Two apextrin-like proteins mediate extracellular and

Two apextrin-like proteins mediate extracellular and intracellular bacterial recognition in amphioxus Guangrui Huanga,b,c,1, Shengfeng Huanga,1, Xinyu Yana, Ping Yanga, Jun Lia, Weiya Xua, Lingling Zhanga, Ruihua Wanga, Yingcai Yua, Shaochun Yuana, Shangwu Chena, Guangbin Luob,c, and Anlong Xua,b,2 a State Key Laboratory of Biocontrol, Open Laboratory for Marine Functional Genomics of the State High-Tech Development Program, Guangdong Province Key Laboratory for Pharmaceutical Functional Genes, School of Life Sciences, Sun Yat-Sen University, Guangzhou, Guangdong 510006, People’s Republic of China; bSchool of Basic Medical Sciences, Beijing University of Chinese Medicine, Beijing 100029, People’s Republic of China; and cDepartment of Genetics and Genome Sciences, Case Western Reserve University School of Medicine, Cleveland, OH 44106

Animals exploit different germ-line-encoded proteins with various domain structures to detect the signature molecules of pathogenic microbes. These molecules are known as pathogen-associated molecular patterns (PAMPs), and the host proteins that react with PAMPs are called pattern recognition proteins (PRPs). Here, we present a novel type of protein domain structure capable of binding to bacterial peptidoglycan (PGN) and the minimal PGN motif muramyl dipeptide (MDP). This domain is designated as apextrin C-terminal domain (ApeC), and its presence was confirmed in several invertebrate phyla and subphyla. Two apextrin-like proteins (ALP1 and ALP2) were identified in a basal chordate, the Japanese amphioxus Branchiostoma japonicum (bj). bjALP1 is a mucosal effector secreted into the gut lumen to agglutinate the Grampositive bacterium Staphylococcus aureus via PGN binding. Neutralization of secreted bjALP1 by anti-bjALP1 monoclonal antibodies caused serious damage to the gut epithelium and rapid death of the animals after bacterial infection. bjALP2 is an intracellular PGN sensor that binds to TNF receptor-associated factor 6 (TRAF6) and prevents TRAF6 from self-ubiquitination and hence from NF-κB activation. MDP was found to compete with TRAF6 for bjALP2, which released TRAF6 to activate the NF-κB pathway. BjALP1 and bjALP2 therefore play distinct and complementary functions in amphioxus gut mucosal immunity. In conclusion, discovery of the ApeC domain and the functional analyses of amphioxus ALP1 and ALP2 allowed us to define a previously undocumented type of PRP that is represented across different animal phyla. pattern recognition receptor

recognized outside the cells by cell-surface receptors, whereas others are recognized intracellularly after the PAMPs have entered the cell. For example, LPS is recognized by cell-surface receptor TLR4, and muramyl dipeptide (MDP) is recognized by NOD2 in the intracellular space (1, 2). Pathogen recognition by sensor PRPs may initiate coordinated immune responses at the molecular, cellular, tissue, and organ levels (8). The NF-κB family of transcription factors plays a central role in regulating the responses triggered by many PRPs, including TLRs, NLRs, and RLRs (9, 10). TNF receptor-associated factor 6 (TRAF6) is a crucial intermediate signal transducer for activating NF-κB in response to the signaling of TLRs, NLRs, RLRs, and TNF receptors (11, 12). During signaling, TRAF6 is activated by upstream adaptors [e.g., Myeloid differentiation primary response gene (88) (MYD88) and Tolllike receptor adaptor molecules] and kinases (e.g., IL-1 receptor-associated kinases and receptor-interacting serine-threonine kinases) (10, 11) and acts as an E3 ubiquitin ligase to catalyze Lys-63–linked polyubiquitination on itself as well as on IκB kinase γ (13). TGF-beta–activated kinase 1 (TAK1) and its binding proteins TAB2 and TAB3 are then recruited to TRAF6; TAK1 then activates downstream IκB kinases, ultimately leading to NF-κB activation (14, 15). TRAF6-depedent NF-κB activation is not only important to the vertebrate immune system but is also Significance Microbial specific pathogen-associated molecular patterns (PAMPs) constitute a key feature by which a host organism detects the presence of microbes and mounts specific immune responses. Here, we report the discovery of two proteins (bjALP1 and 2) that interact with muramyl dipeptide, a panbacterial specific PAMP via a novel pattern recognition domain ApeC. Our studies have revealed that bjALP1 is a secreted immune effector, whereas bjALP2 functions as an intracellular pattern recognition receptor (PRR), both having an important role in protecting the host from microbial pathogens. Specifically, bjAPL1 functions in the extracellular space to reduce the harmful effect of pathogenic microbes, whereas bjALP2 functions as a PRR that serves as a sentinel for intracellular bacterial invasion.

| immune response | NF-κB signaling pathway

T

he innate immune system plays an essential role in the first line of host defense against pathogenic microbes. Microbes express signature molecules known as pathogen-associated molecular patterns (PAMPs), such as lipopolysaccharide (LPS) in Gram-negative bacteria, peptidoglycan (PGN) and lipoteichoic acid (LTA) in Gram-positive bacteria, mannan in yeast, double-stranded RNA in viruses, and unmethylated CpG motifs in bacteria, which are essential for survival and hence difficult for microbes to change via adaptive evolution (1, 2). The innate immune system relies on germ-line-encoded pattern recognition proteins (PRPs) to detect and react with PAMPs. A variety of PRPs based on different protein domain structures with distinct PAMP specificities have been discovered in animals, including Toll-like receptors (TLRs), nucleotide-binding oligomerization domain (NOD)-like receptors (NLRs), RIG-I–like receptors (RLRs), scavenger receptors, C-type lectin receptors (CLRs), peptidoglycan recognition proteins (PGRPs), Gramnegative bacteria-binding proteins (GNBPs), siglecs, beta2-integrins, collectins, and ficolins (3–7). Some PRP families serve as dedicated sensors (e.g., TLRs and RLRs), and the others are exclusive effectors (e.g., collectins and ficolins), or both (e.g., CLRs and PGRPs). Sensor PRPs can be intracellular sensors (e.g., NLRs and RLRs), extracellular sensors (e.g., PGRPs and GNBPs), or both (e.g., TLRs) (3–7). Some PAMPs are

www.pnas.org/cgi/doi/10.1073/pnas.1405414111

Author contributions: G.H., S.H., G.L., and A.X. designed research; G.H., X.Y., P.Y., J.L., L.Z., and R.W. performed research; A.X. directed the project; G.H., S.H., W.X., Y.Y., S.Y., S.C., G.L., and A.X. analyzed data; and G.H., S.H., G.L., and A.X. wrote the paper. The authors declare no conflict of interest. This article is a PNAS Direct Submission. Freely available online through the PNAS open access option. Data deposition: The sequences reported in this paper have been deposited in the GenBank database (accession nos. KM017614 and KM017615) and the Pfam database (accession no. PF16977, which will be available in the v28.0 release). 1

G.H. and S.H. contributed equally to this work.

2

To whom correspondence should be addressed. Email: [email protected].

This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10. 1073/pnas.1405414111/-/DCSupplemental.

PNAS Early Edition | 1 of 6

IMMUNOLOGY AND INFLAMMATION

Edited by Max D. Cooper, Emory University, Atlanta, GA, and approved August 14, 2014 (received for review March 24, 2014)

indispensable for the Toll and Imd antimicrobial signaling pathways in Drosophila (16, 17). Amphioxus represents the basal living chordate lineage that diverged from two other lineages (urochordates and vertebrates) ∼550 Mya (18). This chordate invertebrate underwent no wholegenome duplications, shares extensive genomic conservation with vertebrates, and is one of the best proxies for understanding the chordate ancestral immune gene system (19–22). Amphioxus has not be found to contain molecular hallmarks of adaptive immunity, including antigen receptors and MHC molecules, but it has a much stronger innate immune gene system, which contains nearly 10 times more receptor and mediator genes than in humans and is tightly regulated during immune responses (22, 23). To date, several amphioxus PRPs have been verified as antimicrobial effectors, including a CLR, a ficolin, and a C1qlike protein (24–26). Functional analyses show that amphioxus also relies on the conserved TLR–MyD88–NF-κB and TRAF6– NF–κB/IκB pathways to regulate immune responses (27–29). Here, we report two proteins from the amphioxus Branchiostoma japonicum (bj) that contain the ApeC domain, a novel pattern recognition domain for MDP, a bacterial-specific PAMP. The proteins share high similarity with the C-terminal regions of the apextrin proteins that were first identified in sea urchins (30, 31). We designate this domain apextrin C-terminal domain (ApeC) and named them apextrin-like protein 1 and 2 (bjALP1 and bjALP2). Functional analyses showed that bjALP1 acted as an extracellular effector for bacterial agglutination, whereas bjALP2 could mediate intracellular bacterial recognition and NF-κB activation. We therefore demonstrate that these ApeCcontaining proteins represent a previously unidentified family of immune response proteins that have important roles in antimicrobial defense. Results ALPs and the Novel ApeC Domain. In this study, we cloned two full-length cDNA sequences from amphioxus B. japonicum (bj), encoding a 504-aa protein (bjALP1) and a 462-aa protein (bjALP2). The two proteins share 54% sequence identity and both consist of a signal peptide, an N-terminal spacer region, and a C-terminal region of ∼200 aa (Figs. S1 and S2). A database search revealed that putative ApeC-containing proteins were presence in many invertebrate species, but no clear or unambiguous homologs were found in Caenorhabditis elegans, fruit flies, urochordates, or vertebrates (Fig. S3A). Thus, it seems that bjALP1 and 2 are membersof a group of ApeC-containing proteins that are uniquely prevalent in the invertebrate phyla. Phylogenic analysis shows that there are no cross-(sub)phylum orthologs found between ALPs (Fig. S3B). Gene Expression Patterns of bjALP1 and bjALP2. A quantitative realtime PCR (qRT-PCR) analysis showed that in adult animals bjALP1 mRNA was more highly expressed than bjALP2 in all six tissues examined, including gill, hepatic cecum, intestine, skin, muscle, and ovary (Fig. 1 A and B). Both genes showed the highest expression in the hepatic cecum and skin, whereas bjALP1 also had substantial expression in the gill and intestine. Section in situ hybridization confirmed the mRNA distribution of both genes in the digestive tract, skin, and ovary and the higher expression level of bjALP1 than of bjALP2 (Fig. 1 C and D). We further found that in the gut of adult animals the mRNA expression of both ALP genes was highly unregulated during acute antibacterial responses. BjALP1 showed an astounding 3,700-fold induction at 12 h after Gram-negative Vibrio anguillarum infection and 2,759-fold induction at 24 h after Grampositive Staphylococcus aureus infection, whereas the peak upregulation of bjALP2 was approximately 10 times weaker than that of bjALP1 (Fig. 1 E and F). These data suggest that bjALP1 is more strongly expressed and induced compared with bjALP2 and that both genes might be involved in gut mucosal immunity. 2 of 6 | www.pnas.org/cgi/doi/10.1073/pnas.1405414111

Fig. 1. Expression patterns of amphioxus ALPs. (A and B) Expression of bjALP1 and bjALP2 mRNA in different tissues was determined by qRT-PCR analysis. The data are expressed as a ratio to the mRNA expression level of β-actin and are plotted as the means ± SD. (C) In situ hybridization of bjALP1 in adult female sections. (D) In situ hybridization of bjALP2 in adult female sections; the sense probe was used as the negative control. e, endostyle; g, gill; hc, hepatic cecum; i, intestine; m, muscle; nc, notochord; o, ovary; s, skin; sc, spinal cord. (E and F) The relative expression level of ALP mRNA in hepatic ceca and intestine after challenge with V. anguillarum and S. aureus was determined by qRT-PCR analysis. The data are shown as a ratio to the ALP mRNA expression level after the injection of PBS and are plotted as the means ± SD; *P < 0.05, **P < 0.01, ***P < 0.001.

Bacterial Agglutination Mediated by bjALP1 and bjALP2. This extremely high expression and induction during infection led us to suspect that bjALP1 encodes a secreted mucosal effector. We purified recombinant His-tagged bjALP1 and bjALP2 proteins and incubated them with 10 different bacterial strains. After incubation, the bacterial pellets were washed and analyzed by Western blotting with anti-His monoclonal antibodies, which showed that both recombinant proteins could bind to Grampositive bacteria with different affinities but not to Gram-negative bacteria (Fig. 2A). We then incubated the bjALP1 and bjALP2 recombinant proteins with FITC-labeled bacteria to determine the bacterial aggregation effect. The results showed that the recombinant proteins caused significant aggregation of the Gram-positive bacteria Enterococcus faecium and S. aureus, but not Gram-negative E. coli (Fig. 2B). Quantification of the diameter of green bacterial puncta suggested that the aggregation effect was stronger for S. aureus than for E. faecium (Fig. S4A). Intriguingly, however, we could not detect a significant bactericidal activity from recombinant bjALP1 (Fig. S4 B and C). bjALP1 and bjALP2 Bind PGN Through the ApeC Domain. We used enzyme-linked immunosorbent assay (ELISA) to evaluate the Huang et al.

could only bind efficiently to Gram-positive and not to Gramnegative bacterium (Fig. 2A).


Distribution of the Endogenous bjALP1 and bjALP2 Proteins. We produced anti-bjALP1 and anti-bjALP2 monoclonal antibodies (mAbs) using the peptides 412 GEVPDGNYDR 421 and 375 DGIYNRNTEI384, respectively (Fig. S6). We performed immunogold analysis of hepatic ceca of amphioxus with these antibodies and found that gold particles localize bjALP1 mainly in mucus, with some scattered signals in cytoplasm (Fig. 4 A and B), whereas for bjALP2 most of the gold particles were distributed near plasma membrane in the cytoplasm side (Fig. 4 C and D and

Fig. 2. Amphioxus ALPs bind and aggregate microbes. (A) The binding of microorganisms by recombinant ALP proteins. Living microbial strains were incubated with ALP proteins in PBS buffer, and the washed pellets were subjected to SDS/PAGE and detection by Western blotting with an anti-His antibody. (B) Aggregation of microbes by amphioxus ALPs. The targeted proteins were incubated with FITC-labeled E. faecium, S. aureus, or E. coli (1 × 107 cfu/mL) in PBS for 1 h at room temperature and were examined using fluorescence microscopy.

interaction between the bjALP1 and bjALP2 recombinant proteins and different bacterial cell wall components. Both bjALP1 and bjALP2 could interact with soluble PGN from S. aureus and E. coli but not with soluble LPS, LTA, zymosan, or mannan (Fig. 3A). Pull-down assays confirmed that both recombinant proteins directly bound with insoluble PGN from S. aureus and E. coli in a dose-dependent manner (Fig. 3B). Indeed, both bjALP1 and 2 could bind specifically to MDP, the minimal PGN motif (Fig. 3C). The interaction was investigated further by using mutant bjALP1 or 2 that either lacks the ApeC domain (bjALP1ΔC, bjALP2ΔC) or the N-terminal spacer domain (bjALP1ΔN, bjALP2ΔN). The results showed that only the bjALP1ΔN and bjALP2ΔN recombinant proteins could bind with PGN and MDP (Fig. 3 D and Fig. S5), suggesting that ApeC is required for PGN/MDP recognition. Taken together, we concluded that both bjALP1 and bjALP2 are capable of agglutinating Gram-positive bacteria via interaction between the ApeC domain and the bacterial cell-wall component PGN/MDP. However, because the MDP moieties are exposed at the cell surface of Gram-positive but not of Gram-negative bacterium (32), bjALP1 or bjALP2 Huang et al.

Fig. 3. Amphioxus ALPs directly interacted with PGN. (A) ELISA analysis of the interaction between recombinant ALP proteins and microbial components. Plates were coated with 20-μg components then were washed and incubated with bjALPs at 4 °C overnight followed by detection with an anti-His antibody. One representative experiment of six is shown. Background absorbance without protein was subtracted. (B) Pull-down analysis of the binding of ALP proteins to PGN from S. aureus and E. coli. P, pellet protein; T, total protein. (C ) ELISA analysis of the interaction between recombinant ALP proteins and MDP. One representative experiment of six is shown. Background absorbance without protein was subtracted. (D) ELISA analysis of the interaction between the truncated bjALP1 mutants and PGN. One representative experiment of six is shown. Background absorbance without protein was subtracted. *P < 0.05, **P < 0.01, ***P < 0.001.


Fig. S7 A and B). In the excrement of these animals, which mainly consisted of mucus and S. aureus, we observed a large amount of bjALP1 surrounding the bacteria (Fig. 4E). In contrast, no bjALP2 was detected in the excrement (Fig. 4F). Western blotting confirmed the presence of bjALP1 and the absence of bjALP2 in the excrement of these bacterially fed animals (Fig. S7C). BjALP1 Protected Amphioxus from the Harmful Effect of Bacterial Infections. Considering the abundance of bjALP1 in the feces

of infected animals, the massive induction during bacterial infection, and the bacterial agglutination effect, we suspected that bjALP1 could be an indispensable extracellular effector of gut mucosal immunity. The anti-bjALP1 mAb is specific for the ApeC domain (412–421 aa) of bjALP1 and may interfere with the interaction between bjALP1 and PGN owing to steric hindrance. ELISA verified that the addition of the anti-bjALP1 mAb did reduce the binding of recombinant bjALP1 to MDP in a dose-dependent fashion (Fig. S8). Therefore, anti-ALP1 mAb can be used to block the function of endogenous bjALP1 proteins in the gut of amphioxus. As shown in Fig. 5 A and B, oral feeding of the anti-ALP1 mAb significantly decreased the survival rates of adult amphioxus after bacterial infection. Tissue section examination showed that without the protection of extracellular bjALP1 the gill of infected amphioxus was completely destroyed by S. aureus infection within 72 h (Fig. 5 C–F). bjALP2 Could Function As an Intracellular Pattern Recognition Receptor.

Although bjALP2 proteins were present within gut epithelial cells, they could not be detected in the excrement of infected amphioxus. Also, a full-length bjALP2-enhanced GFP (EGFP) fusion protein, when ectopically expressed in HEK293T cells, was detected within the cells but not in the culture medium (Fig. S7D). The signal peptide of bjALP2 is clearly functional because fusing the bjALP2 signal peptide to the EGFP protein resulted in the secretion of EGFP (Fig. S7D). Nonetheless, replacing bjALP2’s original signal peptide with the signal peptides of Igκ and preprotrypsin also failed to induce bjALP2 secretion (Fig. S7E). Therefore, it seems that certain feature(s) of bjALP2

Fig. 5. BjALP1 plays an important role in the anti-bacterial immune response of amphioxus. (A) Survival analysis of amphioxus following feeding with S. aureus with or without the anti-bjALP1 antibody. (B) Survival analysis of amphioxus following feeding with V. anguillarum with or without the anti-bjALP1 antibody. Purified IgGs from preimmunized mouse serum were used as the control. (C–F) Section H&E staining of amphioxus after feeding with normal IgG (C), the anti-bjALP1 antibody (D), S. aureus and normal IgG (E), and S. aureus and anti-bjALP1 antibody (F). All animals were collected at 72 h after oral infection. Ten animals were used for each treatment. The representative results of three experiments are shown. g, gill; hc, hepatic cecum; m, muscle; nc, notochord; o, ovary; s, skin; sc, spinal cord; sp, spermary. *P < 0.05, **P < 0.01.

render its secretary signal peptide ineffective, leading to its cytoplasmic retention despite having a signal peptide. bjALP2 Could Modulate the Activation of the TRAF6–NF-κB Pathway.

Fig. 4. Immunogold analysis of bjALPs. (A and B) Immunogold electron microscope (EM) localization of bjALP1 in the hepatic cecum of amphioxus. (C and D) Immunogold EM localization of bjALP2 in the hepatic cecum of amphioxus. (E) Immunogold EM localization of bjALP1 in the excrement of S. aureus-fed amphioxus. (F) Immunogold EM localization of bjALP2 in the excrement of S. aureus-fed amphioxus. B, bacteria; L, lysosome; Mi, mitochondria; Mu, mucus; N, nucleus; PM, plasma membrane.

4 of 6 | www.pnas.org/cgi/doi/10.1073/pnas.1405414111

A sequence analysis revealed a potential TRAF6-binding motif in the N-terminal spacer region of bjALP2 (Fig. 6A). We suspected that when retained in the cytoplasm bjALP2 might interact with amphioxus TRAF6 (bjTRAF6). Using luciferase reporter assays in HEK293T cells, we found that full-length bjALP2 strongly inhibited NF-κB activation induced by amphioxus MyD88 (bjMyd88) and bjTRAF6 in a dose-dependent manner (Fig. 6B). Using two mutant constructs, bjALP2ΔC (lacking the ApeC domain) and bjALP2ΔN (lacking the N-terminal spacer region), we further showed that the NF-κB inhibitory activity of bjALP2 is mainly determined by its N-terminal spacer region (Fig. 6C). We then analyzed the subcellular location of bjALP2 and bjTRAF6. When coexpressed with the myc-tagged bjTRAF6 protein in HeLa cells, the bjALP2-EGFP fusion protein colocalized with myc-tagged bjTRAF6 in the cytoplasm (Fig. 6D). We then constructed bjALP2ΔT6, a mutant without the TRAF6binding motif. When overexpressed in HeLa cells, the subcellular localizations of bjALP2ΔT6 and bjALP2 were indistinguishable (Fig. S9 A and B), suggesting that the TRAF6-binding motif has no obvious impact on the subcellular location of bjALP2. However, bjALP2ΔT6-EGFP failed to colocalize with myc-tagged bjTRAF6, suggesting that bjALP2 could not interact with bjTARF6 without the TRAF6-binding motif (Fig. S9C). Indeed, coimmunoprecipitation (Co-IP) assays confirmed that bjALP2 Huang et al.

interacted with bjTRAF6 when overexpressed in 293T cells (Fig. 6E). Furthermore, mutation of the TRAF6-binding motif not only abolished the interaction between bjALP2 and bjTRAF6 but also relieved the inhibition of bjALP2 on NF-κB activation (Fig. 6 E and F). The self-ubiquitination of TRAF6 is key event in NF-κB activation (13, 15). Thus, to determine whether bjALP2 can interfere with the self-ubiquitination of bjTRAF6, we created two plasmid constructs containing Flag-tagged bjTRAF6 and HA-tagged bjUbiquitin, respectively. When these two constructs were cotransfected into HEK293T cells with the EGFP-tagged bjALP2 construct we recorded a significant reduction of the ubiquitination level of bjTRAF6 (Fig. 6G). MDP Could Relieve the Inhibition of bjALP2 on bjTRAF6. In mammals, whereas MDP is a PAMP that is recognized by NOD2, an intracellular pattern recognition receptor, MDP cannot cross the plasma membrane of mammalian cells efficiently through diffusion (33). Interestingly, S. aureus, a microbial pathogen for amphioxus, could get into the intracellular space of the amphioxus gut epithelium shortly after being administered orally (34). Together, these observations have suggested that MDP enters the cells as part of an intact microbe and hence serves as an excellent indicator for microbe intracellular invasion in amphioxus. In this context, our finding that bjALP2 also physically interacted with bjTRAF6 prompted us to speculate whether the interaction of bjALP2 with bjTRAF6, and accordingly the effect of TRAF6-mediated NF-κB activation, could be affected by MDP. Because we showed that bjALP2 is able to recognize PGN/MDP through its C-terminal ApeC domain, we sought to ascertain whether MDP can prevent bjALP2 from suppressing bjTRAF6. We therefore performed Co-IP assays with the presence of MDP (during plasmid transfection) in 293T cells, which showed that the binding between bjALP2 and bjTRAF6 was weakened by the addition of MDP in a dose-dependent manner (Fig. 6H). Using luciferase reporter assays, we observed that the Huang et al.

presence of MDP did reverse the inhibitory effect of bjALP2 on bjTRAF6-dependent NF-κB activation in a dose-sensitive fashion (Fig. 6I). In summary, our functional data indicate that the recognition of PGN/MDP by cytoplasmic bjALP2 freed bjTRAF6 from the binding with bjALP2 and released bjTRAF6, which subsequently activated the downstream NF-κB pathway. Discussion and Conclusions Apextrin and its homologs have been repeatedly documented since 1998, but their structures and functions have remained elusive (30, 31, 35–42). Here, we demonstrate that ALPs contain ApeC, a novel protein domain structure that is conserved in several invertebrate lineages and capable of recognizing the bacterial cell wall component PGN and its moiety MDP. Without the crystal structure of ApeC we cannot rule out the possibility that there

Fig. 7. Proposed model of the synergic function of BjALP1 and BjALP2. During the immune response a large amount of BjALP1 proteins are secreted to form a mucosal cushion with other mucosal effectors to protect the epithelium surface, whereas intracellular BjALP2 regulates the immune response and homeostasis by monitoring the PGN/MDP concentration.



Fig. 6. BjALP2 interacts with bjTRAF6 via the TRAF6-binding motif and inhibits the ubiquitination of TRAF6 to suppress NF-κB activation. (A) Putative TRAF6-binding motifs identified in amphioxus ALPs. (B) Luciferase reporter assay showing the BjALP2 inhibition of NF-κB activation induced by bjMyD88 and bjTRAF6. (C) The influence of truncated bjALP2 mutants on NF-κB activation as analyzed by luciferase reporter assay. (D) BjALP2 colocalizes with amphioxus TRAF6 in HeLa cells. HeLa cells were cotransfected with EGFP-fused bjALP2 and myctagged pCMV-bjTRAF6 and were stained with an anti-myc antibody and an Alexa Fluor 532 secondary antibody. (E ) Co-IP assay showing that bjALP2, but not the TRAF6-binding motif-deleted mutant, interacts with amphioxus TRAF6 when overexpressed in 293T cells. (F) Luciferase reporter assay showing that the TRAF6-binding motifdeleted mutant of bjALP2 did not inhibit NF-κB activation. (G) Co-IP assay showing that bjALP2 suppressed the ubiquitination of amphioxus TRAF6 when coexpressed in 293T cells. The data are representative of at least three independent experiments. (H) Co-IP assay showing that MDP diminished the binding of bjALP2 with bjTRAF6. (I) Luciferase reporter assay showing that MDP relieved the inhibition by bjALP2 of bjTRAF6-induced NF-κB activation. *P < 0.05, **P < 0.01.

are distant ApeC homologs in vertebrates and several other invertebrates (C. elegans, Drosophila, and urochordates). Regardless, ApeC is an ancient domain type that can be dated back to the common ancestor of eumetazoa. Comprehensive functional analyses revealed the role of two amphioxus ALPs and their ApeC domains in gut mucosal immune responses (Fig. 7). Briefly, a huge amount of ALP1 is secreted into the gut lumen to protect the epithelium, together with other extracellular effectors. ALP1 aggregates and immobilizes Gram-positive bacteria and hence most likely facilitates the killing and excreting of these bacteria. ALP2, however, is retained within epithelial cells by an unknown mechanism that likely arose through adaptive evolution. ALP2 binds with TRAF6 and prevents TRAF6 from self-ubiquitination and hence from NF-κB activation. Endoplasmic PGN/MDP competes with TRAF6 for ALP2, which releases a sufficient amount of free TRAF6 to activate the NF-κB pathway. Based on the current information, we can conceive a possible interplay between the two amphioxus ALPs. During bacterial infection in the gut a large amount of PGN and MDP may enter the cytoplasm of gut epithelial cells via the pinocytosis of PGN/MDP or phagocytosis of the bacteria. PGN/MDP binds with ALP2 and releases a sufficient amount of free TRAF6 to activate the NF-κB pathway. In return, NF-κB most likely induces the expression of ALP1. If so, ALP1 will help to quickly clear out the bacteria in the gut lumen and hence reduce the amount of PGN/MDP that ALP2 can access. ALP2 will then rebind TRAF6 to suppress NF-κB. As 1. Medzhitov R (2007) Recognition of microorganisms and activation of the immune response. Nature 449(7164):819–826. 2. Akira S, Uematsu S, Takeuchi O (2006) Pathogen recognition and innate immunity. Cell 124(4):783–801. 3. Kim YS, et al. (2000) Gram-negative bacteria-binding protein, a pattern recognition receptor for lipopolysaccharide and beta-1,3-glucan that mediates the signaling for the induction of innate immune genes in Drosophila melanogaster cells. J Biol Chem 275(42):32721–32727. 4. Areschoug T, Gordon S (2008) Pattern recognition receptors and their role in innate immunity: Focus on microbial protein ligands. Contrib Microbiol 15:45–60. 5. Dziarski R, Gupta D (2006) The peptidoglycan recognition proteins (PGRPs). Genome Biol 7(8):232. 6. Fujita T, Matsushita M, Endo Y (2004) The lectin-complement pathway—its role in innate immunity and evolution. Immunol Rev 198:185–202. 7. Mukhopadhyay S, Gordon S (2004) The role of scavenger receptors in pathogen recognition and innate immunity. Immunobiology 209(1-2):39–49. 8. Hayden MS, Ghosh S (2011) NF-κB in immunobiology. Cell Res 21(2):223–244. 9. Li Q, Verma IM (2002) NF-kappaB regulation in the immune system. Nat Rev Immunol 2(10):725–734. 10. Kumar H, Kawai T, Akira S (2011) Pathogen recognition by the innate immune system. Int Rev Immunol 30(1):16–34. 11. Xie P (2013) TRAF molecules in cell signaling and in human diseases. J Mol Signal 8(1):7. 12. Hayden MS, Ghosh S (2012) NF-κB, the first quarter-century: Remarkable progress and outstanding questions. Genes Dev 26(3):203–234. 13. Lamothe B, et al. (2007) Site-specific Lys-63-linked tumor necrosis factor receptorassociated factor 6 auto-ubiquitination is a critical determinant of I kappa B kinase activation. J Biol Chem 282(6):4102–4112. 14. Wang C, et al. (2001) TAK1 is a ubiquitin-dependent kinase of MKK and IKK. Nature 412(6844):346–351. 15. Kanayama A, et al. (2004) TAB2 and TAB3 activate the NF-kappaB pathway through binding to polyubiquitin chains. Mol Cell 15(4):535–548. 16. Cha GH, et al. (2003) Discrete functions of TRAF1 and TRAF2 in Drosophila melanogaster mediated by c-Jun N-terminal kinase and NF-kappaB-dependent signaling pathways. Mol Cell Biol 23(22):7982–7991. 17. Lemaitre B, Hoffmann J (2007) The host defense of Drosophila melanogaster. Annu Rev Immunol 25(1):697–743. 18. Delsuc F, Brinkmann H, Chourrout D, Philippe H (2006) Tunicates and not cephalochordates are the closest living relatives of vertebrates. Nature 439(7079):965–968. 19. Holland LZ, et al. (2008) The amphioxus genome illuminates vertebrate origins and cephalochordate biology. Genome Res 18(7):1100–1111. 20. Putnam NH, et al. (2008) The amphioxus genome and the evolution of the chordate karyotype. Nature 453(7198):1064–1071. 21. Louis A, Roest Crollius H, Robinson-Rechavi M (2012) How much does the amphioxus genome represent the ancestor of chordates? Brief Funct Genomics 11(2):89–95. 22. Huang S, et al. (2008) Genomic analysis of the immune gene repertoire of amphioxus reveals extraordinary innate complexity and diversity. Genome Res 18(7):1112–1126. 23. Huang S, et al. (2011) The evolution and regulation of the mucosal immune complexity in the basal chordate amphioxus. J Immunol 186(4):2042–2055.

6 of 6 | www.pnas.org/cgi/doi/10.1073/pnas.1405414111

low NF-κB activity reduces the expression of ALP1, new homeostasis will be achieved at the end of the antibacterial response. In conclusion, the discovery of the novel ApeC domain (Pfam accession no. PF16977) and the functional information of two amphioxus ALPs define ALPs as a previously unidentified type of PRP. Furthermore, we demonstrate that amphioxus ALPs can serve as extracellular effector PRPs and intracellular sensor PRPs, which provides new insight into the importance and diversified function of ALPs in different animal phyla or subphyla. Materials and Methods Twenty amphioxi were maintained in 100 mL filtered seawater, and S. aureus and V. anguillarum were added to the seawater to a final concentration of ∼1 × 107 cfu/mL with or without 0.5 μg/mL anti-bjALP1 antibody. Dead amphioxi were counted at the indicated time points after bacterial infection. Purified IgGs from preimmunized mouse serum were used as the control. All primers used in this study are shown in Table S1. ACKNOWLEDGMENTS. This work was supported by project 2013CB917800 of the Major National Scientific Research Project, project 2013CB835300 of the National Basic Research Program (973), projects 2008AA092601 and 2012AA092201 of the State High-Tech Development Project (863), key projects 91231206 and 30730089 and projects 31171193 and 30901305 of National Natural Science Foundation, and projects from the Commission of Science and Technology of Guangdong Province and Guangzhou City and from Sun Yat-sen University Science Foundation. The data analysis was also supported by the Guangdong Province Key Laboratory of Computational Science and the Guangdong Province Computational Science Innovative Research Team.

24. Huang H, et al. (2011) Functional characterization of a ficolin-mediated complement pathway in amphioxus. J Biol Chem 286(42):36739–36748. 25. Yu Y, et al. (2008) A novel C1q family member of amphioxus was revealed to have a partial function of vertebrate C1q molecule. J Immunol 181(10):7024–7032. 26. Yu Y, et al. (2007) A short-form C-type lectin from amphioxus acts as a direct microbial killing protein via interaction with peptidoglycan and glucan. J Immunol 179(12): 8425–8434. 27. Yuan S, et al. (2013) The archaic roles of the amphioxus NF-κB/IκB complex in innate immune responses. J Immunol 191(3):1220–1230. 28. Yuan S, et al. (2009) Genomic and functional uniqueness of the TNF receptorassociated factor gene family in amphioxus, the basal chordate. J Immunol 183(7): 4560–4568. 29. Yuan S, et al. (2009) An amphioxus TLR with dynamic embryonic expression pattern responses to pathogens and activates NF-kappaB pathway via MyD88. Mol Immunol 46(11-12):2348–2356. 30. Haag ES, Raff RA (1998) Isolation and characterization of three mRNAs enriched in embryos of the direct-developing sea urchin Heliocidaris erythrogramma: Evolution of larval ectoderm. Dev Genes Evol 208(4):188–204. 31. Haag ES, Sly BJ, Andrews ME, Raff RA (1999) Apextrin, a novel extracellular protein associated with larval ectoderm evolution in Heliocidaris erythrogramma. Dev Biol 211(1):77–87. 32. Schleifer KH, Kandler O (1972) Peptidoglycan types of bacterial cell walls and their taxonomic implications. Bacteriol Rev 36(4):407–477. 33. Girardin SE, et al. (2003) Nod2 is a general sensor of peptidoglycan through muramyl dipeptide (MDP) detection. J Biol Chem 278(11):8869–8872. 34. Yang P, et al. (2014) Origin of the phagocytic respiratory burst and its role in gut epithelial phagocytosis in a basal chordate. Free Radic Biol Med 70:54–67. 35. Miller DJ, et al. (2007) The innate immune repertoire in cnidaria—ancestral complexity and stochastic gene loss. Genome Biol 8(4):R59. 36. Hayward DC, et al. (2011) Differential gene expression at coral settlement and metamorphosis—a subtractive hybridization study. PLoS ONE 6(10):e26411. 37. Grasso LC, et al. (2008) Microarray analysis identifies candidate genes for key roles in coral development. BMC Genomics 9:540. 38. Meyer E, Aglyamova GV, Matz MV (2011) Profiling gene expression responses of coral larvae (Acropora millepora) to elevated temperature and settlement inducers using a novel RNA-Seq procedure. Mol Ecol 20(17):3599–3616. 39. David E, Tanguy A, Pichavant K, Moraga D (2005) Response of the Pacific oyster Crassostrea gigas to hypoxia exposure under experimental conditions. FEBS J 272(21): 5635–5652. 40. Estévez-Calvar N, Romero A, Figueras A, Novoa B (2011) Involvement of pore-forming molecules in immune defense and development of the Mediterranean mussel (Mytilus galloprovincialis). Dev Comp Immunol 35(10):1017–1031. 41. Dheilly NM, Haynes PA, Bove U, Nair SV, Raftos DA (2010) Comparative proteomic analysis of a sea urchin (Heliocidaris erythrogramma) antibacterial response revealed the involvement of apextrin and calreticulin. J Invertebr Pathol 106(2):223–9. 42. Huang G, et al. (2007) Profile of acute immune response in Chinese amphioxus upon Staphylococcus aureus and Vibrio parahaemolyticus infection. Dev Comp Immunol 31(10):1013–1023.

Huang et al.

Supporting Information Huang et al. 10.1073/pnas.1405414111 SI Materials and Methods Cultivation and Immune Stimulation of Amphioxus and Preparation of Tissue Samples. Adults of Chinese amphioxus (Branchiostoma

japonicum) were obtained from Qingdao, China then maintained in large tanks with aeration and supplied with fresh seawater daily. Tissues were separated under an optical microscope (Nikon). For immune stimulation, live bacteria in PBS suspension (1 × 108 cfu/mL) were injected into the coelom of amphioxus (15 μL per animal). PBS was injected as the negative control. Hepatic ceca and intestine from five animals per treatment were collected at 2, 4, 8, 12, 24, and 48 h after injection and were frozen as experimental samples using liquid nitrogen. Total RNA was extracted using TRIzol (Invitrogen), and the sample was subjected to DNase I treatment (Promega) according to the manufacturer’s protocols. Subsequently, reverse transcription was performed using the PrimeScript RT reagent kit (Takara). For oral infection, amphioxus was maintained in filtered seawater and Staphylococcus aureus or Escherichia coli was added to the seawater to a final concentration of ∼1 × 107 cfu/mL. Isolation of the Full-Length cDNA of Branchiostoma japonicum Apextrin-Like Proteins. Expressed sequence tags with sequence

similarity to Apextrin were identified from a suppression subtractive hybridization library constructed by our laboratory (1). To obtain complete cDNA sequences, 3′ and 5′ RACE was performed using the GeneRacer Kit (Invitrogen) with genespecific primers according to the manufacturer’s instructions. The amplified fragments were cloned into the pGEX-Teasy vector (Promega) and were sequenced using an ABI PRISM 3730 DNA analyzer. Two full-length cDNA sequences consisting of 504 and 462 aa residues were obtained and were designated Branchiostoma japonicum apextrin-like proteins 1 and 2 (bjALP1 and bjALP2), respectively. Quantitative Real-Time PCR Analysis of the Gene Expression Patterns for bjALPs. Quantitative real-time PCR was performed on a

LightCycler 480 real-time PCR system (Roche) using the SYBR Premix Ex Taq II kit (Takara) in a 10-μL reaction system. A total of 45 PCR cycles were run, and the melting curve was recorded. The data were quantified using the 2−ΔΔCt method based on the Ct values of bjALPs and bjβ-actin. All of the samples were analyzed in three replicates, and the results are expressed as the mean ± SD. Section in Situ Hybridization. A 514-bp fragment of bjALP1 and

a 532-bp fragment of bjALP2 were cloned into pGEM-Teasy (Promega). Digoxigenin-labeled sense and antisense probes for bjALP1 and bjALP2 were synthesized using a DIG RNA Labeling Kit (Roche). The animals were dehydrated with gradated ethanol, embedded in paraffin, and sliced into 8-μm transverse histological sections. After drying at 42 °C overnight the sections were dewaxed, rehydrated, washed, and digested by proteinase K. The sense or antisense probes were added to the sections and hybridized overnight at 42 °C. After high-stringency washing the sections were immunodetected using an NBT/5-bromo-4-chloro3- indolyl phosphate stock solution (Roche). Sequence Retrieval, Alignment, and Domain Prediction. The protein sequences of the Branchiostoma floridae and Heliocidaris erythrogramma apextrins were used to perform BLASTP searches against all of the available databases at the National Center for Biotechnology Information (NCBI) and the Joint Genome Institute Huang et al. www.pnas.org/cgi/content/short/1405414111

(JGI, http://genome.jgi-psf.org/Brafl1/Brafl1.home.html). Domain structure predictions were performed on the main page of the SMART website (http://smart.embl-heidelberg.de). For sequence analysis, the ALP protein sequences were first aligned using ClustalX 1.83 (2) and were manually corrected using GeneDoc (3). A minimum-evolution (ME) tree was built by using Mega v4.1, with 1,000 bootstrap tests and handling gaps by pairwise deletion. Preparation of Recombinant Proteins. PCR fragments encoding amino acids 20–504, 20–297, and 298–504 of bjALP1 and 21–462, 21–257, and 258–462 of bjALP2 were inserted into plasmid pET-32a (+) (Novagen). The recombinant expression plasmids were introduced into E. coli BL21 (DE3) for protein expression. All six fusion proteins in inclusion bodies were denatured in 8 M urea and were renatured in three dialysis steps as previously described (4). Each dialysis step was performed at 4 °C for at least 12 h followed by desalting with Tris-buffered saline (TBS) buffer (50 mM Tris·Cl and 150 mM NaCl, pH 8.0) using a Sephadex G-25 column. The concentration of the purified protein was determined using a Bio-Rad Protein Assay dye reagent and BSA as a standard. Microbial Binding Assay. A total of 5 μg of purified recombinant proteins were incubated with microbes in 1 mL of TBS buffer at 4 °C overnight with end-over-end mixing. The mixtures were centrifuged at 12,000 × g for 1 min at 4 °C, and the pellets were washed five times with 1 mL of TBS buffer. To denature the binding proteins, the washed pellets were suspended in 100 μL of TBS buffer [50 mM Tris·Cl (pH 7.5) and 150 mM NaCl] and 20 μL of 6× loading buffer and boiled at 100 °C for 10 min. Western blotting was performed with an anti-His mouse monoclonal antibody (Sigma) to validate the binding proteins. Microbial Aggregation Assay. Fluorescein isothiocyanate (FITC) (isomer I; Amersco) was used to label bacteria. Suspensions of bacteria were mixed with 50 μL of FITC solution (10 mg/mL in DMSO) and gently agitated at room temperature in the dark for 3 h. After washing five times with TBS, FITC-labeled bacteria were mixed with 10 μg ALP proteins in 1 mL of TBS and were incubated at room temperature in the dark for 2 h. Bacterial aggregation was examined immediately using fluorescence microscopy (Carl Zeiss). Analysis of ALPs Binding with Microbial Cell-Wall Components.

ELISA was used to analyze the the binding of ALP proteins with soluble microbial cell wall components as previously described (5). In brief, LPS, lipoteichoic acid, peptidoglycans (PGNs), mannan, zymosan (Sigma), and muramyl dipeptide (MDP) (Invivogen) in carbonate buffer (0.1 M NaHCO3 and 2.5 mM Na2CO3, pH 9.6) were used to coat 96-well microtiter plates (TPP) for 3 h at 37 °C (20 μg each well, PGN and zymosan are ultrasonically solubilized). After washing with TBS three times, the wells were blocked with TBST (0.05% Tween 20 in TBS) containing 10% (wt/vol) BSA overnight at 4 °C. Various concentrations of ALP proteins were then added, and the mixtures were incubated for 2 h at 37 °C. Binding proteins were detected using a mouse anti-His mAb followed by incubation with HRP-labeled anti-mouse IgG (Bio-Rad). The absorbance was read at 450 nm immediately after being developed with the 3,3,5,5′-tetramethyl-benzidine substrate (Sigma). The assay was repeated three times. The insoluble PGN-binding activities of ALP proteins were detected by pull-down assays as previously described (5). In brief, insoluble PGNs were incubated at 4 °C 1 of 8

for 1 h with 5 μg of ALP proteins in TBS. The samples were centrifuged at 15,000 × g for 10 min and pellets were washed with the TBS, centrifuged, and suspended in 100 μL of TBS buffer and 20 μL of 6× loading buffer and boiled at 100 °C for 10 min. Western blotting was performed with an anti-His mouse monoclonal antibody (Sigma) to validate the binding proteins.

with anti-ALP1 or anti-ALP2 antibody. Washes were followed by 1 h of incubation at room temperature in goat anti-mouse secondary antibody labeled with 20-nm gold particles. For excrement immungold assay, 1 × 109 S. aureus in 100 mL filtered seawater were used to orally infect amphioxus for 4 h; the excrement was collected and prepared for EM as described above.

Antimicrobial Activity Assays. Thirty milliliters of 1.0% warm nutrient agar mixed with S. aureus were poured into a 90-mm plate. The pores are 0.5 cm in diameter perforated with perforex. Then, the targeted protein or antibiotic in 20 μL of TBS was added to the pores at a final concentration of 1 μg/μL. The plates were incubated at 37 °C for 16 h. A transparent ring around the pores indicated antibacterial activity. The growth curves of S. aureus were tested as follows. Two single colonies were picked up and transferred into 1 mL of Luria–Bertani (LB) broth. A volume of 50 μL of cell suspension was mixed with 200 μg targeted proteins and added to 1 mL LB broth. Each sample was incubated at 200 rpm at 37 °C and OD600 was measured every 1 h.

H&E Staining. Animals were fixed with 4% (wt/vol) paraformaldehyde in PBS overnight at 4 °C. Samples were embedded in paraffin following dehydration in ethanol. Tissues were cut into 8-μm sections and stored for staining. For H&E (Sigma) staining, sections were freed of paraffin and stained.

Preparation of Antibodies. Anti-bjALP1 and Anti-bjALP2 mouse monoclonal antibodies were prepared against the epitopes 412 GEVPDGNYDR421 and 375DGIYNRNTEI384, respectively, by Abmart Company using the SEAL technology. Expression Plasmids. For the study of subcellular localization, fulllength bjALP2 was inserted into the pEGFP-N1 vector (Clontech). For the reporter assays and coimmunoprecipitation (Co-IP) test, full-length bjALP2 was fused with a Flag tag on the C terminus and cloned into pcDNA3.1 (+) (Clontech). PCR fragments encoding amino acids 1–257 and 258–462 of bjALP2 were fused with a Flag tag on the C terminus, cloned into pcDNA3.1 (+), and designated as bjALP2ΔC and bjALP2ΔN, respectively. The bjMyD88 sequences were cloned into pCMV-HA (Clontech), bjTRAF6 (TNF receptor-associated factor 6) sequences were cloned into pCMVMyc and pCMV-Flag (Clontech), and bjUbiquitin was cloned into pCMV-HA (Clontech). Immunofluorescence Imaging. HeLa cells seeded onto coverslips

(10 × 10 mm) in a 24-well plate were transfected with 400 ng of the indicated expression plasmids. Twenty-four hours later, the cells were fixed for 15 min in a 4% (wt/vol) formaldehyde solution (in PBS), washed three times in PBST (0.05% Tween-20 in PBS), and blocked with 5% (wt/vol) BSA in PBST at room temperature for 1 h. The cells were then incubated with 1 μg/mL of a mAb for 1 h, washed three times in PBST, and incubated with the secondary antibody at 4 °C overnight. After triple washing in PBS, cells were labeled with 0.2 μg/mL DAPI in PBS for 5 min and were washed three times in PBS. The cells were then mounted in MOWIOL R4-88 Reagent (Calbiochem) and photographed with a Carl Zeiss Axiovision 4 microscope. Immunogold Assay. Hepatic ceca were fixed for EM in a mixture of

Survival Analysis. Twenty amphioxi were maintained in 100 mL filtered seawater and S. aureus and Vibrio anguillarum was added to the seawater to a final concentration of ∼1 × 107 cfu/mL with or without 0.5 μg/mL bjALP1 antibody. Dead amphioxi were counted at indicated time points after bacterial oral infection. Purified IgGs from preimmunized mouse serum were used as control. Transfection and Reporter Assays. HEK293T cells were plated in 48-well plates (at 5 × 104 cells per well) and transfected 24 h later with a mixture of DNA as previously descripted (6). The mixed DNA contained the indicated amounts of expression vector, 5 ng of pRL-TK Renilla luciferase reporter plasmid (Promega), 50 ng of the pNFκB-Luc (Clontech), and the complementary empty vector. Luciferase activities were measured by the DualLuciferase Reporter Assay System (Promega). Luciferase activities were normalized to Renilla luciferase activities and are expressed as the fold stimulation relative to that measured in cells transfected with empty vector. To determine the effect of MDP on NF-κB activities, variant amounts of MDP were cotransfected with indicated plasmids to HEK293T cells. Each experiment was performed at least in triplicate and was repeated at least twice in all cases. Co-IP. HEK293T cells in six-well dishes were transfected with 6 μg

DNA plasmids (3 μg/each expression vector). At 36 h posttransfection, whole-cell extracts were prepared in IP lysis buffer [50 mM Tris (pH 7.4), 150 mM NaCl, 1% Nonidet P-40, 0.5% deoxycholic acid sodium salt, and a protease inhibitor mixture from Roche] and incubated with primary antibodies (1 μg antiFlag; Sigma) at 4 °C for 4 h then incubated with Protein G Sepharose (Roche) at 4 °C overnight. The next day, the mixture was washed three times with lysis buffer. The analysis was conducted using SDS/PAGE followed by Western blotting with anti-Myc (1:7,000) and anti-Flag (1:1,000) mouse monoclonal antibody (Sigma). To assess the impact of MDP on the interaction of bjALP2 and TRAF6, an indicated amount of MDP was added to the cell lysate of HEK293T cells that transfected with bjALP2 and TRAF6 plasmids.

glutaraldehyde and paraformaldehyde in seawater overnight and embedded in Epon resin. Sections (60 nm) were collected on 200-mesh nickel grids. Samples were incubated overnight at 4 °C

Statistical Analysis. Quantitative data are presented as the mean ± SD and compared statistically by two-tailed Student t test. A P value of < 0.05 was considered statistically significant.

1. Huang G, et al. (2007) Profile of acute immune response in Chinese amphioxus upon Staphylococcus aureus and Vibrio parahaemolyticus infection. Dev Comp Immunol 31(10):1013–1023. 2. Takeda K, Kaisho T, Akira S (2003) Toll-like receptors. Annu Rev Immunol 21:335–376. 3. Gerold G, Zychlinsky A, de Diego JL (2007) What is the role of Toll-like receptors in bacterial infections? Semin Immunol 19(1):41–47. 4. Yu XQ, Tracy ME, Ling E, Scholz FR, Trenczek T (2005) A novel C-type immulectin-3 from Manduca sexta is translocated from hemolymph into the cytoplasm of hemocytes. Insect Biochem Mol Biol 35(4):285–295.

5. Yu Y, et al. (2007) A short-form C-type lectin from amphioxus acts as a direct microbial killing protein via interaction with peptidoglycan and glucan. J Immunol 179(12): 8425–8434. 6. Yuan S, et al. (2009) An amphioxus TLR with dynamic embryonic expression pattern responses to pathogens and activates NF-kappaB pathway via MyD88. Mol Immunol 46(11–12):2348–2356.

Huang et al. www.pnas.org/cgi/content/short/1405414111

2 of 8

Fig. S1. Amino acid sequences of bjALP1 (A) and bjALP2 (B). The putative signal peptide is in italics, the boxed sequence indicates the predicted TRAF6 binding motif, and the underlined sequence indicates the APT-C region. The asterisk indicates the stop codon. The GenBank accession number of the corresponding nucleotide sequence is KM017614.

Fig. S2. Sequence alignment of the apextrin C-terminal (ApeC) domain. The eight conserved cysteine residues in character are underlined. am, Acropora millepora; bf, Branchiostoma floridae; he, Heliocidaris erythrogramma; hm, Hydra magnipapillata; is, Ixodes scapularis; mg, Mytilus galloprovincialis; nv, Nematostella vectensis; pl, Pomatoceros lamarckii; sk, Saccoglossus kowalevskii; sp, Strongylocentrotus purpuratus. The sequence used in the alignment and phylogenetic analysis is indicated as follows: bf002594828 (XP_002594828); bf002604578 (XP_002604578); mg10750 (AEK10750); hm00217052 (XP_00217052); hm002158419 (XP_002158419); hm002165606 (XP_002165606); nv001635757 (XP_001635757); nv001627111 (XP_001627111); am091848 (EF091848); he28440 (AAC28440); sp001179344 (XP_001179344); sp789518 (XP_789518); sp786957 (XP_786957); sk002735920 (XP_002735920); sk002735923 (XP_002735923); pl11402 (ADB11402); and is002400395 (XP_002400395).


3 of 8

Fig. S3. Domain architecture and phylogeny of apextrin-like proteins. (A) ApeC domain-containing ALP proteins in amphioxus and ApeC domain-containing ALPs selected from other animal phyla. (B) The ME tree of the ApeC domain protein sequences of various ALPs from different phyla or subphyla.


4 of 8

Fig. S4. Antibacterial activity test of ALP proteins against S. aureus. (A) Box plot showing the diameters of green puncta in bacteria aggregation assays. Box plot explanation: upper horizontal line of box, 75th percentile; lower horizontal line of box, 25th percentile; horizontal bar within box, median; upper horizontal bar outside box, 90th percentile; lower horizontal bar outside box, 10th percentile. Circles represent outliers. *P < 0.05 versus TRX control. (B) A photograph of a plate of a representative agar plate assay with 1 μg/μL ampicillin (Amp), 0.2 μg/μL TRX-bjCTL1, 1 μg/μL TRX, 1 μg/μL TRX-bjALP1, or 1 μg/μL TRX-bjALP2. Individual recombinant proteins or ampicillin were applied to pores within the transparent agar plates with microbes. Then, the plates were incubated at 37 °C for 16 h. The conversion of the otherwise transparent agar into an opaque yellowish appearance provided an indicator of the active growth of the bacteria. (C) The results of the growth curve assay. *P < 0.05.

Fig. S5. ELISA analysis of the interaction between the truncated bjApextrin1 mutants and PGN and MDP. One representative experiment of six is shown. Background absorbance without protein was subtracted.

Fig. S6. Western blot analysis of bjALPs in different tissues. Gill, hepatic cecum, and intestine were collected from 10 amphioxi and suspended in 100 μL of PBS buffer and 20 μL of 6× SDS/PAGE loading buffer and the tissues were lysed by boiling at 100 °C for 10 min. Western blotting was performed with anti-ALP1 or anti-ALP2 mouse monoclonal antibodies.


5 of 8

Fig. S7. Immunogold EM analysis and Western blot of biALPs. (A and B) Immunogold EM analysis of bjALP2 in the hepatic cecum (A) and intestine (B) of amphioxus. N, nucleus. (C) Western blot of bjALPs in bacteria excrement. 1 × 109 S. aureus in 100 mL filtered seawater were used to orally infect amphioxus for 4 h. The excrement was collected and suspended in 100 μL of PBS buffer and 20 μL of 6× SDS/PAGE loading buffer and then boiled at 100 °C for 10 min. Western blotting was performed with anti-ALP1 or anti-ALP2 mouse monoclonal antibodies. (D) Western blot analysis of bjALP2-GFP fusion protein overexpressed in HEK293T cell. Anti-GFP antibody was used. FL, full-length ALP2; SP, signal peptide of ALP2. (E) Western blot analysis of mature bjALP2 peptides fused with different signal peptides overexpressed in HEK293T cells. Anti-bjALP2 antibody was used. CL, cell lysates; M, culture media.


6 of 8

Fig. S8. ELISA analysis showed that anti-bjALP1 antibody can block the interaction between recombinant ALP proteins and MDP. Background absorbance without protein was subtracted.

Fig. S9. Subcellular location of bjALP2 and its mutant. (A) Subcellular location of EGFP-fused bjALP2. (B) Subcellular location of EGFP-fused bjALP2ΔT6. (C) BjALP2ΔT6 is not colocalized with amphioxus TRAF6 in HeLa cells. HeLa cells were cotransfected with EGFP-fused bjALP2ΔT6 and Myc-tagged pCMV-bjTRAF6 and were stained with an anti-Myc antibody and an Alexa Fluor 532 secondary antibody.


7 of 8

Table S1. Primer list Gene

Forward primer (5′–3′)

bjALP1 bjALP2 β-actin bjALP1 bjALP1 bjALP2 bjALP2

3′RACE 5′RACE 3′RACE 5′RACE

bjALP1 bjALP2 bjALP1 bjALP1ΔC bjALP1ΔN bjALP2 bjALP2 ΔC bjALP2 ΔN bjALP2 bjALP2 ΔC bjALP2 ΔN

Reverse primer (5′–3′)

Quantitative real-time PCR primers CATGGCGGTTTGGGGGTGTA ACAGCTCTAGCGCCTCGGGAAC TGCCAAACCCAGTCCGCTCC GGGATGACCACTGCGAGGAT TCTGGCATCATACCTTCTACAA TCTGTGTCATCTTTTCCCTGTT RACE primers 5′–3′ TTCCGTTGGGACGATGAAGATG CGTCATCTTCATCGTCCCAACGGAA TTCCGGTGGGACGATGAAGAC CGTCTTCATCGTCCCACCGGAA Cloning primers for in situ hybridization gcggtttgggggtgtact ccaggttcaggttgttgtcc ggaacgtcagaacattcacca cctgctccatcgtcgtcc Cloning primers for pET-32a (+) CGCGGATCCgcaccgaccgaagtcgtaca CCGCTCGAGagagtagtagcagtagtg CCGCTCGAGctgtttggtaaaccgttggagg CGCGGATCCcaaggtgaagttgcggctgt CGCGGATCCgctccgaccgagctgcaag CCGCTCGAGcgaatagtaacagtagtgga CCGCTCGAGttctttagtgtagggcagaagg CGCGGATCCagcctgcaggaggtggcag Cloning primers for pcDNA3.1 (+) and/or pEGFP-N1 CGCGGATCCatgaaattcgcgctgctc CCGCTCGAGcgaatagtaacagtagtgga CCGCTCGAGttctttagtgtagggcagaagg CGCGGATCCagcctgcaggaggtggcag


Amplicon size, bp 131 107 113

532 514 1,473 852 639 1,326 729 633 1,386 789 633

8 of 8