RESEARCH ARTICLE

The Role of the Pax1/9 Gene in the Early Development of Amphioxus Pharyngeal Gill Slits XIN LIU1, GUANG LI1,2, XIAN LIU1, 1,2,3 AND YI‐QUAN WANG * 1

State Key Laboratory of Cellular Stress Biology, School of Life Sciences, Xiamen University, Xiamen, Fujian, China 2 Shenzhen Research Institute of Xiamen University, Shenzhen, Guangdong, China 3 Collaborative Innovation Center of Deep Sea Biology, Xiamen, Fujian, China

ABSTRACT

J. Exp. Zool. (Mol. Dev. Evol.) 324B:30–40, 2015

The pharynx is a major characteristic of chordates. Compared with vertebrates, amphioxus has an advantage for the study of pharynx development, as embryos lack neural crest, and the pharynx is mainly derived from endoderm cells. The Pax1/9 subfamily genes have essential roles in vertebrate pharyngeal patterning, but it is not known if the Pax1/9 gene has similar functions in amphioxus pharynx development. To answer this question, we examined the Pax1/9 gene expression pattern in amphioxus embryos at different developmental stages, and observed morphological changes following Pax1/9 knockdown. RT‐qPCR analysis indicated that Pax1/9 expression was initiated during early neurula stage and rapidly peaked during mid‐neurula stage. Furthermore, in situ hybridization analysis showed that Pax1/9 transcripts were localized exclusively in the most endodermal region of the developing pharynx in early neurula stage embryos; however, Pax1/9 expression was strikingly down‐regulated in the region where gill slits would form from the fusion of endoderm and ectoderm in subsequent developmental stages and was maintained in the border regions between adjacent gill slits. Knockdown of Pax1/9 function using both morpholino and siRNA approaches led to embryonic defects in the first three gill slits, and fusion of the first two gill slits. Moreover, the expression levels of the pharyngeal marker genes Six1/2 and Tbx1/10 were reduced in Pax1/9 knockdown embryos. From these observations, we concluded that the Pax1/9 gene has an important role in the initial differentiation of amphioxus pharyngeal endoderm and in the formation of gill slits, most likely via modulation of Six1/2 and Tbx1/10 expression. J. Exp. Zool. (Mol. Dev. Evol.) 324B:30–40, 2015. © 2014 Wiley Periodicals, Inc. How to cite this article: Liu X, Li G, Liu X, Wang YQ. 2015. The role of the Pax1/9 gene in the early development of amphioxus pharyngeal gill slits. J. Exp. Zool. (Mol. Dev. Evol.) 324B:30–40.

The segmented pharynx is a general feature of chordates, and its development represents an interesting paradigm for deciphering the molecular basis of organogenesis during evolution (Graham and Smith, 2001; Graham, 2008). In vertebrates, the pharynx is segmented by a series of pharyngeal pouches and arches (Graham, 2003). The pharyngeal pouches are segmental diverticula in the pharyngeal endoderm and form pharyngeal gill slits during embryonic development (Grevellec and Tucker, 2010). Together with the neural crest, some of the pouches generally give rise to a variety of essential glands, such as the thymus and parathyroid glands (Graham, 2001). The pharyngeal arches consist of a core of mesoderm

Grant sponsor: National Natural Science Foundation of China; grant numbers: 31101631, 31071110; grant sponsor: Natural Science Foundation of Fujian Province of China; grant number: 2011J05097; grant sponsor: Scientific and Technical Innovation Committee of Shenzhen, China; grant number: CXZZ20120614164555920. Conflicts of interest: None.
Correspondence to: Yi‐Quan Wang, School of Life Sciences, Xiamen University, Xiangan District, Xiamen, Fujian 361102, China. E‐mail: [email protected] Received 31 March 2014; Accepted 26 August 2014 DOI: 10.1002/.22596 Published online in Wiley Online Library (wileyonlinelibrary.com).

© 2014 WILEY PERIODICALS, INC.

ROLE OF Pax1/9 IN THE EARLY DEVELOPMENT OF AMPHIOXUS PHARYNX and neural crest, and produce a number of craniofacial structures, including the jaw and hyoid bones (Graham et al., 2005; Graham and Richardson, 2012). Vertebrate pharyngeal morphogenesis can be regarded as the result of neural crest integration and pharyngeal endoderm regionalization (Graham, 2001). Neural crest cells are believed to be of great importance in pharyngeal pouch and arch development, as they generate skeletal and connective tissue derivatives (Couly et al., 2002); however, it is now apparent that the pharyngeal endoderm plays a more prominent role in directing the formation of particular components of the pharyngeal arches and pouches, and serves a major expression site for a number of developmentally important genes in early embryos (Veitch et al., '99; Piotrowski and Nüsslein‐Volhard, 2000; Trainor and Krumlauf, 2001). Two of those genes, Pax1 and Pax9, which belong to the Pax1/ 9 gene subfamily, are expressed in the pharynx of various vertebrate species and have crucial roles in pharynx development (Lang et al., 2007). In teleost medaka, the two genes are expressed in the endodermal pharyngeal pouches (Mise et al., 2008), while in chick they are expressed dorsally within the pouches (Peters et al., '95; Müller et al., '96); in mouse, Pax1 and Pax9 expression also overlaps the pharyngeal pouches. Loss of Pax1 function in mouse leads to the development of a hypoplastic thymus via impaired thymocyte development (Neubüser et al., '95; Wallin et al., '96; Su and Manley 2000; Su et al., 2001), and Pax9 mutant mice lack the derivatives of the third and fourth pouches, including the thymus and parathyroid glands. Moreover, mouse Pax9 is also expressed in the mesenchyme of the maxillary and mandibular arches and is involved in craniofacial and tooth development (Peters et al., '98). Amphioxus belongs to the subphylum Cephalochordata, a Vertebrata sister group, and is a promising model animal for examining the genetic basis of pharynx development (Holland et al., 2004; Koop and Holland 2008). Because its genome has not undergone the large‐scale gene duplication that co‐occurred with the early appearance of vertebrates, knockdown of a single amphioxus gene can give a discrete phenotype (Dehal and Boore 2005; Putnam et al., 2008). Moreover, the pharynx of the small, transparent amphioxus embryos is composed of only two cell layers, an inner endoderm and an outer ectoderm, and its primary gill slits form ventrally on the right side in an anterior– posterior order through fusion of the pharyngeal endoderm and ectoderm. Additionally, the absence of neural crest in amphioxus greatly facilitates our understanding of the contribution of pharyngeal endoderm to pharynx development (Escriva et al., 2002; Holland et al., 2004). In amphioxus embryos, the pre‐duplicated Pax1/9 gene is strongly expressed in the endoderm of the developing pharynx (Holland et al., '95; Kozmik et al., 2007; Chen et al., 2010), so functional studies of this ancient gene could give us greater insight into the role of pharyngeal endoderm in pharynx patterning.

31

To investigate the function of the Pax1/9 gene in developing amphioxus embryos, we employed two parallel techniques: antisense morpholino oligonucleotide (MO)‐ and synthetic duplex small interfering RNA (siRNA)‐mediated knockdown. Following microinjection, we observed the morphology of the embryonic pharyngeal region under light and electronic microscopes, and detected expression of several pharynx marker genes through whole‐mount in situ hybridization and quantitative PCR.

MATERIALS AND METHODS Animals Adult amphioxi (Branchiostoma belcheri) were cultured in the laboratory as previously described (Zhang et al., 2007; Li et al., 2012). For gene knockdown experiments, we collected ripe eggs and sperm separately via heat‐shock‐induced spawning as described (Li et al., 2013). Before heat‐shock‐induced spawning, we cultured the mature animals at approximately 22°C for a minimum of 5 days. We then cleaned the containers and amphioxi using fresh seawater and placed them in a 27°C water bath in the afternoon before a spawning‐induction day. At approximately 11:00 on the spawning‐induction day, we separated the amphioxi from the sand, placed them into individual plastic cups pre‐filled with 20 mL filtered seawater and then arranged the cups in a 27°C water bath. After the light turned off at 13:30, we examined each cup every 30 min. Once the amphioxi released gametes, the eggs were immediately microinjected prior to fertilization. Pax1/9 Knockdown Experiments MO‐ and siRNA‐mediated Pax1/9 knockdown experiments were carried out. For MO‐mediated knockdown experiments, we designed and synthesized a Pax1/9 antisense morpholino oligonucleotide (MO) (50 ‐GTCATGATGAATATGGAGCAAACAT‐30 ) (Gene Tools, Corvallis, OR, USA) targeting the translation start site within the Pax1/9 mRNA sequence (DQ991501). Additionally, a standard MO (50 ‐CCTCCTACCTCAGTTACAATTTATA‐30 ) was synthesized to serve as a negative control. For siRNA‐mediated knockdown experiments, two synthetic duplex small interfering RNAs (siRNA) targeting different regions of the coding sequence were designed and synthesized (Invitrogen, Carlsbad, CA, USA). Their sequences were as follows: Pax1/9 siRNA‐1, sense r(GCCCAUCAACGCUAACGAATT) and antisense r(UUCGUUAGCG UUGAUGGGCTT); Pax1/9 siRNA‐2, sense r(GGUCCAGUUUCAGGU GCAATT) and antisense r(UUGCACCUGAAACUGGACCTT). A control siRNA was also synthesized using sense r(UUCUCCGAACG UGUCACGUTT) and antisense r(ACGUGACACGUUCGGAGAATT). All siRNAs were resuspended in nuclease‐free water. Microinjection of unfertilized amphioxus eggs was performed as previously described (Liu et al., 2013). The injection solution contained 2% glycerol (Sangon Co., Shanghai, China), 2 mg/mL Texas Red dextran (Invitrogen Co.) and 7 mM MO or 1.4 mg/ml J. Exp. Zool. (Mol. Dev. Evol.)

32

LIU ET AL.

siRNA, and approximately 2 pl of solution was injected into each unfertilized egg. Following the injection, the eggs were fertilized and those that were injected well were raised in an incubator at 25°C with 90% humidity (Liu et al., 2013). Some of the embryos were then photographed using an inverted microscope (Olympus, Tokyo, Japan) at early neurula, early larval and 3‐day larval stages. In Vitro Translation Assay To test the efficiency of Pax1/9 MO, a portion of the Pax1/9 sequence including the MO‐targeted region was amplified using Pax1/9‐F and Pax1/9‐R primers (Table 1) and then cloned into the pCS2 þ MT vector. In vitro translation of Pax1/9‐pCS2 þ MT recombinant plasmid was assayed in combination with Pax1/9 or control MO using the TnT Quick Coupled Transcription/ Translation System (Promega, Madison, WI, USA); translation in the absence of MO was used as a negative control. Following in vitro translation reactions, the solutions were subjected to western blotting using anti‐Myc antibody (Sigma, Louis, MO, USA) to determine if the Pax1/9 MO effectively blocked translation from the Pax1/9‐pCS2 þ MT plasmid. Real‐Time Quantitative PCR Detection To detect the temporal expression of Pax1/9, total RNA was isolated from amphioxus embryos at various stages of development using the RNeasy mini kit (Qiagen, Valencia, CA, USA), and

cDNA was synthesized using the PrimescriptTM RT reagent kit (Takara, Shiga, Japan). To test the efficiency of Pax1/9 siRNAs, total RNA was also extracted from embryos injected with Pax1/9 siRNA or control siRNA at different embryonic stages, and used for cDNA synthesis. Real‐time quantitative PCR (RT‐qPCR) reactions were then carried out using a Rotor‐Gene 6000 Real‐ Time System (Corbett Robotics, San Francisco, CA, USA), using SYBR Green (Promega Co.) for fluorescence detection. Samples were denatured at 94°C for 2 min followed by 40 cycles of 15 sec at 94°C, 15 sec at 60°C and 20 sec at 72°C. Three replicates of each reaction were performed for each sample, using Gapdh expression as an internal control. The RT‐qPCR data were analyzed using the comparative Ct method (Livak and Schmittgen, 2001). Similarly, we also determined the expression levels of Six1/2, Eya and Tbx1/10 genes in the injected embryos, and the primer sequences are listed in Table 1. All statistical analyses were performed using SPSS version 16.0. Histological Observation To prepare histological sections of the gill slits, we collected 3‐day larvae and fixed them in 4% paraformaldehyde dissolved in 4‐morpholinepropanesulfonic acid buffer. After double embedding in agar‐paraffin, samples were cut into 5‐mm serial sections, stained with hematoxylin and eosin, and observed under a microscope. Histological sections of embryos were also

Table 1. Primers used for gene cloning and RT‐qPCR. Primer namea

Sequence (50 ‐30 )

Product

Pax1/9‐F Pax1/9‐R Pax1/9‐RT‐F Pax1/9‐RT‐R Six1/2‐RT‐F Six1/2‐RT‐R Eya‐RT‐F Eya‐RT‐R Tbx1/10‐RT‐F Tbx1/10‐RT‐R Gapdh‐RT‐F Gapdh‐RT‐R Pax1/9‐F2 Pax1/9‐R2 Six1/2‐F Six1/2‐R Eya‐F Eya‐R Tbx1/10‐F Tbx1/10‐R

ccatcgatggatccaagtactggaggc ccatcgatggaagcccagaatgtcact cccatcaacgctaacgaaca gagacgaaactactgactgct accacgacccaagtcagtaa tcggctccagggtttcta gaccctcatcactacaactgg atgggagaactctggcact atcgtgttccattccctc atctcctccattctcaagc ggtggaaaggtcctgctctc ctggatgaaagggtcgttaatgg gcgggtaccaccatgatgaatatggagcaaac tggcactagtttaggaggaagaagcggatg atgctaccttcgttcgggttcac gtggctgaggcaagtgttagg atgaagctgtataatgcctatgcc ccaactttctcatccagtccac atggaagccaacagtccgctct tcaccgcggatctaagtcgtag

Pax1/9‐pCS2 þ MT

a

The letter F denotes the forward primer and R denotes the reverse primer. ClaI restriction sites are underlined.

J. Exp. Zool. (Mol. Dev. Evol.)

Pax1/9‐RT Six1/2‐RT Eya‐RT Tbx1/10‐RT Gapdh‐RT Pax1/9‐pGEM‐T Six1/2‐pGEM‐T Eya‐pGEM‐T Tbx1/10‐pGEM‐T

ROLE OF Pax1/9 IN THE EARLY DEVELOPMENT OF AMPHIOXUS PHARYNX

33

generated following whole‐mount in situ hybridization (WISH) to detect Pax1/9, Six1/2 and Tbx1/10 expression signals. The embryos were embedded in agar‐paraffin, cut into 5‐mm serial sections and then observed under a microscope. For scanning electron microscopy (SEM) observations, 3‐day larvae were fixed with 2.5% glutaraldehyde overnight at 4°C. The embryos were then washed with PBS, dehydrated in a series of solutions of increasing ethanol content to absolute ethyl alcohol, washed in tert‐butanol and stored overnight at 4°C. The next day, samples were critically dried and mounted on aluminum stubs using double‐sided carbon tabs. After sputter‐coating with platinum, the samples were visualized using a JSM‐6390 scanning electron microscope (JEOL, Tokyo, Japan). Gene Expression Detection Using Whole‐Mount In Situ Hybridization To detect the spatial expression pattern of Pax1/9 in developing amphioxus embryos by whole‐mount in situ hybridization (WISH), we amplified the full coding sequence of the gene with primer pairs (Table. 1) and subcloned it into the pGEM‐T easy vector. The correctly recombined vector was then used to synthesize digoxigenin‐labeled antisense RNA probes using the riboprobe1 in vitro transcription system according to the manufacturer’s instructions (Promega Co.). To analyze pharynx defects in Pax1/9 knockdown embryos, we used WISH to examine the expression patterns of three pharyngeal marker genes, Six1/2, Tbx1/10 and Eya. Partial coding sequences of these genes were amplified with primer pairs (Table. 1), subcloned into the pGEM‐T easy vector and then used to synthesize RNA probes. Embryos were fixed at different stages, separately hybridized with riboprobes and photographed using an inverted microscope (Olympus, Tokyo, Japan) as previously described (Qian et al., 2013).

RESULTS Pax1/ 9 Expression in Amphioxus Embryos To analyze Pax1/9 gene function, we examined its temporal and spatial expression pattern in the developing amphioxus embryo. As shown in Figure 1A, Pax1/9 was not expressed in unfertilized eggs and mid‐gastrulas, and this gene expression was not initiated until the early neurula stage. Shortly thereafter, the expression rapidly increased, peaked at the mid‐neurula stage, then decreased and was maintained at a relatively low level in developing larvae (Fig. 1A). WISH analysis also indicated that Pax1/9 gene expression signal was not detected at the mid‐gastrula stage (Fig. 1B), consistent with the RT‐qPCR analysis. Clear Pax1/9 expression was first observed in the pharyngeal endoderm at the early neurula stage (Fig. 1C and D). Concomitant with neurula development, the gene was still strongly expressed in the pharyngeal endoderm and downregulated in the region where the gill slit primordium would subsequently form (Fig. 1E–G). At

Figure 1. Temporal and spatial Pax1/9 expression pattern in developing amphioxus embryos. (A) RT‐qPCR analysis of Pax1/9 gene expression at nine different embryonic stages. (B–J) Pax1/9 in situ hybridization in amphioxus embryos. (B) Mid‐gastrula. (C) Early neurula. (D) Cross‐section through level d of image C. (E) Mid‐neurula. (F and G) Cross‐sections through levels f and g of image E. (H) Early larval stage. (I and J) Cross‐sections through levels i and j of image H. Asterisks in E and F indicate the pharyngeal endoderm that will form the first gill slit primordium, and in H and I they indicate the gill slit primordia of early larvae. With the exception of D, F, G, I and J, all panels are from the lateral view. Scale bars are 50 mm in panels B, C and E, 100 mm in panel H and 20 mm in other panels.

the early larval stage, the signal was still distinct in the pharyngeal endoderm, with the exception of the region containing the first three gill slit primordia (Fig. 1H and I). Additionally, the gene was clearly expressed in the border region between the gill slit primordia (Fig. 1H and J). Morpholino Oligonucleotide‐Mediated Knockdown of Pax1/9 Function In vitro translation studies indicated that the protein could be successfully translated from the Pax1/9‐pCS2 þ MT vector J. Exp. Zool. (Mol. Dev. Evol.)

34 (Fig. 2A, lane 1); however, this translation was inhibited when Pax1/9 MO was added to the reaction (Fig. 2A, lane 2). In contrast, such inhibition did not occur in the presence of control MO (lane 3). Together, these results demonstrate that Pax1/9 MO could effectively block the translation of target gene transcripts. To knock down Pax1/9 gene function in amphioxus embryos, we injected either control or Pax1/9 MO into unfertilized eggs and examined the embryos at different developmental stages following fertilization. At the early neurula stage, the normally developing embryos that had been injected with control MO had a slightly enlarged anterior part of the archenteron where the pharynx would later form, but no pharyngeal apparatus were visible in this region under the microscope (Fig. 2D). Until the early larval stage, the first and second gill slit primordia developed on the ventral side of the body, and the third slit appeared slightly later (Fig. 2E). The first three gill slits then formed and gradually moved toward the right‐ ventral region of the body at the 3‐day larval stage (Fig. 2I and J). Embryos injected with Pax1/9 MO had survival rates of 91.67%, 85.93% and 75.60%, respectively, at the aforementioned

LIU ET AL. developmental stages. These values were nearly equal to those observed for embryos injected with control MO (92.08%, 86.33% and 81.64%, respectively) (Fig. 2B). At the early neurula or early larval stages, most of the knockdown embryos (92.98% and 85.77%, respectively) developed normally, similar to those injected with control MO (90.95% and 87.5%, respectively) (Fig. 2C–G). However, when the first three gill slits developed at the 3‐day larval stage, most of the Pax1/9 MO injected embryos (63.33%) presented smaller gill slits and fusion of the first two gill slits (Fig. 2C,H,K, and L). In contrast, the control MO injected group did not present malformed gill slits at this stage (Fig. 2C,H,I, and J). siRNA‐Mediated Knockdown of Pax1/9 Gene Expression For experiments involving siRNA, we first evaluated the efficiency of two siRNAs for knockdown of the target gene expression using RT‐qPCR. The results of this analysis indicated that Pax1/9 mRNA levels in embryos injected with either Pax1/9 siRNA‐1 or siRNA‐2 were reduced by 85.73% and 73.14%, respectively, in contrast with embryos injected with control

Figure 2. MO‐mediated Pax1/9 gene knockdown in amphioxus embryos. (A) In vitro activity of Pax1/9 MO. All lanes contained Pax1/9‐ pCS2þMT expression plasmid DNA. Lane 1: Pax1/9‐pCS2þMT expression plasmid (negative control); lane 2: Pax1/9‐pCS2þMT expression plasmid plus Pax1/9 MO; lane 3: Pax1/9‐pCS2þMT expression plasmid plus control MO. (B) Embryo survival rates following injection with control or Pax1/9 MO. (C) Ratios of normally developed embryos following injection with control or Pax1/9 MO. (D‐G) Images of embryos injected with control MO (D and E) or Pax1/9 MO (F and G) at the early neurula (D and F) and early larval (E and G) stages. (H) Ratios of different phenotypes in embryos injected with either control or Pax1/9 MO observed at the 3‐day larval stage. (I–L) Images of embryos injected with control MO (I and J) or Pax1/9 MO (K and L) observed at the 3‐day larval stage. In B, C and H, n denotes the total number of embryos observed. Asterisks in D and F denote archenteron. e, endostyle; p, gill slit primordium; g, gill slit. Scale bars are 100 mm in E, G, I, and K, and 50 mm in D, F, J, and L. J. Exp. Zool. (Mol. Dev. Evol.)

ROLE OF Pax1/9 IN THE EARLY DEVELOPMENT OF AMPHIOXUS PHARYNX siRNA (Fig. 3A), demonstrating that both siRNAs effectively knocked down the target gene expression. Following microinjection, we also observed embryos at the three aforementioned developmental stages. Generally, the embryos that had been injected with control siRNA (Fig. 3D,E,K, and L) were morphologically similar to those in the control MO experiments. The survival rates of the embryos injected with Pax1/9 siRNA‐1 or siRNA‐2 were nearly equal to those for embryos injected with control siRNA at each stage (Fig. 3B). At the neurula and early larval stages, most of the embryos injected with Pax1/9 siRNA‐1 (92.66% and 82.38%, respec-

35

tively) or siRNA‐2 (92.07% and 88.85%, respectively) developed normally, similarly to embryos injected with control siRNA (91.12% and 85.67%, respectively) (Fig. 3C–I). However, most of the embryos injected with siRNA‐1 or siRNA‐2 (74.58% or 71.43%, respectively) showed abnormal development at the 3‐day larval stage (Fig. 3J); additionally, the defective embryos presented smaller gill slits and the fusion of front two gills (Fig. 3M–P). As injection of Pax1/9 siRNA‐1 resulted in a stronger reduction of Pax1/9 expression levels than that induced by Pax1/9 siRNA‐2 injection and led to more obvious defects than those caused

Figure 3. siRNA‐mediated knockdown of Pax1/9 expression in amphioxus embryos. (A) Relative expression level of Pax1/9 gene in control siRNA, Pax1/9 siRNA‐1 and Pax1/9 siRNA‐2 injected embryos at the neurula stage. Double asterisks indicate a significant difference between embryos injected with control or Pax1/9 siRNA. (B) Embryo survival rates following injection of control siRNA, Pax1/9 siRNA‐1 or Pax1/9 siRNA‐2. (C) Ratios of normally developed embryos following injection with either control siRNA, Pax1/9 siRNA‐1 or Pax1/9 siRNA‐2. (D–I) Embryos injected with either control siRNA (D and E), Pax1/9 siRNA‐1 (F and G) or Pax1/9 siRNA‐2 (H and I) at the early neurula (D, F and H) and early larval stages (E, G and I). (J) Ratios of different phenotypes in embryos injected with control siRNA, Pax1/9 siRNA‐1 or Pax1/9 siRNA‐2 at the 3‐day larval stage. (K–P) Embryos injected with either control siRNA (K and L), Pax1/9 siRNA‐1 (M and N) or Pax1/9 siRNA‐2 (O and P) at the 3‐day larval stage. In B, C and J, n denotes the total number of embryos. Asterisks in D, F and H indicate archenteron. e, endostyle; p, gill slit primordium; g, gill slits. Scale bars are 100 mm in E, G, I, K, M and O, and 50 mm in D, F, H, L, N and P. J. Exp. Zool. (Mol. Dev. Evol.)

36 by injection of Pax1/9 MO, we used Pax1/9 siRNA‐1 for all subsequent knockdown experiments. To analyze the functional lifespan of Pax1/9 siRNA‐1, we examined Pax1/9 mRNA expression levels at five embryonic stages in embryos that had been injected with control siRNA or Pax1/9 siRNA‐1. This analysis indicated that relative expression levels of Pax1/9 mRNA were significantly reduced in the Pax1/9 siRNA‐1 injected embryos from the early neurula stage to the 3‐day larval stage, demonstrating the efficiency of Pax1/9 siRNA‐1 at these developmental stages (Fig. 4A). Additionally, WISH analysis also indicated that the spatial distribution of Pax1/9 mRNA in Pax1/9 siRNA‐1 injected embryos was similar to that observed for control siRNA injected embryos (Fig. 4B–G).

Figure 4. Pax1/9 siRNA‐1 injection affected Pax1/9 gene expression in amphioxus embryos. (A) Relative expression levels of Pax1/9 gene at five embryonic stages in embryos injected with control siRNA or Pax1/9 siRNA‐1. Double asterisks indicate a significant difference between control siRNA and Pax1/9 siRNA‐1 injected embryos, and a single asterisk indicates a striking difference. (B–G) Distribution of Pax1/9 mRNA in early neurula (B and D), mid‐neurula (C and E) and early larval stage (F and G) embryos injected with control siRNA or Pax1/9 siRNA‐1. All the panels are of the lateral view. Scale bars are 50 mm in panel B, C, D and E, and 100 mm in the others. J. Exp. Zool. (Mol. Dev. Evol.)

LIU ET AL. Morphological Observation of Knockdown Embryos Embryos that had been injected with control siRNA were morphologically well‐developed at the 3‐day larval stage when observed using an inverted microscope, but those injected with Pax1/9 siRNA‐1 displayed defective features. For more detailed observation, we examined those 3‐day larvae under SEM. The results of this analysis indicated that the larvae that had been injected with control siRNA normally developed the first three gill slits on the ventral right side (Fig. 5A). Each gill slit was surrounded by a thick lip, and two adjacent slits were separated by a border. In contrast, the larvae from embryos injected with Pax1/9 siRNA lost the second gill slit, and the remaining primordium was fused to the first gill. The first and the third gill slits of those larvae were also small and had thick lips surrounding them (Fig. 5B). Additionally, we made serial transverse sections of 3‐day stage larvae, and observation of these sections indicated that control siRNA‐injected embryos developed into larvae with fully perforated primary gill slits that were round with thick lips (Fig. 5C,I, and J). The border between two adjacent gill slits was composed of a thin inner endoderm and outer ectoderm (Fig. 5F). In contrast, both the splits and lips of the first three gills of larvae from embryos injected with Pax1/9 siRNA‐1 were smaller than that of the control group (Fig. 5K,P, and R). In particular, the second gill primordium was extremely small and was not punched (Fig. 5P). Sections from the region between the first two gill slits (Fig. 5L–O) also displayed fusion of the two gills in Pax1/ 9 siRNA‐injected embryos. Pharynx Marker Gene Expression In amphioxus, Six1/2 and Eya were strongly expressed in the gill slit primordia, and Tbx1/10 was expressed in the border region between two adjacent gill slits (Mahadevan et al., 2004; Kozmik et al., 2007). WISH analysis indicated that Six1/2 expression was reduced in the pharyngeal endoderm of Pax1/9 knockdown embryos at the early neurula and mid‐neurula stages (Fig. 6A–D) and was downregulated in the three gill slit primordia at the early larval stage (Fig. 6E,E1,E2,F,F1, and F2). In contrast, Eya gene expression at the neurula stages was not affected by Pax1/9 knockdown (Fig. 6G–J), but its expression was slightly reduced in the early larval stages (Fig. 6K and L). Additionally, at the early neurula stage, Tbx1/10 expression was slightly reduced in the pharynx region of embryos injected with siRNA (Fig. 6M and N) and its expression disappeared in the border region at the mid‐ neurula stage (Fig. 6O and P). Expression was also reduced at the border region by early larval stages (Fig. 6Q,Q1,R, and R1). Furthermore, RT‐qPCR analysis also indicated that Six1/2 and Tbx1/10 expression levels in Pax1/9 siRNA‐1 injected embryos were significantly reduced in contrast with those of the control siRNA‐injected embryos at the mid‐neurula and early larval stages (Fig. 6S and U). However, at the early neurula stage, we did not detect any notable reductions in their expression, as their

ROLE OF Pax1/9 IN THE EARLY DEVELOPMENT OF AMPHIOXUS PHARYNX

37

Figure 5. SEM and microscope observations of amphioxus larvae with Pax1/9 gene knockdown. (A and B) SEM images of 3‐day larvae following injection of control siRNA (A) or Pax1/9 siRNA‐1 (B). The embryo heads were oriented to the right. (C–R) Transverse sections through the same stage larvae injected with either control siRNA (C–J) or Pax1/9 siRNA‐1 (K–R). The mouth openings are oriented to the left. C and K are sections through the first gill slit. D–H are five serial sections through the border region between the first and second gill slits. L‐O are four serial sections through the border region between the first and second gill slits. I and P are transverse sections through the second gill slit. Q is a section through the region between the second and third gill slits. J and R are sections through the third gill slit. b, border; m, mouth; l, lip; p, gill slit primordia; s, split. Scale bars indicate 20 mm.

strong expression in the dorsal mesoderm was unaffected by Pax1/9 knockdown. Additionally, Pax1/9 knockdown did not affect Eya expression at either early neurula or mid‐neurula stage, but Eya levels were reduced at the early larval stage (Fig. 6T).

DISCUSSION The Pax1/9 Gene is Required for Amphioxus Gill Slit Development The expression of Pax1/9 genes in the pharynx is a highly conserved feature of chordates (Holland et al., '95; Peters et al., '95; Ogasawara et al., '99; Mise et al., 2008), suggesting a fundamental role for Pax1/9. In mouse, both Pax1 and Pax9 single mutants have defective pharyngeal derivatives (Wallin et al., '96; Peters et al., '98), but their function is unknown in amphioxus. In this study, we found that the zygotic expression of amphioxus Pax1/9 initiated at the early neurula stage and rapidly increased to a peak at the mid‐neurula stage. Spatially, the gene expression was mainly limited to the pharyngeal endoderm

where gill slits later formed. At the early neurula stage when gill slits or their primordia were indistinguishable in histological sections, Pax1/9 gene was expressed throughout the entire pharyngeal region. However, shortly thereafter, this even expression was reduced in the ventral pharyngeal endoderm although no cell differentiation could be clearly observed in the pharyngeal region. With embryonic development, the ventral pharyngeal endoderm would thicken segmentally and these cells would then later form gill slit primordia. Meanwhile, Pax1/9 expression disappeared in the thickened area but persisted in the dorsal region and borders between adjacent primordia. These observations revealed that Pax1/9 expression was required for early pharyngeal endoderm development before gill slit primordium development, but halted in primordial cells once they differentiated from the endoderm. However, we did not observe any apparent morphological defects in Pax1/9 knockdown embryos at either the early neurula, mid‐neurula or even early larval stages. Until the 3‐day larval stage, the hypoplastic second gill slit and the fusion of two anterior J. Exp. Zool. (Mol. Dev. Evol.)

38

LIU ET AL.

Figure 6. Expression of pharynx marker genes in embryos injected with either control siRNA or Pax1/9 siRNA‐1. (A–R) WISH analysis of early neurula, mid‐neurula and early larval stages examining Six1/2, Eya and Tbx1/10 expression in embryos that had been injected with control siRNA or Pax1/9 siRNA. Abbreviations: b, border. Scale bars indicate 100 mm in E, F, K, L, Q, and R, and 50 mm in the others. (E1 and E2) Cross‐sections through levels E1 and E2 of image E. (F1 and F2) Cross‐sections through levels F1 and F2 of image F. (Q1 and R1) Cross‐ sections through levels Q1 and R1 in image Q and R, respectively. Scale bars indicate 20 mm in E1, E2, F1, F2, Q1 and R1. (S–U) Relative expression levels of Six1/2, Eya and Tbx1/10 at early neurula, mid‐neurula and early larval stages in embryos injected with Pax1/9 siRNA or control siRNA. Double asterisks indicate a significant difference between control siRNA and Pax1/9 siRNA‐1 injected embryos, and a single asterisk indicates striking difference.

primordia were observed in both the MO‐ and siRNA‐mediated knockdown experiments. Obviously, malformed larvae were produced by Pax1/9 gene knockdown, although the appearance of defective features did not match the dynamic pattern of Pax1/9 expression. Based on the above observations, we presumed that Pax1/9 might regulate gill slit primordium development by modulating genes expressed in the primordium. The Action of the Pax1/9 Gene in Early Pharyngeal Development It is known that Six1, Eya1 and Tbx1 are expressed in pharyngeal pouches or arches in the early developing embryos of vertebrates, and they are considered to be pharyngeal marker genes (Garg et al., 2001; Xu et al., 2002; Kochilas et al., 2003; Laclef et al., 2003; Ataliotis et al., 2005). Similarly, Six1/2, Eya and Tbx1/10 homologous genes were also expressed in the amphioxus pharyngeal endoderm at the early neurula stage (Mahadevan et al., 2004; Kozmik et al., 2007). In this report, we found that the expression of Six1/2 and Tbx1/10 genes was reduced in the pharyngeal endoderm of early developing amphioxus embryos in response to loss of Pax1/9 function, suggesting that Pax1/9 is required for the expression J. Exp. Zool. (Mol. Dev. Evol.)

of these genes and may have a role in the development of pharyngeal endoderm upstream of Six1/2 and Tbx1/10. Spatially, the expression patterns of both Six1/2 and Tbx1/ 10 partially overlapped with that of Pax1/9 in the pharyngeal region of early neurula stage embryos. Initially, all three genes are expressed in the early pharyngeal endoderm, then, as the embryos develop, Six1/2 expression is focused on the gill slit primordia, and Tbx1/10 is expressed in the ventral brachial arch mesoderm and endoderm (Mahadevan et al., 2004; Kozmik et al., 2007). Pax1/9 expression was continued in the undifferentiated endodermal region but was lost in the primordia. Together, these observations lead us to speculate that amphioxus Pax1/9 plays an important role in the initial development of the pharyngeal endoderm by modulating the expression of Six1/2 and Tbx1/10, consequently affecting the formation of gill slits and the border region. Previous studies have demonstrated functional interactions between Pax1 or Pax9 and Eya1 in mouse pharyngeal pouch development, although their positional relationship is disputed (Xu et al., 2002; Blackburn and Manley, 2004; Zou et al., 2006). This study revealed that in Pax1/9 knockdown embryos, Eya

ROLE OF Pax1/9 IN THE EARLY DEVELOPMENT OF AMPHIOXUS PHARYNX expression was unaffected at the early neurula and mid‐neurula stages, suggesting that the expression of Eya is independent of Pax1/9 in amphioxus.

ACKNOWLEDGMENTS The authors thank Dr. Jr‐Kai Yu at the Institute of Cellular and Organismic Biology, Academia Sinica, Taiwan for the gift of the pCS2 þ MT reporter vector.

LITERATURE CITED Ataliotis P, Ivins S, Mohun TJ, Scambler PJ. 2005. XTbx1 is a transcriptional activator involved in head and pharyngeal arch development in Xenopus laevis. Dev Dyn 232:979–991. Blackburn CC, Manley NR. 2004. Developing a new paradigm for thymus organogenesis. Nat Rev Immunol 4:278–289. Chen L, Zhang QJ, Wang W, Wang YQ. 2010. Spatiotemporal expression of Pax genes in amphioxus: insights into Pax‐related organogenesis and evolution. Sci China Life Sci 53:1031–1040. Couly G, Creuzet S, Bennaceur S, Vincent C, Le Douarin NM. 2002. Interactions between Hox‐negative cephalic neural crest cells and the foregut endoderm in patterning the facial skeleton in the vertebrate head. Development 129:1061–1073. Dehal P, Boore JL. 2005. Two rounds of whole genome duplication in the ancestral vertebrate. PLoS Biol 3:e314. Escriva H, Holland ND, Gronemeyer H, Laudet V, Holland LZ. 2002. The retinoic acid signaling pathway regulates anterior/posterior patterning in the nerve cord and pharynx of amphioxus, a chordate lacking neural crest. Development 129:2905–2916. Garg V, Yamagishi C, Hu T, et al. 2001. Tbx1, a DiGeorge syndrome candidate gene, is regulated by sonic hedgehog during pharyngeal arch development. Dev Biol 235:62–73. Graham A. 2001. The development and evolution of the pharyngeal arches. J Anat 199:133–141. Graham A. 2003. Development of the pharyngeal arches. Am J Med Genet 119A:251–256. Graham A. 2008. Deconstructing the pharyngeal metamere. J Exp Zool (Mol Dev Evol) 310B:336–344. Graham A, Smith A. 2001. Patterning the pharyngeal arches. BioEssays 23:54–61. Graham A, Okabe M, Quinlan R. 2005. The role of the endoderm in the development and evolution of the pharyngeal arches. J Anat 207:479–487. Graham A, Richardson J. 2012. Developmental and evolutionary origins of the pharyngeal apparatus. EvoDevo 3:24–32. Grevellec A, Tucker AS. 2010. The pharyngeal pouches and clefts: development, evolution, structure and derivatives. Semin Cell Dev Biol 21:325–332. Holland ND, Holland LZ, Kozmik Z. 1995. An amphioxus Pax gene, AmphiPax1, expressed in embryonic endoderm, but not in mesoderm: implications for the evolution of class I paired box genes. Mol Mar Biol Biotechnol 4:206–214.

39

Holland LZ, Laudet V, Schubert M. 2004. The chordate amphioxus: an emerging model organism for developmental biology. Cell Mol Life Sci 61:2290–2308. Kochilas LK, Potluri V, Gitler A, Balasubramanian K, Chin AJ. 2003. Cloning and characterization of zebrafish Tbx1. Gene Expr Patterns 3:645–651. Kozmik Z, Holland ND, Kreslova J, et al. 2007. Pax‐Six‐Eya‐Dach network during amphioxus development: conservation in vitro but context specificity in vivo. Dev Biol 306:143–159. Koop D, Holland LZ. 2008. The basal chordate amphioxus as a simple model for elucidating developmental mechanisms in vertebrates. Birth Defects Res C 84:175–187. Laclef C, Souil E, Demignon J, Maire P. 2003. Thymus, kidney and craniofacial abnormalities in Six1 deficient mice. Mech Dev 120:669–679. Lang D, Powell SK, Plummer RS, Young KP, Ruggeri BA. 2007. PAX genes: roles in development, pathophysiology, and cancer. Biochem Pharmacol 73:1–14. Li G, Yang X, Shu ZH, Chen XY, Wang YQ. 2012. Consecutive spawnings of chinese amphioxus, Branchiostoma belcheri, in captivity. PLoS ONE 7:e50838. Li G, Shu ZH, Wang YQ. 2013. Year‐round reproduction and induced spawning of Chinese amphioxus, Branchiostoma belcheri, in laboratory. PLoS ONE 8:e75461. Liu X, Li G, Feng J, Yang X, Wang YQ. 2013. An efficient microinjection method for unfertilized eggs of Asian amphioxus Branchiostoma belcheri. Dev Genes Evol 223:269–278. Livak KJ, Schmittgen TD. 2001. Analysis of relative gene expression data using real‐time quantitative PCR and the 2DDCT method. Methods 25:402–408. Mahadevan NR, Horton AC, Gibson‐Brown JJ. 2004. Developmental expression of the amphioxus Tbx1/10 gene illuminates the evolution of vertebrate branchial arches and sclerotome. Dev Genes Evol 214:559–566. Mise T, Iijima M, Inohaya K, Kudo A, Wada H. 2008. Function of Pax1 and Pax9 in the sclerotome of medaka fish. Genesis 46:185– 192. Müller TS, Ebensperger C, Neubüser A, et al. 1996. Expression of avian Pax1 and Pax9 is intrinsically regulated in the pharyngeal endoderm, but depends on environmental influences in the paraxial mesoderm. Dev Biol 178:403–417. Neubüser A, Koseki H, Balling R. 1995. Characterizations and developmental expression of Pax9, a paired‐box containing gene related to Pax1. Dev Biol 170:701–716. Ogasawara M, Wada H, Peters H, Satoh N. 1999. Developmental expression of Pax1/9 genes in urochordate and hemichordate gills: insight into function and evolution of the pharyngeal epithelium. Development 126:2539–2550. Peters H, Doll U, Niessing J. 1995. Differential expression of the chicken Pax‐1 and Pax‐9 gene: in situ hybridization and immunohistochemical analysis. Dev Dyn 203:1–16.

J. Exp. Zool. (Mol. Dev. Evol.)

40 Peters H, Neubuser A, Kratochwil K, Balling R. 1998. Pax9‐deficient mice lack pharyngeal pouch derivatives and teeth and exhibit craniofacial and limb abnormalities. Genes Dev 12:2735–2747. Piotrowski T, Nüsslein‐Volhard C. 2000. The endoderm plays an important role in patterning the segmented pharyngeal region in zebrafish (Danio rerio). Dev Biol 225:339–356. Putnam NH, Butts T, Ferrier DEK, et al. 2008. The amphioxus genome and the evolution of the chordate karyotype. Nature 453:1064–1072. Qian GH, Li G, Chen XY, Wang YQ. 2013. Characterization and embryonic expression of four amphioxus Frizzled genes with important functions during early embryogenesis. Gene Expr Patterns 13:445–453. Su D, Ellis S, Napier A, Lee K, Manley NR. 2001. Hoxa3 and Pax1 regulate epithelial cell death and proliferation during thymus and parathyroid organogenesis. Dev Biol 236:316–329. Su D, Manley NR. 2000. Hoxa3 and Pax1 transcription factors regulate the ability of fetal thymic epithelial cells to promote thymocyte development. J Immunol 164:5753–5760.

J. Exp. Zool. (Mol. Dev. Evol.)

LIU ET AL. Trainor PA, Krumlauf R. 2001. Hox genes, neural crest cells and branchial arch patterning. Curr Opin Cell Biol 13:698–705. Veitch E, Begbie J, Schilling T, Smith MM, Graham A. 1999. Pharyngeal arch patterning in the absence of neural crest. Curr Biol 9:1481– 1484. Wallin J, Eibel H, Neubüser A, et al. 1996. Pax1 is expressed during development of the thymus epithelium and is required for normal T‐cell maturation. Development 122:23–30. Xu PX, Zheng WM, Laclef C, et al. 2002. Eya1 is required for the morphogenesis of mammalian thymus, parathyroid and thyroid. Development 129:3033–3044. Zhang QJ, Sun Y, Zhong J, et al. 2007. Continuous culture of two lancelets and production of the second filial generations in the laboratory. J Exp Zool (Mol Dev Evol) 308:464–472. Zou D, Silvius D, Davenport J, et al. 2006. Patterning of the third pharyngeal pouch into thymus/parathyroid by Six and Eya1. Dev Biol 293:499–512.