain in vertebrates. It has recently been claimed that the HDs of Six3 and

    other members of the Six family have a common recognition sequence,

    TGATAC. However, a different recognition sequence including the typical

    TAAT core motif, which has not yet been fully dened, has also been pro-

    posed for the Six3 HD in mice. Our study of the zebrash orthologue

    six3a, which has an identical HD, shows that it binds in vitro to multiple

    TAAT-containing sites within its promoter region. Comparison of the dif-

    ferent binding afnities for these sequences identies three high-afnity

    sites with a common TAATGTC motif. Notably, this new recognition

    sequence, which is supported by our analysis of the inuence of single-

    nucleotide substitutions on the DNA-binding afnity, is distinct from all

    of the DNA-binding specicities previously described in surveys of HDs.

    In addition, our comparison of Six3a HD binding to the novel TAATGTC

    motif and the common recognition sequence of Six family HDs

    (TGATAC) shows very similar afnities, suggesting two distinct DNA-

    binding modes. Transient reporter assays of the six3a promoter in zebrash

    embryos also indicate that the three high-afnity sites are involved in auto-

    regulation. In support of this, chromatin immunoprecipitation experiments

    show enrichment of Six3a binding to a six3a promoter fragment containing

    two clustered high-afnity sites. These ndings provide strong evidence that

    the TAATGTC motif is an important target sequence for vertebrate Six3

    proteins in vivo.


    ChIP, chromatin immunoprecipitation; EGFP, enhanced green fluorescent protein; EMSA, electrophoretic mobility shift assay; GFP, green

    fluorescent protein; GST, glutathione-S-transferase; HD, homeodomain; hpf, hours postfertilization.

    FEBS Journal 277 (2010) 17611775 2010 The Authors Journal compilation 2010 FEBS 1761

  • tions 2, 3 and 58 in the N-terminal arm. Also, an

    arginine at position 5 is important in most HDs

    [4,5,9]. Similarly, the recognition helix makes specic

    contacts with several nucleotides in the core motif, and

    its residues at positions 47, 50 and 54 also specify two

    adjacent nucleotides 3 of the TAAT core [5,10]. Forexample, HDs containing Lys50 and Gln50 have

    been shown to bind specically to TAATCC and

    TAATGG, respectively [1113]. Recent studies indicate

    that the sequence recognition also depends on a few

    additional anking nucleotides, and this variation in

    specicity may include more than 60 distinct DNA-

    binding activities [14].

    The Six Sine oculis family proteins also have a con-served Six domain of 115119 amino acids involved in

    proteinprotein interactions [1,15], and can be subdi-

    vided into three subfamilies, Six1 2, Six4 5, andSix3 6, on the basis of their HD sequence divergenceand characteristic tetrapeptides in the N-terminal arm

    [16]. The absence of Arg5 in their N-terminal arms

    may explain, in part, why regulatory DNA sequences

    that bind Six1 2 and Six4 5 do not contain the TAATcore [17]. Although the Six3 6 proteins also lack Arg5,their HDs are more distinct from those of the members

    of the other two subfamilies, and various studies have

    suggested that their DNA-binding specicity is differ-

    ent [1,16,18]. In a previous investigation of murine

    Six3, a common TAAT core motif was identied by

    in vitro binding site selection from a randomized pool

    of oligonucleotides, and autoregulatory binding sites

    containing the TAAT core were also identied in the

    promoter of the Six3 gene [18]. However, more recent

    studies have provided evidence that Six3 6 proteinshave similar in vitro DNA-binding specicities to those

    of the other Six family members [10,14]. Further analy-

    sis of the binding afnities of functional Six3 target

    sites and how they function in vivo may help to clarify

    uncertainties regarding the recognition sequences of

    these HD proteins.

    Two orthologues of the murine Six3 gene, six3a and

    six3b, are present in the zebrash, Danio rerio, owing

    to the extra genome duplication that occurred before

    the teleost radiation [15,19]. An additional Six3-like

    gene in zebrash, six7, was probably generated by an

    independent gene duplication event [20]. Several stud-

    ies of Six3 homologues in mouse, sh and Xenopus

    have demonstrated that these genes are essential for

    forebrain and eye development, and their importance

    is also reected in human mutant phenotypes [2125].

    In these processes, Six3 proteins have been shown to

    act both as transcriptional activators and repressors,

    and as regulators of cell proliferation through interac-

    tions with the cell cycle inhibitor Geminin [26,27].

    Studies on Six3 proteins in zebrash have contrib-

    uted to our understanding of their functional roles in

    forebrain and eye development [22,2830], and how

    they can act as transcriptional repressors through

    interactions with members of the Groucho family of

    corepressors [31]. Relatively little is known about the

    regulation of zebrash six3 genes during development

    [32,33], but essential cis-regulatory elements have been

    identied in one of the gene homologues in medaka

    sh [34].

    We have investigated the signicance of the high

    density of TAAT sequences present in the zebrash

    six3a promoter region. Our comparison of the relative

    binding afnities of these potential target sites for the

    Six3a HD identied several strong binding sites that

    dened the sequence TAATGTC as a recognition

    motif. Results from chromatin immunoprecipitation

    (ChIP) experiments and transient reporter assays of

    the six3a promoter in zebrash embryos supported the

    functional role of these high-afnity sites in mediating

    autoregulation. Hence, it is also likely that many of

    the target genes of vertebrate Six3 proteins are recog-

    nized on the basis of high-afnity binding to sequence

    elements containing this motif.


    A 3.6 kb promoter region of six3a recapitulates

    early embryonic expression

    The genomic region upstream of the translational start

    site in zebrash six3a is syntenic with a 4.5 kb pro-

    moter region of the orthologous medaka (Oryzia latipes)

    gene olSix3.2, which contains cis-regulatory elements

    responsible for its spatiotemporal regulation in embryos

    [34]. Additional evidence that the corresponding

    promoter region of zebrash six3a contains cis-acting

    elements required for early expression in the eyes and

    forebrain was obtained from transient expression assays

    with injected reporter constructs [33].

    To analyse the signicance of the six3a promoter

    region, we fused a 3.6 kb genomic fragment to the

    ORF of an enhanced green uorescent protein (EGFP)

    reporter gene in a Tol2 vector (Fig. 1A), and used this

    construct to establish transgenic lines of zebrash (see

    Experimental procedures). From 53 founders crossed

    to wild-type sh, we identied three transgenic lines

    with EGFP expression comparable to that of the

    endogenous six3a gene (data not shown). The trans-

    genic line Tg(3.6S3a:EGFP) was chosen for direct

    comparison of EGFP expression with the spatial distri-

    bution of endogenous six3a transcripts by in situ

    hybridization. At 12 h postfertilization (hpf), EGFP

    New recognition sequence for Six3 C. S. Suh et al.

    1762 FEBS Journal 277 (2010) 17611775 2010 The Authors Journal compilation 2010 FEBS

  • expression was detected in the optic vesicles and ros-

    tral brain, where six3a transcripts were also shown to

    accumulate (Fig. 1B). These results conrm that the

    3.6 kb promoter region included in the six3a:EGFP

    transgene contains regulatory sequences sufcient to

    drive expression mimicking early six3a endogenous


    Differences in Six3a HD binding to TAAT core

    motifs within its promoter region

    The promoters of murine Six3 and human SIX3

    contain autoregulatory binding sites [18,35]. In the case

    of the murine Six3 promoter, it has been shown that

    negative autoregulation involves clustered TAAT core

    motifs and interaction with Groucho-related corepres-

    sors [18]. Sequence analysis of the 3.6 kb promoter

    region of six3a revealed enrichment and clustering of

    the same sequence motif (Fig. 2). Within this promoter

    region, the common core motif is present at 43 posi-

    tions in both orientations (TAAT or ATTA). The ratio

    between TAAT and ATTA on the coding strand is

    25 : 18 (Fig. S1). In initial studies of several of the 18

    ATTA sites by electrophoretic mobility shift assays

    (EMSAs), we observed the strongest shift for the a1

    site (data not shown). Therefore, a1 was selected as a

    reference for comparisons of the binding afnities of

    the 18 different ATTA-containing sites. In this study,

    we used a biotin-labelled 27 bp DNA fragment con-

    taining a1 as a probe in EMSAs, and tested the inu-

    ence of this on HD complex formation in the presence

    of excess amounts of unlabelled fragments representing

    the individual a1a18 sites (Fig. 2B). Whereas a 200-

    fold excess of unlabelled a1 competitor almost

    completely prevented the formation of probeHD

    complexes in EMSAs, only a few of the other sites

    were able to compete signicantly under the same

    conditions. Notably, the competitor representing a2,

    which is located a short distance ( 10 bp) upstreamof a1 (Fig. S1), also caused a strong reduction in

    formation of the probeHD complex.

    In addition to their clustering and high binding

    afnities for the Six3a HD, the anking nucleotides of

    the core ATTA motifs in a1 (G1T2C3A4T5T6A7G8G9)

    and a2 (G1A2T3A4T5T6A7T8G9) have common Gs at

    positions upstream (G1) and downstream (G9). Taking

    into consideration that the binding specicities of HDs

    have been shown to depend mainly on the two nucleo-

    tides 5 to the ATTA core motif [1014], the commonG1 was likely to be important. However, three addi-

    tional sites (a6, a9, and a11), which have a G in the

    same 5-position relative to the ATTA core, showedmuch weaker binding, indicating that other nucleotide

    positions also inuence the binding afnity (Fig. 2B).

    To address the functional importance of the G1 nucle-

    otide, an inspection of the 25 TAAT sites was per-

    formed, and this identied four sites (t2, t13, t15, and

    t17) containing a G in this 5-position. Among thesesites, only t15 showed similar binding afnity to the

    Six3a HD as a1 and a2 (Fig. 2C). This further indi-

    cated that anking nucleotides other than G1 have sig-

    nicant inuence on the binding afnity. Notably,

    three ATTA sites (a10, a14, and a18) without the G1anking nucleotides also bound quite strongly to the

    Six3a HD (Fig. 2B). However, among the 21 TAAT

    sites lacking the G1 nucleotide, none showed signicant

    competition with the a1 probe in EMSAs (Fig. S2).

    Hence, these comparative analyses showed that the

    frequency of high-afnity sites was signicantly higher

    among the G1-containing sites. An additional compari-

    son of the relative strengths of the various high-afnity

    binding sites conducted with lower amounts of

    competitor still showed strongest binding for the

    three G1-containing sites, a1, a2, and t15 (Fig. S3).

    six3a promoter region



    730 bp3635 bp


    B C DTg (3.6S3a:EGFP) six3aEGFP

    Fig. 1. A 3.6 kb promoter region of the

    six3a gene is sufficient to recapitulate its

    early expression. (A) Schematic representa-

    tion of the six3a promoter region including

    the 5-UTR, fused to the coding region ofEGFP. Arrows indicate the position of the

    transcription start site and the initiation

    codon. (B) Lateral view of EGFP expression

    in a Tg(six3a:EGFP) embryo at 12 hpf.

    (C) Detection of EGFP transcripts in a

    Tg(six3a:GFP) embryo by in situ

    hybridization (lateral view of 12 hpf stage).

    (D) Detection of endogenous six3a

    transcripts at 12 hpf (lateral view).

    C. S. Suh et al. New recognition sequence for Six3

    FEBS Journal 277 (2010) 17611775 2010 The Authors Journal compilation 2010 FEBS 1763

  • Therefore, we aimed to investigate the functionality of

    these sites in vivo and the relative importance of the

    different nucleotide positions anking their ATTA core


    Deletion analysis of the six3a promoter indicates

    autoregulatory binding sites

    To determine whether any of the strong Six3a

    HD-binding sites identied by EMSA might have a

    function in vivo, we made several promoterreporter

    constructs with small deletions of regions containing

    particular sites. These constructs, which were made

    from the construct pS3aPG used to make the trans-

    genic line (Fig. 1A; see Experimental procedures), were

    tested in transient reporter assays based on microinjec-

    tion into fertilized eggs and measuring the number of

    EGFP-expressing cells at 12 hpf (Fig. 3). Notably,

    when six3a mRNA and pS3aPG were coinjected, we

    observed a more than two-fold increase in the number

    of EGFP-expressing cells as compared with injection of

    pS3aPG alone. This indicated that overexpression

    of Six3a caused an increase in EGFP expression

    through binding to one or more sites within the

    promoter region in the pS3aPG reporter construct.

    Consistent with the expression pattern of the endoge-










    x 200


    a2 t6 a9 a11

    t2 t13



    x 200











    x 200


















    17 16 15 14 13 12 1110 9 8 7 6 5 4 3 2




    six3a promoter region (pS3aP)

    Fig. 2. Distribution and relative binding affinities of potential Six3a HD target sites within the six3a promoter...


View more >