A screen for co-factors of Six3

A screen for co-factors of Six3

Mechanisms of Development 117 (2002) 103–113 www.elsevier.com/locate/modo A screen for co-factors of Six3 Kristin Tessmar 1, Felix Loosli, Joachim Wi...

481KB Sizes 1 Downloads 30 Views

Mechanisms of Development 117 (2002) 103–113 www.elsevier.com/locate/modo

A screen for co-factors of Six3 Kristin Tessmar 1, Felix Loosli, Joachim Wittbrodt* Developmental Biology Programme, EMBL-Heidelberg, Meyerhofstrasse 1, 69012 Heidelberg, Germany Received 30 November 2001; received in revised form 23 May 2002; accepted 23 May 2002

Abstract The vertebrate Six3 gene a homeobox gene of the Six-family, plays a crucial role in early eye and forebrain development. Here we report the isolation of candidate factors that interact with Six3 in a yeast two-hybrid screen. Among these are two basic helix loop helix (bHLH) domain containing proteins. Biochemical analysis reveals that the bHLH proteins ATH5, ATH3, NEUROD as well as ASH1 interact specifically with XSix3. By defining the interacting domains we show that the bHLH domain of NEUROD interacts with the SIX domain of XSix3. The co-expression of the interacting molecules during late retina determination/differentiation suggests a new role for Six3 and the respective interaction partner also in these late steps of eye development. q 2002 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Six3; NeuroD; ath5; ath3; ash1; Eye; Retina; Determination; Differentiation; Basic helix loop helix; Xenopus; Yeast two hybrid

1. Introduction Homologous genes function in the eyes of Drosophila as well as in those of vertebrates. The Six-/sine oculis family of transcription factors, comprising the Six1/2; Six3/6 and Six4/ 5 subfamilies (Seo et al., 1999, reviewed in Kawakami et al., 2000) is one of the prominent examples for the conservation of gene function. Drosophila sine oculis (Six1/2 subclass) is essential for the development of the larval and adult eye and can induce ectopic compound eyes in cooperation with eyes absent (eya) (Cheyette et al., 1994; Pignoni et al., 1997; Serikaku and O’Tousa, 1994). Drosophila Optix/Six3 (Six3/6 subclass) is involved in the development of the adult compound eye from the eye imaginal disc (Seimiya and Gehring, 2000; Seo et al., 1999). In vertebrates, members of the Six3/6 class are expressed in the anterior central nervous system (CNS) starting during neurulation in anterior neuroectoderm. Their expression is maintained later in the anlagen of the forebrain, the eye anlage, lens placode, olfactory and hypothalamic primordium (Loosli et al., 1998; Seo et al., 1998) and they function during the determination of the retina anlage and retinal proliferation (Loosli et al., 1999; Zuber et al., 1999). In the eye, Six3 is expressed early during retinal determination as well as later during differentiation of the neuroretina (Bovolenta et al., * Corresponding author. Tel.: 1 49-6221-387-576; fax: 149-6221-387166. E-mail address: [email protected] (J. Wittbrodt). 1 Present address: Department of Biochemistry and Biophysics, University of California, 513 Parnassus Avenue, San Francisco, CA 94143, USA.

1998; Kobayashi et al., 1998; Loosli et al., 1998; Oliver et al., 1995; Seo et al., 1998; Zhou et al., 2000). Consistent with its various expression domains, vertebrate Six3 gain-of-function phenotypes range from an enlargement of the forebrain (Kobayashi et al., 1998) and eye to the induction of ectopic optic cups (Loosli et al., 1999). In human patients, mutations in the human Six3 homologue cause holoprosencephaly, a condition leading to the failure of forebrain separation into left and right hemispheres (Pasquier et al., 2000; Wallis and Muenke, 2000; Wallis et al., 1999). Lossof-function experiments in medaka (Oryzias latipes) indicate a dual role for Six3 in early eye development: in forebrain and retina determination as well as in proximo-distal retinal patterning (Carl et al., 2002). It seems unlikely that these diverse effects are mediated by a single transcription factor alone. Rather, specific co-factors and regulatory networks are likely to functionally integrate Six3 at its various expression sites and stages, including eye development. In Drosophila melanogaster the interaction between sine oculis and eyes absent (EYA) has been shown to be necessary for compound eye formation (Chen et al., 1997). Conservation of the interaction between different EYA proteins and members of the SIX1/2 and SIX4/5 subfamilies has been reported, but an interaction with SIX3 was not detected (Heanue et al., 1999; Ohto et al., 1999). We searched for SIX3 interacting proteins by a yeast twohybrid screen. Here we describe the isolation of interacting factors, the expression patterns of which overlap with Six3 during eye and brain development. Among those are basic helix loop helix (bHLH) domain proteins. Their interaction

0925-4773/02/$ - see front matter q 2002 Elsevier Science Ireland Ltd. All rights reserved. PII: S 0925-477 3(02)00185-5

104

K. Tessmar et al. / Mechanisms of Development 117 (2002) 103–113

with SIX3 was further characterised, which revealed that the conserved SIX and bHLH domains interact with each other. Glutathione S-transferase (GST) pull-down analyses with different members of the bHLH domain containing protein family reveal the specificity of this interaction and show that SIX3 interacts only with a subset of the bHLH domain proteins. One of the strongest interactors is NEUROD. In Xenopus, NeuroD and Six3 are co-expressed during cell fate determination and differentiation in the neuroretina, suggesting a novel role for SIX3 in later eye development.

2. Results 2.1. Interaction candidates for Six3 with distinct spatially and temporally overlapping expression profiles Using the entire Six3 open reading frame of the teleost medaka (OlSix3) as bait for a Xenopus yeast two-hybrid library, we selected interactors that are conserved between these species, and possibly among vertebrates. We initially isolated 380 clones that were positive after retransformation and showed no auto-activation. Among those, 75% were highly specific in their interaction with OlSIX3 and neither interacted with mouse SIX2 (mSIX2) nor with chicken SIX6 (cSIX6). Among all 380 clones, we isolated 24 at least twice, independently. Here we describe those genes

(Table 1) that show co-expression with XSix3 as revealed in whole-mount in situ hybridisation analysis in Xenopus laevis (Fig. 1). Our expression analysis complementing the previously published expression data of XclaudinA (Brizuela et al., 2001) and of Xath1 (Kim et al., 1997) indicates that the isolated genes can be divided into two groups with respect to their co-expression with XSix3. The majority shows a temporally extended co-expression with XSix3 (Fig. 1D–I), starting at neurula stages (stage 17) and lasting well into eye differentiation stages (stage 24 and later). The remaining candidates are co-expressed only later during development, starting with stage 24 or later (Fig. 1K–N). A summary of the different candidates isolated, their co-expression with XSix3 and their previously assigned function(s) is given in Table 1. Two of the genes that are strongly co-expressed with XSix3 only during later development show either no significant homology to any gene in the database so far, namely clone #16 (Fig. 1N), or have no designated function yet, as in the case of scoco (Fig. 1M). Both of them are highly specific in their interaction with SIX3, when tested in yeast cells. Among the genes that show early and late co-expression, we isolated XESG1, one of the GROUCHO homologs in Xenopus laevis (Fig. 1G). In accordance with this, an interaction of Groucho3 and SIX proteins of all three subfamilies has recently been described in zebrafish (Kobayashi et al., 2001).

Table 1 SIX3 interaction candidates Clone ID

Accession number of BLAST hit

Co-expression with XSix3

Known function(s)

Xesg1

U18775

Early to late; eye and brain

Homologous to bmal2

AY005163

Early to late; eye and brain

Homologous to Xswi/snf related

U85614, NM_009211

Early to late; eye and brain

Xmaskin

AF200212

Early to late; eye and brain

Homologous to zinc-finger gene 326 in mouse XclnA

NM_018759, AB012725

Early to late; eye and brain

AF359435

Unknown/homologous to chmp1

XM_047526

Transiently early, stage 24 and older in nasal placode region, not in eye anlage Very weak staining around stage 24 in eye region; late strong in eye and brain

Unknown #16 (19AA high homology to ATPase inhibitor) Homologous to short coiled coil gene scoco

BC004955E-value: 7e-09

Only late, eye and brain

Non-DNA binding corepressor (Jimenez et al., 1997); see also (Kobayashi et al., 2001) bHLH domain containing transcription factor, circadian clock system (Okano et al., 2001; Pando et al., 2001) Chromatin remodelling enzyme (Fry and Peterson, 2001) Local translational control of cell division (Groisman et al., 2000) Unknown, associated with nuclear matrix (Lee et al., 1998) Tetraspan transmembrane protein in tight junctions, cell adhesion (Brizuela et al., 2001; Kollmar et al., 2001) Vesicle trafficking protein/ nuclear matrix associated protein affecting chromatin structure and cell-cycle progression (Howard et al., 2001; Stauffer et al., 2001) Unknown

AB015335

Xstriatin

X99326

Stage 24 very weak staining in dorsal diencephalic region; late eye and brain Early very weak, late strong- always eye and brain

Unknown; binds to ARL1 (Van Valkenburgh et al., 2001) Impairment of dendritic, but not axonal growth, calmodulin binding (Bartoli et al., 1999)

K. Tessmar et al. / Mechanisms of Development 117 (2002) 103–113

105

Fig. 1. Comparison of the expression domains of XSix3 and its potential interacting partners. Whole-mount in situ analysis at different developmental stages, lateral views; anterior is to the left. Clone names are indicated in the lower left corner of each picture, and stages in the lower right corner. (A–C) XSix3 expression at stage 17 (inset in (A)), 21 (A) 24 (B) and 37. (D–I) The first two columns show candidate interactors expressed at early and later stages; staining was performed at stages 23/24 and at 17 (small insets). Note that all these genes, except for XclaudinA (I), are expressed in the prospective eye and forebrain region from stage 17 onwards. (K–N) The third column depicts the expression of genes that are co-expressed with XSix3 only later in development (stage 24 or later). Expression domains are shown at stages 37 and 24 or 28 (small insets).

Other genes that show an early and late co-expression with XSix3 are transcriptional regulators, e.g. a chromatinassociated protein (Fig. 1D), which interacted specifically with OlSIX3 and not with mSIX2 or cSIX6. Interestingly, we also isolated two transcription factors containing a bHLH domain. One shows highest homology to the bmal2

gene in other vertebrates, and is co-expressed with XSix3 especially in the eye and brain (Fig. 1E). The other protein was Xenopus atonal1 (XATH1). Previous studies as well as our own whole-mount in situ analysis could not detect any expression of Xath1 in XSix3-expressing areas (Kim et al., 1997, data not shown). However, math1, the murine homo-

106

K. Tessmar et al. / Mechanisms of Development 117 (2002) 103–113

2.2. Interaction of bHLH domain transcription factors with Six3

Fig. 2. XSIX3 interacting with different members of the atonal-related protein family. Pull-down experiments of the GST::XSIX3 fusion protein with different members of the ARP family. Pull-down of XATH3 (lane 1, molecular weight < 40 kDa), XNEUROD (lane 2, < 45 kDa) and XATH5 (lane 3, < 15 kDa) with GST::XSIX3 and for controls with GST alone (lanes 4–6, respectively). Lanes 7–9 show the proteins translated by TNT (input), identity as indicated.

log shows expression in the inner nuclear layer of the eye (Huda Zoghbi, personal communication) where it is coexpressed with mSix3 (Kawakami et al., 1996; Marquardt et al., 2001; Oliver et al., 1995). We thus cannot exclude that the expression of Xath1 in the Xenopus eye has been overlooked so far or that this interaction is relevant only in organisms other than Xenopus laevis. In the yeast assay, XATH1 strongly interacted with OlSIX3 and showed some weaker interaction with mSIX2 (data not shown), suggesting that this interaction is also relevant for other Six family members.

XATH1 is a member of the atonal-related protein family (ARP), which consists of three different subfamilies – termed ATO, NEUROD and NEUROGENIN (Hassan and Bellen, 2000). bHLH proteins closely related to XATH1, namely ATH5/3 and NEUROD have been shown to play an important role during retina and brain differentiation and are prominently coexpressed with XSix3 (Brown et al., 2001; Cepko, 1999; Hutcheson and Vetter, 2001; Kanekar et al., 1997; Kay et al., 2001; Liu et al., 2001; MatterSadzinski et al., 2001; Morrow et al., 1999; Perron et al., 1999; Takebayashi et al., 1997; Wang et al., 2001). Therefore, we investigated whether SIX3 interacts with any of these factors that function in retinal differentiation by pull-down assays with GST-fusion constructs. We found that XSIX3 interacts with members of both, the ATO and NEUROD subfamily, but not with NGNR-1, a member of the neurogenin subfamily (Figs. 2 and 3). We confirmed these results by reciprocally fusing the GST domain to the respective interacting bHLH domain protein and probing the purified proteins for their interaction with XSIX3 (Fig. 4). To test whether the SIX3–bHLH protein interaction is limited to ARPs, we investigated the interaction of XSIX3 with other bHLH family members. We tested XASH1, a member of the achaete–scute family, XESR1, a member of the enhancer of split-related family, and XMAX2, a more distantly related bHLH protein that contains a zincfinger and belongs to the Myc/Max/Mad network of transcriptional regulators. We found that XASH1 interacted, albeit weaker (Fig. 3). This is an interesting finding because Xash1 is specifically coexpressed with Xsix3 at stage 22 in

Fig. 3. XSIX3 interaction with a subset of bHLH domain containing proteins. Pull-down experiments of the GST::XSIX3 fusion protein with different bHLH domain containing proteins. GST::XSIX3 incubated with XNEUROD (lanes 2 and 8, 45 kDa), XMAX2 (lane 4, < 16 kDa), XASH1 (lane 10, < 26 kDa), XESR1 (lane 12, 30 kDa) or XNGNR-1 (lane 14, 40 kDa), showing that only XNEUROD and – weaker – XASH1 interact significantly stronger than the respective band in the control incubations with GST alone (lanes 1, 3, 7, 9, 11, 13). Lanes 5, 6, 15–18 show the respective TNT input.

K. Tessmar et al. / Mechanisms of Development 117 (2002) 103–113

Fig. 4. Interaction of different ARPs with XSIX3. Reverse GST fusions were used to re-confirm interactions in pull-down experiments. GST was fused to XATH3, XNEUROD and XATH5. Incubation of any of these GST fusion proteins with XSIX3 (lanes 1, 3, 5, apparent size: < 35 kDa) leads to stronger interaction than the corresponding controls with GST alone (lanes 2, 4, 6). Lane 7 shows the XSIX3 TNT input.

the presumptive diencephalon and at stage 23 in the eye anlagen (Ferreiro et al., 1993). However, we did not find interaction of either XESR1 or XMAX2 in GST pull-down assays (Fig. 3) indicating that the interaction with SIX3 is not a general feature of bHLH domain proteins. Taken together, our data indicate that XSIX3 specifically interacts with members of the ATO and NEUROD subfamilies of ARPs and with a member of the ASH family, all shown to function as positive regulators of retina cell determination/differentiation (reviewed in Cepko, 1999). Among the ARPs coexpressed with Six3 (see below), NeuroD and XATH5/3 showed the strongest interaction with XSIX3. We therefore focussed the following analysis on Xath5 and in more detail on NeuroD. 2.3. XNeuroD and XSix3 are co-expressed in the dorsal diencephalon and in the eye In order to further determine the time frame in which the interaction between XSIX3 and ARPs might play a role in retina development, we chose one prominent member of this group, NeuroD that resembles the expression of Xath3 and Xath5 in the developing retina (Kanekar et al., 1997; Perron et al., 1999) and performed double labelling whole-mount in situ hybridisation studies (Fig. 5). The expression patterns of ARPs are well described (Kanekar et al., 1997; Perron et al., 1999) so we focussed on the co-expression of Six3 and Xath5 as well as NeuroD in the developing retina. The expression of Xath5 (Kanekar et al., 1997; Perron et al., 1999) prominently overlaps with that of Six3 in the developing retina (Fig. 5A–C). At stage 28, coexpression is detected in the retinal ganglion cell layer and in bipolar cells. XNeuroD starts to be expressed around stage 14. Coexpression with XSix3 is not detected prior to stage 24 (Fig. 5D–F). At stage 24, expression of XNeuroD can be detected in the optic vesicle and in the dorsal region of the diencephalon, presumably in the pineal organ, both regions where XSix3 is expressed (Fig. 5G–I and Ghanbari et al., 2001). Interestingly, only isolated, scattered cells in the eye coexpress XNeuroD and XSix3 at this stage (Fig. 5G, H). During subsequent stages, the expression domain of

107

XNeuroD extends, covering almost the entire retina at stage 28 and vanishes from the central region almost completely until stage 37. Since the expression of XSix3 is confined mainly to the prospective retinal ganglion cell layer and inner nuclear layer of the neuroretina until stage 37 (Fig. 5M–O), the strongest co-expression of XNeuroD and XSix3 can be observed around stage 28 (arrowhead in Fig. 5L). The genes remain co-expressed in the marginal regions of the stage 37 retina (arrowheads in Fig. 5O). The RGC layer expresses only XSix3 at this stage. In summary, we observed that XNeuroD and XSix3 are coexpressed in two domains i.e. the pineal organ and the developing eye from stage 24 onwards. Notably, this is when retinal cells are still being determined, and are just about to become post-mitotic and start differentiation (Chang and Harris, 1998; Holt et al., 1988). The expression domains of both genes separate again towards the end of differentiation. 2.4. The SIX domain of XSIX3 is sufficient to interact with XNEUROD In order to define the domains necessary for the XNEUROD–XSIX3 interaction, we generated deletion constructs with domains of the XSIX3 protein fused to GST (Fig. 6B). The resulting proteins were tested for their interaction with XNEUROD in a GST pull-down experiment (Fig. 6A). We found that the N-terminus and C-terminus of XSIX3 did not interact with XNEUROD (Fig. 6A, lanes 6 and 7), suggesting that both domains alone are not sufficient for the interaction. In contrast, four other constructs, containing the SIX domain (SD) alone or in combination with other domains, did interact with XNEUROD (Fig. 6A, lanes 2–5). Thus, the SD of XSIX3 is sufficient for the interaction with XNEUROD, whereas the N- and C-termini are not required for the interaction. 2.5. Both the N-terminus and the bHLH domain of XNEUROD can interact with XSIX3 We next tested which region(s) of the XNEUROD protein can interact with XSIX3. For this we constructed several GST::XNEUROD fusion proteins (Fig. 7B). These were tested in GST pull-down assays for their interaction with XSIX3 (Fig. 7A). GST-fusions of either the N-terminus or the bHLH domain of XNEUROD, as well as a fusion construct that contained both of these domains, were able to interact with the full-length XSIX3 protein (Fig. 7A, lanes 1–3), arguing that both the N-terminus and the bHLH domain are sufficient for the interaction. This shows that the SIX domain and the bHLH domain, both highly conserved in evolution, are sufficient to interact with the full-length XNEUROD or XSIX3 protein, respectively. The less conserved N-terminus of XNEUROD also contributes to this interaction.

108

K. Tessmar et al. / Mechanisms of Development 117 (2002) 103–113

Fig. 5. Co-expression of XSix3 and XneuroD. Double whole-mount in situ hybridisation detecting XSix3 (blue) and Xath5 (red, A–C) and XNeuroD (red, D–O). Anterior to the left, dorsal (A, C, D, G, J, M), lateral (B, E, H, K, N) and frontal views (F, I). Stages as indicated. (A–C) Full overlap of expression Xath5 and XSix3 detectable only in embryos weakly stained for XSix3. Compare A and J for XSix3. (D–F) No overlap of expression detected at stage 21. (G–I) Coexpression first detectable at stage 24. Note that at this stage only scattered individual cells express XNeuroD in a broad XSix3 domain (arrow in inset in H). In addition, the presumptive pineal organ expresses both genes (arrowhead in H and K, compare this to Fig. 1B). (J–L) At stage 28, co-expression broadens, comprising almost the entire retina (arrowhead in L). (M–O) At stage 37, co-expression becomes restricted to marginal regions (arrowheads in O). (L, O) Transversal sections at levels indicated in J and M, dorsal to the left, inset shows NeuroD single staining.

K. Tessmar et al. / Mechanisms of Development 117 (2002) 103–113

Fig. 6. XNEUROD interacts with the SIX domain of XSIX3 (A) Pull-down experiment to determine the domains of SIX3 interacting with XNEUROD. Different domains of XSIX3 fused to GST were tested for their interaction with XNEUROD. All constructs containing the SIX domain interacted with XNEUROD, the N-terminus plus SIX domain (GST::XSIX3N-term 1 SD, lane 2), SIX domain plus HOMEO domain and C-terminus (GST::XSIX3SD 1 HD 1 C-term, lane 3), SIX domain plus HOMEO domain and N-terminus (GST::XSIX3N-term 1 SD 1 HD, lane 4) and the SIX domain alone (GST::XSIX3SD, lane 5). No interaction in GST control (lane 1), N-terminus (GST::XSIX3N-term, lane 6) and C-terminus (GST::XSIX3C-term, lane 7). Lane 8 shows the TNT protein input. (B) Schematic representation of the different fusion constructs and their interaction with XNEUROD. Corresponding lane in A is indicated.

109

coexpression only at early stages, suggesting that the same interactors mediate early and late functions of SIX3. Two categories of candidate interactors could be identified: those that show early and late co-expression, and genes that show only late co-expression in the developing forebrain and/or eye. Among the early and late co-expressing candidates, we isolated XESG1, a member of the groucho family of transcriptional co-repressors. Members of the GROUCHO family have recently been reported to interact with SIX protein family members (Kobayashi et al., 2001). However, since members of all SIX subfamilies showed this interaction, GROUCHO is not likely to contribute alone to the functional diversity that different SIX-subfamily members display in gain-of-function studies (Loosli et al., 1999). Apart from the early and late co-expressing candidates, we isolated factors that are co-expressed with XSix3 only later in development. This hints at a specific function of Six3 during later development. Among those candidates are several proteins of unknown function albeit interesting coexpression patterns, e.g. scoco, as well as some members of the atonal-related protein class and related bHLH transcription factors, which we investigated in more detail as they are known to play key roles during retinal differentiation.

3. Discussion 3.1. A screen for interacting partners of Six3 Although it has been shown that Six3 has important functions in the development of eye and forebrain structures (Loosli et al., 1999; Wallis et al., 1999), an interaction network in which it is integrated to convey these functions is largely unknown. In this manuscript, we report the isolation of 11 interacting proteins from a yeast two-hybrid screen. These numbers represent a substantial enrichment for clones expressed within the Six3 expression domain, when compared to expression patterns of randomly picked clones (Wittbrodt and Wittbrodt, unpublished observation), and thus are a strong indication that our screen was successful in identifying interacting partners of SIX3. All these, except one, are co-expressed with XSix3 at a early (neurulation) and/or later stage 24 or later) stages in the developing brain and or eye of Xenopus laevis. Interestingly, none of the interacting proteins identified shows

Fig. 7. XSIX3 interacts with the N-terminus and bHLH domain of XNEUROD. (A) Pull-down experiment to determine the domains of XNEUROD interacting with SIX3. Different domains of XNEUROD fused to GST were tested for their interaction with XSIX3. The N-terminus alone (GST::XNDN-term, lane 1), N-terminus together with bHLH domain (GST:: XNDN-term 1 bHLH, lane 2) and the bHLH domain alone (GST::XNDbHLH, lane3) interacted with XSIX3 but not the GST control (lane 4). Lane 5 shows the TNT protein input. (B) Schematic representation of the different fusion constructs and their interaction with XSIX3. Corresponding lane in A is indicated.

110

K. Tessmar et al. / Mechanisms of Development 117 (2002) 103–113

3.2. A possible late role for Six3 in retina differentiation XSIX3 can specifically interact with XNEUROD, XATH3 and XATH5. All these genes are involved in determination/differentiation processes of retina cells and form a regulatory network among each other and with XNGNR-1. In this network, XNGNR-1 regulates XNEUROD, XATH3 and XNEUROD regulate each other’s expressions, and XNEUROD regulates XATH5 (Kanekar et al., 1997; Kay et al., 2001; Koyano-Nakagawa et al., 1999; Liu et al., 2001; Perron et al., 1999). While we find an interaction of SIX3 with XNEUROD, XATH3 and XATH5, we show that, under the same conditions, SIX3 does not interact with XNGNR-1, a member of the neurogeninin subfamily of ARPs. Xngnr-1 is expressed at more peripheral regions than XNeuroD, Xath3 and Xath5 in the ciliary marginal zone of Xenopus (Perron et al., 1998, 1999) and plays a role in the initiation retinogenesis. SIX3 interacts specifically with those factors that act later than XNGNR-1 (Perron et al., 1998) during the subsequent steps of retinal differentiation. In agreement with a late function, we observe that coexpression of XNeuroD or Xath5 and XSix3 in the eye starts around the time when cell fate determination is still in progress, but differentiation is initiated. At the end of differentiation, their coexpression vanishes, suggesting that their interaction with SIX3 is important for the determination and/or differentiation of distinct cell types in the retina. This finding is in accordance with the notion that in a conditional inactivation of murine Pax6 in retinal progenitor cells, co-expression of Six3 and NeuroD coincides with the exclusive generation of amacrine cells. Therefore, Six3 is discussed to permit amacrine cell fate in the presence of NeuroD (Marquardt et al., 2001). Xath3 and Xath5 show expression patterns similar to XNeuroD in the developing neuroretina (Kanekar et al., 1997; Perron et al., 1999), suggesting that these proteins likewise form part of the determination/differentiation network of the eye. The interaction of SIX3 with a specific combination of ARPs thus may specify distinct cell types of the neuroretina. On the other hand, differentiation requires a stop of proliferation on the one hand and the expression of cell type specific differentiation genes. Therefore, the ARP/ XSIX3 interaction should initiate, directly or indirectly, a proliferation-stop signal. As Six3 on its own has been shown to stimulate proliferation (Kobayashi et al., 1998; Loosli et al., 1999) it is tempting to speculate that it is the interaction of SIX3 with XNEUROD, XATH3 or XATH5 that abolishes its proliferative activity, to promote differentiation in those cells of the retina that co-express these ARPs. One further aspect of our study is the question of evolutionary conservation of these interactions. By performing a cross-species screening experiment, we had already selected interaction partners that were presumably conserved between two different vertebrates. Our domain-mapping experiments further support this conservation. Since the

conserved bHLH domain of XNEUROD interacted with XSIX3, and XSIX3 interacted with XNEUROD via the conserved SIX domain, it is reasonable to speculate that similar interactions might take place in other, non-vertebrate organisms. On the other hand, the interaction between the N-terminus of NEUROD and SIX3 appears to be a specific feature of the NEUROD subfamily, since no conserved domain could be detected at the amino acid sequence level. The interaction of XSIX3 with the non-atonal related bHLH protein XASH1, but not with XESR1 or the more distantly related protein XMAX2, clearly indicates specific interactions with other non-ARP bHLH transcription factors, the relevance of which will be addressed in future experiments. 4. Experimental procedures 4.1. Animals Fertilised eggs were obtained from pigmented and albino Xenopus laevis injected with 500 U of human chorionic gonadotropin (Sigma) to induce egg laying. Embryos were de-jellied in 2% cysteine (pH 7.5) and staged according to Nieuwkoop and Faber (1994). 4.2. Yeast two-hybrid analysis In order to construct the baits, different OlSix3 fragments and the complete mSix2 open reading frame (ORF) were amplified by polymerase chain reaction (PCR) such that an artificial EcoRI restriction site at the 5 0 end and a BglII restriction site at the 3 0 end flanked the product. Following digestion with EcoRI and BglII, the fragments were cloned in frame with the Gal4BD into pGBDU-C3, containing the Ura3 gene (James et al., 1996). The constructs were verified by sequencing. The primer sequences are as follows: Complete OlSix3::Gal4BD:upper primer: 5 0 -ggaattcgttttcagagctccgctt-3 0 , lower primer: 5 0 -gaagatctctgacatccaagtcagagtca-3 0 complete mSix2::Gal4BD: upper primer: 5 0 ggaattcctgcccaccttcggcttca-3 0 , lower primer: 5 0 -gaagatctctggagcccaggtccacaag-3 0 The complete cSix6 bait was provided by J. Lopez-Rios. A yeast two-hybrid library, 145 mg, generated by S. Pierce (D. Kimelman’s laboratory) and M. Chen (J.A. Cooper’s laboratory) (Yost et al., 1998) was used for transformation into the PJ69-4A yeast strain as described in the ‘High Efficiency Transformation Protocol’ {(Agatep et al., 1998); see also http://www.umanitoba.ca/faculties/medicine/biochem/gietz/method.html} yielding 1.5 £ 10 7 transformed yeast cells in the case of harbouring the complete OlSix3 gene as a bait construct. The vector used for this library was f1-VP16, containing theLeu2 gene. The

K. Tessmar et al. / Mechanisms of Development 117 (2002) 103–113

complexity of the library was given with 7.5 £ 10 6independent clones. For all other cases of yeast transformations, we used the ‘Quick and Easy TRAFO Protocol’{(Gietz, 1994); see also http://www. umanitoba. ca/faculties/medicine/biochem/ gietz/Quick.html}. 6.7 £ 10 5 transformed yeast cells were plated onto a SD- LEU 2, URA 2, ADE 2 9 cm plate and grown at 308C for 7 days. The plates were daily checked for growing colonies. Single growing colonies were transferred to SD-LEU 2 plates and, once grown bigger, replicated onto SD- LEU 2, URA 2, HIS 2containing 3 mM 3aminotriazol (3-AT) plates to reconfirm the interaction. Yeast cells growing on both SD- LEU 2, URA 2, ADE 2 and SD- LEU 2, URA 2, HIS 2plates were forced to lose the bait plasmid by exposing them several times (4–5 times) to 1 mg/ml 5-Fluoroorotat (5-FOA) until no growth on SD- URA 2 plates could be observed. Interactions were reconfirmed by remating these cells with a PJ69-4a strain containing the bait plasmid. To test for autoactivation, cells were remated with cell containing the empty pGBDU-C3 vector. Reconfirmed, non-autoactivating prey plasmids were isolated with the ‘RPM Yeast Plasmid Isolation Kit’ (#2069-400 from Q*Biogene) with two modifications. Yeast cells were grown for 3–4 days and DNA was eluted in 40 ml. After reisolation from Escherichia coli cells the interaction was reconfirmed in yeast with the plasmids yielded from these preparations. 4.3. GST pull-down constructs In order to construct the GST::XNEUROD and GST::XSIX3 fusion constructs, the various XNeuroD or XSix3 fragments were amplified by PCR such that an artificial EcoRI restriction site at the 5 0 end and a XhoI restriction site at the 3 0 end flanked the product. Primer sequences are as follows:

For XNeuroD-fusions: Complete XGST::NEUROD: upper primer: 5 0 -acggaattctcaccaaatcgtatggagagaat-3 0 , lower primer: 5 0 -cgactcgagatttaaaggagtgtcgattgg-3 0 ; N-terminus: upper primer: 5 0 -acggaattctcaccaaatcgtatggagagaat3 0 , lower primer: 5 0 -cgactcgagctacactttaaatcgctccacccg3 0 ; N-terminus 1 bHLH domain: upper primer: 5 0 acggaattctcaccaaatcgtatggagagaat-3 0 , lower primer: 5 0 cgactcgagctaagaaagagcccagatgtagtt-3 0 ; bHLH domain: upper primer: 5 0 -acggaattctcagacgcatgaaggcaaacgcc-3 0 , lower primer: 5 0 -cgactcgagctaagaaagagcccagatgtagtt-3 0 For XSix3-fusions: Complete GST::XSIX3: upper primer: 5 0 -caggaattctggtgttcaggtcccctctagag-3 0 , lower primer: 5 0 -ggcctcgagccataggagccctgatctgcc-3 0 ; N-terminus: upper primer: 5 0 caggaattctggtgttcaggtcccctctagag-3 0 lower primer: 5 0 cgactcgagctacattgacagctcgtcctgggc-3 0 ; N-terminus 1 SD: upper primer: 5 0 -caggaattctggtgttcaggtcccctctagag-

111

3 0 , lower primer: 5 0 -cgactcgagctagtcccagatggttcggggcag3 0 ; N-terminus 1 SD 1 HD: upper primer: 5 0 caggaattctggtgttcaggtcccctctagag-3 0 , lower primer: 5 0 cgactcgagctaggcggccgctctgtccctctg-3 0 ; SD: upper primer: 5 0 -acggaattctcttccacctgccgagcctgaac-3 0 , lower primer: 5 0 cgactcgagctagtcccagatggttcggggcag-3 0 ; SD 1 HD 1 Cterminus: upper primer: 5 0 -acggaattctcttccacctgccgagcctgaac-3 0 , lower primer: 5 0 -ggcctcgagtcatacgtcacattcagagtc-3 0 ; C-terminus: upper primer: 5 0 -acggaattctcgctaaaaacaggcttcagcac-3 0 , lower primer: 5 0 -ggcctcgagtcatacgtcacattcagagtc-3 0 . Following digestion with EcoRI and XhoI, the fragments were cloned in frame with GST into pGEX–KG obtained from Amersham Pharmacia Biotech. The in-frame fusion was verified by sequencing. 4.4. In vitro glutathione S-transferase -fusion protein pulldown assays GST and GST fusion proteins were isolated from the E. coli strain BL21 (DE3) 1 RP after induction with 0.5 mM IPTG followed by 4–6 h of incubation at 308C. The cells were harvested in cold phosphate-buffered saline (PBS) containing 1 mM phenyl-methyl-sulfonyl-fluoride (PMSF) and 1 mg/ml lysozyme. After sonication, Triton X-100 was added to a final concentration of 1%. After removal of DNA via centrifugation, the supernatant was aliquoted and stored at 2808C. Volumes (100–600 ml) corresponding to approximately even amounts of individual GST/GST fusion protein were coupled to washed glutathione sepharose beads (20 ml suspension), washed twice with PBS and once with IPbuffer (50 mM Tris–HCl pH 7.5; 150 mM NaCl; 0.5% NP40; 5 mM EGTA pH 8.0; 5 mM EDTA pH 8.0; 20 mM NaF) both buffers containing 1.5% bovine serum albumin (BSA) and 1 mM PMSF. Coupled beads were incubated with 20 ml of in vitro translated protein (using a SP6 TNT coupled transcription/ translation system with Sp6 polymerase (Promega) and 35S-methionine according to manufacturer’s instructions) in IP-buffer containing complete proteinase inhibitor cocktail (Roche) and 1.5% BSA for 2 h at 48C with shaking. The beads were washed four times with IP-buffer, resuspended in 20 ml of electrophoresis sample buffer and heated (958C) for 5 min. Following SDS polyacrylamide gel electrophoresis, the gel was checked for roughly equal GST/GST-fusion protein amounts by Coomassie staining. Non-GSTproteins were visualised by audioradiography. 4.5. Whole-mount in situ analysis Whole-mount in situ hybridisations were performed as described in standard protocols (Harland, 1991) with a change in the hybridisation temperature, which was increased to 658C.Vibratome sections were done following standard procedures at a thickness of 40 mm (Loosli et al., 1998).

112

K. Tessmar et al. / Mechanisms of Development 117 (2002) 103–113

Acknowledgements The authors thank Mike Zuber, Shin-ichi Ohnuma, and Bill Harris, the lab members of the Harris, Nebreda and Bouwmeester labs for constructs, reagents, support and help. We thank Detlev Arendt, Florian Raible, Filippo del Bene and Matthias Carl for critically reading the manuscript and all the members of the Wittbrodt lab for support, Andreas Jenny and Javier Lopez-Rios for helpful suggestions on the yeast two-hybrid screen, Kristian Tersar for GST fusion constructs and Annette Krone for technical assistance. This work was supported through a scholarship of the German National Scholarship Foundation (K.T.) and through grants from the European Commission and the HFSPO (J.W). References R. Agatep, R.D. Kirkpatrick, D.L. Parchaliuk, R.A. Woods, R.D. Gietz. Transformation of Saccharomyces cerevisiae by the lithium acetate/ single-stranded carrier DNA/polyethylene glycol (LiAc/ss-DNA/PEG) protocol. Technical Tips Online (http://tto.trends.com). Bartoli, M., Ternaux, J.P., Forni, C., Portalier, P., Salin, P., Amalric, M., Monneron, A., 1999. Down-regulation of STRIATIN, a neuronal calmodulin-binding protein, impairs rat locomotor activity. J. Neurobiol. 40, 234–243. Bovolenta, P., Mallamaci, A., Puelles, L., Boncinelli, E., 1998. Expression pattern of cSix3, a member of the six/sine oculis family of transcription factors. Mech. Dev. 70, 201–203. Brizuela, B.J., Wessely, O., De Robertis, E.M., 2001. Overexpression of the Xenopus tight-junction protein CLAUDIN causes randomization of the left–right body axis. Dev. Biol. 230, 217–229. Brown, N.L., Patel, S., Brzezinski, J., Glaser, T., 2001. Math5 is required for retinal ganglion cell and optic nerve formation. Development 128, 2497–2508. Carl, M., Loosli, F., Wittbrodt, J., 2002. Six3 is essential for the formation and patterning of the vertebrate eye. Development (in press). Cepko, C.L., 1999. The roles of intrinsic and extrinsic cues and bHLH genes in the determination of retinal cell fates. Curr. Opin. Neurobiol. 9, 37–46. Chang, W.S., Harris, W.A., 1998. Sequential genesis and determination of cone and rod photoreceptors in Xenopus. J. Neurobiol. 35, 227–244. Chen, R., Amoui, M., Zhang, Z., Mardon, G., 1997. Dachshund and eyes absent proteins form a complex and function synergistically to induce ectopic eye development in Drosophila. Cell 91, 893–903. Cheyette, B.N.R., Green, P.J., Martin, K., Garren, H., Hartenstein, V., Zipursky, S.L., 1994. The Drosophila sine oculis locus encodes a homeodomain-containing protein required for the development of the entire visual system. Neuron 12, 977–996. Ferreiro, B., Skoglund, P., Bailey, A., Dorsky, R., Harris, W.A., 1993. XASH1, a Xenopus homolog of achaete–scute: a proneural gene in anterior regions of the vertebrate CNS. Mech. Dev. 40, 25–36. Fry, C.J., Peterson, C.L., 2001. Chromatin remodeling enzymes: who’s on first? Curr. Biol. 11, R185–R197. Ghanbari, H., Seo, H.C., Fjose, A., Brandli, A.W., 2001. Molecular cloning and embryonic expression of Xenopus Six homeobox genes. Mech. Dev. 101, 271–277. Gietz, R.D., Woods, R.A., 1994. High efficiency transformation in yeast. In: Johnston, J.A. (Ed.). Molecular Genetics of Yeast: Practical Approaches, Oxford University Press, Oxford, pp. 121–134 (invited book chapter). Groisman, I., Huang, Y.S., Mendez, R., Cao, Q., Theurkauf, W., Richter, J.D., 2000. CPEB, maskin, and cyclin B1 mRNA at the mitotic appa-

ratus: implications for local translational control of cell division. Cell 103, 435–447. Harland, R.M., 1991. In situ hybridization: an improved whole-mount method for Xenopus embryos. Methods Cell Biol. 36, 685–695. Hassan, B.A., Bellen, H.J., 2000. Doing the MATH: is the mouse a good model for fly development? Genes Dev. 14, 1852–1865. Heanue, T.A., Reshef, R., Davis, R.J., Mardon, G., Oliver, G., Tomarev, S., Lassar, A.B., Tabin, C.J., 1999. Synergistic regulation of vertebrate muscle development by Dach2, Eya2, and Six1, homologs of genes required for Drosophila eye formation. Genes Dev. 13, 3231–3243. Holt, C.E., Bertsch, T.W., Ellis, H.M., Harris, W.A., 1988. Cellular determination in the Xenopus retina is independent of lineage and birth date. Neuron 1, 15–26. Howard, T.L., Stauffer, D.R., Degnin, C.R., Hollenberg, S.M., 2001. CHMP1 functions as a member of a newly defined family of vesicle trafficking proteins. J. Cell Sci. 114, 2395–2404. Hutcheson, D.A., Vetter, M.L., 2001. The bHLH factors Xath5 and XNeuroD can upregulate the expression of XBrn3d, a POU-homeodomain transcription factor. Dev. Biol. 232, 327–338. James, P., Halladay, J., Craig, E.A., 1996. Genomic libraries and a host strain designed for highly efficient two-hybrid selection in yeast. Genetics 144, 1425–1436. Jimenez, G., Paroush, Z., Ish-Horowicz, D., 1997. Groucho acts as a corepressor for a subset of negative regulators, including Hairy and Engrailed. Genes Dev. 11, 3072–3082. Kanekar, S., Perron, M., Dorsky, R., Harris, W.A., Jan, L.Y., Jan, Y.N., Vetter, M.L., 1997. Xath5 participates in a network of bHLH genes in the developing Xenopus retina. Neuron 19, 981–994 (published erratum appears in Neuron 21(5) (1998) 1221). Kawakami, K., Ohto, H., Takizawa, T., Saito, T., 1996. Identification and expression of six family genes in mouse retina. FEBS Lett. 393, 259– 263. Kawakami, K., Sato, S., Ozaki, H., Ikeda, K., 2000. Six family genesstructure and function as transcription factors and their roles in development. Bioessays 22, 616–626. Kay, J.N., Finger-Baier, K.C., Roeser, T., Staub, W., Baier, H., 2001. Retinal ganglion cell genesis requires lakritz, a zebrafish atonal homolog. Neuron 30, 725–736. Kim, P., Helms, A.W., Johnson, J.E., Zimmerman, K., 1997. XATH-1, a vertebrate homolog of Drosophila atonal, induces a neuronal differentiation within ectodermal progenitors. Dev. Biol. 187, 1–12. Kobayashi, M., Nishikawa, K., Suzuki, T., Yamamoto, M., 2001. The homeobox protein Six3 interacts with the Groucho corepressor and acts as a transcriptional repressor in eye and forebrain formation. Dev. Biol. 232, 315–326. Kobayashi, M., Toyama, R., Takeda, H., Dawid, I.B., Kawakami, K., 1998. Overexpression of the forebrain-specific homeobox gene six3 induces rostral forebrain enlargement in zebrafish. Development 125, 2973– 2982. Kollmar, R., Nakamura, S.K., Kappler, J.A., Hudspeth, A.J., 2001. Expression and phylogeny of claudins in vertebrate primordia. Proc. Natl Acad. Sci. USA 98, 10196–10201. Koyano-Nakagawa, N., Wettstein, D., Kintner, C., 1999. Activation of Xenopus genes required for lateral inhibition and neuronal differentiation during primary neurogenesis. Mol. Cell Neurosci. 14, 327–339. Lee, J.Y., Kambe, M., Hayashi, M., Takenaga, K., 1998. Cloning and characterization of a novel zinc finger protein that associates with nuclear matrix. DNA Cell Biol. 17, 849–858. Liu, W., Mo, Z., Xiang, M., 2001. The Ath5 proneural genes function upstream of Brn3 POU domain transcription factor genes to promote retinal ganglion cell development. Proc. Natl Acad. Sci. USA 98, 1649– 1654. Loosli, F., Ko¨ ster, R.W., Carl, M., Krone, A., Wittbrodt, J., 1998. Six3, a medaka homologue of the Drosophila homeobox gene sine oculis is expressed in the anterior embryonic shield and the developing eye. Mech. Dev. 74, 159–164.

K. Tessmar et al. / Mechanisms of Development 117 (2002) 103–113 Loosli, F., Winkler, S., Wittbrodt, J., 1999. Six3 overexpression initiates the formation of ectopic retina. Genes Dev. 13, 649–654. Marquardt, T., Ashery-Padan, R., Andrejewski, N., Scardigli, R., Guillemot, F., Gruss, P., 2001. Pax6 is required for the multipotent state of retinal progenitor cells. Cell 105, 43–55. Matter-Sadzinski, L., Matter, J.M., Ong, M.T., Hernandez, J., Ballivet, M., 2001. Specification of neurotransmitter receptor identity in developing retina: the chick ATH5 promoter integrates the positive and negative effects of several bHLH proteins. Development 128, 217–231. Morrow, E.M., Furukawa, T., Lee, J.E., Cepko, C.L., 1999. NeuroD regulates multiple functions in the developing neural retina in rodent. Development 126, 23–36. Nieuwkoop, P.D., Faber, J., 1994. Normal Table of Xenopus laevis (Daudin), Garland Publishing Inc, New York, NY. Ohto, H., Kamada, S., Tago, K., Tominaga, S.I., Ozaki, H., Sato, S., Kawakami, K., 1999. Cooperation of six and eya in activation of their target genes through nuclear translocation of Eya. Mol. Cell Biol. 19, 6815– 6824. Okano, T., Yamamoto, K., Okano, K., Hirota, T., Kasahara, T., Sasaki, M., Takanaka, Y., Fukada, Y., 2001. Chicken pineal clock genes: implication of bmal2 as a bidirectional regulator in circadian clock oscillation. Genes Cells 6, 825–836. Oliver, G., Mailhos, A., Wehr, R., Copeland, N.G., Jenkins, N.A., Gruss, P., 1995. Six3, a murine homologue of the sine oculis gene, demarcates the most anterior border of the developing neural plate and is expressed during eye development. Development 121, 4045–4055. Pando, M.P., Pinchak, A.B., Cermakian, N., Sassone-Corsi, P., 2001. A cell-based system that recapitulates the dynamic light-dependent regulation of the vertebrate clock. Proc. Natl Acad. Sci. USA 98, 10178– 10183. Pasquier, L., Dubourg, C., Blayau, M., Lazaro, L., Le Marec, B., David, V., Odent, S., 2000. A new mutation in the six-domain of Six3 gene causes holoprosencephaly. Eur. J. Hum. Genet. 8, 797–800 (in process citation). Perron, M., Kanekar, S., Vetter, M.L., Harris, W.A., 1998. The genetic sequence of retinal development in the ciliary margin of the Xenopus eye. Dev. Biol. 199, 185–200. Perron, M., Opdecamp, K., Butler, K., Harris, W.A., Bellefroid, E.J., 1999. X-ngnr-1 and Xath3 promote ectopic expression of sensory neuron markers in the neurula ectoderm and have distinct inducing properties in the retina. Proc. Natl Acad. Sci. USA 96, 14996–15001. Pignoni, F., Hu, B., Zavitz, K.H., Xiao, J., Garrity, P.A., Zipursky, S.L., 1997. The eye-specification proteins SO and EYA form a complex and

113

regulate multiple steps in Drosophila eye development. Cell 91, 881– 891. Seimiya, M., Gehring, W.J., 2000. The Drosophila homeobox gene optix is capable of inducing ectopic eyes by an eyeless-independent mechanism. Development 127, 1879–1886. Seo, H.C., Curtiss, J., Mlodzik, M., Fjose, A., 1999. Six class homeobox genes in drosophila belong to three distinct families and are involved in head development. Mech. Dev. 83, 127–139. Seo, H.C., Drivenes, O., Ellingsen, S., Fjose, A., 1998. Expression of two zebrafish homologues of the murine Six3 gene demarcates the initial eye primordia. Mech. Dev. 73, 45–57. Serikaku, M.A., O’Tousa, J.E., 1994. sine oculis is a homeobox gene required for Drosophila visual system development. Genetics 138, 1137–1150. Stauffer, D.R., Howard, T.L., Nyun, T., Hollenberg, S.M., 2001. CHMP1 is a novel nuclear matrix protein affecting chromatin structure and cellcycle progression. J. Cell Sci. 114, 2383–2393. Takebayashi, K., Takahashi, S., Yokota, C., Tsuda, H., Nakanishi, S., Asashima, M., Kageyama, R., 1997. Conversion of ectoderm into a neural fate by ath3, a vertebrate basic helix–loop–helix gene homologous to Drosophila proneural gene atonal. EMBO J. 16, 384–395. Van Valkenburgh, H., Shern, J.F., Sharer, J.D., Zhu, X., Kahn, R.A., 2001. ADP-ribosylation factors (ARFs) and ARF-like 1 (ARL1) have both specific and shared effectors: characterizing ARL1-binding proteins. J. Biol. Chem. 276, 22826–22837. Wallis, D., Muenke, M., 2000. Mutations in holoprosencephaly. Hum. Mutat. 16, 99–108. Wallis, D.E., Roessler, E., Hehr, U., Nanni, L., Wiltshire, T., RichieriCosta, A., Gillessen-Kaesbach, G., Zackai, E.H., Rommens, J., Muenke, M., 1999. Mutations in the homeodomain of the human Six3 gene cause holoprosencephaly. Nat. Genet. 22, 196–198. Wang, S.W., Kim, B.S., Ding, K., Wang, H., Sun, D., Johnson, R.L., Klein, W.H., Gan, L., 2001. Requirement for math5 in the development of retinal ganglion cells. Genes Dev. 15, 24–29. Yost, C., Farr 3rd, G.H., Pierce, S.B., Ferkey, D.M., Chen, M.M., Kimelman, D., 1998. GBP, an inhibitor of GSK-3, is implicated in Xenopus development and oncogenesis. Cell 93, 1031–1041. Zhou, X., Hollemann, T., Pieler, T., Gruss, P., 2000. Cloning and expression of xSix3, the Xenopus homologue of murine Six3. Mech. Dev. 91, 327–330. Zuber, M.E., Perron, M., Philpott, A., Bang, A., Harris, W.A., 1999. Giant eyes in Xenopus laevis by overexpression of XOptx2. Cell 98, 341–352.