Lrrn1 Is Required for Formation of the Midbrain–Hindbrain Boundary And

http://www.diva-portal.org

This is the published version of a paper published in Developmental Biology.

Citation for the original published paper (version of record):

Tossell, K., Andreae, L C., Cudmore, C., Lang, E., Muthukrishnan, U. et al. (2011) Lrrn1 is required for formation of the midbrain-hindbrain boundary and organiser through regulation of affinity differences between midbrain and hindbrain cells in chick Developmental Biology, 352(2): 341-352 https://doi.org/10.1016/j.ydbio.2011.02.002

Access to the published version may require subscription.

N.B. When citing this work, cite the original published paper.

Permanent link to this version: http://urn.kb.se/resolve?urn=urn:nbn:se:umu:diva-57023 Developmental Biology 352 (2011) 341–352

Contents lists available at ScienceDirect

Developmental Biology

journal homepage: www.elsevier.com/developmentalbiology

Lrrn1 is required for formation of the midbrain–hindbrain boundary and organiser through regulation of afﬁnity differences between midbrain and hindbrain cells in chick

Kyoko Tossell a,1, Laura C. Andreae b, Chloe Cudmore a, Emily Lang a, Uma Muthukrishnan c, Andrew Lumsden b, Jonathan D. Gilthorpe c, Carol Irving a,⁎ a Department of Cell and Developmental Biology, University College London, Gower Street, London, WC1E 6BT, UK b MRC Centre for Developmental Neurobiology, Kings College London, London, SE1 1UL, UK c Umeå Centre for Molecular Medicine, Umeå University, 901 87 Umeå, Sweden article info abstract

Article history: The midbrain–hindbrain boundary (MHB) acts as an organiser/signalling centre to pattern tectal and Received for publication 26 November 2009 cerebellar compartments. Cells in adjacent compartments must be distinct from each other for boundary Revised 1 February 2011 formation to occur at the interface. Here we have identified the leucine-rich repeat (LRR) neuronal 1 (Lrrn1) Accepted 3 February 2011 protein as a key regulator of this process in chick. The Lrrn family is orthologous to the Drosophila tartan/ Available online 18 February 2011 capricious (trn/caps) family. Differential expression of trn/caps promotes an affinity difference and boundary Keywords: formation between adjacent compartments in a number of contexts; for example, in the wing, leg and eye Lrrn1 imaginal discs. Here we show that Lrrn1 is expressed in midbrain cells but not in anterior hindbrain cells. Lrrn1 Tartan is down-regulated in the anterior hindbrain by the organiser signalling molecule FGF8, thereby creating a Capricious differential affinity between these two compartments. Lrrn1 is required for the formation of MHB — loss of Boundary function leads to a loss of the morphological constriction and loss of Fgf8. Cells overexpressing Lrrn1 violate Midbrain–hindbrain boundary organiser the boundary and result in a loss of cell restriction between midbrain and hindbrain compartments. Lrrn1 also regulates the glycosyltransferase Lunatic Fringe, a modulator of Notch signalling, maintaining its expression in midbrain cells which is instrumental in MHB boundary formation. Thus, Lrrn1 provides a link between cell affinity/compartment segregation, and cell signalling to specify boundary cell fate. © 2011 Elsevier Inc.Open access under CC BY license.

Introduction this expression interface results in a corresponding shift in the position of the MHB (Broccoli et al., 19990; Katahira et al., 2000; Millet The MHB is an organising centre that is crucial for the formation of et al., 1999). A number of genes have been identified with expression tectum and cerebellum from midbrain and hindbrain, respectively domains at the MHB, forming a large signalling network that gen- (Chi et al., 2003; Marin and Puelles, 1994; Martinez et al., 1995; erates the complexity required for production of a stable organiser Reifers et al., 1998). It arises at Hamburger Hamilton stage (HH)10 in signal. Fgf8 is the best candidate for providing the MHB organiser the chick when morphological constrictions begin to appear along the signal, as recombinant FGF8 protein can mimic organiser tissue grafts length of the neural tube, sub-dividing it into smaller units upon when inserted into the neural tube (Crossley et al., 1996; Irving and which patterning signals can bestow specific regional identities Mason, 2000). Fgf8 is first expressed in a broad domain at the MHB (Lumsden and Krumlauf, 1996). from HH8−. As the boundary forms, Fgf8 becomes restricted to a tight The formation of the MHB is a complex process, involving the domain on the posterior (hindbrain) side of the boundary (Shamim integration of numerous signalling pathways. Early in development et al., 1999). Similarly, Wnt1 is broadly expressed in midbrain but the position of the future boundary is demarcated by the expression becomes progressively restricted to a stripe on the anterior (mid- borders of the homeobox transcription factors, Otx2 and Gbx2, which brain) side of the boundary (McMahon and Bradley, 1990). These and abut at the interface between midbrain and hindbrain (Millet et al., other genes become locked in a regulatory network that maintains and 1996, 1999; Wassarman et al., 1997). Experimental manipulation of restricts the organiser at the boundary, through mutual positive and negative feedback loops (Canning et al., 2007; Wurst and Bally-Cuif, 2001; Ye et al., 2001). ⁎ Corresponding author. Formation of the MHB also requires the action of the Notch E-mail address: [email protected] (C. Irving). 1 Present address: Department of Biology, University of Leicester, 1 University Road, signalling pathway (Tossell et al., submitted for publication). The Leicester, LE1 7RH, UK. Notch modifier, Lunatic Fringe (LFng), is a glycosyltransferase that

0012-1606 © 2011 ElsevierOpen Inc. access under CC BY license. doi:10.1016/j.ydbio.2011.02.002 342 K. Tossell et al. / Developmental Biology 352 (2011) 341–352 together with the distribution of Notch ligands, Delta and Serrate,is adhesion here as in other systems, or whether they act as ligands in an instrumental in determining where Notch signalling is active (Wu and as yet unknown pathway. However, no evidence of a receptor has Rao, 1999). At the MHB, we have recently shown that the border of been found to date (Andreae et al., 2007; Milan et al., 2005; Shishido LFng expression coincides with the Otx2:Gbx2 border, where the MHB et al., 1998). will form, and that the integrity of this border of LFng expression is The closest vertebrate orthologues of trn and caps are members of instrumental for boundary formation. Furthermore, cells ectopically the LRR family: Lrrn1, Lrrn2 and Lrrn3 (Taguchi et al., 1996; Taniguchi expressing activated Notch are excluded from the r1/2 domain et al., 1996). Lrrn1 has recently been identified in chick, as a 3′ EST (metencephalon), and instead are clustered at the MHB and r2/3 (2B10) isolated from a subtracted hindbrain cDNA library, with boundaries, where activated Notch promotes boundary cell fate expression in the midbrain and posterior hindbrain from the r2/3 (Tossell et al., submitted for publication). boundary, but specifically absent from the MHB at HH10 (Christiansen Cellular affinity or adhesive differences between cells in adjacent et al., 2001). More recently, a full length chick Lrrn1 cDNA has been compartments are necessary to help stabilise separate domains identified (Andreae et al., 2007; Garcia-Calero et al., 2006). Lrrn1 has to allow a boundary to form at their interface. For example, either 12 LRRs flanked by characteristic cysteine-rich repeats, an immuno- cells within compartments could have a high affinity for each other, or globulin-like (Ig-like) domain and a fibronectin type III domain. The cells in adjacent compartments could repel each other, thereby intracellular domain contains two endocytic sorting motifs and a PDZ preventing intermixing. In the hindbrain adhesive differences ligand binding motif. between rhombomere compartments drive cell sorting, and together Intriguingly, Lrrn1 has a dynamic expression pattern in the CNS, with cell plasticity, lead to a stable interface which is necessary for with specific down-regulation at boundaries, the timing of which boundary formation (Guthrie and Lumsden, 1991; Irving et al., 1996; correlates with the activation of signalling molecules there (Andreae Wizenmann and Lumsden, 1997). Ephrin/Eph receptor signalling is et al., 2007). Lrrn1 expression demarcates the position of the anterior important for rhombomere compartment-specific cell sorting (Cooke boundary of the zona limitans intrathalamica (ZLI), becoming down- et al., 2001; Mellitzer et al., 1999; Xu et al., 1999). At the interface of regulated just before the onset of the signalling molecule Shh within Ephrin/Eph expression domains, Notch signalling promotes the the ZLI. In the hindbrain, Lrrn1 is down-regulated at rhombomere segregation of boundary cells from rhombomere compartments and boundaries as they become morphologically distinct. Similarly, at the inhibits neurogenesis (Cheng et al., 2004; Pan et al., 2004). These MHB the down-regulation of Lrrn1 in r1/2 correlates with the onset boundary cells are identified by their elongated morphology, fan- of MHB organiser gene expression and boundary formation (Andreae shaped arrangement, low rate of proliferation, lack of neurogenesis et al., 2007). and the expression of a number of molecular markers (Guthrie and Due to the striking expression pattern of Lrrn1 with respect to Lumsden, 1991; Trokovic et al., 2005). At the MHB, the midbrain and boundary regions in the CNS and the role of caps/trn in boundary hindbrain form such compartments that do not mix, are lineage formation, we hypothesised that Lrrn1 may be involved in establishing restricted and form a boundary at their interface that displays similar the compartment boundary between midbrain and hindbrain at the characteristics to hindbrain boundary cells (Jungbluth et al., 2001; MHB. Here we show that the posterior border of Lrrn1 expression Langenberg and Brand, 2005; Trokovic et al., 2005; Zervas et al., correlates with the posterior midbrain border at the MHB, where it is co- 2004). Notch signalling is implicated in the specification of bound- expressed with LFng and Otx2. Reducing Lrrn1 function using ary cell fate at the MHB (Tossell et al., submitted for publication). Morpholinos or the overexpression of truncated proteins disrupts this However, the mechanisms that prevent cell mixing across this border and results in loss of the organiser, as assessed by Fgf8 boundary remain unknown. expression. Misexpression of Lrrn1 across the boundary coupled with In Drosophila, capricious (caps) and tartan (trn) are members of cell lineage labelling reveals that the border of Lrrn1 expression is the LRR family. Both are transmembrane proteins characterised by a important for making an affinity boundary between midbrain and long extracellular domain containing 14 LRRs and a relatively short hindbrain cells. Furthermore, disrupting the border by ectopically intracellular domain. Caps was originally identified due to its role in expressing Lrrn1 in anterior hindbrain causes midbrain and hindbrain neuromuscular target recognition where it mediates the interaction cells to mix and disrupts organiser activity. Lrrn1 regulates LFng between innervating axons and their targets, possibly through expression and is itself regulated by FGF8, providing a link between homophilic adhesion. Caps is expressed in both innervating axon compartment segregation, boundary formation and organiser signal- and target muscle, and has been shown to bind homophilically in a ling. Although the mechanism of action of this gene family is currently cell aggregation assay (Shishido et al., 1998). Subsequently, caps has unknown, here we show that Lrrn1 does not act as a homophilic also been found to regulate axonal targeting in the visual system adhesion molecule to prevent cell mixing. (Shinza-Kameda et al., 2006). In the leg imaginal disc, caps and trn may act to promote cell mobility within the epithelium, as over- Materials and methods expression leads to cell invasion of ectopic territories whilst down- regulation coincides with reduced mobility (Sakurai et al., 2007). Chick embryos Similarly in the eye, Mao et al. have recently proposed that caps and trn regulate adhesive properties that allow ommatidial organisation Fertile chick eggs (Brown Bovan Gold; Henry Stewart & Co.) were from the epithelial sheet (Mao et al., 2008). incubated at 38 °C and staged according to Hamburger and Hamilton Caps and trn are expressed in a compartment restricted fashion (Hamburger and Hamilton, 1992). at the DV boundary of the wing imaginal disc where they prevent cell mixing between compartments (Milan et al., 2001). Both are In ovo electroporation of DNA constructs regulated by the selector gene Apterous (Ap) and their expression only in dorsal cells mediates an affinity difference between dorsal and Hamburger and Hamilton stage (HH) 8–9 chick embryos were ventral cells, preventing cell mixing between adjacent compartments electroporated using fine platinum electrodes and an ElectroSquare (Milan and Cohen, 2003; Milan et al., 2001, 2005). Forced expression Porator™ ECM 830 (BTX) on the following settings; 25 V, 4 pulses, of caps or trn in ventral cells is sufficient to cause cells to sort to 50 ms duration, interval 950 ms. DNA was injected into the neural the dorsal compartment (Milan et al., 2001). Furthermore, in an Ap tube; electrodes were placed on the vitelline membrane either side of mutant background where the DV boundary is lost, expression of caps the neural tube. or trn is sufficient to rescue the affinity boundary (Milan et al., 2001). A full length chicken Lrrn1 cDNA, under the control of the chicken It is unclear whether trn and caps function through homophilic ß-actin promoter, was linked with green fluorescent protein (GFP) by K. Tossell et al. / Developmental Biology 352 (2011) 341–352 343 an internal ribosomal entry site (IRES-GFP): Lrrn1-IRESGFP (1 μg/μl). Results Construct Lrrn1T-GFP carries a truncation immediately after the predicted transmembrane domain at amino acid 655, which is fused Spatio-temporal expression of Lrrn1 at the MHB to the N-terminus of GFP. Consequently it lacks the intracellular portion of the protein including the putative C-terminal PDZ In order to investigate the expression of Lrrn1 in the chick MHB in interaction and endocytosis motifs (1 μg/μl). Mouse LFng-IRESGFP more detail, we performed a series of whole mount in-situ hybridisa- (1 μg/μl) was a kind gifft from O. Cinquin. An anti-sense Morpholino tions on embryos between the 4 and 25 somite stages (HH8−15), oligonucleotide was targeted to a region surrounding the ATG over the time period that the midbrain and hindbrain compartments initiation codon of the Lrrn1 mRNA in order to block translation of form. Lrrn1 is expressed in the neural plate from HH5 (Andreae et al., the LRRN1 protein. The Lrrn1 Morpholino was tagged with fluorescein 2007). We detected strong expression of Lrrn1 throughout the neural isothiocyanate (FITC) to enable detection within the embryo and plate at HH8 when the neural folds have formed. From stage HH9 facilitate uptake of the uncharged molecule by electroporation. Lrrn1 was down-regulated in the MHB region (defined as the region surrounding the constriction where Fgf8 is broadly expressed prior to boundary formation), with high levels remaining both anteriorly and In situ hybridisation and immunohistochemistry posteriorly in the neural tube (Fig. 1A). Down-regulation was reminiscent of that of LFng, which becomes down-regulated in the In situ hybridisation used digoxygenin (DIG) and FITC labelled MHB from HH9 (Fig. 1B). We have previously shown that this border probes as previously described (Irving and Mason, 2000). Both probes of LFng expression coincides with the Otx2:Gbx2 border (Tossell et al., were added simultaneously. Alkaline phosphatase (AP)-conjugated submitted for publication), which defines where the MHB will form anti-DIG and anti-FITC antibodies (Roche) were added sequentially. (Fig. 1C; Millet et al., 1996; Wassarman et al., 1997). The MHB DIG probes were detected using NBT:BCIP (Roche); FITC probes were organiser marker, Fgf8, is expressed in anterior hindbrain, with an detected using FAST TR/Naphtol AS-MX solution (Sigma). Embryos anterior border of expression that coincides with Gbx2 and abuts the were fixed in 4% paraformaldehyde for 20 min before immunohisto- Otx2-positive midbrain domain (Hidalgo-Sanchez et al., 1999) chemistry as previously described (Irving et al., 2002). Anti-EGFP (Fig. 1D). Double in-situ hybridisation of Fgf8 and Lrrn1 revealed antibody (rabbit polyclonal: Clontech) was used at 1:1000. Secondary that Lrrn1 was being down-regulated within this Fgf8-positive domain anti-rabbit HRP was used at 1:200. Embryos were flatmounted or (Fig. 1E). sectioned as previously described (Irving and Mason, 2000). Down-regulation became increasingly marked as development proceeded and at HH11− there was an absence of Lrrn1 in the posterior midbrain vesicle and anterior hindbrain (Fig. 1F). Flatmount Iontophoresis analysis of the dissected neural tube revealed that Lrrn1 was also absent from rhombomere boundaries at this stage as has previously In ovo electroporation was carried out with the appropriate GFP been reported at later stages (Fig. 1G) (Andreae et al., 2007). The constructs at HH8+,9, then DiI (Molecular Probes; D-282) was posterior border of Lrrn1 expression at HH11 coincided with the immediately applied by iontophoresis (Nittenberg et al., 1997). posterior border of LFng and Otx2:Gbx2 (Figs. 1H,I; Andreae et al., 2007). Double in-situ hybridisation analysis of Fgf8 and Lrrn1 revealed that the two genes were expressed in complementary, non-over- Bead implantation lapping, domains at the MHB and in posterior r1 (Fig. 1J). Later in development Lrrn1 expression could be seen in the midbrain, with a FGF8b was introduced on heparin-coated acrylic beads. Control decreasing gradient of expression towards the MHB where it beads were soaked in PBS. Beads were implanted as previously remained down-regulated. In the hindbrain Lrrn1 was weakly described (Irving and Mason, 2000). expressed in r1, then strongly expressed in posterior rhombomeres (Fig.1K). The down-regulation at the MHB coincided with that of LFng (Figs. 1L,M). Therefore, Lrrn1 also appears to define the extent of the Cell sorting assay midbrain compartment.

Assays were performed using human embryonic kidney 293T FGF8 represses expression of Lrrn1 at the MHB (HEK293T) cells as described previously (Karaulanov et al., 2006). Cells were transiently transfected at 50–70% confluence in a 6-well Fgf8 is first expressed broadly in the MHB region at the 4 somite plate with 200 ng of pCAß-eGFPm5 (Yaneza et al., 2002) and 1.8 μg stage (HH8), shortly before Lrrn1 becomes down-regulated there. At per well of the specific expression vector being assayed (see below), the 11 somite stage (HH10+) Fgf8 is restricted to a tight domain at the using Fugene HD (Roche). A ratio of 6 μl Fugene HD:2 μg DNA resulted MHB. Lrrn1 was absent from this domain but continued to be in a transfection efficiency of 80–100% as judged by eGFP fluores- expressed in the midbrain and hindbrain on either side. As Lrrn1 cence. After 24 h, cells were trypsinised, mixed with an equal number appeared to be down-regulated at the MHB concurrent with the of control cells transfected with 200 ng of pCAß-mRFP1 and 1.8 μg per appearance of Fgf8, we investigated whether FGF8 was responsible for well of pUC19 and allowed to aggregate in complete medium in this dynamic expression. We introduced a local source of ectopic FGF8 bovine serum albumin (BSA)-coated 48-well plates (coated with 1% protein unilaterally into the midbrain on heparin-coated acrylic beads (w/v) BSA in PBS, overnight at 4 °C) on an orbital shaker (120 R.P.M.). at 7–8 somites, and examined the effect on Lrrn1 expression. Lrrn1 After 48 h, cell aggregates were examined on an inverted microscope was dramatically down-regulated in the midbrain tissue surrounding (Zeiss Axiovert with YFP and CY3 filter sets) and photographed with a the FGF8 bead (Fig. 2A and Table 1) as compared to the control contra- Zeiss Axiocam HS camera. Images were pseudocoloured and super- lateral side or to control beads soaked in PBS (Fig. 2B and Table 1). imposed in Image J. The expression vectors used were: pCS2+ FGF8 beads implanted in the hindbrain were able to repress Lrrn1 −xFLRT3 and pCS2+−xFLRT3ΔLRR (positive and negative controls, throughout the entire hindbrain (Fig. 2C). FGF8 induces Fgf8 respectively. Karaulanov et al., 2006), Flag-mLrrn1, N-terminally Flag- expression in the midbrain (Crossley et al., 1996; Martinez et al., tagged mouse Lrrn1 in pCDNA3.1(+) (Haines et al., 2005). CAß- 1999; Shamim et al., 1999). In agreement with this we observed mRFP1 was made by cloning a BamHI-EcoRI fragment of pRSETB- strong induction of Fgf8 around FGF8 beads (positive control) (Fig. 2D mRFP1 (Campbell et al., 2002) in to pCAß. and Table 1). No induction of Fgf8 was observed following 344 K. Tossell et al. / Developmental Biology 352 (2011) 341–352

Fig. 1. Expression of Lrrnl during chick MHB development. Wholemount in-situ hybridization of Lrrnl, LFng, Otx2 (red), Gbx2, Fgf8 (red). Neural tubes in (G,J) are opened in a flatmount preparation. (A) At HH9 Lrrnl begins to be down-regulated in the MHB region with high levels of expression remaining in the anterior and posterior neural tube. (B) LFng is also down-regulated in the MHB region at HH9. (C) The MHB is defined by the expression interface of Otx2 and Gbx2. (D) Fgf8 is expressed in the anterior-most domain of Gbx2 expression at the MHB. (E) Double in-situ hybridisation of Lrrnl (blue) and Fgf8 (red) show that Lrrnl is down-regulated in the Fgf8 domain. (F) At HH11− there is clear down- regulation of Lrrnl in MHB. (G)Absence of Lrrnl in MHB and also in rhombomere boundaries in hindbrain at HH11. (H) LFng expression mirrors that of Lrrnl at HH11 and is down- regulated at the MHB. (I) The expression border of Otx2 and Gbx2 is anterior to the morphological constriction at HH12 and coincides with the posterior border of LFng expression (black arrows). (J) Double in-situ of Fgf8 (red) and Lrnn1 (blue) shows complementary expression at the MHB at HH11−. (K) Transverse view of Lrrnl at HH15 shows clear down- regulation in MHB and a decreasing gradient in posterior midbrain towards the MHB. (L) LFng (red) and Lrnn1 (blue) are expressed coincidently in midbrain and hindbrain and are both absent at the MHB. (M) High magnification view of MHB in (L) showing coincident expression of LFng (red) and Lrrnl (blue). MHB: mid-hindbrain boundary, r2: rhombomere 2; r3: rhombomere 3: r4: rhombomere 4. implantation of PBS beads (Table 1). These results demonstrate that Electroporation of the truncated construct Lrrn1T-GFP into the FGF8 downregulates Lrrn1 expression and suggest that FGF8 may act neural tube of chick embryos at 5–7 somites (HH8+,9 — before the normally to define the limits of Lrrn1 expression in the posterior MHB has formed) resulted in an apparent loss of the morphological midbrain and anterior hindbrain, through repression in r1. constriction at the MHB 24 h later. The neural tube was much straighter on the electroporated side compared to the contra-lateral Lrrn1 is required for boundary and organiser formation at the MHB control side (Fig. 3B,C,I). Using Otx2 and Gbx2 to mark the interface of midbrain and hindbrain compartments, we saw a caudal shift in Gbx2 The posterior border of LFng expression in the midbrain is expression and a fuzzy interface between the two expression domains important for boundary formation at the MHB, as perturbation of (Fig. 3B,C; n=9/10) as compared to the control contra-lateral side or the LFng expression border through ectopic expression in anterior control embryos electroporated with GFP alone (Fig. 3B,C and data not hindbrain leads to a shift in position of the markers Otx2:Gbx2, and shown; GFP control n=0/7). We repeated these experiments using subsequently in the boundary itself (Tossell et al., submitted for Fgf8 as a molecular marker of the organiser itself, as recombinant FGF8 publication). In order to investigate the involvement of Lrrn1 in protein can mimic organiser grafts in vivo (Crossley et al., 1996; Irving compartition and boundary formation at the MHB we sought to block and Mason, 2000). Fgf8 expression at the MHB was either severely Lrrn1 function during development of the MHB boundary region, down-regulated, or in some cases, completely absent on the electro- using a number of strategies. Firstly, we designed a construct porated side (Fig. 3E–G; n=8/10). In stark contrast we saw no effect on containing a truncation of the intracellular domain, including the the control contra-lateral side of electroporated embryos or in controls putative C-terminal PDZ interaction and endocytosis motifs, that electroporated with a GFP construct alone (Fig. 3E–G and data not could potentially work as dominant negative protein (Lrrn1T-GFP). shown; GFP control n=0/10). Secondly, we used an anti-sense Morpholino oligonucleotide targeted In order to confirm that the organiser was lost/disrupted through to Lrrn1 mRNA in order to block translation of the Lrrn1 protein blocking Lrrn1 function we used an alternative strategy. We electro- following electoporation in vivo (Fig. 3A). porated anti-Lrrn1 Morpholino into embryos at HH8+,9. Fgf8 K. Tossell et al. / Developmental Biology 352 (2011) 341–352 345

Fig. 2. FGF8 represses Lrrnl expression at the MHB. (A) Lrrnl expression is down-regulated (black arrowheads) around an FGF8 bead inserted into midbrain. (B) Control contra- lateral side of the same embryo showing normal Lrnn1 expression throughout the midbrain. (C) Lrnn1 expression is down-regulated along the length of the hindbrain by an FGF8 bead inserted into hindbrain (black arrowheads). (D) Positive control: Fgf8 expression is up-regulated around an FGF8 bead inserted into midbrain (black arrowheads). Red arrowheads mark the position of the bead. expression was found to be absent consistently, 24 h later on the side of compartments, at the time of its normal down-regulation in anterior the embryo where the Morpholino was detected using the FITC tag (Fig. hindbrain. We electroporated an expression vector encoding full 3H–J; n=8/19). Furthermore, the morphology of the neural tube at the length Lrrn1 cDNA into the neural tube at HH8+, 9 and analysed MHB was again perturbed, being much straighter than the control side embryos 24 h later, when the MHB had formed. Cells ectopically (Fig. 3H–J). A standard fluorescent control Morpholino had no effect on expressing Lrrn1 were visualised by co-expression of GFP from a Fgf8 expression or MHB morphology (Fig. 3K–M; n=10/10). bicistronic message linked by an internal ribosome entry sequence These data suggest that Lrrn1 is required for MHB boundary (Lrrn1-IRESGFP, Fig. 4A). formation. Interfering with Lrrn1 function using a construct lacking We used the molecular markers of midbrain and hindbrain com- the intracellular domain or by Morpholino-mediated knockdown partments, Otx2 and Gbx2, to analyse the effect of Lrrn1 misexpression elicited loss of a morphological constriction and downregulation of on compartition. When Lrrn1 expression extended into the anterior the organiser gene, Fgf8. This strongly suggests that the mutant hindbrain, crossing the compartment boundary, the expression border construct acts as a dominant negative, and that the actions of Lrrn1 at of Otx2 and Gbx2 was shifted caudally when compared to the control the MHB are dependent on the presence of an intact intracellular side, which was normal (Fig. 4A,B). Otx2 positive cells were seen in domain. rhombomere one. The morphological constriction of the MHB also shifted caudally on the electroporated side as compared to the control Ectopic expression of Lrrn1 across the MHB causes loss of restriction of contra-lateral side, and became less well defined (Fig. 4A,B). Opening MHB gene expression borders and a shift in the morphological boundary the neural tube in a flatmount preparation confirmed the shift in expression of Otx2 and Gbx2, and revealed that cells were mosaically Previous studies in Drosophila have shown that caps and trn expressing Otx2 and Gbx2 within the anterior hindbrain (Fig. 4C,D; restrict cell movement across boundaries. Their ectopic expression n=5/5). We repeated these experiments using Wnt1 and Fgf8 as across the border at both the DV boundary in the wing disc and tarsus molecular markers of the organiser itself. Wnt1 is normally restricted 5-pretarsus boundary in the leg disc causes cell invasion of adjacent to a tight band of expression in anterior MHB in midbrain, Otx2 positive territories (Milan et al., 2001, 2005; Sakurai et al., 2007). The cells, whereas Fgf8 is restricted to a tight band of expression in mechanism(s) involved remains unknown, but may be a direct result posterior MHB in hindbrain, Gbx2 positive cells. In embryos ectopically of a change in cell adhesive properties conferred by caps and trn. expressing Lrrn1 across the boundary, Wnt1 expression was seen To investigate the hypothesis that Lrrn1 confers a differential extending caudally into anterior hindbrain. The border of expression affinity property to cells in neighbouring compartments at the MHB, between Fgf8 and Wnt1 was fuzzy on the electroporated side of the we ectopically expressed Lrrn1 across the midbrain and hindbrain embryo, and Wnt1 cells were observed mixing with Fgf8-positive cells (Fig. 4E–H; n=4/5).

Table 1 These results show that ectopic Lrrn1 expression results in a loss of FGF8 represses Lrrn1 expression. restriction of midbrain (Otx2) and hindbrain (Gbx2) markers and subsequently of MHB organiser genes Wnt1 and Fgf8. Moreover, they FGF8 bead PBS bead suggest that coordinated down-regulation of Lrrn1 from the MHB is Lrrn1 5/7 0/9 required for boundary formation and appropriate speciﬁcation of the Fgf8 2/2 0/2 organiser. 346 K. Tossell et al. / Developmental Biology 352 (2011) 341–352

Fig. 3. Lrrn1 is required for MHB boundary formation and organiser gene expression. (A) Schematic diagram of electroporation strategy into HH8+ or 9 chick embryos, and analysis 24 h later. (B–G) Electroporation of Lrrn1T-GFP. (B) Overlay of Lrrn1TGFP fluorescence reveals the position of electroporated cells. (C,D) In-situ hybridisation of Otx2 (red) and Gbx2 (blue) mark the midbrain and hindbrain compartments respectively. The neural tube appears flat on the electroporated right hand side and a diminished morphological constriction can be seen at the MHB (black arrows). (D) Flat mount preparation of the neural tube reveals that the sharp boundary between Otx2 and Gbx2 expressing cells is lost on the electroporated right hand side — compare to the control left hand side (black double arrow). (E) Overlay of Lrrn1TGFP fluorescence. (F,G) In-situ hybridisation of Fgf8 (red). Fgf8 is absent at the MHB on the electroporated right hand side of the neural tube. (G) Flat mount preparation of neural tube confirms the absence of Fgf8 at the MHB on the electroporated side compared to the control contra-lateral side (black arrow). (H-J) Lrrn1 Morpholino (MO). (H) Fgf8 expression (blue) is absent from the electroporated side of the neural tube at the MHB (black arrow) but remains expressed in other locations e.g. nasal placode. (I) High magnification confirms loss of Fgf8 only on the electroporated side compared to the control contra-lateral side (black double arrow). (J) Anti-FITC antibody (red) reveals the position of the Morpholino within the neural tube. (K-M) control Morpholino. (K) Fgf8 is expressed normally at the MHB (black arrow) and other locations. (L) Fgf8 is expressed equally on both sides of neural tube in control embryos. (M) Anti-FITC antibody (red) reveals the position of the control Morpholino within the neural tube. MHB; midbrain–hindbrain boundary. K. Tossell et al. / Developmental Biology 352 (2011) 341–352 347

Fig. 4. Ectopic expression of Lrrn1 across the MHB causes mixing at gene expression borders and a shift in the MHB. (A–H) Electroporation of Lrrnl-IRESGFP across the MHB interface (A) Overlay of Lrrnl-IRESGFP fluorescence reveals the position of electroporated cells. (B–D) In-situ hybridisation of Otx2 (red) and Gbx2 (blue). (B) The MHB morphological constriction and the expression border of Otx2 and Gbx2 are shifted caudally on the electroporated left hand side as compared to the control side (C) Flatmount preparation confirms the shift in expression of Otx2 and Gbx2 (white double arrow). (D) High magnification of box in C reveals mosaic expression of Obc2 and Gbx2 within anterior hindbrain (arrows). (F–H) In-situ hybridisation of Fgf8 (red) and Wntl (blue). Wntl marks anterior MHB in midbrain (Otx2-positive) cells FgfB marks posterior MHB in hindbrain (Gbx2-positive) cells (E) Overlay of Lrrnl-IRESGFP fluorescence. (F) Wntl expression extends caudally into anterior hindbrain on the electroporated left hand side as compared to control contra-lateral side (white double arrow) (G) The border of expression between Fgf8 and Wntl is fuzzy on the electroporated side (white double arrow). (H) High magnification of box in G reveals that Wntl-positive cells mix with Fgf8-positive cells and extend ectopically into the hindbrain (arrows). MHB: midbrain-hindbrain boundary.

Cell mixing at the midbrain–hindbrain compartment boundary labelled midbrain cells were seen extending throughout the MHB region and into r1. The extent of the DiI label was as extensive as the Blocking Lrrn1 function at the MHB resulted in loss of a clear ectopic Lrrn1-IRESGFP domain, indicating that cells were free to move border of Otx2:Gbx2 expression, loss of a morphological boundary and within the Lrrn1-positive domain but remained within it (Fig. 5H–K; loss of Fgf8 “organiser” expression. Ectopic expression of Lrrn1 n=14/17). This data suggests that it is unlikely that Lrrn1 confers a resulted in a caudal shift in the morphological boundary which cell-fate change on hindbrain cells to that of a midbrain fate. Rather, it correlated with a shift in the Otx2:Gbx2 interface. This suggests that is more likely that the boundary of expression of Lrrn1 is important for differential expression of Lrrn1 at the MHB is required for proper determining how cells segregate at the MHB, and that the role of Lrrn1 boundary positioning and formation. We hypothesised that Lrrn1 may at the MHB is to maintain the midbrain compartment and prevent cell confer a cell surface affinity property to cells that is required to mixing between domains. prevent cell mixing between midbrain and hindbrain compartments. Down-regulation of Lrrn1 in hindbrain cells may be required to separate midbrain and hindbrain cells. Alternatively, the role of Lrrn1 Lrrn1 regulates Notch signalling by positioning the LFng boundary may be to specify midbrain fate; ectopic Lrrn1-expressing cells may induce midbrain-specific markers Otx2 and Wnt1, and the cell mixing In the Drosophila wing disc, trn and fng are independently observed may be due to the mosaic nature of electroporation in the activated by Ap to regulate two distinct processes required for neural tube. To test between these two possibilities, we used DiI boundary formation: cell affinity differences (trn), and signalling labelling to trace the movement of cells immediately following (fng), between dorsal and ventral compartments (Milan et al., 2001). electroporation with Lrrn1-IRESGFP (Fig. 5A). A small number of We sought to investigate a possible relationship between LFng and midbrain cells just anterior to the Otx2:Gbx2 interface, which marks Lrrn1 at the MHB. To test this we misexpressed Lrrn1-IRESGFP across the molecular boundary, were labelled with DiI by iontophoresis and the midbrain–hindbrain region at HH8+,9 and analysed LFng analysed 24 h later. The DiI was positioned at (x), 6/8th of (y), the expression 24 h later. LFng was up-regulated to high levels through- distance along the neural tube from diencephalon to hindbrain out the domain where Lrrn1 was ectopically expressed, both within its constriction (Fig. 5A,B,G). In control embryos electroporated with a normal domains of expression and also throughout the metenceph- GFP expression vector before DiI labelling, the progeny of labelled alon (r1/2) where it is normally absent (Fig. 6,A–C; n=4/6). Both the cells could be seen in a tight cluster in the anterior MHB and clearly control, contra-lateral side and embryos electroporated with a GFP within the electroporated domain (Fig. 5C–F; n=18/18). In sharp control showed normal LFng expression with a clear absence of contrast, however, in embryos electroporated with Lrrn1-IRESGFP, DiI expression in r1/2 (n=6/6; data not shown). 348 K. Tossell et al. / Developmental Biology 352 (2011) 341–352

Fig. 5. Misexpression of Lrnn1 across posterior midbrain and anterior hindbrain allows cells to move across the MHB. (A) Schematic representation of electroporation followed immediately by iontophoresis of Dil label (x) at 6/8th of (y), the distance along the neural tube from diencephalon to hindbrain constriction (just anterior to Otx2:Gbx2 interface) at HH9. (B,G) Position of Dil label at Ohr. (B–F) GFP control; dorsal view. (C) Neural tube 24 h later. (D) Cells labelled with Dil after 24 h are clustered in posterior midbrain and do not cross the MHB boundary. (E) Control GFP expressing cells are continuous throughout posterior midbrain, MHB and anterior hindbrain, (F) Merged view of Dil labelled cells and GFP expressing cells. (G–K) Lrrnl-IRESGFP; dorsal view. (H) Neural tube 24 h later. (I) Cells labelled with Dil are observed extensively in both midbrain and hindbrain compartments 24 h after labelling. (J) Lrnn1-IRESGFP expressing cells are seen in both midbrain and crossing the MHB into hind-brain. (K) Merged view shows Dil labelled cells within the Lrrnl- electroporated domain in both midbrain and hindbrain. MHB: midbrain–hindbrain boundary; black arrows indicate position of Dil label.

To test whether LFng was able to regulate Lrrn1 expression in vitro, work from Stephen Cohen's group found no evidence of reciprocally, we electroporated LFng-IRESGFP throughout the midbrain homophilic interactions but favour the hypothesis that caps and trn and anterior hindbrain and analysed Lrrn1 expression. We did not act by recognising an as yet unknown ligand on cells in the dorsal observe any changes in Lrrn1 expression on the side of the embryos compartment of the wing disc (Milan et al., 2005; Shinza-Kameda et al., ectopically expressing LFng (n=9/9). Lrrn1 was expressed in its normal 2006). domains of expression within midbrain and hindbrain, and remained To test the possibility that Lrrn1 may promote the integrity of the absent from the metencephalic domain (Fig. 6D–F). Both the midbrain compartment through a homophilic cell adhesion mecha- experimental and the control contra-lateral sides of each embryo nism, we assayed the ability of Lrrn1 to promote cell aggregation in looked the same, as did embryos electroporated with GFP control HEK293T cells. A closely related LRR family member, fibronectin- (n=6/6; data not shown). The LFng border determines where Notch is leucine-rich transmembrane 3 (FLRT3), has recently been shown to activated at the midbrain–hindbrain interface (Tossell et al., submitted promote the sorting out of transfected from non-transfected cells in for publication). Therefore, through regulating LFng expression, Lrrn1 this assay, a process which requires the LRR but not the intracellular may determine the boundary at which Notch is activated in the neural domain (Karaulanov et al., 2006). tube, thus providing a link between compartmentalisation due to Cells were transfected with both Lrrn1 and eGFP expression affinity differences and inter-compartmental boundary formation. vectors at a ratio that ensured effective co-transfection of the two plasmids (9:1). These were then mixed with mRFP1 transfected cells Lrrn1 does not promote cell aggregation in transfected cells and allowed to form aggregates (Karaulanov et al., 2006). In contrast to FLRT3/eGFP transfected cells, which sorted in to small, clearly Previous studies have provided conflicting data about the role of defined clusters over a 48 h period (Fig. 7A,E,I), Lrrn1/GFP transfected tartan-like LRR proteins as homophilic cell adhesion molecules. Whilst cells formed a mixed population with RFP transfected cells and did not Shinza-Kameda et al. find that caps promotes homophilic cell adhesion exhibit any detectable sorting activity (Fig. 7C,G,K). Identical results K. Tossell et al. / Developmental Biology 352 (2011) 341–352 349

Fig. 6. Lrrnl regulates LFng expression at the midbrain-hindbrain border (A–C) Misexpression of Lrrnl-IRESGFP or (D–F) LFng-IRESGFP. (A,D) Overlay of live fluorescence to reveal position of electroporated cells. (B,C) In-situ hybridisation of LFng (blue). (B) LFng is up-regulated on the electroporated left hand side compared to the control contra-lateral side (black arrows). (C) Flat mount preparation of neural tube reveals up-regulation of LFng within the ectopic domain of Lrnn1 detected by anti-GFP antibody (brown) (black double arrow). (E,F) In-situ hybridisation of Lrrnl (blue). No change in Lrrnl expression is seen following misexpression of LFng-IRESGFP in the neural tube. (F) Flat mount preparation reveals equal Lrrnl expression on both sides of neural tube. Anti-GFP antibody (brown) reveals position of ectopic LFng-IRESGFP-expressing cells. MHB; midbrain–hindbrain boundary. were obtained with Lrrn1-IRESGFP, Lrrn1T-GFP, Lrrn2 and Lrrn3 (data Fgf8), with some mixing between midbrain and hindbrain cells. Cell not shown). Therefore, Lrrn1 does not appear to promote homophilic labelling confirmed that cell mixing was due to a lack of restriction of cell adhesion in the same way as FLRT3, suggesting that the affinity midbrain cells, which were able to cross into anterior hindbrain differences between Lrrn1 expressing midbrain cells and Lrrn1 non- within the ectopic Lrrn1 domain. The shift in position of the boundary expressing hindbrain cells is modulated indirectly, by an as yet is probably due to the cell mixing observed. This suggests that Lrrn1 unidentified mechanism. acts to restrict cells at the boundary, by mediating an affinity difference between midbrain and hindbrain cells that prevents cells Discussion from mixing at their interface. Similarly, overexpression of trn at the tarsus5/pretarsus boundary of the Drosophila leg disc causes cells to Lrrn1 is the vertebrate orthologue of Drosophila trn, a gene involved violate compartmental restriction and cross into the adjacent domain, in DV boundary formation in the wing disc through specification of suggesting that Lrrn1 may perform a similar function to trn at dorsal cell affinity properties (Milan et al., 2001, 2005). Because the vertebrate boundaries (Sakurai et al., 2007). expression border of Lrrn1 at the MHB correlates with that of Otx2, which defines the midbrain compartment, we investigated the Integration of segregation, boundary formation and organiser signalling hypothesis that Lrrn1 may play a similar role to trn in boundary at the MHB formation at the MHB in the vertebrate CNS. In common with the DV boundary, Notch signalling is important for the formation of the MHB, At the DV boundary in the Drosophila wing disc, affinity differences where it specifies boundary cell fate (Tossell et al., submitted for between D and V cells are initially regulated by caps/trn independently publication). Furthermore, the co-expression of Lrrn1 and LFng in the of signalling at the boundary, which is mediated by Notch. The selector midbrain compartment is strikingly similar to that seen with trn/fng in gene Ap directs both cell affinity and cell signalling at the DV boundary the dorsal compartment of the Drosophila wing disc. through independent regulation of both caps/trn and fng (Milan and Cohen, 2003; Milan et al., 2001). We have found that Lrrn1 upregulates The role of Lrrn1 in cell restriction and compartition at the MHB LFng in the hindbrain and, therefore, it may determine the border of LFng expression at the posterior midbrain where they are co-expressed. Both loss and gain of function experiments resulted in a disruption In a similar manner to the DV boundary in the wing disc, the LFng of the MHB. Blocking Lrrn1 function at the MHB resulted in a loss of border (in concert with the expression border of the Notch ligand Ser1) the morphological constriction at the MHB and in some cases, plays an important role at the MHB to determine where Notch is complete loss of the organiser gene, Fgf8; probably as a secondary activated (Tossell et al., submitted for publication). Unlike the situation consequence of loss of the boundary. Ectopic expression of Lrrn1 in the wing disc, however, the regulation of LFng by Lrrn1 may provide a across the MHB region resulted in a shift of the boundary (identified mechanism for linking compartmentalisation, due to affinity differ- by midbrain markers Otx2 and Wnt1, and hindbrain markers Gbx2 and ences, and signalling, at inter-compartmental boundaries. 350 K. Tossell et al. / Developmental Biology 352 (2011) 341–352

Fig. 7. Lrrnl does not act through homophilic adhesion. HEK293T cells were cotransfected with an eGFP expression vector and constructs for (A) xFLRT3, (B) xFLRT3ΔLRR (C) Flag- mLrnn1 or empty pCS2+ vector (D) aid mixed in a 1:1 ratio with control cells transfected with an mRFP1 expression vector (E–H). After 48h, xFLRT3- expressing cefis (positive control) sorted into a small, clearly defined dusters. No sorting was observed with xFLRT3ΔLRR (negative control), which lacks the LRR domain of xFLRT3 necessary for homophilic interaction aid cell sorting. Sorting was not observed with Flag-tagged mLrnn1, empty vector or mRFP1 cells as can be seen in the merged nages (I–L). None of the constructs affected cell proliferation or cell attachment and confluent cell monolayers were seen in all cases under phase contrast illumination (M–P). The Flag-mLrrn1 construct was highly expressed in HEX293T cells and generated a protein of approximately 100 KDa, consistent with the size of the glycosylated form of mLrrn1 when analysed by western blot (Q).

It is possible that a vertebrate orthologue of Ap also functions to and maintain MHB organiser genes (Adams et al., 2000; Matsunaga regulate both LFng and Lrrn1 at the MHB, providing a mechanism for et al., 2002). Furthermore, mouse and zebraﬁsh mutant analyses have linking compartmentalisation, due to afﬁnity differences, and signal- revealed that Lmx1b is essential for the proper formation of the MHB ling, at inter-compartmental boundaries. Indeed, one family member, organiser, and later tectum and cerebellum development (Guo et al., Lmx1b, is expressed broadly in the forebrain and midbrain at HH9, 2007; O'Hara et al., 2005). becoming restricted to midbrain at HH10 with a posterior boundary of We have found a close correlation between the onset of Fgf8 expression that coincides with Otx2 and Wnt1 (Yuan and Schoenwolf, expression and the down-regulation of Lrrn1 at the MHB, and indeed 1999). In the chick, Lmx1b acts upstream of Wnt1 and Fgf8 to induce have shown that FGF8 protein represses Lrrn1 expression in midbrain. K. Tossell et al. / Developmental Biology 352 (2011) 341–352 351

Thus, a complex interaction is becoming apparent that ensures the Andreae, L.C., Peukert, D., Lumsden, A., Gilthorpe, J.D., 2007. Analysis of Lrrn1 expression and its relationship to neuromeric boundaries during chick neural precise positioning of the boundary between midbrain and hindbrain. development. Neural Dev. 2, 22. Initially, Otx2 and Gbx2 are expressed broadly across anterior and Bottcher, R.T., Pollet, N., Delius, H., Niehrs, C., 2004. The transmembrane protein XFLRT3 posterior neural plate, respectively (Millet et al., 1996; Wassarman forms a complex with FGF receptors and promotes FGF signalling. Nat. Cell Biol. 6, − 38–44. et al., 1997). From 3 somites (HH8 ), MHB genes are expressed in an Broccoli, V., Boncinelli, E., Wurst, W., 1999. The caudal limit of Otx2 expression initially broad domain covering posterior midbrain and anterior positions the isthmic organizer. Nature 401, 164–168. hindbrain. Fgf8 itself is expressed from 4 somites (HH8) (Shamim and Campbell, R.E., Tour, O., Palmer, A.E., Steinbach, P., Baird, G.S., Zacharias, D.A., Tsien, R.Y., fl – Mason, 1999). Correlating with this, a down-regulation of Lrrn1 is 2002. A monomeric red uorescent protein. Proc. Natl Acad. Sci. 99, 7877 7882. Canning, C.A., Lee, L., Irving, C., Mason, I., Jones, C.M., 2007. Sustained interactive Wnt observed shortly after, from 7 somites (HH9) in the Fgf8 domain, at a and FGF signaling is required to maintain isthmic identity. Dev. Biol. 305, 276–286. time when organiser genes are broadly expressed prior to boundary Cheng, Y.C., Amoyel, M., Qiu, X., Jiang, Y.J., Xu, Q., Wilkinson, D.G., 2004. Notch activation formation and before restriction to the boundary. Given that regulates the segregation and differentiation of rhombomere boundary cells in the zebrafish hindbrain. Dev. Cell 6, 539–550. recombinant FGF8 protein represses Lrrn1 it is likely that the normal Chi, C.L., Martinez, S., Wurst, W., Martin, G.R., 2003. The isthmic organizer signal FGF8 is down-regulation of Lrrn1 in anterior hindbrain (r1) is due to required for cell survival in the prospective midbrain and cerebellum. Development repression by FGF8. The domain over which Lrrn1 is repressed by an 130, 2633–2644. Christiansen, J.H., Coles, E.G., Robinson, V., Pasini, A., Wilkinson, D.G., 2001. Screening FGF8 bead is striking. Previously, inductive responses to FGF8 have from a subtracted embryonic chick hindbrain cDNA library: identification of genes not been observed to cross rhombomere boundaries, yet strikingly, expressed during hindbrain, midbrain and cranial neural crest development. Mech. here we observed Lrrn1 repression across a number of rhombomere Dev. 102, 119–133. Cooke, J., Moens, C., Roth, L., Durbin, L., Shiomi, K., Brennan, C., Kimmel, C., Wilson, S., boundaries (Irving and Mason, 2000). Holder, N., 2001. Eph signalling functions downstream of Val to regulate cell sorting and boundary formation in the caudal hindbrain. Development 128, 571–580. How does Lrrn1 function at the MHB? Crossley, P.H., Martinez, S., Martin, G.R., 1996. Midbrain development induced by FGF8 in the chick embryo. Nature 380, 66–68. Garcia-Calero, E., Garda, A.L., Marin, F., Puelles, L., 2006. Expression of Lrrn1 marks the We propose that Lrrn1 provides an affinity cue for cells in the prospective site of the zona limitans thalami in the early embryonic chicken – midbrain, which sharpens and defines the boundary by segregating diencephalon. Gene Expr. Patterns 6, 879 885. Guo, C., Qiu, H.Y., Huang, Y., Chen, H., Yang, R.Q., Chen, S.D., Johnson, R.L., Chen, Z.F., Lrrn1-positive cells from the cells that have turned off Lrrn1 expression Ding, Y.Q., 2007. Lmx1b is essential for Fgf8 and Wnt1 expression in the isthmic in r1. At the same time, Lrrn1 integrates boundary cell determination organizer during tectum and cerebellum development in mice. Development 134, by the Notch signalling pathway, through the regulation of LFng. This 317–325. fi Guthrie, S., Lumsden, A., 1991. Formation and regeneration of rhombomere boundaries leads to a sharp LFng border which acts to de ne the domain of Notch in the developing chick hindbrain. Development 112, 221–229. activation, and subsequently where specialised boundary cells are Haines, B.P., Gupta, R., Jones, C.M., Summerbell, D., Rigby, P.W., 2005. The NLRR gene family positioned (Tossell et al., submitted for publication). and mouse development: modified differential display PCR identifies NLRR-1 as a gene expressed in early somitic myoblasts. Dev. Biol. 281, 145–159. Given the results of our cell aggregation assay it is unlikely that the Haines, B.P., Wheldon, L.M., Summerbell, D., Heath, J.K., Rigby, P.W., 2006. Regulated segregation of midbrain and hindbrain cells occurs through a expression of FLRT genes implies a functional role in the regulation of FGF homophilic adhesion mechanism, although it is possible that Lrrn1 signalling during mouse development. Dev. Biol. 297, 14–25. binds homophilically, but requires the addition of a co-receptor which Hamburger, V., Hamilton, H.L., 1992. A series of normal stages in the development of the chick embryo. 1951. Dev. Dyn. 195, 231–272. is missing from the HEK293 cell line. Previously, aggregation assays Hidalgo-Sanchez, M., Millet, S., Simeone, A., Alvarado-Mallart, R.M., 1999. Comparative have been described for other members of the LRR family, such as caps analysis of Otx2, Gbx2, Pax2, Fgf8 and Wnt1 gene expressions during the formation – (Shinza-Kameda et al., 2006) and FLRT3 (Karaulanov et al., 2006), of the chick midbrain/hindbrain domain. Mech. Dev. 81, 175 178. Irving, C., Flenniken, A., Alldus, G., Wilkinson, D.G., 1996. Cell–cell interactions and which showed that these family members can bind homophilically. segmentation in the developing vertebrate hindbrain. Biochem. Soc. Symp. 62, 85–95. However, although FLRT3 indeed demonstrates the capacity to bind Irving, C., Malhas, A., Guthrie, S., Mason, I., 2002. Establishing the trochlear motor axon – homophilically in our assay, Lrrn1 does not. trajectory: role of the isthmic organiser and Fgf8. Development 129, 5389 5398. Irving, C., Mason, I., 2000. Signalling by FGF8 from the isthmus patterns anterior FLRT3 functions by regulating the levels of cadherins on the hindbrain and establishes the anterior limit of Hox gene expression. Development surface of cells via endocytosis, and has also been identified as a co- 127, 177–186. receptor for FGF (Bottcher et al., 2004; Haines et al., 2006; Karaulanov Jungbluth, S., Larsen, C., Wizenmann, A., Lumsden, A., 2001. Cell mixing between the embryonic midbrain and hindbrain. Curr. Biol. 11, 204–207. et al., 2006). FLRT3 is also expressed at the MHB and it is possible that Karaulanov, E.E., Bottcher, R.T., Niehrs, C., 2006. A role for fibronectin-leucine-rich it regulates both cell adhesion and FGF signalling there. However, co- transmembrane cell-surface proteins in homotypic cell adhesion. EMBO Rep. 7, transfection of Lrrn1 and FLRT3 was unable to block the homophilic 283–290. Katahira, T., Sato, T., Sugiyama, S., Okafuji, T., Araki, I., Funahashi, J., Nakamura, H., 2000. aggregation of FLRT3-expressing cells in our assay (data not shown), Interaction between Otx2 and Gbx2 defines the organizing center for the optic suggesting that it does act independently of FLRT3, perhaps in a tectum. Mech. Dev. 91, 43–52. parallel pathway. Langenberg, T., Brand, M., 2005. Lineage restriction maintains a stable organizer cell fi – We have also been unsuccessful in our attempts to identify a population at the zebra sh midbrain hindbrain boundary. Development 132, 3209–3216. receptor, or ligand, for Lrrn1, using a number of strategies (Andreae et al., Lumsden, A., Krumlauf, R., 1996. Patterning the vertebrate neuraxis. Science 274, 2007). Therefore it is likely that Lrrn1 forms part of a multiprotein 1109–1115. complex with other co-receptors, or binds to either a complex or small Mao, Y., Kerr, M., Freeman, M., 2008. Modulation of Drosophila retinal epithelial integrity by the adhesion proteins capricious and tartan. PLoS ONE 3, e1827. molecule ligand. Indeed, the crystal structure of another LRR protein, Marin, F., Puelles, L., 1994. Patterning of the embryonic avian midbrain after experimental Lingo-1, reveals that it forms a tetramer which acts as a scaffold to inversions: a polarizing activity from the isthmus. Dev. Biol. 163, 19–37. assemble the Nogo receptor complex (Mosyak et al., 2006). Martinez, S., Crossley, P.H., Cobos, I., Rubenstein, J.L., Martin, G.R., 1999. FGF8 induces formation of an ectopic isthmic organizer and isthmocerebellar development via a repressive effect on Otx2 expression. Development 126, 1189–1200. Acknowledgments Martinez, S., Marin, F., Nieto, M.A., Puelles, L., 1995. Induction of ectopic engrailed expression and fate change in avian rhombomeres: intersegmental boundaries as barriers. Mech. Dev. 51, 289–303. We thank Jon Clarke for advice on DiI labelling. This study was Matsunaga, E., Katahira, T., Nakamura, H., 2002. Role of Lmx1b and Wnt1 in funded by the Medical Research Council. mesencephalon and metencephalon development. Development 129, 5269–5277. McMahon, A.P., Bradley, A., 1990. The Wnt-1 (int-1) proto-oncogene is required for development of a large region of the mouse brain. Cell 62, 1073–1085. References Mellitzer, G., Xu, Q., Wilkinson, D.G., 1999. Eph receptors and ephrins restrict cell intermingling and communication. Nature 400, 77–81. Adams, K.A., Maida, J.M., Golden, J.A., Riddle, R.D., 2000. The transcription factor Lmx1b Milan, M., Cohen, S.M., 2003. A re-evaluation of the contributions of Apterous and Notch maintains Wnt1 expression within the isthmic organizer. Development 127, to the dorsoventral lineage restriction boundary in the Drosophila wing. 1857–1867. Development 130, 553–562. 352 K. Tossell et al. / Developmental Biology 352 (2011) 341–352

Milan, M., Perez, L., Cohen, S.M., 2005. Boundary formation in the Drosophila wing: Shishido, E., Takeichi, M., Nose, A., 1998. Drosophila synapse formation: regulation by functional dissection of Capricious and Tartan. Dev. Dyn. 233, 804–810. transmembrane protein with Leu-rich repeats, CAPRICIOUS. Science 280, Milan, M., Weihe, U., Perez, L., Cohen, S.M., 2001. The LRR proteins capricious and Tartan 2118–2121. mediate cell interactions during DV boundary formation in the Drosophila wing. Taguchi, A., Wanaka, A., Mori, T., Matsumoto, K., Imai, Y., Tagaki, T., Tohyama, M., 1996. Cell 106, 785–794. Molecular cloning of novel leucine-rich repeat proteins and their expression in the Millet, S., Bloch-Gallego, E., Simeone, A., Alvarado-Mallart, R.M., 1996. The caudal limit of developing mouse nervous system. Brain Res. Mol. Brain Res. 35, 31–40. Otx2 gene expression as a marker of the midbrain/hindbrain boundary: a study using Taniguchi, H., Tohyama, M., Takagi, T., 1996. Cloning and expression of a novel gene for in situ hybridisation and chick/quail homotopic grafts. Development 122, 3785–3797. a protein with leucine-rich repeats in the developing mouse nervous system. Brain Millet, S., Campbell, K., Epstein, D.J., Losos, K., Harris, E., Joyner, A.L., 1999. A role for Gbx2 in Res. Mol. Brain Res. 36, 45–52. repression of Otx2 and positioning the mid/hindbrain organizer. Nature 401, 161–164. Tossell, K., Kiecker, C., Wizenmann, A., Lang, E., Irving, C., submitted for publication. Mosyak, L., Wood, A., Dwyer, B., Buddha, M., Johnson, M., Aulabaugh, A., Zhong, X., Notch signalling regulates formation of the boundary at the midbrain-hindbrain Presman, E., Benard, S., Kelleher, K., et al., 2006. The structure of the Lingo-1 interface. ectodomain, a module implicated in central nervous system repair inhibition. Trokovic, R., Jukkola, T., Saarimaki, J., Peltopuro, P., Naserke, T., Weisenhorn, D.M., J. Biol. Chem. 281, 36378–36390. Trokovic, N., Wurst, W., Partanen, J., 2005. Fgfr1-dependent boundary cells Nittenberg, R., Patel, K., Joshi, Y., Krumlauf, R., Wilkinson, D.G., Brickell, P.M., Tickle, C., between developing mid- and hindbrain. Dev. Biol. 278, 428–439. Clarke, J.D., 1997. Cell movements, neuronal organisation and gene expression in Wassarman, K.M., Lewandoski, M., Campbell, K., Joyner, A.L., Rubenstein, J.L., Martinez, S., hindbrains lacking morphological boundaries. Development 124, 2297–2306. Martin, G.R., 1997. Specification of the anterior hindbrain and establishment of a O'Hara, F.P., Beck, E., Barr, L.K., Wong, L.L., Kessler, D.S., Riddle, R.D., 2005. Zebrafish normal mid/hindbrain organizer is dependent on Gbx2 gene function. Development Lmx1b.1 and Lmx1b.2 are required for maintenance of the isthmic organizer. 124, 2923–2934. Development 132, 3163–3173. Wizenmann, A., Lumsden, A., 1997. Segregation of rhombomeres by differential Pan, Y., Lin, M.H., Tian, X., Cheng, H.T., Gridley, T., Shen, J., Kopan, R., 2004. Gamma- chemoaffinity. Mol. Cell. Neurosci. 9, 448–459. secretase functions through Notch signaling to maintain skin appendages but is not Wu, J.Y., Rao, Y., 1999. Fringe: defining borders by regulating the notch pathway. Curr. required for their patterning or initial morphogenesis. Dev. Cell 7, 731–743. Opin. Neurobiol. 9, 537–543. Reifers, F., Bohli, H., Walsh, E.C., Crossley, P.H., Stainier, D.Y., Brand, M., 1998. Fgf8 is mutated Wurst, W., Bally-Cuif, L., 2001. Neural plate patterning: upstream and downstream of in zebrafish acerebellar (ace) mutants and is required for maintenance of midbrain- the isthmic organizer. Nat. Rev. Neurosci. 2, 99–108. hindbrain boundary development and somitogenesis. Development 125, 2381–2395. Xu, Q., Mellitzer, G., Robinson, V., Wilkinson, D.G., 1999. In vivo cell sorting in Sakurai, K.T., Kojima, T., Aigaki, T., Hayashi, S., 2007. Differential control of cell affinity complementary segmental domains mediated by Eph receptors and ephrins. required for progression and refinement of cell boundary during Drosophila leg Nature 399, 267–271. segmentation. Dev. Biol. 309, 126–136. Yaneza, M., Gilthorpe, J.D., Lumsden, A., Tucker, A.S., 2002. No evidence for ventrally Shamim, H., Mahmood, R., Logan, C., Doherty, P., Lumsden, A., Mason, I., 1999. migrating neural tube cells from the mid- and hindbrain. Dev. Dyn. 223, 163–167. Sequential roles for Fgf4, En1 and Fgf8 in specification and regionalisation of the Ye, W., Bouchard, M., Stone, D., Liu, X., Vella, F., Lee, J., Nakamura, H., Ang, S.L., midbrain. Development 126, 945–959. Busslinger, M., Rosenthal, A., 2001. Distinct regulators control the expression of the Shamim, H., Mason, I., 1999. Expression of Fgf4 during early development of the chick mid-hindbrain organizer signal FGF8. Nat. Neurosci. 4, 1175–1181. embryo. Mech. Dev. 85, 189–192. Yuan, S., Schoenwolf, G.C., 1999. The spatial and temporal pattern of C-Lmx1 expression Shinza-Kameda, M., Takasu, E., Sakurai, K., Hayashi, S., Nose, A., 2006. Regulation of in the neuroectoderm during chick neurulation. Mech. Dev. 88, 243–247. layer-specific targeting by reciprocal expression of a cell adhesion molecule, Zervas, M., Millet, S., Ahn, S., Joyner, A.L., 2004. Cell behaviors and genetic lineages of the capricious. Neuron 49, 205–213. mesencephalon and rhombomere 1. Neuron 43, 345–357.