Spatial and temporal coordination of expression of ... - CiteSeerX

Genes and Immunity (2005), 1–18 & 2005 Nature Publishing Group All rights reserved 1466-4879/05 $30.00 www.nature.com/gene

FULL PAPER

Spatial and temporal coordination of expression of immune response genes during Pseudomonas infection of horseshoe crab, Carcinoscorpius rotundicauda JL Ding1, KC Tan1,3, S Thangamani1, N Kusuma1, WK Seow1, THH Bui1, J Wang1,4 and B Ho2 Department of Biological Sciences, National University of Singapore, Singapore 117543, Singapore; 2Department of Microbiology, National University of Singapore, Singapore 117543, Singapore 1

Knowledge on how genes are turned on/off during infection and immunity is lacking. Here, we report the coregulation of diverse clusters of functionally related immune response genes in a horseshoe crab, Carcinoscorpius rotundicauda. Expressed sequence tag (EST) clusters for frontline immune defense, cell signalling, apoptosis and stress response genes were expressed or repressed spatio-temporally during the acute phase of Pseudomonas infection. An infection time course monitored by virtual Northern evaluation indicates upregulation of genes in blood cells (amebocytes) at 3-h postinfection, whereas most of the hepatopancreas genes remained downregulated over 72 h of infection. Thus, the two tissues orchestrate a coordinated and timely response to infection. The hepatopancreas probably immunomodulates the expression of other genes and serves as a reservoir for later response, if/when chronic infection ensues. On the other hand, being the first to encounter pathogens, we reasoned that amebocytes would respond acutely to infection. Besides acute transactivation of the immune genes, the amebocytes maintained morphological integrity, indicating their ability to synthesise and store/secrete the immune proteins and effectors to sustain the frontline innate immune defense, while simultaneously elicit complement-mediated phagocytosis of the invading pathogen. Our results show that the immune response against Pseudomonas infection is spatially and temporally coordinated. Genes and Immunity advance online publication, 7 July 2005; doi:10.1038/sj.gene.6364240 Keywords: immune-response gene clusters; Pseudomonas infection; expressed sequence tags (ESTs); transcript profiling; spatial and temporal immune gene expression

Introduction During an infection, the innate immune response in multicellular host organisms is initiated by pathogenrecognition receptors (PRRs), which are germline-encoded proteins capable of recognising conserved pathogen-associated molecular patterns (PAMPs) unique to pathogens.1,2 PRRs are strategically expressed in those cells first exposed to pathogens. Activation of these PRRs leads to the expression of immunomodulated genes that are vital in protecting the host against pathogens. Studies on the genetics of Manduca, Anopheles and Drosophila have generated a great deal of information on innate immune defense, providing some insights on the expression of antimicrobial peptides3,4 and the functions of endopeptidases, serpins and other immune mole-

Correspondence: Professor JL Ding, Department of Biological Sciences, National University of Singapore, 14 Science Drive 4, Singapore 117543, Singapore. E-mail: [email protected], http://www.dbs.nus.edu.sg/Staff/ding.html 3 Current address: The Australian Centre For Necrotrophic Fungal Pathogens, Murdoch University, 6150, Australia. 4 Current address: Institute of Molecular Cell Biology, Singapore. Received 4 April 2005; revised 1 June 2005; accepted 2 June 2005

cules.5 Furthermore, data mining of the Drosophila and Anopheles genomes and DNA microarray analyses6–8 have revealed that the Toll and Imd signalling pathways are responsive to microbial infection and induction of the innate immune responses. The horseshoe crab represents an ancient family of arthropods with 4500 million years of evolutionary history. There are four extant species of horseshoe crabs: Limulus polyphemus (in the Eastern seaboard of USA); Tachypleus tridentatus (in China and Japan) and Tachypleus gigas and Carcinoscorpius rotundicauda in South Asia. The C. rotundicauda is the mud-dwelling species, whose habitat contains very high counts of Gramnegative bacteria. Its ability to thrive under such highly infective conditions attests to its possession of a superior frontline innate immune system.9–16 Like other invertebrates, the horseshoe crab lacks adaptive immunity and relies solely on a very potent innate immune system to combat invading microbes. In the past two decades, the components of the innate immune system of the horseshoe crab have been extensively investigated at the level of individual proteins. This has led to the elucidation of many unique frontline defense molecules such as clotting factors and serine proteases,9–12 lectins,14,17 protease inhibitors,18 antimicrobial peptides13,15

Transcript profiling of immune responsive genes JL Ding et al

2

and other humoral factors.19. Most of these molecules are identified from the amebocytes, the major blood cell type. In contrast, only a few immune-related molecules have been identified from the hepatopancreas, which is the immune-responsive functional equivalent to the insect fat bodies and the vertebrate liver. Despite the discovery of these unique molecules in the horseshoe crab, knowledge on their spatial and temporal gene expression profiles and the presence of other innate immune-related molecules responsive to microbial infection remains elusive. Thus, mapping the ensemble of functionally related immune genes, which may be up-/ downregulated during bacterial infection represents the first step towards elucidating the pathways contributing to innate immunity. The horseshoe crab offers significant advantages over its smaller arthropod counterparts since it possesses large amounts of blood and sizeable tissues, which makes the system readily amenable to physiological and molecular manipulations. Furthermore, as a ‘living fossil’, it is expected to harbour an immensely powerful repertoire of innate immune molecules, which act in frontline defense. In view of these advantages, we sought to examine the functional display of expressed sequence tags (ESTs) in the horseshoe crab, in response to infection by Pseudomonas aeruginosa. As a ubiquitous and potent Gram-negative pathogen that has acquired multiple antibiotic resistance, the P. aeruginosa is a major cause of nosocomial infection and is the epitome of opportunistic human pathogens. Therefore, its elimination remains a critical challenge to the medical industry.15 We have recently shown that the horseshoe crab effectively clears a systemic infection by P. aeruginosa (106 cfu/ml) within 6 h, whereas this infection dosage would have been lethal to mice.14 The development of high-throughput methods of gene identification by EST analysis has become a commonly used approach to identify genes involved in specific biological functions. This is especially so in organisms where genome data is unavailable or limited,20 and has accelerated the pace at which new immune functions can be discovered. A growing number of EST databases from Bombyx mori,21 Galleria mellonella22 and other organisms testify to the importance of this technique. Here we report a suppression subtractive cDNA hybridisation approach, to isolate and identify differentially expressed genes from the horseshoe crab, C. rotundicauda, in response to P. aeruginosa infection. In our study, four subtractive cDNA libraries were constructed from the two major immune responsive tissues: hepatopancreas and amebocytes, which are expected to reflect genes that are expressed or repressed during infection. The kinetic profile of transcription of selected genes responsive to infection was studied. By comparing the sequences of cDNA clones from subtractive libraries with known gene sequences deposited in the GenBank, this study has revealed the diversity of genes (frontline immunity, cell signalling, apoptosis and stress responses), which were invoked by infection. We also show that as the major frontline defense tissue which displays acute increases in gene transcription, the blood cells, amebocytes, were intact and functional during Pseudomonas infection, and that phagocytosis, which was recently demonstrated in this species,16 was essentially occurring to clear the microbial pathogens and apoptotic cells.

Genes and Immunity

Results and discussion To understand differential immune gene expression in amebocytes and hepatopancreas of the horseshoe crab during P. aeruginosa infection, we constructed suppression subtractive cDNA libraries from the RNAs pooled from 3 and 6 h postinfection (hpi). These two time points were chosen based on studies on Drosophila, which showed the occurrence of acute phase expression of antimicrobial genes at 3–6 h postmicrobial infection.4 In our study, we performed suppression subtractive hybridisation to create EST libraries on amebocytes and hepatopancreas samples of infected and saline-treated horseshoe crabs. These samples were used as both tester and driver. To define, the tester refers to the cDNA population containing sequences of interest whereas the driver refers to the cDNA population used to remove genes present in both conditions. Four cDNA libraries were established: amebocyte forward, amebocyte reverse, hepatopancreas forward and hepatopancreas reverse. The ‘forward’ libraries were constructed using cDNAs from tissues of Pseudomonas-infected horseshoe crabs as the tester and cDNAs from tissues of mockinfected horseshoe crabs as the driver. The ‘reverse’ libraries were constructed vice versa. Additionally, due to the rapid change of immune-related genes over the time course of infection, we carried out virtual Northern analyses of gene expression on 3 þ 6 hpi pooled cDNAs as well as cDNAs from seven individual time intervals over 72 h of Pseudomonas-infection. Representative ESTs from each library were used as probes for the Northern analyses. Characterisation of the ESTs The EST clones from the four subtractive cDNA libraries, representing gene expression patterns from the amebocytes (Ame) and hepatopancreas (Hp) were sequenced. The ESTs from forward, F (upregulated genes) and reverse, R (repressed genes) libraries of amebocytes and hepatopancreas are referred to as AmeF/AmeR and HpF/HpR, respectively. Detailed information on each library is summarised in Table 1. From the four libraries, a total of 776 randomly selected clones were sequenced, resulting in the characterisation of 447 ESTs. Of these ESTs, 268 (60%) showed significant BLASTx matches (Evalue p103) to known identified genes. In the forward libraries, 549 clones were sequenced, yielding 333 ESTs (60.7%) with E-values p103. Of these 333 ESTs, 94 (28.2%) were from the amebocyte and 104 (31.2%) were from the hepatopancreas forward libraries that matched previously known sequences in the database. Of 227 clones sequenced from the reverse libraries, 114 ESTs (50%) with E-values p103 were identified by sequence homology. Of these, 42 ESTs (36.8%) from the amebocyte and 28 ESTs (24.6%) from the hepatopancreas reverse libraries showed significant similarities to known sequences in the database. Based on the general Expressed Gene Anatomy Database (EGAD),23 the ESTs were classified into nine broad functional categories (Figure 1). The ESTs putatively associated with the immune-related gene clusters (frontline immunity, cell signalling, apoptosis and stress responses), together with their respective NCBI dbEST database accession numbers are shown in Tables 2a, b and 3a, b. The remaining functional clusters of ESTs from

Transcript profiling of immune responsive genes JL Ding et al

Table 1 Functional groups of ESTs from P. aeruginosa-challenged horseshoe crab amebocytes and hepatopancreas forward and reverse

3

libraries Amebocyte

Total number of clones sequenced E-value p103 (%) Nonredundant clonesa Redundancy (%)b

Hepatopancreas

Total

Forward

Reverse

Forward

Reverse

272 64.7 (176/272) 94 46.5

131 46.6 (61/131) 42 31.1

277 56.6 (157/277) 104 33.7

96 55.2 (53/96) 28 47.1

776 57.6 (447/776) 268 40.07

% of nonredundant clones Putative functional groups of ESTs Frontline immunity Cell signalling Apoptosis Stress response Cell cycle and development Cell structure Gene/protein expression Metabolism Others

30 9 2 11 3 7 10 17 11

10 23 10 10 5 7 10 13 12

21 12 4 13 2 9 14 16 9

14 7 11 21 7 4 18 7 11

60c 32 13 33 9 20 33 41 27

a

Nonredundant clones refer to individual genes being represented by one EST sequence each (even if each gene is represented by 41 EST). Number of redundant clones relative to total EST sequenced (for E-value p103). c Actual number of ESTs (total from all four libraries). b

Figure 1 Functional classification of C. rotundicauda ESTs in the acute phase of infection by Pseudomonas aeruginosa. Based on the general Expressed Gene Anatomy Database (EGAD),23 the ESTs are categorised into putative functional groups: frontline immunity, cell signaling, apoptosis, stress response, cell cycle and development, cell structure, gene/protein expression, metabolism and others. Genes and Immunity

4

Genes and Immunity

AmeF105 AmeF81 AmeF49

HpF286 HpF249 HpF219 HpF22 HpF282 HpF18

Stress response CK086649 CK086654 CK086665

(b) Frontline immunity CK086930 CK086940 CK086988 CK086942 CK086989 CK086945

AAA37191 AAP44511

Acrogranin (Mus musculus) Progranulin-b (D. rerio)

Hormone/growth factors CK086966 CK086978

HpF40 HpF4

A31254

T18544 XP_345912 S45677 AAP33177 XP_111429 Q05826

Cell signalling Effectors/modulators and intracellular transducers CK086980 HpF125 Variant-surface-glycoprotein phospholipase C (T. brucei)

a2-macroglobulin (L. polyphemus) Hemolectin CG7002-PA, similar to (R. norvegicus) Kazal/proteinase inhibitor (P. leniusculus) Peritrophin 1 (M. configurata) Serine protease inhibitor EIC, similar to (M. musculus) Transcription factor NF-M (G. gallus)

Catalase (C. jejuni) Glutathione-S-transferase (B. micropus) Selenoprotein W (D. rerio)

Q59296 AAD15991 NP_840072

NP_150602 NP_150603 NP_150606 XP_135065.2

Cytochrome c oxidase subunit I (L. polyphemus) Cytochrome c oxidase subunit II (L. polyphemus) Cytochrome c oxidase subunit III (L. polyphemus) SAG (sensitive to apoptosis gene)/Ring finger protein 7 (M. musculus)

AmeR198 AmeF57 AmeR189 AmeR204

P07087 P80957 P03997 B27257 AAB34362 AAH43981 A55533 PC1315 JC5147 P14213 NP_058896 P81281

Apoptosis CK086806 CO038879 CK086807 CK086799

Antilipopolyssacharide factor (T. tridentatus) Big defensin (T. tridentatus) Coagulogen (C. rotundicauda) P03997 Coagulogen II precursor (T. tridentatus) Factor C (C. rotundicauda) Interleukin enhancer binding factor 2, similar to (X. laevis) Intracellular coagulation inhibitor type 2 (T. tridentatus) Large granule L2 chain (T. tridentatus) Tachycitin precursor (T. tridentatus) Tachyplesin I precursor (T. tridentatus) Tissue factor pathway inhibitor (R. norvegicus) Transglutaminase substrate 8.6 kDa (T. tridentatus)

Putative match accession

NP_113553

AmeF48 AmeF72 AmeF230 AmeF113 AmeF287 AmeF142 AmeF267 AmeF219 AmeF155 AmeF93 AmeR235 AmeF238

(a) Frontline immunity CK086627 CK086629 CK086630 CK086631 S77063 CK086634 CK086636 CK086637a CK086643 CK086644 CK086826 CK086646

Putative match (organism)

Cell signalling Effectors/modulators and intracellular transducers CK086817 AmeR50 Phosphoinositide-3-kinase adaptor protein M. musculus)

Clone

dbEST accession

1e-10 5e-10

2e-04

1e-107 4e-12 1e-09 2e-15 1e-16 5e-16

5e-10 5e-03 3e-20

1e-75 4e-34 4e-70 9e-31

9e-03

1e-51 2e-38 3e-82 2e-84 1e-109 8e-79 9e-61 2e-01 7e-45 2e-20 9e-10 1e-40

E-value

39%/66 34%/97

32%/103

87%/217 53%/52 36%/101 35%/139 44%/96 47%/99

96%/28 85%/20 54%/88

84%/173 80%/85 81%/132 94%/58

30%/103

92%/102 89%/79 99%/140 86%/160 99%/186 69%/198 90%/129 89%/19 96%/83 100%/45 43%/65 86%/81

Identity/ alignment length (aa)

5 2

2

5 3 6 2 2 2

2 2 2

— 2 — —

—

2 2 27 12 2 3 2 2 5 4 — 4

— —

—

— — — — — —

— — —

2 — 6 2

2

— — — — — — — — — — 2 —

Forward Reverse

Redundancy

Table 2 Subtractive ESTs from the (a) amebocytes and (b) hepatopancreas with multiple redundancies grouped according to their putative functions as outlined in the ‘Expressed Gene Anatomy Database’ protocol23

Transcript profiling of immune responsive genes JL Ding et al

— 2 — — — — 48%/99 59%/159 52%/177 73%/92 89%/193 41%/56 HpF51 HpR156 HpF11 HpF24 HpF102 HpF135 Stress response CK086953 CK087141 CK086954 CK086956 CK086957 CK086962

Copper chaperone for superoxide dismutase (R. norvegicus) Glutathione-dependent formaldehyde dehydrogenase (U. hardwickii) Glutathione-S-transferase m1 (M. musculus) Heat-shock protein 40 (M. musculus) Heat-shock protein 70 (M. sexta) Prostaglandin-D synthase (H. sapiens)

Q9JK72 P80467 P10649 AAH40747 AAF09496 NP_055300

1e-15 3e-52 1e-48 2e-37 2e-94 7e-06

2 — 2 4 5 2

— 7 4 4 — 2 — — — 2 61%/111 85%/173 80%/226 80%/140 48%/111 HpF292 HpR216 HpR215 HpR187 HpF55 Apoptosis CK086948 CK087172 CK087171 CK087155 CK086958

Amine oxidase (H. sapiens) Cytochrome c oxidase subunit I (L. polyphemus) Cytochrome c oxidase subunit II (L. polyphemus) Cytochrome c oxidase subunit III (L. polyphemus) Hexokinase A (D. Melanogaster)

JH0817 NP_150602 NP_150603 NP_150606 NP_727350

1e-33 4e-77 3e-90 1e-59 2e-20

Forward Reverse

Identity/ alignment length (aa) Clone dbEST accession

Table 2 Continued

Putative match (organism)

Putative match accession

E-value

Redundancy

Transcript profiling of immune responsive genes JL Ding et al

the amebocytes and hepatopancreas are catalogued in Supplementary Tables 1a,b and 2a,b. In the scope of this study, only four functional clusters of ESTs (1) frontline immunity, (2) cell signalling, (3) apoptosis and (4) stress response will be discussed. Generally, the forward libraries of both the amebocytes and hepatopancreas contained higher numbers of frontline immunity genes than their corresponding reverse libraries. In contrast, both the amebocyte and hepatopancreas reverse libraries contained more apoptosis-related genes than the forward libraries. The cell signalling and stress response genes were more highly represented in the amebocyte and hepatopancreas reverse libraries, respectively (Figure 1).

5

Frontline immunity In total, 60 immune response ESTs from the forward and reverse libraries showed significant matches to known genes in the database (Table 1). A total of 27 ESTs from the amebocyte forward library encode coagulogen (AmeF230). Such abundant representation strongly indicates that the coagulogen gene (a functional homologue of spaetzle in Drosophila) is highly expressed during infection. The role of coagulogen in host defense and hemostasis in the horseshoe crab is well-documented.24 This is strongly linked to our observation of the infection-induced upregulation in both the amebocytes and hepatopancreas of Factor C (AmeF287 and HpF301), a key regulator of lipopolysaccharide (LPS) mediated immune response, which is known to trigger the blood coagulation cascade culminating in the conversion of coagulogen to coagulin. The coagulin gel clot is stabilised by transglutaminase (HpF170 and AmeF98), which is secreted upon LPS-stimulation. Consequently, the insoluble gel clot traps the invading microbes. Transglutaminase catalyses the crosslinking reactions between coagulin and amebocyte cell surface antigens, proxins.25 Interestingly, apoptosis induces the expression of transglutaminase26 that promotes the formation of insoluble clots to minimise cytoplasmic leakages of apoptotic cells.27 Notably, the transglutaminase substrate in the amebocytes (AmeF238) showed early transcriptional activation (Figures 2 and 3) indicating that it probably underwent crosslinkage, being catalysed by transglutaminase. The multimeric forms of transglutaminase substrate have been postulated to form a physical barrier against the spread of the invading pathogen and to possess bactericidal activity.28 Hemocyanin (HMC) is an abundant oligomeric protein constituting B90% of the total plasma protein of the horseshoe crab. The HMC EST in the amebocyte (AmeF45) was transcribed within 3 hpi, suppressed between 6 and 12 hpi and then escalated over 24–72 h of infection (Figure 4a). This is consistent with our earlier observation14 that the horseshoe crab effectively clears the Pseudomonas infection after 6 h, hence, the subsequent recovery of HMC expression. As an abundant protein, it is conceivable that the prolonged and relatively high level of transcription of the HMC gene during 24–72 hpi would help to replenish and maintain the high level of the HMC protein. Besides being an important respiratory protein, the HMC plays a crucial role in frontline innate immune response during which it is proteolytically processed by serine proteases to activate its prophenoloxidase (PPO) activity.29 Our recent study has shown Genes and Immunity

6

Genes and Immunity

Q16983 NP_000450

Diuretic hormone receptor precursor (A. domesticus) TEK tyrosine kinase (H. sapiens)

Receptors CK086667 CK086824

AmeF201 AmeR214

A46584 NP_446432 S13184 NP_004086 AAH23898 AAH45391 NP_005331 NP_002700 BAA81712 NP_062216 NP_055892 NP_006612 AAD45400 AAH42263 AAC38852

Cell signalling Effectors/modulators and intracellular transducers CK086666 AmeF189 Adenylyl cyclase-associated protein homolog (R. norvegicus) CK086811 AmeR229 ADP-ribosylation factor related protein 1 (R. norvegicus) CK086741 AmeF290 Calcium-binding protein (B. lanceolatum) CK086717 AmeF276 Eukaryotic translation initiation factor 4E binding protein 1 (H. sapiens) CK086718 AmeF69 Eukaryotic initiation translation factor 4g M. musculus) CK086816 AmeR65 MAP kinase-interacting serine/threonine kinase 2, similar to (D. rerio) CK086818 AmeR70 Protein kinase C inhibitor (H. sapiens) CK086680 AmeF150 Protein phosphatase 1b (H. sapiens) CK086819 AmeR222 Protein tyrosine kinase (E. fluviatilis) CK086820 AmeR13 Regulator of G-protein signalling 7 (R. norvegicus) CK086823 AmeR209 Sterile a and TIR motif containing protein 1 (H. sapiens), SARM CK086821 AmeR224 S-adenosylhomocysteine hydrolase-like 1 (H. sapiens) CK086682 AmeF107 Ser/thr protein phosphatase catalytic subunit (H. sapiens) CK086699 AmeF21 Soc2 (X. laevis) CK086822 AmeR95 Slimb (D. melanogaster)

Putative match accession

T18545 AAP59456 BAB03398 AAF71303 AAF44663 P80476 XP_345912 CAB64684 P02284 NP_511095 A53120 AAN10061 Q99538 NP_060830 NP_780433 NP_667344 AAG33867 P21902 AAA42064 P82151 BAA84188 BAA85250 Q05187

AmeF244 AmeF295 AmeF89 AmeR169 AmeR112 AmeF45 AmeF47 AmeF288 AmeF248 AmeF82 AmeF210 AmeF153 AmeF202 AmeF75 AmeF249 AmeF179 AmeR218 AmeF231 AmeF111 AmeR167 AmeR159 AmeF259 AmeF98

(a) Frontline immunity CK086626 CK086673 CK086674 CK086808 CK086833 CK086632 CK086633 CK086661 CK086662 CK086663 CK086635 CK086675 CK086676 CK086638 CK086677 CK086640 CK086798 CK086641 CK086642 CK086800 CK086801 CK086645 CK086647

Putative match (organism)

a´2-macroglobulin (L. polyphemus) Cathepsin B precursor (A. ventricosus) Cysteine-rich protease inhibitor (M. musculus) Dual oxidase (C. elegans) Golgi membrane protein (H. sapiens) Hemocyanin AA6 chain (A. australis) Hemolectin CG7002-PA, similar to (R. norvegicus) Histone H2A, putative (A. aquaticus) Histone H2B (P. granatina) Histone H3 (D. melanogaster) Intracellular coagulation inhibitor LICI precursor (T. tridentatus) Kunitz-like protease inhibitor precursor (A. caninum) Legumain precursor (H. sapiens) Lipopolysaccharide specific response-5 protein (H. sapiens) Methionyl aminopeptidase 1 (M. musculus) Nuclear transcription factor NFX2 (H. sapiens) Platelet-activating factor acetylhydrolase Ib-a´ subunit (S. scrofa), PAFAH Proclotting enzyme precursor (T. tridentatus) Salivary proline-rich protein (R. norvegicus) Tachylectin I (T. tridentatus) Tachylectin-5A (T. tridentatus) Tachystatin A2 (T. tridentatus) Transglutaminase (T. tridentatus)

Clone

dbEST accession

7e-77 2e-34

4e-33 4e-15 1e-13 7e-10 2e-35 5e-57 7e-17 9e-92 7e-38 3e-17 5e-42 6e-80 2e-43 1e-47 2e-68

2e-47 5e-10 4e-16 8e-11 3e-03 1e-44 1e-04 5e-27 2e-45 2e-68 2e-19 3e-26 3e-12 5e-20 2e-66 2e-36 7e-63 7e-03 1e-07 9e-78 3e-73 9e-30 1e-92

E-value

53%/77 46%/135

46%/134 63%/55 25%/179 50%/77 35%/219 64%/154 66%/53 98%/160 54%/151 49%/79 55%/159 79%/167 90%/87 69%/131 92%/133

73%/124 35%/121 40%/92 46%/66 24%/69 66%/128 40%/50 93%/66 94%/99 99%/133 38%/112 43%/130 32%/119 48%/92 68%/170 53%/112 72%/160 31%/87 35%/124 95%/135 63%/183 85%/67 68%/220

Identity/alignment length (aa)

1 —

1 — 1 1 1 — — 1 — — — — 1 1 —

1 1 1 — — 1 1 1 1 1 1 1 1 1 1 1 — 1 1 — — 1 1

— 1

— 1

— — — 1 1 — 1 1 1 1 — — 1

— — — 1 1 — — — — — — — — — — — 1 — — 1 1 — —

Forward Reverse

Redundancy

Table 3 Subtractive ESTs from the (a) amebocytes and (b) hepatopancreas with single redundancy grouped according to their putative functions as outlined in the ‘Expressed Gene Anatomy Database’ protocol23

Transcript profiling of immune responsive genes JL Ding et al

AmeF162 AmeF227 AmeR114 AmeF182 AmeF112 AmeF253 AmeF123 AmeR177 AmeF224 AmeR165 AmeR240

HpF189 HpF238 HpF66 HpF193 HpF276 HpF192 HpF160 HpF226 HpF302 HpR14 HpR151 HpR201 HpF54 HpF301 HpF302

Stress response CK086648 CK086650 CK086804 CK086655 CK086658 CK086657 CK086660 CK086835 CK086664 CK086809 CK086856

(b) Frontline immunity CK086929 CK086931 CK086981 CK086982 CK086983 CK086932 CK087197 CK086933a CK086936 CK087123 CK087137 CK087165 CK086984 CK086935 CK086936 BAA88518 AAC78603 AAN10061 PC1315 BAA34642 Q05187 AAL99079

Hemolectin (D. melanogaster) HLA class III protein Dom3z (H. sapiens) Kunitz-like protease inhibitor precursor (A. caninum) Large granule L2 chain (T. tridentatus) Prophenoloxidase activating enzyme-I precursor (H. diomphalia) Transglutaminase (T. tridentatus) Zinc proteinase Mpc1 (P. camtschaticus)

NP_506864 NP_080097

BAA86911 CAB51918 P04069 O97578 Q26636 NP_038512 BAB47146 CAA47396 AAQ57208 Q41651 P54985 JC4536 NP_741839 AAB34362 AAQ57208

NP_504339 AAA58676 AAH42347 NP_506112 AAD31042 AAF09496 AAO52675 AAG47843 AAG45936 NP_464365 NP_033037

AAN76709 AAF71303

Putative match accession

26/29 kDa proteinase, homologue of Sarcophaga (P. americana) Apolipophorin precursor protein (L. migratoria) Carboxypeptidase B (A. astacus) Cathepsin C (C. familiaris) Cathepsin L precursor (S. peregrina) Complement component C2 (M. musculus) Complement component C3 (B. belcheri) Complement component C4 (O. aries) Cyclophilin-FK506-binding protein 7 (H. sapiens) Cyclophilin-peptidyl-prolyl cis–trans isomerase (V. faba) Cyclophilin - Peptidyl-prolyl cis–trans isomerase (B. germanica) Cystatin precursor (T. tridentatus) EB module containing serine protease inhibitor family (C. elegans) Factor C (C. rotundicauda) FK506-binding protein 7-cyclophilin (H. sapiens)

78 kDa glucose-regulated protein precursor (X. laevis) Chaperonin-like protein (H. sapiens) Chaperonin subunit 2, similar to (X. laevis) Glutathione-S-transferase, C-terminal family member (C. elegans) Heat-shock protein 70 (C. gigas) Heat-shock protein 70 (M. sexta) Heat-shock protein 90a (A. mexicanus) Na+/K+ ATPase alpha subunit (C. sapidus) Protein disulfide isomerase (B. mori) Tetracycline resistance protein, similar to (L. monocytogenes) UV excision repair protein RAD23 homolog B (M. musculus)

Acheron (M. sexta) Deoxyribonuclease I precursor (R. catesbeiana)

Putative match (organism)

Cell signalling Effectors/modulators and intracellular transducers CK086967 HpF134 Calcineurin-like phosphoesterase (C. elegans) /metallo-phosphoesterase CK086968 HpF6 Calcineurin substrate (M. musculus)

HpF212 HpF44 HpF232 HpR217 HpF149 HpF170 HpF294

AmeR1 AmeF277

Apoptosis CK086803 CK086653

CK086939 CK086937 CK086985 CK087173a CK086943 CK086946 CK086991

Clone

dbEST accession

Table 3 Continued

7e-32 2e-12

4e-17 2e-13 3e-26 3e-01 1e-09 5e-36 6e-20

5e-04 2e-08 1e-15 8e-62 2e-15 1e-05 2e-38 9.9e-01 9e-03 2e-44 3e-47 7e-31 7e-09 1e-131 9e-03

3e-41 4e-17 2e-70 5e-09 6e-89 1e-75 1e-67 9e-43 3e-45 2e-08 4e-46

9e-14 8e-11

E-value

58%/105 60%/53

30%/149 55%/65 43%/149 89%/19 44%/68 58%/131 42%/140

28%/87 46%/58 60%/63 67%/108 75%/44 29%/112 40%/189 37%/51 41%/36 61%/140 82%/105 52%/100 26%/134 97%/207 41%/36

89%/87 76%/50 84%/158 50%/54 78%/210 94%/150 85%/146 62%/132 57%/150 30%/122 65%/152

45%/79 42%/145

Identity/alignment length (aa)

1 1

1 1 1 — 1 1 1

1 1 1 1 1 1 1 1 1 — — — 1 1 1

1 1 — 1 1 1 1 — 1 — —

— 1

— —

— — — 1 — — —

— — — — — — — — — 1 1 1 — — —

— — 1 — — — — 1 — 1 1

1 —

Forward Reverse

Redundancy

Transcript profiling of immune responsive genes JL Ding et al

7

Genes and Immunity

Continued

Genes and Immunity

HpF143 HpF247

HpF239 HpF41 HpF186 HpR211 HpF94 HpR40 HpR54 HpF20 HpF155 HpF108 HpF241 HpR235 HpR168

Apoptosis CK087003 CK087022

Stress response CK086947 CK086949 CK086950 CK087170 CK086951 CK087127 CK087129 CK086955 CK086960 CK086961 CK086964 CK087179 CK087147

Aldehyde dehydrogenase class 1 (X. laevis) Amine sulfotransferase (O. cuniculus) Carbonyl reductase 1 (C. griseus) Carbonyl reductase 4 (R. norvegicus) Catalase (C. jejuni) Copper-induced metallothionein (T. thermophila) Glutathione-S-transferase D8 (A. gambiae) Glutathione-S-transferase 11 (C. elegans) Multidrug resistance protein (C. elegans) Muscle LIM protein (E. scudderiana) Protein disulfide isomerase (D. melanogaster) Selenium-binding protein-like protein (A. thaliana) Thioredoxin family member, predicted CDS (C. elegans)

Glucose transporter 1 (D. melanogaster) 3’-5’ RNA exonuclease (H. sapiens)

Gephyrin (G. gallus) Transient receptor potential locus C protein precursor (D. melanogaster)

HpF47 HpF218

Receptors CK086970 CK086979

NP_079471 BAA85874 AAK26171 CAD89005 AAL71878

G-protein-coupled receptor induced protein (H. sapiens) JNK/SAPK-associated protein-1 (M. musculus) Phosphatidylinositol-3 phosphate 3- phosphatase adaptor subunit (H. sapiens) TRAF4 protein (D. rerio) Trypsin type 4 (L. salmonid)

BAA76411 BAA24994 BAB62840 NP_872613 Q59296 AAK84951 AAG45165 NP_508625 NP_509658 AAF23406 AAA85099 AAM96954 NP_503954

NP_523878 NP_149100

AAD49748 P36951

AAH48224 AAD37797

Putative match accession

Putative match (organism)

Granulin, similar to (X. laevis) Miple (D. melanogaster)

HpF206 HpF174 HpF141 HpR202 HpR152

Clone

Hormone/growth factors CK086972 HpF226 CK086941 HpF127

CK086971 CK086973 CK086976 CK087166 CK087138

dbEST accession

Table 3

6e-28 2e-09 2e-32 1e-07 1e-75 1e-06 5e-12 3e-04 5e-11 4e-22 3e-10 9e-10 2e-33

1e-67 3e-37

1e-11 2e-12

6e-14 3e-09

4e-17 1e-22 2e-42 2e-11 2e-24

E-value

59%/88 36%/87 64%/107 50%/52 94%/27 31%/92 31%/128 40%/57 60%/52 75%/57 57%/47 86%/29 55%/118

70%/179 100%/78

40%/83 53%/54

32%/131 39%/76

51%/76 32%/239 36%/231 33%/119 37%/151

Identity/alignment length (aa)

1 1 1 — 1 — — 1 1 1 1 1 —

1 1

1 1

1 1

1 1 1 — —

— — — 1 — 1 1 — — — — — 1

— —

— —

— —

— — — 1 1

Forward Reverse

Redundancy

Transcript profiling of immune responsive genes JL Ding et al

8

Transcript profiling of immune responsive genes JL Ding et al

9

Figure 2 Virtual Northern analysis of some representative genes taken from pooled cDNAs of 3 þ 6 h postinfection (hpi). Representative genes from each of the four major functional clusters of ESTs are analysed in (a) forward libraries and (b) reverse libraries. The following genes were studied under each functional group: (1) Frontline immunity: coagulogen II precursor (AmeF113); transglutaminase (TGase) substrate 8.6 kDa (AmeF238); tachycitin precursor (AmeF155); kazal proteinase inhibitor (HpF219); platelet-activating factor acetylhydrolase1b-a (PAFAH1b-a, AmeR218); tissue factor pathway inhibitor (TFPI, AmeR235) and peptidyl-prolyl cis–trans isomerase cyclophilin (CYP, HpR14). (2) Cell signaling: Variant surface glycoprotein phospholipase C (VSG Plp C, HpF125); tumour necrosis receptor associated factor 4 (TRAF4, HpR202). (3) Apoptosis: cytochrome c oxidase subunits I, II and III (COX I, COX II and COX III); sensitive to apoptosis gene (SAG, AmeR204) also known as RING finger 7 immunomodulators. (4) Stress response: glutathione S-transferase m1 (GST m1, HpF11); amine sulfotransferase (HpF41); copper chaperone for superoxide dismutase (Cu-SOD, HpF51); thioredoxin family member (TRX, HpR168). N and I denote samples from the naı¨ve and infected tissues, respectively. The ribosomal protein L3 (RpL3) was used as a loading control.

that in the presence of Gram-negative bacteria or LPS, the HMC/PPO is activated into phenoloxidase, PO (Jiang et al, unpublished). Indeed, we have also identified a prophenoloxidase activating enzyme, PPOA (HpF149), which was strongly upregulated in the hepatopancreas from 3 hpi to reach a peak at 6 hpi (Figure 4b). The resulting PPO activity of HMC catalyses the oxidation of phenols to quinones, thereby inducing melanisation to

oxidatively kill the invading microbe.29 Furthermore, HMC also releases a C-terminal peptide that has antimicrobial activity.30 Proteases are essential for a variety of immune processes, including hemostasis and clot resolution,31 complement activation,32 inflammation33 and tissue remodelling.34 In particular, the role of serine proteases in immunity has been thoroughly studied.11 Interestingly, Genes and Immunity

Transcript profiling of immune responsive genes JL Ding et al

10

Figure 3 Kinetic profile of transcription of genes in forward and reverse libraries of (a) amebocytes and (b) hepatopancreas. Virtual Northern analysis of the representative ESTs was performed following Pseudomonas infection at 3, 6, 12, 24, 48 and 72 h postinfection (hpi). The ESTs are as described in Figure 2. In addition, the following ESTs were examined: Kunitz-like protease inhibitor precursor, Kunitz (AmeF153, HpF232); tachycitin (AmeF155); hemocyanin (AmeF45); complement component 4 (C4, HpF226); prophenoloxidase activating enzyme (PPOA, HpF149); calcineurin substrate (CaN, HpF6). The COX I, II and III from both the amebocytes and hepatopancreas showed repression, although AmeF57 (COX II) was isolated from a forward library, possibly due to a false positive EST clone. This possibility is supported by the consistent trend of the downregulation of all COX members, including AmeF57 (see Figure 4). Arrows indicate the isoforms studied for that particular EST. Ribosomal protein L3 (RpL3) was used as a loading control.

in this work, we have identified several serine and cysteine proteases, specifically from the forward libraries of both tissues: a 26/29 kDa proteinase (HpF189), Factor C (AmeF287, HpF301), carboxypeptidase B (HpF66), cathepsins B, C and L (AmeF295, HpF193 and HpF276, respectively) and legumain precursor (AmeF202). Factor C is a well-characterised serine protease that plays a pivotal role in LPS-recognition during Gram-negative bacterial infection.9–11,13,35,36 Ironically, at the other extreme, proteases that act beyond their narrowly constituted physiological function can impose serious damage to host tissues. Thus, hosts Genes and Immunity

have evolved regulatory processes to inactivate proteases associated with the pathological processes and the infective microbes. Most important of these regulatory molecules are protease inhibitors.18,36,37 From the forward libraries, we have identified a well-known protease inhibitor, a2-macroglobulin (AmeF244, HpF286). Although earlier studies on the L. polyphemus a2-macroglobulin have shown its expression only in the amebocytes,38 our finding of the a2-macroglobulin ESTs in both the amebocyte and hepatopancreas libraries indicates a more widespread spatial expression of this gene during Pseudomonas infection. To confirm this

Transcript profiling of immune responsive genes JL Ding et al

11

Figure 4 Comparison of the expression profiles of genes in (a) amebocyte forward, (b) hepatopancreas forward, (c) amebocyte reverse and (d) hepatopancreas reverse libraries. The hybridisation signals from Figure 3 were densitometrically scanned and normalisation was made against ribosomal protein L3 gene (RpL3). The annotations for the ESTs are as described in Figures 2 and 3.

observation, the tissue distribution of expression of a2macroglobulin was examined by RT-PCR, which showed that it is indeed expressed in amebocytes, hepatopancreas, heart, intestine and stomach (Supplementary Figure 1). Our finding of other protease inhibitor ESTs such as Kazal-proteinase inhibitor, HpF219 (Figure 2a), intracellular coagulation inhibitor precursor (AmeF210), intracellular coagulation inhibitor type 2 (AmeF267), Kunitz-like protease inhibitor precursors (AmeF153, HpF232) (Figures 3 and 4), cysteine-rich protease inhibitor (AmeF89) and serine protease inhibitor (HpF282) further reiterates the importance of protease inhibitors in regulating infection-induced proteolytic activation of serine- and cysteine-proteases in innate immune defense. Some of these inhibitors19,39–41 have not been previously reported in the horseshoe crab. Another protease inhibitor, a tissue factor pathway inhibitor, TFPI (AmeR235), was isolated from the amebocyte reverse library (Figure 2b). TFPI, which is composed of three

Kunitz-type protease inhibitor domains, represents the only physiological inhibitor of hemostasis.42 The TFPI identified in this study differs from the members of the serpin superfamily (limulus intracellular coagulation inhibitor, LICI-1,2,3) of coagulogen inhibitors43 and is also distinctive from the LICI (AmeF210) found in this work (Table 2a). Hence, the C. rotundicauda TFPI-like protease inhibitor represents a novel type of anticoagulant. Therefore, downregulation of TFPI (AmeR235) (Figure 2b) would increase tissue-localised hemostasis, indicating that it plays an important role in the early phase of infection, to entrap and confine the invading pathogen to the infected tissue. Besides the above-mentioned factors such as coagulogen, proteases and protease inhibitors that regulate infectioninduced hemostasis, the histones H2A (AmeF288), H2B (AmeF248) and H3 (AmeF82) were present in the amebocyte forward library. This concurs with recent evidence suggesting that histones possess antimicrobial activities44 and endotoxin binding capabilities.45 Genes and Immunity

Transcript profiling of immune responsive genes JL Ding et al

12

PAFAH (AmeR218), a platelet-activating factor acetylhydrolase, plays a role in hydrolysing a plateletactivating factor (PAF),46 which mediates inflammatory activity.47 Thus, the Pseudomonas infection-mediated suppression of PAFAH was probably required for PAF to accumulate to an effective level for the production of antimicrobial proinflammatory cytokines. Therefore, the downregulation of PAFAH in the acute phase of infection (Figure 2b) suggests a mechanism that the host might employ to enable PAF-mediated control of pathologic inflammation. Cyclophilin (CYP) comprises a ubiquitous family of proteins,48 which plays a role in immunity by binding to the immunosuppressant drug cyclosporin A, resulting in the blockage of cytokine gene transcription in activated T cells.49 Two CYP ESTs (HpR14, HpR151) were isolated from the hepatopancreas reverse library. Downregulation of the CYP, for example, HpR14 (Figures 2b and 3b) during the acute phase of Pseudomonas infection indicates that CYP participates in the modulation of other genes involved in the host innate immune response. Previously, the complement system was thought to exist only in deuterostomes. Here, we have identified ESTs encoding complement components C3 (HpF160), C2/Factor B (HpF192) and C4 (HpF226) from the hepatopancreas forward library. Invertebrate C3 isolated from the sea urchin is upregulated during LPS injection.50 Some of the known roles for C3 are opsonisation of the invading microbes,51 formation of membrane attack complex32 and clearance of apoptotic cells.52 Our EST data suggest the involvement of complement system in frontline immune defense against Pseudomonas infection. Recently, we have isolated and extensively characterised the cDNAs of complement factors and their corresponding proteins, C3 and C2/Bf, from the horseshoe crab, C. rotundicauda16 (an ancient protostome). It is therefore conceivable that a formidable and complex host–pathogen interaction mechanism already exists 4500 million years and that the vertebrate complement system evolved from that of the early invertebrates. Cell signalling During microbial infection, the recognition of PAMPs by host cell Toll/Toll-like receptors and IMD triggers complex signalling cascades to the nucleus, leading to the expression of effector molecules. NFkB and MAPK pathways are major cytoplasmic regulators of innate immunity.53,54 Recent evidence indicates that G-protein signalling through the inositol-1,4,5-trisphosphate pathway potentially plays a role in the amebocyte degranulation upon LPS stimulation.55 In addition, we observed that Factor C (AmeF287 and HpF301) was upregulated in amebocytes and hepatopancreas. Factor C released from degranulated amebocytes9,24 as well as extracellular Factor C secreted by the hepatopancreas36 probably activates the G protein pathway.56 On the other hand, downregulation of AmeR13, a regulator of G-protein signalling 7 (RGS7), which promotes GTPase activity, indicates prolonged activity of G proteins for the signalling action of Factor C.36 However, in view of the inositol phosphate pathway, the appearance of a variant surface glycoprotein phospholipase C, VSG PlpC (HpF125) (Figure 2a) is intriguing as this enzyme cleaves a glycosylphosphatidylinositol anchor of the VSG, a Trypanosome coat protein.57 Therefore, the involvement of

Genes and Immunity

VSG PlpC as a potential cell signalling molecule in the horseshoe crab during Pseudomonas infection remains uncertain. The calcineurin (CaN) homologue in the vertebrates is a calcium- and calmodulin-dependent protein serine/ threonine phosphatase, which plays an essential role in T-cell receptor-mediated signal transduction leading to the transcriptional activation of cytokines.58 Thus, upregulation of both of the calcineurin-like phosphoesterase/ metallo-phosphoesterase (HpF134) and CaN substrate (HpF6) (Figures 3 and 4 and Table 3b) in response to Pseudomonas infection is indicative of a parallel existence of T-cell receptor-like mediation of signal transduction in this ancient species and that these genes possibly interact during Pseudomonas infection. The horseshoe crab coagulogen and Drosophila spaetzle are proposed to be structurally and functionally equivalent,59 with the latter functioning in a signal transduction cascade upstream of Drosophila Toll.36,60 Therefore, coagulogen, which appeared in 27 ESTs, was most highly expressed and probably serves multiple roles as a direct pathogen recognition protein in frontline immune defense, acting as a coagulant to immobilise invading pathogens, as well as priming cell signalling events during infection. Although cell signalling is expected to regulate many crucial defense mechanisms, a sizeable proportion (23%) of the cell signalling ESTs was present in the amebocyte reverse library at 3–6 hpi (Figure 1c). It is conceivable that the cell signalling genes were repressed during this acute phase, and may only be de-repressed at a later phase of infection. Supporting this hypothesis was the downregulation of several cell signalling ESTs as they appeared in the reverse libraries of both the hepatopancreas and amebocytes (Tables 2 and 3). Firstly, MAPK (AmeR65), a major arm in intracellular signalling, was suppressed. Secondly, PAFAH (AmeR218), which hydrolyses and attenuates the bioactivity of the PAF, a mediator of inflammation,46 was repressed (Figure 2b), indicating the suppression of cell signalling47 and the accumulation of PAF, which mediates inflammatory response. Thirdly, SARM (AmeR209), which codes for sterile a and Toll interleukin-1 receptor (TIR) motifcontaining protein 1, and TRAF4 (HpR202), a signaltransducing adaptor protein, were also downregulated (Figures 2b and 4d). Our discovery of TRAF4 and SARM provides a strong indication on the existence of a complex network of Toll-like receptor signalling pathway in this species. Nevertheless, the presence of TRAF4 (Figures 2b, 3b and 4d) and SARM (Table 3a) in the reverse libraries consistently suggests substantial suppression of cell signalling genes at this acute phase of infection. It is plausible that, with some exceptions, most of the cell signalling genes may not need to be upregulated until after the pre-existing high levels of extracellular frontline innate immune proteins are exhausted from the initial antimicrobial combat during the early phase of infection. Equally likely are the post-translational modifications like phosphorylation/dephosphorylation that could change the signalling protein profile rather than the expression/repression of signalling genes. Apoptosis Pathogen-mediated apoptosis may be viewed as a means to disable the host from properly mounting an immune

Transcript profiling of immune responsive genes JL Ding et al

defense. On the other hand, the host can also bring about apoptosis to minimise the spread of infection from infected cells to healthy cells, since apoptosis of the infected host cell permits other cells to ingest the apoptotic bodies containing internalised pathogens.61 Therefore, whether the occurrence of apoptosis during Pseudomonas infection of the horseshoe crab is pathogenor host-mediated is of particular interest. Recent studies have reported that apoptosis induced by many stimuli such as TNF-a depends on the integrity of the mitochondrial respiratory chain,62 and that its role in the generation of reactive oxygen species (ROS) can lead to apoptosis.63 In this study, 13 ESTs related to apoptosis genes were identified, mostly from the reverse libraries. Of these, cytochrome c oxidases (COX) were found in both tissues. COX is a terminal enzyme of the mitochondrial respiratory chain,64 which plays dual functions in energy generation and apoptosis. COX I, II and III proteins have been previously isolated from the Atlantic horseshoe crab, L. polyphemus.65 As abrogation of host cell apoptosis is often beneficial for the pathogen,66 it can be postulated that P. aeruginosa has mediated the repression of the COX I, II and III genes (Figures 2b, 3 and 4c and d), a mechanism which the pathogen has employed for its survival in the host. At the same time, the invading Pseudomonas has established various effects like latent infection61 and inhibition of the immune cell oxidative burst, although the latter is important in the host innate immune defense against pathogens.67 It appears that the host has counterbalanced the pathogen-mediated repression of COX genes by upregulation of another mitochondrial EST, the amine oxidase (HpF292), which generates H2O2, causing oxidative damage68 and apoptosis.69 Another gene, sensitive to apoptosis gene, SAG, which is a member of the zinc RING finger family of proteins, is evolutionarily conserved in diverse organisms and has been shown to function as an antioxidant to protect cells from metal ion- or ROS-induced apoptosis.70 Hence, SAG plays an antiapoptotic function. In the 3 þ 6 hpi pooled cDNA library, SAG (AmeR204) appeared downregulated (Figure 2b and 3a). Virtual Northern analysis of separate time points, however, showed the two major isoforms of SAG to be strongly represented at 3 hpi and downregulated at 6 hpi (Figures 3a and 4c). Taken together, this observation indicates that, as an antiapoptotic factor, the SAG isoforms probably played opposing roles at different time points of Pseudomonas infection, preventing host cell apoptosis at 3 hpi and causing apoptosis of the acutely infected cells at 6 hpi. Thus, in view of the expression of amine oxidase and SAG 1 and 2 on the one hand and the repression of COX enzymes on the other, there appears to be a fine balance of forces between the up- and/or downregulation of apoptosis and immune cell oxidative damage during host–pathogen interaction, which is beneficial for the survival of either the host or the pathogen as the modulation of immune response progresses. Stress response As many as 33 ESTs (Table 1) were found to show high homology to known stress-related genes. During an infection, the host generates ROS as a cytotoxic process.71 However, the host employs detoxification mechanisms to protect itself against excessive ROS. Four upregulated

ESTs show high homology to genes involved in detoxification (Tables 2 and 3): (i) copper chaperone for superoxide dismutase, Cu-SOD (HpF51, see Figure 2a); (ii) catalase (AmeF105); (iii) amine sulfotransferase (HpF41) and (iv) glutathione S-transferases (GSTs): HpF11 (Figures 2a, 3b and 4b), HpF20, AmeF81, AmeF182. SOD converts free radicals into H2O2, which is further detoxified into nontoxic components by catalase. Amine sulfotransferase and the family of GST enzymes facilitate the coupling of glutathione to endoand xeno-biotics, which bear electrophilic functional groups, hence detoxifying them.72 Thus, in response to P. aeruginosa invasion, GSTs were expressed to cope with oxidative stress due to pathogen-induced release of free radicals. Other than GST, another glutathione-dependent ROS-scavenger, selenoprotein W (AmeF49), was isolated from the amebocyte forward library. Selenoproteins are known to protect cells from oxidative stress.73 Contrary to expectation, thioredoxin, TRX (HpR168) was found in the hepatopancreas reverse library (Figure 2). TRX is known to play a role in homeostasis, detoxification and regulation of cytokine expression through modulation of NFkB upon induction by LPS.74 Its promoter contains regulatory elements that are responsible for oxidative stress.75 It is conceivable that the downregulation of TRX implies a tight control of this gene during Pseudomonas infection. It is also plausible that pre-existing TRX protein was sufficient to maintain the redox potential of the host cell. Perhaps, other stress-response genes like Cu-SOD, GSTs and amine sulfotransferase played a major role in protecting cells from oxidative stress during this acute phase of infection. Another interesting cluster of stress-response ESTs was heat-shock proteins. Hsp 90a (AmeF123), Hsp 70s (AmeF112, AmeF253, HpF102) and Hsp 40 (HpF24) were identified in the forward libraries. The Hsps 90, 70 and 40 are molecular chaperones, which assist protein folding, solubilise aggregated proteins and degrade damaged proteins.76 Stress-related oxidative damage increases chaperone levels. Thus, a high representation of Hsp ESTs (Tables 2 and 3) in the immune-responsive tissues of the horseshoe crab clearly indicates infectioninduced Hsp response to stress.

13

Kinetic profile of transcriptional response to Pseudomonas infection Although virtual Northern analysis of 3 þ 6 hpi pooled cDNAs yielded some general clues to the expression of genes over the acute phase of infection (Figure 2), it was insufficient to reveal the kinetic profile of gene expression, since the true levels of transcripts at 3 and 6 hpi may be compromised through pooling of reciprocal representations of the mRNAs at the two time points of infection. Thus, based on the premise that the immunerelated genes undergo rapid change over the time course of infection, we carried out virtual Northern analyses of gene expression in amebocyte and hepatopancreas from naı¨ve and Pseudomonas-challenged animals over 72 h of infection, using representative ESTs from each forward and reverse library (Figures 3 and 4). In the forward libraries, changes in gene expression upon bacterial challenge were apparent, while in the reverse libraries, there was repression of transcription of various genes. A consistent increase in the amebocyte gene expression was detected at 3 hpi in the forward libraries (Figure 4a and Genes and Immunity

Transcript profiling of immune responsive genes JL Ding et al

14

b). It is also notable that while some genes from the same functional group in the same tissue were induced, others were repressed. Recently, Ng et al14 demonstrated that the horseshoe crab rapidly clears a systemic infection by 106 cfu/ml P. aeruginosa within 6 hpi. This suggests that the pre-existing innate immune defense proteins in the plasma acted immediately in frontline antimicrobial combat, while transcriptional upregulation of the corresponding genes was imminently mounted within 3 h of infection, such that nascent immune response proteins may be translated to replenish the diminishing store of pre-existing proteins and effectors in order to sustain the immune defense. Interestingly, the hepatopancreas did not display a distinctive profile of transcriptional activation. Of the ESTs examined, only the genes encoding GST and PPOA were upregulated (Figure 4b), while the remaining genes were either only slightly activated later or were unresponsive and showed general suppression of transcription (Figure 4d) to this regime of Pseudomonas challenge over the 72 h period. Ameboyctes are intact and functional during Pseudomonas infection Phagocytosis is essential in host defense against microbial pathogens and in the clearance of apoptotic cells. Both microbial and apoptotic cells are delivered on a common route for degradation. We have recently demonstrated the role of amebocytes in the phagocytosis of pathogens.16 The importance of the amebocytes to such frontline innate immune defense was exemplified by its ability to remain largely intact and granulated (Figure 5) during the course of infection. Thus, we propose that in response to infection, the spatial and

temporal difference in expression and interactions of gene clusters, namely, frontline immune defense, cell signalling, apoptosis and stress response allow for the compensation of/gain-of-functions to overcome the infection. Since the amebocytes of the bacteria-challenged horseshoe crab showed increased gene transcription and continued to exhibit morphological integrity indicating their physiological and biochemical competency, the amebocytes would be expected to synthesise the cognate proteins and retain their ability to store and/ or secrete those newly synthesised proteins. Hence, exemplary to expectation, being the first to encounter the invading pathogens, the amebocytes would phagocytose the foreign bodies, rapidly transcribe their immune response genes and remain mostly intact to continue synthesising a formidable array of innate immune molecules to sustain the antimicrobial action. Consistent with our recent discovery of a primitive yet complex opsonic complement defense system in the horseshoe crab,16 we propose that during Pseudomonas infection, the amebocytes actively and acutely express immune-related genes and simultaneously invoke a complement-mediated phagocytosis of the invading pathogen.

Conclusion The primary goal of establishing the horseshoe crab EST database was to identify potential genes involved in the immune response against Gram-negative bacterial infection. As a ‘living fossil’, which lacks adaptive immunity, the horseshoe crab is an ideal experimental model as it has sustained an exceptionally powerful innate immune

Figure 5 The amebocytes sampled from horseshoe crab after infection with 106 cfu/ml of P. aeruginosa: (a) naı¨ve (0 h), (b) 3 hpi and (c) 6 hpi. The blood cells remained largely intact and granulated, with some showing vacuolation at 6 hpi. Consistent with being highly transcriptionally active, the amebocytes appear to maintain their morphological integrity, indicating that they are fully functional during the acute phase of infection. Genes and Immunity

Transcript profiling of immune responsive genes JL Ding et al

system. It offers a unique isolated system to study the genetics of innate immunity during host–pathogen interaction. Thus, mapping and clarifying the functional immunogenomics, particularly those that show homology in structure and function to human counterparts would be requisite to offering solutions to clinical problems associated with Pseudomonas-induced inflammation and sepsis. A total of 60 frontline immune response genes that represent previously characterised ESTs have been identified from both the amebocyte and hepatopancreas cDNA libraries. Among these genes, coagulogen was most highly expressed, suggesting its important role in innate immunity. Various complement components, members of the Toll-like receptor signalling pathway, apoptosis and stress response were elucidated. The amebocytes appeared to be acutely responsive to Pseudomonas infection while the hepatopancreas probably acted as a long term ‘immune-reserve’. The high levels of pre-existing frontline innate immune proteins in the plasma were probably summoned into immediate antimicrobial combat. While the frontline immunity genes and a limited number of cell signalling genes in the amebocytes underwent an upsurge in transcriptional activation, the transcription of the corresponding genes in the hepatopancreas seemed to remain somewhat repressed or delayed. It is not surprising that the majority of amebocyte cell signalling genes were downregulated (Figure 1c) at the early phase of infection since desensitisation might have occurred. Such spatial and temporal expression of the host genes that were differentially up-/downregulated during Gram-negative bacterial infection highlights the intricate complexity of the host–pathogen interaction and antimicrobial response. We propose that the participation and interplay of functionally related clusters of genes in different tissues encoding innate immunity, cell signalling, apoptosis and stress response occur during a systemic infection by a dose (106 cfu/ml) of P. aeruginosa, to successfully clear the pathogen within 6 hpi.14 This dose of P. aeruginosa would have been lethal to mice. Many of these pathogen-responsive genes could be developed as strategic antimicrobial candidates. Our findings contribute towards the understanding and resolution of the pathogen-recognition gene clusters; how they are exploited and/or modulated during the acute phase of infection; and how different immune-responsive tissues coordinate their actions in a concerted and timely manner. These findings will ultimately offer the much needed insights into such systems in the vertebrates and humans.

Materials and methods Bacterial infection P. aeruginosa was used as an inoculum as it is a known opportunistic and clinically significant pathogen.15 The horseshoe crab, C. rotundicauda, was collected from the Kranji estuary in Singapore. Infection of the horseshoe crab was performed as described by Ng et al.14 Briefly, P. aeruginosa ATCC 27853 was cultured overnight in tryptic soy broth (Difco) at 371C. Bacteria were pelleted at 5000 g for 5 min at 41C, washed in saline and resuspended in the original culture volume

in saline. Serial dilution and colony count were performed to determine colony-forming units (cfu). Following bacterial titration, the horseshoe crabs were injected with a sublethal dose of 1  106 cfu/ml of P. aeruginosa. Crude plasma was collected by cardiac puncture into 100 mM PMSF and centrifuged at 150 g for 15 min to separate the cell-free plasma from amebocytes. The hepatopancreas was removed via ventral excision at various hours postinfection (3, 6, 9, 12, 24, 48, 72 hpi). Each time point constitutes three experimental animals. The amebocytes and hepatopancreas were stored at 801C for extraction of total RNA. Naı¨ve tissue samples were obtained under the same conditions after mock infection with pyrogen-free saline as an inoculum.

15

RNA isolation Total RNA was isolated using TRIzolt reagent (Invitrogen) and mRNA was purified using the Oligotex-direct mRNA mini kit (Qiagen). The quality of the total RNA from the different tissue samples was assessed by electrophoresis in 1.2% formaldehyde agarose gel. mRNA samples were stored at 801C until use. Construction of cDNA libraries by suppression subtractive cDNA hybridisation Double-stranded cDNA was synthesised with SMARTt cDNA synthesis kit (Clontech). Briefly, the first strand cDNA synthesis reaction consist of 100 ng mRNA, 1 mM 30 SMART CDS Primer II A, 1 mM SMART II A Oligo, 1  first-strand buffer, 2 mM dithiothreitol, 50 U Powerscript reverse transcriptase and 1 mM dNTP. The reaction was performed at 421C for 1 h followed by a 10-fold dilution of the resulting cDNA in TE buffer. Second strand cDNA utilised 4 ml of the first strand diluted cDNA as template, 1  Advantage 2 PCR buffer, 0.2 mM dNTP, 0.2 mM 50 PCR Primer II A, and 1  Advantage 2 polymerase mix. PCR was performed as follows: 951C (1 min), proceeding to 15–19 cycles of 951C (15 s), 651C (30 s), and 681C (6 min) in a DNA gradient cycler AC 1234 Thermocycler (MJ Research, Inc.). CHROMA-SPIN columns were used for nucleotide removal and size-fractionation of the cDNA. The double-stranded cDNA was blunt-ended by digestion with RsaI. Suppression subtractive cDNA hybridisation was performed with PCR-Select cDNA subtraction kit (Clontech). Briefly, the tester cDNA was divided into two aliquots. Each aliquot was ligated to either adaptor 1 or 2R. The adaptor 1-ligated and adaptor 2R-ligated tester cDNAs were separately denatured at 981C for 90 s and then hybridised at 681C for 8 h with excess driver cDNA. The two primary hybridisation samples were then combined without denaturing. Simultaneously, freshly denatured driver cDNA was added to further enrich differentially repressed genes and the mixture was hybridised at 681C for 16 h. Subtraction was performed at a 1 : 40 ratio of tester to driver cDNA. Following this, the secondary hybridisation mixture was subjected to two rounds of suppression PCR to selectively enrich differential transcripts. The first round of PCR with primer 1 (50 -CTA ATA CGA CTC ACT ATA GGG C-30 ) was carried out at 751C for 5 min and 941C for 25 s, followed by 27 cycles of 941C for 30 s, 661C for 30 s and 721C for 90 s. A second round of PCR using nested primers 1 (50 -TCG AGC GGC CGC CCG GGC AGG T-30 ) and 2R (50 -AGC GTG GTC GCG GCC GAG GT-30 ) were employed to further enrich Genes and Immunity

Transcript profiling of immune responsive genes JL Ding et al

16

differential cDNAs. To perform this, primary PCR products were diluted 10-fold and used as a template for 12 cycles of PCR at 941C for 30 s, 681C for 30 s and 721C for 90 s each. Subtracted cDNAs were ligated into pGEM-T Easy vector (Promega) and transformed into Escherichia coli TOP10 competent cells. The blue-white colony selection method was applied to select positive clones. DNA sequencing and analysis of ESTs Randomly selected clones were sequenced unidirectionally using T7 primer (50 -TAA TAC GAC TCA CTA TAG GG-30 ) in Big Dye Terminator (Ver 3.1) reactions with an ABI Prism model 3100A sequencer. Vector sequences were manually removed with the DNAMAN ver. 4.15 software (Lynnon BioSoft), thereby leaving the insert for BLASTx nonredundant searches for matches to known sequences in GenBank.77 A sequence is considered to be significantly matched when the E-value is p103 and the alignment length is more than 10 amino acids.23,78 Virtual Northern analysis Virtual Northern analysis enables the study of gene transcription using minimal amounts of mRNAs, which are reverse-transcribed and amplified by RT-PCR into cDNAs.79 The analysis of the transcription of genes before and after infection using mRNAs pooled from 3 þ 6 hpi was first performed to offer a general profile on genes which may be up- or downregulated by Pseudomonas infection. Furthermore, to define the kinetics of expression of various functional groups of ESTs over the 72 h of infection, the transcription of selected members of each functional group of immune responsive genes was monitored over individual time points (0, 3, 6, 12, 24, 48 and 72 hpi). Each time point used mRNAs pooled from three infected animals. To ensure unbiased amplification of the cDNAs for all tissues taken at each time point, the mRNAs were subjected to the same number of cycles of RT-PCR. To this end, the number of PCR cycles was predetermined by several rounds of optimisation of the PCR conditions. To analyse the expression of selected clones, cDNAs were synthesised as described above. The doublestranded cDNA (400 ng) was resolved in a 1% agarose gel and transferred onto a nylon membrane (Hybond N þ , Amersham) by capillary action. DNA probe fragment was obtained by PCR amplification of the insert using Clontech nested PCR primers 1 and 2R, followed by RsaI digestion to remove the cDNA subtraction adapters. The probe fragment was [32P]labelled with Rediprimet II labelling kit (Amersham). For loading controls, the Northern membrane was simultaneously hybridised with endogenous gene coding for ribosomal protein L3 (RpL3). Hybridisation was carried out at 681C overnight in 1% SDS, 6  SSC (0.9 M sodium chloride and 0.1 M trisodium citrate), 100 mg/ml calf thymus DNA (Sigma), 0.1% Ficoll 400 (Sigma), 0.1% polyvinylpyrrolidone and 0.1% bovine serum albumin. The membrane was subjected to stringency washes in 2  SSC 0.1% SDS at room temperature, 1  SSC 0.1% SDS at 371C and 0.2  SSC 0.1% SDS at 421C, respectively. The nylon membrane was then exposed to X-ray film (Kodak) and developed using a Kodak M35 X-OMAT Processor.

Genes and Immunity

Analysis of the amebocytes The horseshoe crab was infected with P. aeuruginosa as described above. At 0 h (naı¨ve), 3 and 6 hpi, blood was collected by cardiac puncture, directly into tubes containing a fixative (10% formalin in 0.5 M NaCl) at the final ratio of 1 : 5 (v/v). After 30 min of fixation in an ice bath, the samples were centrifuged at 150 g for 10 min. The supernatant was removed, and the amebocytes were resuspended in fresh fixative and viewed under bright field microscopy.

Acknowledgements This work was supported by a grant (03/1/21/17/227) from the Agency of Science, Technology and Research (A*STAR), Singapore. We thank Ms Kaitian Peng (an A*STAR-funded undergraduate scholar of the Imperial College, London, UK) for help with sequencing some ESTs.

References 1 Beutler B. Innate immunity: an overview. Mol Immunol 2004; 40: 845–859. 2 Medzhitov R, Janeway Jr CA. How does the immune system distinguish self from nonself? Semin Immunol 2000; 12: 185–188; discussion 257–344. 3 Engstrom Y. Induction and regulation of antimicrobial peptides in Drosophila. Dev Comp Immunol 1999; 23: 345–358. 4 Lemaitre B, Reichhart JM, Hoffmann JA. Drosophila host defense: differential induction of antimicrobial peptide genes after infection by various classes of microorganisms. Proc Natl Acad Sci USA 1997; 94: 14614–14619. 5 Irving P, Troxler L, Heuer TS, Belvin M, Kopczynski C, Reichhart JM et al. A genome-wide analysis of immune responses in Drosophila. Proc Natl Acad Sci USA 2001; 98: 15119–15124. 6 De Gregorio E, Spellman PT, Rubin GM, Lemaitre B. Genomewide analysis of the Drosophila immune response by using oligonucleotide microarrays. Proc Natl Acad Sci USA 2001; 98: 12590–12595. 7 De Gregorio E, Spellman PT, Tzou P, Rubin GM, Lemaitre B. The Toll and Imd pathways are the major regulators of the immune response in Drosophila. EMBO J 2002; 21: 2568–2579. 8 Dimopoulos G, Christophides GK, Meister S, Schultz J, White KP, Barillas-Mury C et al. Genome expression analysis of Anopheles gambiae: responses to injury, bacterial challenge, and malaria infection. Proc Natl Acad Sci USA 2002; 99: 8814–8819. 9 Ding JL, Navas MAr, Ho B. Two forms of factor C from the amoebocytes of Carcinoscorpius rotundicauda: purification and characterisation. Biochim Biophys Acta 1993; 1202: 149–156. 10 Ding JL, Navas III MA, Ho B. Molecular cloning and sequence analysis of factor C cDNA from the Singapore horseshoe crab, Carcinoscorpius rotundicauda. Mol Mar Biol Biotechnol 1995; 4: 90–103. 11 Ding JL, Wang LH, Ho B. Current genome-wide analysis on serine proteases in innate immunity. Curr Genomics 2004; 5: 147–155. 12 Iwanaga S, Kawabata S, Muta T. New types of clotting factors and defense molecules found in horseshoe crab hemolymph: their structures and functions. J Biochem (Tokyo) 1998; 123: 1–15. 13 Tan NS, Ho B, Ding JL. High-affinity LPS binding domain(s) in recombinant factor C of a horseshoe crab neutralizes LPSinduced lethality. FASEB J 2000; 14: 859–870.

Transcript profiling of immune responsive genes JL Ding et al

17 14 Ng ML, Tan SH, Ho B, Ding JL. The C-reactive protein: a predominant LPS-binding acute phase protein responsive to Pseudomonas infection. J Endotoxin Res 2004; 10: 163–174. 15 Yau YH, Ho B, Tan NS, Ng ML, Ding JL. High therapeutic index of factor C Sushi peptides: potent antimicrobials against Pseudomonas aeruginosa. Antimicrob Agents Chemother 2001; 45: 2820–2825. 16 Zhu Y, Thangamani S, Ho B, Ding JL. The ancient origin of the complement system. EMBO J 2005; 24: 382–394. 17 Kawabata S, Iwanaga S. Role of lectins in the innate immunity of horseshoe crab. Dev Comp Immunol 1999; 23: 391–400. 18 Armstrong PB. The contribution of proteinase inhibitors to immune defense. Trends Immunol 2001; 22: 47–52. 19 Iwanaga S. The molecular basis of innate immunity in the horseshoe crab. Curr Opin Immunol 2002; 14: 87–95. 20 Aaronson JS, Eckman B, Blevins RA, Borkowski JA, Myerson J, Imran S et al. Toward the development of a gene index to the human genome: an assessment of the nature of highthroughput EST sequence data. Genome Res 1996; 6: 829–845. 21 Mita K, Morimyo M, Okano K, Koike Y, Nohata J, Kawasaki H et al. The construction of an EST database for Bombyx mori and its application. Proc Natl Acad Sci USA 2003; 100: 14121–14126. 22 Seitz V, Clermont A, Wedde M, Hummel M, Vilcinskas A, Schlatterer K et al. Identification of immunorelevant genes from greater wax moth (Galleria mellonella) by a subtractive hybridization approach. Dev Comp Immunol 2003; 27: 207–215. 23 White O, Kerlavage AR. TDB: new databases for biological discovery. Methods Enzymol 1996; 266: 27–40. 24 Iwanaga S. The limulus clotting reaction. Curr Opin Immunol 1993; 5: 74–82. 25 Osaki T, Kawabata S. Structure and function of coagulogen, a clottable protein in horseshoe crabs. Cell Mol Life Sci 2004; 61: 1257–1265. 26 Volokhina EB, Hulshof R, Haanen C, Vermes I. Tissue transglutaminase mRNA expression in apoptotic cell death. Apoptosis 2003; 8: 673–679. 27 Melino G, Piacentini M. ‘Tissue’ transglutaminase in cell death: a downstream or a multifunctional upstream effector? FEBS Lett 1998; 430: 59–63. 28 Tokunaga F, Yamada M, Miyata T, Ding YL, HiranagaKawabata M, Muta T et al. Limulus hemocyte transglutaminase. Its purification and characterization, and identification of the intracellular substrates. J Biol Chem 1993; 268: 252–261. 29 Nagai T, Osaki T, Kawabata S. Functional conversion of hemocyanin to phenoloxidase by horseshoe crab antimicrobial peptides. J Biol Chem 2001; 276: 27166–27170. 30 Lee SY, Lee BL, Soderhall K. Processing of an antibacterial peptide from hemocyanin of the freshwater crayfish Pacifastacus leniusculus. J Biol Chem 2003; 278: 7927–7933. 31 Furie B, Furie BC. Molecular and cellular biology of blood coagulation. N Engl J Med 1992; 326: 800–806. 32 Reid KB, Porter RR. The proteolytic activation systems of complement. Annu Rev Biochem 1981; 50: 433–464. 33 Cohn Z. The role of proteases in macrophage physiology. In: Reich E, Rifkin DB, Shaw E (eds). Proteases and Biological Control. Cold Spring Harbr Press: Cold Spring Harbor, 1975. 34 Werb Z. Proteases and matrix degradation. In: Kelly WN, Harris ED, Sledge RS (eds). Textbook of Rheumatology. Saunders: Philadelphia, 1993. 35 Wang J, Tan NS, Ho B, Ding JL. Modular arrangement and secretion of a multidomain serine protease. Evidence for involvement of proline-rich region and N-glycans in the secretion pathway. J Biol Chem 2002; 277: 36363–36372. 36 Wang LH, Ho B, Ding JL. Transcriptional regulation of limulus Factor C: repression of an NF B motif modulates its responsiveness to bacterial lipopolysaccharide. J Biol Chem 2003; 278: 49428–49437. 37 Saravanan T, Weise C, Sojka D, Kopacek P. Molecular cloning, structure and bait region splice variants of alpha2-macroglobulin from the soft tick Ornithodoros moubata. Insect Biochem Mol Biol 2003; 33: 841–851.

38 Iwaki D, Kawabata S, Miura Y, Kato A, Armstrong PB, Quigley JP et al. Molecular cloning of Limulus alpha 2macroglobulin. Eur J Biochem 1996; 242: 822–831. 39 Armstrong PB, Quigley JP. Proteinase inhibitory activity released from the horseshoe crab blood cell during exocytosis. Biochim Biophys Acta 1985; 827: 453–459. 40 Nakamura T, Tokunaga F, Morita T, Iwanaga S. Interaction between lipopolysaccharide and intracellular serine protease zymogen, factor C, from horseshoe crab (Tachypleus tridentatus) hemocytes. J Biochem (Tokyo) 1988; 103: 370–374. 41 Donovan MA, Laue TM. A novel trypsin inhibitor from the hemolymph of the horseshoe crab Limulus polyphemus. J Biol Chem 1991; 266: 2121–2125. 42 Golino P. The inhibitors of the tissue factor: factor VII pathway. Thromb Res 2002; 106: V257–V265. 43 Miura Y, Kawabata S, Wakamiya Y, Nakamura T, Iwanaga S. A limulus intracellular coagulation inhibitor type 2. Purification, characterization, cDNA cloning, and tissue localization. J Biol Chem 1995; 270: 558–565. 44 Richards RC, O’Neil DB, Thibault P, Ewart KV. Histone H1: an antimicrobial protein of Atlantic salmon (Salmo salar). Biochem Biophys Res Commun 2001; 284: 549–555. 45 Augusto LA, Decottignies P, Synguelakis M, Nicaise M, Le Marechal P, Chaby R. Histones: a novel class of lipopolysaccharide-binding molecules. Biochemistry 2003; 42: 3929–3938. 46 Arai H, Koizumi H, Aoki J, Inoue K. Platelet-activating factor acetylhydrolase (PAF-AH). J Biochem (Tokyo) 2002; 131: 635–640. 47 Tjoelker LW, Stafforini DM. Platelet-activating factor acetylhydrolases in health and disease. Biochim Biophys Acta 2000; 1488: 102–123. 48 Montague JW, Hughes Jr FM, Cidlowski JA. Native recombinant cyclophilins A, B, and C degrade DNA independently of peptidylprolyl cis–trans-isomerase activity. Potential roles of cyclophilins in apoptosis. J Biol Chem 1997; 272: 6677–6684. 49 Matsuda S, Koyasu S. Mechanisms of action of cyclosporine. Immunopharmacology 2000; 47: 119–125. 50 Clow LA, Gross PS, Shih CS, Smith LC. Expression of SpC3, the sea urchin complement component, in response to lipopolysaccharide. Immunogenetics 2000; 51: 1021–1033. 51 Underhill DM, Ozinsky A. Phagocytosis of microbes: complexity in action. Annu Rev Immunol 2002; 20: 825–852. 52 Nauta AJ, Daha MR, van Kooten C, Roos A. Recognition and clearance of apoptotic cells: a role for complement and pentraxins. Trends Immunol 2003; 24: 148–154. 53 Khush RS, Leulier F, Lemaitre B. Drosophila immunity: two paths to NF-kappaB. Trends Immunol 2001; 22: 260–264. 54 Kim DH, Feinbaum R, Alloing G, Emerson FE, Garsin DA, Inoue H et al. A conserved p38 MAP kinase pathway in Caenorhabditis elegans innate immunity. Science 2002; 297: 623–626. 55 Solon E, Gupta AP, Gaugler R. Signal transduction during exocytosis in Limulus polyphemus granulocytes. Dev Comp Immunol 1996; 20: 307–321. 56 Ariki S, Koori K, Osaki T, Motoyama K, Inamori K, Kawabata S. A serine protease zymogen functions as a pattern-recognition receptor for lipopolysaccharides. Proc Natl Acad Sci USA 2004; 101: 953–958. 57 Gruszynski AE, DeMaster A, Hooper NM, Bangs JD. Surface coat remodeling during differentiation of Trypanosoma brucei. J Biol Chem 2003; 278: 24665–24672. 58 Crabtree GR, Clipstone NA. Signal transmission between the plasma membrane and nucleus of T lymphocytes. Annu Rev Biochem 1994; 63: 1045–1083. 59 Bergner A, Muta T, Iwanaga S, Beisel HG, Delotto R, Bode W. Horseshoe crab coagulogen is an invertebrate protein with a nerve growth factor-like domain. Biol Chem 1997; 378: 283–287. 60 Mizuguchi K, Parker JS, Blundell TL, Gay NJ. Getting knotted: a model for the structure and activation of Spatzle. Trends Biochem Sci 1998; 23: 239–242. 61 Grassme H, Jendrossek V, Gulbins E. Molecular mechanisms of bacteria induced apoptosis. Apoptosis 2001; 6: 441–445.

Genes and Immunity

Transcript profiling of immune responsive genes JL Ding et al

18 62 Newmeyer DD, Ferguson-Miller S. Mitochondria: releasing power for life and unleashing the machineries of death. Cell 2003; 112: 481–490. 63 Dussmann H, Kogel D, Rehm M, Prehn JH. Mitochondrial membrane permeabilization and superoxide production during apoptosis. A single-cell analysis. J Biol Chem 2003; 278: 12645–12649. 64 Brunori M, Wilson MT. Electron transfer and proton pumping in cytochrome oxidase. Biochimie 1995; 77: 668–676. 65 Lavrov DV, Boore JL, Brown WM. The complete mitochondrial DNA sequence of the horseshoe crab Limulus polyphemus. Mol Biol Evol 2000; 17: 813–824. 66 Hasnain SE, Begum R, Ramaiah KV, Sahdev S, Shajil EM, Taneja TK et al. Host–pathogen interactions during apoptosis. J Biosci 2003; 28: 349–358. 67 Bogdan C, Rollinghoff M, Diefenbach A. Reactive oxygen and reactive nitrogen intermediates in innate and specific immunity. Curr Opin Immunol 2000; 12: 64–76. 68 Hauptmann N, Grimsby J, Shih JC, Cadenas E. The metabolism of tyramine by monoamine oxidase A/B causes oxidative damage to mitochondrial DNA. Arch Biochem Biophys 1996; 335: 295–304. 69 Malorni W, Giammarioli AM, Matarrese P, Pietrangeli P, Agostinelli E, Ciaccio A et al. Protection against apoptosis by monoamine oxidase A inhibitors. FEBS Lett 1998; 426: 155–159. 70 Duan H, Wang Y, Aviram M, Swaroop M, Loo JA, Bian J et al. SAG, a novel zinc RING finger protein that protects cells from

71 72 73 74 75 76 77

78 79

apoptosis induced by redox agents. Mol Cell Biol 1999; 19: 3145–3155. Nappi AJ, Vass E, Frey F, Carton Y. Superoxide anion generation in Drosophila during melanotic encapsulation of parasites. Eur J Cell Biol 1995; 68: 450–456. Wilce MC, Parker MW. Structure and function of glutathione S-transferases. Biochim Biophys Acta 1994; 1205: 1–18. Chen J, Berry MJ. Selenium and selenoproteins in the brain and brain diseases. J Neurochem 2003; 86: 1–12. Hirota K, Nakamura H, Masutani H, Yodoi J. Thioredoxin superfamily and thioredoxin-inducing agents. Ann NY Acad Sci 2002; 957: 189–199. Nishinaka Y, Masutani H, Nakamura H, Yodoi J. Regulatory roles of thioredoxin in oxidative stress-induced cellular responses. Redox Rep 2001; 6: 289–295. Soti C, Csermely P. Molecular chaperones and the aging process. Biogerontology 2000; 1: 225–233. Altschul SF, Madden TL, Schaffer AA, Zhang J, Zhang Z, Miller W et al. Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res 1997; 25: 3389–3402. Anderson I, Brass A. Searching DNA databases for similarities to DNA sequences: when is a match significant? Bioinformatics 1998; 14: 349–356. Endege WO, Steinmann KE, Boardman LA, Thibodeau SN, Schlegel R. Representative cDNA libraries and their utility in gene expression profiling. Biotechniques 1999; 26: 542–548, 550.

Supplementary information accompanies the paper on Genes and Immunity website (http://www.nature.com/gene).

Genes and Immunity