Expression and distribution of penaeidin antimicrobial peptides are regulated by haemocyte reactions in microbial challenged shrimp Marcelo Mun ˜ oz 1 , Franck Vandenbulcke 2 , Denis Saulnier 3 and Evelyne Bache ` re 1 1 IFREMER/CNRS/Universite ´ de Montpellier, ÔDe ´ fense et Re ´ sistance chez les Inverte ´ bre ´ s MarinsÕ, Montpellier, France; 2 Laboratoire d’Endocrinologie des Anne ´ lides, Groupe de Neuroimmunite ´ des Hirudine ´ es, Universite ´ des Sciences et Technologies de Lille, France; 3 IFREMER, Centre Oce ´ anologique du Pacifique, Taravao, Tahiti, Polyne ´ sie Franc¸ aise Penaeidins are a family of antimicrobial peptides constitu- tively produced and stored in the haemocytes of penaeid shrimp. In response to microbial stimulation, they are released into the blood circulation and they further attach to shrimp cuticle surfaces through a chitin-binding property. In the present paper, we have analysed their expression, regu- lation and distribution in shrimp tissues in response to experimental microbial challenge. We have shown that penaeidinmRNAandproteinarerestrictedtogranular haemocytes and that their expression and distribution are regulated through dramatic changes in haemocyte popula- tions, both circulating and infiltrating shrimp tissues. Two distinct phases in the immune reactions were evidenced: (a) a migration of haemocytes towards the infection site within the first 12 h following microbial injection, with a local and massive release of peptides; (b) the appearance into the blood circulation and tissues of a haemocyte population displaying increased penaeidin-transcriptional activity, which may correspond to a systemic reaction involving haemocyte proliferation process. Finally, in vitro confrontation of hae- mocytes and bacteria revealed that penaeidins are released from granular haemocytes by a novel phenomenon of intracellular degranulation, probably followed by the lysis of the cells. Furthermore, penaeidins were shown covering bacterial surfaces suggesting that the peptides could be involved in opsonic activity. Penaeidin-positive bacteria were observed to be phagocytosed mainly by hyaline cells, a population that does not express penaeidins. Keywords: antimicrobial peptide; crustacea; innate immu- nity; penaeid shrimp; phagocytosis. Antimicrobial peptides are major components of innate immunity that have been conserved in evolution and found in different phyla of the plant and animal kingdom. Although these immune effectors share common character- istics (small size and cationic character) and similarities in structural patterns or motifs [1], one striking feature is their great diversity in terms of amino acid sequences, anti- microbial activities and modes of action. Moreover, depending on their distribution, antimicrobial peptide expression appears to be regulated by different tissue- specific pathways [2] and these effectors may consequently participate in either a local or a systemic reaction. Antimi- crobial peptides are produced in phagocytic cells of vertebrates [3] and invertebrates [4–6], and in various tissues such as epithelia of mammals and insects [7,8], or insect fat body [9]. Peptides are produced constitutively and stored in circulating cells, where they can act intracellularly against phagocytosed microorganisms as shown in human for defensins [3] and in a bivalve mollusc for mytilin [6]. Peptides can also be released by exocytosis upon microbial stimulation [5,10]. In various epithelia of invertebrates [2] and vertebrates [11], antimicrobial peptides are either produced constitutively or induced in response to infection or inflammation, and participate in a local antimicrobial reaction. Finally, antimicrobial peptide expression in fat body cells is induced in response to infection and peptides are secreted into body fluids, which characterizes the acute or systemic reaction in insects [12]. In Crustacea, penaeidins are a unique family of antimicrobial peptides originally isolated and characterized in the shrimp Penaeus vannamei. In previous works, three members of the penaeidin family, penaeidin (Pen)-1, -2 and -3 were purified in their mature and active form (5.48–6.62 kDa) and cloned from the haemocytes of experimentally uninfected shrimp [13]. Penaeidins were shown to be constitutively expressed in haemocytes and mature peptides were localized in the cytoplasmic granules of the granular haemocyte population of unchallenged animals. Regarding penaeidin gene expression and peptide distribution, first data suggested that in response to a microbial challenge, penaeidin transcription is not up-regulated in shrimp haemocytes, but relative penaeidin concentrationinshrimpplasmawasshowntoincrease upon stimulation [14]. Penaeidins, which present in their amino acid sequences a chitin-binding motif [15] were demonstrated to bind to shrimp cuticle surfaces in response to microbial challenges [14]. Thus, we speculated Correspondence to E.Bache ` re, UMR 5098, ÔDe ´ fense et Re ´ sistance chez les Inverte ´ bre ´ sMarinsÕ, CC 80, 2 place Euge ` ne Bataillon – 34095 Montpellier, France. Fax:+33467144622,Tel.:+33467144710, E-mail: ebachere@ifremer.fr Abbreviations: DIG, digoxigenin; NGS, normal goat serum; ISH, in situ hybridization; ICC, immunocytochemistry. (Received 17 December 2001, revised 9 April 2002, accepted 16 April 2002) Eur. J. Biochem. 269, 2678–2689 (2002) Ó FEBS 2002 doi:10.1046/j.1432-1033.2002.02934.x that penaeidin could be released from granular haemo- cytes by regulated exocytosis as demonstrated previously for the antimicrobial peptide-mediated immune response in Limulus [5]. The purpose of the current study was to define the regulation and distribution of penaeidin expression in shrimp during immune response considering Pen-3, the most abundant and representative member of the family [13,16]. We demonstrate that penaeidins are expressed exclusively in shrimp haemocytes and that experimental microbial infection induces great changes in haemocyte populations. Through in situ hybridization and immuno- histochemical analyses, haemocytic reactions were high- lighted as an important component of the immune response ) involved in the distribution of the antimicro- bial peptides. Finally, to define the cellular mechanisms of penaeidin release, haemocytes were challenged with bac- teria in vitro, which gave new insights into haemocyte functions and involvement of penaeidins in shrimp defence. MATERIALS AND METHODS Animals and immune challenge Juvenile shrimp (8–10 g) P. vannamei (Crustacea, Deca- poda) in intermoult stage were obtained from a farm in the province of Guayas (Ecuador) and from the French Polynesia IFREMER laboratory. Shrimp microbial challenge was performed by injecting, into the last abdo- minal segment, a suspension (50 lL; 10 8 cells/animal) of heat-killed (100 °C, 10 5 Pa, 20 min) microorganisms including bacteria, Aerococcus viridans, Vibrio alginolyt- icus and fungal spores of Fusarium oxysporum.Haemo- lymph and tissues were collected at different times (from 0 to 72 h) post-injection as described previously [14]. Unchallenged shrimps (i.e. shrimp at time 0 h) were used as controls. Northern blot analyses Penaeidin-specific and ribosomal probes were amplified by PCR on, respectively, pen-3a cDNA clone (GenBank accession number Y14926) and an 18 S rRNA genomic DNA clone (a gift from T. Spears, Florida State University, USA) as described previously [14]. The probes were radiolabelled by random priming using the Ready-to-go DNA labelling kit (Amersham Pharmacia Biotech). Total RNA from shrimp haemocytes and tissues was prepared according to the method of Trizol reagent (BRL Life technologies). Two or 10 lg total RNA were fraction- ated on denaturating 1% agarose gel containing 17% formaldehyde, and then transferred to Hybond-N filter membranes (Amersham Pharmacia Biotech) by vacuum blotting. Membranes were hybridized at 55 °Cfor12 hwith 32 P-labelled pen-3a cDNA fragment in a solution containing 50% formamide, 5 · NaCl/Cit, 8 · Denhardt’s solution, 50 m M sodium phosphate pH 6.5, 0.1% SDS and 100 lgÆmL )1 denatured salmon sperm DNA. Filters were washed twice in 2 · NaCl/Cit, 0.1% SDS at room tem- perature and twice in 1 · NaCl/Cit, 0.1% SDS, first at room temperature, then at 65 °C followed by autoradiography. After stripping, the membranes were hybridized under identical conditions with 32 P-labelled 18 S ribosomal DNA probe and subjected to further autoradiography. Penaeidin transcript and 18 S rRNA signals were quantified using the STORM TM system (Molecular Dynamics). Tissue preparation for histology Tissues from juvenile shrimp were fixed in a solution containing 22% formalin, 31.5% ethanol and 11.5% glacial acetic acid. After dehydration, tissues were embedded in Paraplast and 8 lm sections were cut, mounted on poly L -lysin coated slides and stored at 4 °C until use. Haemolymph was collected under 1 vol. anti-aggregant modified Alsever solution buffer [14]. Then, cells were fixed for 10 min by addition of 1 vol. ice-cold 4% paraformal- dehyde in 100 m M NaCl/P i containing 10% saccharose. Cells were centrifuged on slides for 5 min at 200 g in a cytospin (Cyto-tek centrifuge, Miles Scientific) and stored at )20 °C until use. Phagocytosis assay Haemolymph was collected as described above and imme- diately centrifuged (800 g, 10 min). Supernatant was elim- inated and haemocytes were incubated at room temperature with bacteria (V. alginolyticus)attheratioof20bacteriaper haemocyte in modified Hanks’ balanced salt solution supplemented with 6 m M CaCl 2 and 13 m M MgCl 2 .At various incubation times (0, 1, 3, 5, 20, 30, 45 and 60 min), cells were fixed and treated for ultrastructural analyses and immunodetection as described below. In situ hybridization Probes. A plasmid containing pen-3a cDNA (GenBank accession number Y14926) was used as template for the preparation of the probes. Digoxigenin (DIG)-UTP- labelled and [ 35 S]UTP-labelled antisense and sense ribo- probes were generated from linearized cDNA plasmids by in vitro transcription using RNA labelling kits, T3 RNA polymerase (Roche) and [ 35 S]UTP (Amersham). Hybridization. DIG-labelled riboprobes ( 40–100 ng per section) were hybridized to tissue sections as described previously [17]. For cytocentrifuged cells, the protocol of hybridization was adapted, i.e. cytocentrifuged cells were incubatedfor10minin100m M glycine, 200 m M Tris/ HClpH7.4,immersed5mininNaCl/P i and fixed in 100 m M phosphate buffer containing 4% paraformalde- hyde and 5 m M MgCl 2 . After the postfixation step, cell preparations were washed 5 min in phosphate buffer, incubated for 10 min in 0.25% anhydride acetic prepared in 100 m M triethanolamine pH 8, and briefly washed in 2 · NaCl/Cit. Samples were then rinsed in distilled water, dehydrated by graded alcohol and air dried at room temperature. DIG-labelled riboprobes (40–100 ng per slide) and 35 S-labelled riboprobes (100 ng or 1 · 10 6 c.p.m. per slide) were diluted in hybridization buffer containing 50% form- amide, 10% dextran sulfate, 10 · Denhardt’s solution, 0.5 mgÆmL )1 tRNA from Escherichia coli,100m M dithio- threitol and 0.5 mgÆmL )1 salmon sperm DNA. Hybridiza- tion was carried out overnight at 55 °C in a humid chamber. Ó FEBS 2002 Antimicrobial peptide expression in shrimp (Eur. J. Biochem. 269) 2679 Slides were then washed twice (2 · 15 min) with 2 · NaCl/ Cit, treated with RNase A (20 mgÆmL )1 in 2 · NaCl/Cit) for 10 min at 37 °C and consecutively rinsed 2 · 10 min in 0.1 · NaCl/Cit containing 0.07% 2-mercaptoethanol at 55 °C. The probes labelled with DIG-UTP were revealed using alkaline phosphatase-conjugated antibodies as des- cribed previously [17]. Detection and quantification of the 35 S-labelled probes After hybridization step, the slides were rinsed in 0.1 · NaCl/Cit at 20 °C, briefly immersed in graded alcohol and air dried at room temperature. Hybridization signal was visualized using autoradiography. Samples were coated by dipping in LM1 Amersham liquid emulsion, immediately dried and exposed for a 4-day period. At the end of the exposure period, the autoradiograms were developed in D19b (Kodak), fixed in 30% sodium thiosulfate (10 min at room temperature), stained with 1% Toluidine blue and mounted with Xam (Merck). Quantification of the radio- labelling at the cellular level was performed using an Axiophot Zeiss microscope and a Biocom quantification system as established [18]. Controls Control for in situ hybridization consisted in replacing antisense riboprobe with sense riboprobe. RNase control sections were obtained by adding a preincubation step with 10 lgÆmL )1 RNase A prior to hybridization. Immunodetection of penaeidins Whole animal. Eight micrometer-thick paraffin sections were re-hydrated and treated as follow: (a) 10 min at 20 °C in 150 m M NaCl, 100 m M Tris/HCl pH 7.4 buffer (NaCl/ Tris); (b) NaCl/Tris containing 1% normal goat serum (NGS), 1% BSA (NaCl/Tris/NGS/BSA) and 0.1% Triton X-100, 30 min at room temperature; (c) incubation with anti-penaeidin IgG (3 lgÆmL )1 ) diluted in NaCl/Tris/NGS/ BSA, overnight at room temperature; (d) 3 · 10 min in NaCl/Tris; (e) 1 nm colloidal gold-labelled goat anti-rabbit IgG (Amersham) diluted 1 : 100 in the incubation buffer, 3 h at room temperature; (f) 3 · 10 min in NaCl/Tris; (g) equilibration 2 · 5 min in 200 m M citrate buffer pH 7.4; (h) silver amplification performed with the IntensSE TM kit according to the manufacturer’s instructions (Amersham), 12 min at 20 °C; (i) 2 · 2 min in distilled water. Then, paraffin sections were mounted in XAM (Merck) and observed using a Zeiss Axioskop light microscope. Immuno- dection was also performed by using Texas red-tagged goat anti-rabbit serum (Jackson Immunoresearch) as described below. Circulating haemocytes. Cytocentrifuged haemocytes were equilibrated for 10 min in NaCl/Tris before perme- abilization with 0.1% Triton X-100 in NaCl/Tris for 30 min at room temperature. One hour preincubation was performed in the presence of 1% NGS and 1% BSA to block nonspecific antibody binding. Rabbit anti-penaeidin polyclonal antibody purified IgG (1.5 lgÆmL )1 ) [14], was applied for 12–16 h at room temperature in NaCl/Tris/ NGS/BSA. After washing three times (10 min) in NaCl/ Tris, cells were incubated for 2 h at room temperature with 1 : 100 Texas red-tagged goat anti-rabbit antiserum (Jackson Immunoresearch). The slides were washed 3 · 10 min in NaCl/Tris, mounted in glycerol containing 25% NaCl/Tris and 0.1% p-phenylenediamine and exam- ined using a laser scanning microscope (TCS NT) equipped with a Leica (DMIRBE, Inc.) inverted micro- scope and an argon/krypton laser. Texas red signal was detected by exciting samples at 568 nm. Images were acquired as single transcellular optical sections and averaged over 16 scans per frame. Positive or negative cells were subsequently counted. Controls were incubations of anti-penaeidin IgG pread- sorbed by purified recombinant penaeidin-3 [19]. Electron microscopy Ultrastructural microscopy. After haemolymph collection under 1 vol. modified Alsever solution buffer, cells were fixed for 1 h at 4 °Cin0.1 M NaCl/P i pH 7.4, containing 2% glutaraldehyde, 4% paraformaldehyde and 10% sucrose. Cell pellets were obtained by 10 min centrifuga- tion at 800 g. The pellets were rinsed in NaCl/P i , postfixed in 1% OsO 4 for 1 h, dehydrated in graded acetone solutions and embedded in Embed 812 Kit. Ultrathin sections (80–90 nm thick) were cut from the blocks, collected onto 200 mesh copper grids, double-stained with uranyl acetate and lead citrate and examined with a Jeol JEM 100 CX. Immunogold labelling. Immunogold detection of penaei- dins was performed on circulating cells but also on tissues. Haemocytes and dissected tissues were fixed for 1 h at 4 °C in a mixture of 4% paraformaldehyde, 1% glutaraldehyde, 10% sucrose in 100 m M NaCl/P i , pH 7.4. Cells and tissues werepostfixedin1%OsO 4 for 3–5 min and dehydrated in graded alcohol before embedding in LR white (TAAB Laboratories). Semi-thin sections (1 lm thick) were collected on alcohol- washed glass slides, and penaeidin immunostaining was performed using a gold-tagged secondary antibody and silver amplification as described above. Ultrathin (90 nm-thick) sections from embedded pellets or tissues were collected on nickel grids. Sections were treated 8 min in 10% H 2 O 2 , 10 min in distilled water, 30 min in NaCl/Tris/NGS/BSA and then incubated for 36 h at 4 °Cwith3lgÆmL )1 rabbit anti-penaeidin IgGs in NaCl/Tris/NGS/BSA. Grids were washed three times for 10 min with NaCl/Tris/NGS/BSA and incubated for 2 h at room temperature in 10 nm colloidal gold-labelled goat anti-rabbit IgGs (Amersham) diluted 1 : 100 in NaCl/ Tris/NGS/BSA. Grids were then washed three times for 10 min with NaCl/Tris, postfixed for 3 min in NaCl/Tris containing 1% glutaraldehyde and washed twice for 5 min with distilled water. Sections were stained for 15minwith2.5%uranylacetateandexaminedwitha Jeol JEM 100 CX. Statistical analyses The data were analysed using Fisher PLSD test (P < 0.05) at 95% confidence level with STATVIEW SE + GRAPHICS TM program. 2680 M. Mun ˜ oz et al. (Eur. J. Biochem. 269) Ó FEBS 2002 RESULTS Tissue localization of penaeidins in nonstimulated animals Penaeidins are known to be constitutively expressed in shrimp haemocytes and penaeidin transcripts were also detected by Northern blot in different tissues of nonstimu- lated animals [14]. In the present paper, the origin of penaei- din mRNA and peptide localization in shrimp tissues were determined by both in situ hybridization analyses (ISH) using pen-3 antisense and sense RNA probes and immunocyto- chemistry (ICC) at optical and electron microscopy levels. Among the different tissues analysed, penaeidin mRNAs were detected in circulating haemocytes in blood vessels and sinuses and in cells present within most tissues. The shape of the positive cells suggests that they are infiltrating haemo- cytes. A high number of cells containing penaeidin tran- scripts was detected in heart and epigastric haematopoietic nodule (also named lymphoid organ) (Fig. 1A), in blood vessels irrigating gills and hepatopancreas (Fig. 1B and C), and to a lesser extent in all the shrimp tissues such as haematopoietic tissue (Fig. 1D), brain, subcuticular epithe- lia or midgut caecum (data not shown). According to penaeidin sense probe hybridization used as control, for which no signal was observed (Fig. 1E and F), the detection of penaeidin transcripts with antisense riboprobe was shown to be specific for the tissues analysed. In addition, pretreat- ment of sections with RNaseA before hybridization abol- ished the positive staining providing further evidence of the signal specificity (data not shown). Antibody used in this study was a rabbit antiserum directed against recombinant Pen-3a [14]. The high degree of homology between the different penaeidin forms [13] implies that the antibody recognizes different isoforms. Consequently, we qualified any immune positive signal as related to the presence of penaeidins. When the specific anti- penaeidin antibodies were preincubated with purified recombinant penaeidins [19], penaeidin immunostaining was no longer observed providing evidence of the specificity of the reaction (data not shown). Regarding penaeidin distribution, the peptides were shown to be localized in circulating haemocytes but also in cells located in gills, heart, brain, subcuticular epithelium, epigastric haemato- poietic nodule, midgut, midgut caecum and muscle, where strong labelling was observed (Fig. 2A, B and C). In order Fig. 2. Immunodetection of penaeidins in tissue sections of nonstimulated shrimp. Positive cells (arrows)areshownonsemithinsectionsof midgut (A), midgut caecum (B) and in muscle (C). Ultrastructural distribution of penaeidin immune reactivity was performed using a 10-nm gold particle-conjugated secondary antibody. As shown in intestine, numerous gold particles are present in electron dense granules of two subtypes of infiltrating hae- mocytes, namely large-granule haemocytes (D, E) and small-granule haemocytes (F, G). No labelling was seen throughout cytoplasm and nucleus. bl, Basal lamina; ep, epithelia; mf, muscular fibers; n, nucleus; star, lumen of the intestine. Bar ¼ 10 lm(A,B,C),1lm (D,E,FG). Fig. 1. Detection of penaeidin mRNA in non- stimulated shrimp tissues by in situ hybridiza- tion. Labelling appears in most tissues and is particularly obvious in epigastric haemato- poietic nodule (A), gills (B), hepatopancreas (C) and haematopoietic tissue (D). The shape of the positive cells evokes haemocytes (arrows), infiltrating tissue (A, D) or free- haemocytes in blood vessels (B, C). In a neg- ative control consisting of sections hybridized with pen-3a sense riboprobes, no labelling was observed as shown for gills (E) and hepato- pancreas (F). gf, Gill filaments; dt, digestive tubule. Bar ¼ 10 lm. Ó FEBS 2002 Antimicrobial peptide expression in shrimp (Eur. J. Biochem. 269) 2681 to confirm the localization of penaeidins and to determine the nature of the positive cells, immunogold labelling was performed. Penaeidin storage was confirmed to be restricted to granular haemocytes, with large granules or small granules, located and infiltrating all tissues analysed such as brain, subcuticular epithelia, epigastric haematopoietic nodule or midgut (Fig. 2D, E, F and G). The presence of some infiltrating haemocytes without labelling was also found confirming previous data about the presence of different haemocyte populations, expressing vs. not expres- sing penaeidins [14]. Microbial stimulation induces changes in the total number of circulating haemocytes, and in the population of haemocytes expressing penaeidins Previous work showed that microbial challenge induces a decrease of penaeidin mRNA concentration in circulating haemocytes in the first hours following stimulation [14]. In order to define the regulation of penaeidin transcription, we analysed time-course changes in total circulating haemocyte number and haemocyte penaeidin mRNA levels, occurring in response to stimulation. In two independent experiments, shrimp were challenged by injection of heat-killed micro- organisms and haemolymph was collected from five individual animals at different times (0, 6, 12, 48 and 72 h) following injection. A strong decrease in haemocyte total number (from 9 · 10 6 ±7· 10 6 cells to 1.2 · 10 6 ±1.4· 10 6 cells) was observed in the first 12 h following injection, with a significant difference (P < 0.05) at 6 h between stimulated and nonstimulated animals. The number of total haemocytes returned to levels observed for unchallenged animals at 48 h and a significant increase (up to 19.8 · 10 6 ±3· 10 6 haemocytes; P <0.05) of total haemocyte number was observed at our last time point (72 h poststimulation) (Fig. 3). In similar experiments, total RNA was extracted from the circulating haemocytes of 10 animals at the same intervals after injection, and 2 lgof total RNA were analysed by Northern blot. The STORM TM quantifications of penaeidin mRNA and ribosomal hybrid- ization signals were compared at each time post-injection. Analyses revealed a strong decrease in penaeidin mRNA levels for the first 12 h and a return to nonstimulated animal levels at 48 h post-challenge. A threefold increase in penaeidin mRNA levels was noticed at 72 h following challenge (Fig. 3). To better understand such a decrease in penaeidin transcript concentration within the circulating haemocytes after microbial challenge, Northern blot analyses were performed on total RNA extracted from a constant number of haemocytes (1 · 10 6 cells for each individual) at every time post-challenge, instead of constant total RNA quantity (2 lg). Hybridization signals obtained, respectively, for pen-3a transcripts and 18 S rRNA probes were quantified by STORM and analysed separately. Data analysis revealed an important individual variation in both pen-3a transcripts and 18 S rRNA signals with a decrease in pen-3a transcript levels and constant average values with 18 S rRNA during the first 12 h post-challenge (data not shown). However, at 48 h post-challenge, penaeidin mRNA levels appeared to increase slightly and hybridization signals with 18 S rRNA probes were dramatically stronger than those observed for unchallenged animals (Fig. 4). In order to determine whether changes in penaeidin transcript and protein levels could be also associated with changes in the composition of circulating haemocyte populations, the percentage of circulating haemocytes expressing and storing penaeidin was further analysed by ISH and ICC, respectively. Circulating haemocytes from five individual shrimp were collected, counted, fixed and cytocentrifuged on slides at different times (0, 6, 12, 48 and 72 h) after microbial stimulation. As shown before, signi- ficant modifications in total circulating haemocyte number were observed in these experiments (Fig. 5). In nonstimu- lated animals 35 ± 6% of the total haemocyte population expressed penaeidins. This percentage decreased to 19% (± 8%) and 13% (± 8%), respectively, 6 and 12 h after microbial challenge (Fig. 5). Then, the percentage of penaeidin mRNA-positive haemocytes in the total circula- ting population reached 50 ± 3% at 48 h post-challenge, before returning slightly to a mean percentage (39 ± 11%) close to that observed for nonstimulated animals at 72 h Fig. 3. Time-course analysis of total haemocyte number and penaeidin expression (histograms) in circulating haemocytes after microbial chal- lenge. Haemocyte counts were performed on five shrimps at different time intervals after challenge using an haemocytometer. Vertical bars represent mean values of haemocyte numbers at each time point (line). Northern blot analyses were performed on 2 lg of a pool of total RNA extracted from 10 shrimps at each time point. Hybridization signals obtained with 32 P-labelled pen-3a cDNA probe were quantified by the STORM TM system and compared to those obtained with the 18 S rRNA specific probe. The penaeidin/18 S rRNA signal ratios were calculated and the expression level in unchallenged shrimp was normalized to 100. Results are given as percentage expression relative to this level. Fig. 4. Northern blot analysis of total RNAs from constant number of haemocytes from unchallenged shrimp and shrimp 48-h post-challenge. Total RNA was extracted from 1 · 10 6 haemocytes per shrimp, unchallenged (lanes 1–5) and 48 h following challenge (lanes 6–9) and hybridized successively with 32 P-labelled probes specific for pen-3a (top) and specific for 18 S rRNA (bottom). Strong hybridization sig- nals are observed with the 18 S rRNA probe at 48 h post-challenge compared to those observed for unchallenged shrimp. 2682 M. Mun ˜ oz et al. (Eur. J. Biochem. 269) Ó FEBS 2002 post-injection (Fig. 5). Regarding storage of the peptides, the percentage of penaeidin-immunoreactive haemocytes was also established. In nonstimulated animals, the relative number of haemocytes storing penaeidins was similar to that of haemocytes expressing the peptides (37 ± 4% of the total circulating population). At the different times post- challenge, changes similar to those observed with transcript detection occurred in the percentages of penaeidin-positive haemocytes (Fig. 5). During the first 6 and 12 h post- challenge, the percentage of penaeidin-positive haemocytes decreased, respectively, to 24 (± 4%) and 17% (± 4%) of the total number of circulating haemocytes, and increased thereafter to 45 ± 6% (48 h sampling point) (Fig. 5). However, at 72 h post-stimulation, the percentage of penaeidin-immunoreactive haemocytes decreased dramatic- ally to 19 ± 2% of the total circulating population, a percentage inferior to that of haemocytes expressing penaeidin observed at the same time. This last result indicated that, at 72 h post-injection, circulating haemo- cytes display differences both in their penaeidin transcrip- tion activity and their storage ability (Fig. 5). Microbial stimulation induces an increase in haemocyte penaeidin-transcriptional activity at 48–72 h post-challenge To determine whether penaeidin expression could be transcriptionally regulated at the level of the haemocytes or whether changes in penaeidin transcript rates could be only the result of changes in haemocyte populations, penaeidin mRNA content was quantified at the cellular level. Cytocentrifuged haemocytes, collected from shrimp at 0, 6, 12, 48, 72 h post-injection, were probed with 35 S-radiolabelled penaeidin antisense riboprobes. Twenty- five haemocytes from four individual animals were analysed at each time post-injection. Quantification was expressed as Arbitrary Units (AU) corresponding to the number of silver grains counted for every haemocyte by the autoradiography BIOCOM software. Silver grains are produced by contact of 35 S-emission with the autoradiographic emulsion. The number of grains is proportional to the hybridization signal. Background level was measured and subtracted for each slide. According to the quantification of penaeidin mRNA content in every haemocyte expressing the peptides, two groups of haemocytes were distinguished: one group with AU values < 50 and another group with AU values > 50. In nonstimulated animals (time 0) the percentage of haemocytes with AU values > 50 constituted only 5% of the haemocytes analysed (Fig. 6A and B1). At 6, 12 and 48 h post-stimulation, this percentage increased, respect- ively, to 19, 23 and 34% of haemocytes displaying an AU value > 50 (Fig. 6A). At 72 h after microbial stimulation, significant differences appeared (P < 0.05) and haemocytes with AU values > 50 represented  49% of the total haemocytes analysed (Fig. 6A) revealing an important heterogeneity in penaeidin expression levels within circula- ting cell populations (Fig. 6B2). The increase of the percentage of haemocytes with a high level of penaeidin transcriptional activity is concomitant with a decrease of the relative percentage of circulating haemocytes storing penaeidins (Fig. 6C1 and C2). Localization of penaeidin expression and storage in shrimp tissues after microbial challenge In order to investigate the ability of tissues other than haemocytes to express penaeidins and to study the distri- bution of both transcripts and peptides in response to challenge, shrimp tissues were analysed by Northern blot, ISH and ICC at different times post-injection. For Northern blot analyses, total RNA was extracted from gills, midgut, cephalothorax subcuticular epithelium and brain from 10 shrimps at 0, 3, 6, 12, 24, 48, 72 h post- stimulation. As described previously for haemocytes, STORM TM quantified penaeidin and ribosomal hybridization signals were compared for every tissue at each time post- injection (Fig. 7A). Relative penaeidin mRNA levels dra- matically decreased in all the tissues analysed 6 and 12 h after microbial challenge, and increased thereafter in gills and midgut, at 48 or 72 h post-stimulation, up to the level observed in nonstimulated shrimp (Fig. 7B). However, penaeidin mRNA levels remained low in subcuticular epithelium during the 72 h after the challenge in comparison to that observed for control shrimp (Fig. 7B). ISH analyses of the different tissues confirmed that penaeidin transcripts were confined to haemocytes (Fig. 8). Moreover, these observations revealed that, at 6 and 12 h post-challenge, the decrease in penaeidin mRNA levels observed by Northern blot in tissues could be related to a decrease in the number of haemocytes containing transcripts that infiltrated the tissues (Fig. 8B, E and H). At 48 and 72 h, the number of haemocytes expressing penaeidin in stimulated shrimp tissues appeared to be restored (Fig. 8C, F and I) and was similar to that observed in nonstimulated animals (Fig. 8A, D and G). Fig. 5. Time-course analysis of percentages of haemocytes expressing penaeidins and haemocytes storing penaeidins in the circulating popula- tion after microbial challenge. Circulating cells were harvested from five shrimps at different times post-injection (0, 6, 12, 48, 72 h) and fixed in paraformaldehyde. Total haemocyte numbers were established using a haemocytometer (line). The haemocytes were cytocentrifuged onto slides and analysed by in situ hybridization using antisense pen-3a riboprobes labelled with DIG-UTP and by immunocytochemistry using an anti-penaeidin Ig detected by secondary antibody labelled with Texas red. Immunostaining was observed by confocal micro- scopy. The percentage of haemocytes expressing penaeidin corres- ponds to the ratio between the number of penaeidin riboprobe-positive cells and the total number of haemocytes (open bars). The percentage of haemocytes storing penaeidins corresponds to the ratio between immunopositive cells and the total number of haemocytes (black bars). Four hundred cells per slide and three slides per shrimp were counted and each value represents the mean of five shrimps ± SEM. Ó FEBS 2002 Antimicrobial peptide expression in shrimp (Eur. J. Biochem. 269) 2683 Similar observations were obtained with ICC analyses relative to the distribution of penaeidin-stained haemocytes within tissues and following microbial challenge (data not shown). Haemocyte recruitment and penaeidin localization at the site of injection Injection of microorganisms resulted in a dramatic decrease in numbers of both circulating and tissue infiltrating haemocytes within 3 h of injection. To study haemocyte behaviour and changes, sections of the last abdominal segments (site of injection) were analysed at 3, 6 and 72 h by ICC and ISH. Both penaeidin-producing haemocytes and released peptides were therefore localized. Concerning peptide detection and distribution as studied by ICC, the last abdominal segment of untreated animals appeared totally devoid of immunoreactivity (Fig. 9A). Three h post-challenge, some penaeidin-positive haemocytes were observed together with a slight spread of penaeidin immunostaining near the injection site. However, 6 h post-stimulation, an increased number of haemocytes containing penaeidins was seen not only around the injection sites, but also on surrounding subcuticular epithe- lia. Strong penaeidin immunoreactivity was detected around the injection site revealing the presence of released peptides and their binding to cuticular surfaces close to the injection site (Fig. 9B). Such reactivities were observed up to 72 h post-injection, with an increasing number of penaeidin- positive haemocytes and an accumulation of free peptides into the muscle around the site of injection (Fig. 9C). Concerning penaeidin expression, an accumulation of infiltrating haemocytes containing penaeidin transcripts began also to be seen around the injection site 3 h after stimulation (data not shown). At 6 and 72 h, a high concentration of penaeidin-positive haemocytes was reached around the site of injury when very few positive haemocytes were observed in the muscle of nonstimulated animals or in other parts of the tail of injected shrimps (Fig. 9D and E). Confrontation of haemocytes and bacteria The recruitment of penaeidin-positive haemocytes around the site of injection confirmed the importance of haemocytic reactions in response to microbial challenge. However, the Fig. 6. Changes in penaeidin transcriptional activities and penaeidin storage of circulating haemocytes after microbial challenge. (A) Cytocentrifuged haemocytes were hybridized with antisense pen-3a riboprobes labelled with [ 35 S]UTP. The radiolabelling appears as dark silver deposits. Individual haemocyte titration of the level of expression was performed using a BIOCOM system. Results are expressed in arbitrary 3 units (AU). The level of expression was quantified in 25 cells per slide and five slides per shrimp and each value represents the average of four shrimps. Histograms refer to the percentage of hae- mocytes exhibiting more than 50 of AU (black bars) and the percentage of haemocytes showing less than 50 of AU (open bars). (B) Penaeidin mRNA content in cytocentrifuged haemocytes were visualized by silver grains resulting from the contact of 35 S-emission with autoradiographic emulsion. Silver grains are seen in the haemocytes of nonstimulated ani- mals (B1); comparatively, at 72 h after microbial challenge, stronger signals are observed in some haemocytes (B2). (C) Cytocentrifuged haemocytes were investigated for penaeidin content by immunodetection with Texas red-labelled secondary antibody. Strong immunoreactivity is observed in hae- mocytes from nonstimulated shrimp (C1) whereas at 72 h post-challenge haemocytes display weak penaeidin-immunostaining (C2). Bars ¼ 10 lm. 2684 M. Mun ˜ oz et al. (Eur. J. Biochem. 269) Ó FEBS 2002 strong penaeidin reactivity observed at the site of injury did not allow investigation of the close interaction between the haemocytes and the microorganisms and the role of penaeidins in these reactions. To address this question, in vitro analyses were performed. Haemocytes were incuba- ted in the presence of the bacteria V. alginolyticus,andthen treated at 1, 3, 5, 10, 20, 30, 45 and 60 min. incubation for electron microscopy examination and penaeidin immuno- detection (immunogold labelling). Observations of control haemocytes (t 0 ) confirmed penaeidin localization into cytoplasmic granules of granular haemocytes (Fig. 10A). Haemocytes without granules or with only a few small granules, termed hyaline cells, presented no penaeidin immunoreactivity as described previously [14]. In cell preparations exposed to bacteria for 5 min, haemocytes with granules showed slight penaeidin immunoreactivity in the cytoplasm, when deformations of their cytoplasmic granules began to be observed (Fig. 10B, C). At the same incubation time with bacteria (5 min), extracellular bacteria were not reactive to penaeidin-specific antibody Fig. 7. Time-course analysis of penaeidin expression in shrimp tissues after microbial challenge. Ten micrograms of a pool of total RNAs, extracted from tissues of 10 animals at different times (0, 3, 6, 12, 24, 48, 72 h) following microbial injection, were hybridized successively with pen-3a and 18 S rRNA 32 P-labelled DNA probes. (A) Hybridization profiles of midgut, gills and subcuticular epithelium are shown; penaeidin mRNA levels decrease in all the tissues within hours of challenge and increase again after 12–24 h (B) Hybridization signals were quantified STORM TM and the penaeidin/18 S rRNA signal ratio was determined and normalized to 100 in untreated animals. Results, given as percentage expression relative to this level, show great variations in penaeidin transcript content resulting from microbial challenge. Fig. 8. Detection of penaeidin mRNA by in situ hybridization in shrimp tissues after microbial challenge. Positive haemocytes (arrows) were detected in gills (A, B, C), epigastric haemato- poietic nodule (D, E, F) and hepatopancreas (G, H, I). Positive haemocytes are fairly abundant in the tissues of untreated animals (A, D, G), but almost undetectable in tissues 6 h after microbial injection (B, E, H). At 48 h post-challenge, the distribution of penaeidin- positive cells is restored and is quite similar to that observed in unchallenged shrimp but with more intense labelling of haemocytes (C, F, I). Bars ¼ 10 lm. Ó FEBS 2002 Antimicrobial peptide expression in shrimp (Eur. J. Biochem. 269) 2685 (Fig. 10C, D). In haemocytes incubated with bacteria for 10 min, most of the granules showed gross deformation, such as a lost of round shape and electron density, and retraction within the granule membranes causing star- shaped contours (Fig. 10E, F). Immunoreactivity to penaei- dins was evidenced in the cytoplasm of these haemocytes, suggesting the release of granule content within the cell (Fig. 10F). No evidence of degranulation or exocytosis was found in these experiments. After 20 min incubation, extracellular penaeidin-immunoreactive bacteria were seen in the preparation (Fig. 10G, H). Regarding phagocytosis reactions, internalized bacteria were observed after 20 min incubation mainly into hyaline haemocytes (Fig. 10I). Intracellular phagocytosed bacteria observed in hyaline cells were shown to be immuno-positive to penaeidin (Fig. 10I, J) as well as bacteria not yet phagocytosed, suggesting that they had been covered with released penaeidins before their internalization. To a lesser extent, phagocytosed bacteria were also observed into some granular haemocytes in which neither intracellular lysis of their granules nor fusion of granules with phagocytic vacuoles were seen (Fig. 10K). At the same time, a large number of haemocytes with degenerated cytoplasm and nuclei was observed revealing that a phenomenon of lysis has occurred in response to Vibrio contact (Fig. 10L, M). After 45 min incubation, degenerative haemocytes within the preparation were predominant. DISCUSSION Through investigations on penaeidin expression, the aims of the present study were to define shrimp defence mechanisms in response to microbial infections. We applied in vivo experimental infection model in the shrimp P. vannamei to analyse the expression, regulation and production of penaeidins in circulating haemocytes and tissues of the animals. We previously showed that penaeidins are constitutively produced and stored in granular haemocytes of shrimps that have not been experimentally infected, indicating haemocytes as the main site of production of the peptides [14]. Here, we show that in shrimp tissues, the distribution of penaeidin transcripts and proteins is restricted to haemo- cytes either circulating in blood vessels irrigating tissues such as the brain, hepatopancreas or gills, or infiltrating tissues such as subcuticular epithelia or midgut caecum. Penaeidins are solely present in large-granule haemocytes and small- granule haemocytes (also called semigranular cells), and are absent from the hyaline haemocyte population, devoid of granules. In the haematopoietic tissues, penaeidin tran- scripts were clearly visible in a few cells, showing that penaeidin expression occurs in this tissue. This result differs from those obtained in crayfish where the haematopoietic tissue was found to be negative for prophenoloxidase [20], a gene that is expressed in circulating haemocytes [21]. The haematopoietic tissues have been described in crustacean species [22,23] but knowledge of the haematopoietic process remains limited and few data are available on the expression of immune effectors during haemocyte differentiation and maturation. Our observations suggest that penaeidins are expressed either by maturating stem cells or by haemocytes early before leaving the haematopoietic tissues. However, it cannot be excluded that circulating haemocytes expressing penaeidins may return to infiltrate this tissue for some signalling reaction. In invertebrates, little is known about the regulation and expression of antimicrobial peptide encoding genes during the immune response, apart from insects where transcrip- tion is induced in fat body cells and surface epithelia and for which signalling and regulatory pathways controlling Fig. 9. Haemocyte recruitment at the site of microbial injection. Sections were immuno- stained using anti-penaeidin Ig and secondary antibody labelled with Texas red (A, B, C) and sections were hybridized with antisense pen-3a riboprobes labelled with DIG-UTP (D, E). Thelastabdominalsegmentofanunchal- lenged animal is totally devoid of immuno- reactivity (A). Six hours post-injection, numerous haemocytes storing penaeidin are observed around the injury site (arrow) and an intense immunoreactive signal is also detected throughout the tissue (star) (B). At the same time point, a large number of haemocytes expressing penaeidins are present around the injection site and near the subcuticular epi- thelium (D). Seventy-two h after microbial challenge, a large number of penaeidin-storing (C) and expressing haemocytes (E) is observed throughout surrounding tissue. Bar ¼ 20 lm (A, C, E), 10 lm(B,D). 2686 M. Mun ˜ oz et al. (Eur. J. Biochem. 269) Ó FEBS 2002 peptide expression are particularly well characterized [2,24]. In the bivalve mollusc, Mytilus galloprovincialis, antimicro- bial peptides are constitutively expressed and stored in phagocytic haemocytes where they participate in the destruction of engulfed microorganisms [6]. In Limulus, upon microbial stimulation, antimicrobial peptides are released from haemocytes by regulated exocytosis [25]. In shrimp, as previously shown [14], microbial challenge results in a dramatic drop of penaeidin mRNA concentration (relative to 18 S rRNA) in circulating haemocytes in the early hours post-injection with a return to initial levels at 48 h. However, at 72 h post-injection, penaeidin transcript concentration appears to be threefold higher than that observed in unchallenged shrimp. Similar kinetics (a decrease followed by a significant increase) has been observed in the total number of circulating haemocytes as the result of microbial challenge, a phenomenon already described in other crustacean species [26,27]. From our results and data acquired from Northern blot, ISH and ICC analyses, two distinct phases can be distin- guished in the immune response of shrimp to microbial challenge. During the first phase, corresponding to the first 12 h post-challenge, haemocytes constitutively produce penaeidin mRNA and protein. The decrease of penaeidin detection within total circulating populations is the result of a decrease in penaeidin-expressing haemocytes: they leave the blood circulation and most of the shrimp tissues and migrate towards injured tissues. 2 This is in agreement with previous studies on other crustacean species [28]. Massive accumulation of penaeidin-producing haemocytes was seen around the site of injection 6 h post-injection, as well as a massive release of penaeidin which spread into muscle tissue around the injection site, as a local antimicrobial response. As previously demonstrated, penaeidins are released, upon stimulation, from haemocytes into haemolymph where their concentration increases; subsequently, they bind to cuticle surfaces [14]. During the second phase, at about 48–72 h post-chal- lenge, intense penaeidin-labelling is observed in the tissues surrounding injection site as well as on subcuticular surfaces. Moreover, haemocytes displaying high transcriptional activity appear in the blood circulation as evidenced by Fig. 10. In vitro confrontation of haemocytes with V. alginolyticus. Haemocytes were incubated with bacteria, fixed at different time intervals (0, 1, 3, 5, 10, 20, 30 min) and embedded in resin for penaeidin immunostaining using a 10-nm gold particle-conjugated secondary antibody. (A) In control haemocytes (t 0 ), positive cells exhibit numerous gold particles in electron dense round granules. (B, C) After a 5-min incubation with bacteria, the immunoreactive granules loose their round shape and retraction of the granule membranes is seen (B, arrow). All of the bacteria observed are extracellular and are totally devoid of immunoreactivity (C, D). After 20 min of contact with bacteria, penaeidin-positive granules have star-shaped contours (E, F, G, arrow). At the same time point, immunoreactivity is also observed in the preparation outside haemocytes and on bacteria (H). Internalized penaeidin-positive bacteria are observed to a great extent into hyaline cells (I). In these cells that do not express penaeidin, phagocytosed bacteria appear to be penaeidin immunopositive (J). Granular haemocytes display also phagocytic activity (K). At longer time intervals (20 and 30 min), many cells appear as ghosts (L, M) 4 , probably originating from distinct haemocyte populations. b, Bacteria; n, nucleus; mb, plasma membrane; pg, phagosome. Bars ¼ 1 lm. Ó FEBS 2002 Antimicrobial peptide expression in shrimp (Eur. J. Biochem. 269) 2687 [...]... site of microbial injection precluded both clarification of the mechanisms of penaeidin release from haemocytes, and determination of any potential involvement of penaeidin in the elimination of microorganisms via phagocytosis To further address these questions haemocytes were challenged with Vibrio in vitro Regarding penaeidin release, there was no indication of degranulation of granule-containing penaeidin, ... penaeidin, or any migration of granules towards the cell periphery, in contrast with regulated exocytosis reported in Tachypleus [33] Penaeidin containing haemocytes showed striking changes in the shape and morphology of their granules, suggesting a possible release of granule content within the haemocyte cytoplasm This quite original phenomenon appears to be followed by the lysis of the haemocytes and. .. and haemocyte proliferation processes Penaeidins may be involved in local defence reaction through their release by haemocytes and binding to shrimp cuticle surfaces By their antimicrobial activities against Gram-positive bacteria and fungi [19], penaeidins may protect tissues from infections and/ or participate in wound healing process Penaeidins do not display strong antimicrobial activity against Gram-negative... elimination of internalized microbes It is attractive to assume that these two populations of penaeidin- positive haemocytes can contain different classes of penaeidins with various functions, which are impossible to discriminate with the tools available today In conclusion, the expression and distribution of penaeidins in response to microbial challenge are regulated through haemocyte reactions and haemocyte. .. are comparatively penaeidin- poor, are probably intensively produced and released precociously from haematopoietic tissues Such a phenomenon has already been been proposed for Syciona ingentis during the moulting cycle and after bacterial injection [27,31] Concomitantly, as a result of this proliferative process, a dramatic invasion of haemocyte producing penaeidin mRNA and protein is seen in most of. .. they can contribute to their elimination by phagocytic cells by a potential opsonic function Indeed, extracellular bacteria as well as internalized phagocytosed bacteria were seen to be immunoreactive to penaeidins These observations argue in favour of a coating, by released penaeidin, of the bacteria before their internalization into penaeidin- devoid hyaline cells Finally, the diversity, the large distribution. .. large-granules are minimally phagocytic and internalized bacteria are observed late in this cell population These results give new insights into the identification of haemocyte types and their respective function in crustaceans Indeed, in crab and crayfish, using haemocytes previously separated on Percoll gradient, hyaline cells were considered to be primary phagocytic cells [35,36] However, in penaeid shrimp, granular... transcripts and proteins, as carried out here with penaeidins, will be of great benefit to clarify haemocyte lineage and identification of cell types as well as their functions in immune response Our data suggest that different populations of granular haemocytes may exist: (a) one population involved in a phenomenon of lysis with a massive and early release of penaeidins; and (b) another population involved in. .. phagocytosis of bacteria taking place late than hyaline cell phagocytosis No evidence for discharge of granular penaeidin content into bacteriacontaining phagosomes has been observed in shrimp, as demonstrated in human neutrophils for defensins [38] or in mussel haemocytes for mytilins [17] The question remains about the function of these intracellular penaeidins and their potential involvement in the... assume that penaeidin up-regulation in circulating haemocytes reflects an induced proliferation process, similar to results obtained in P japonicus In this species, an increase in the proliferation rate of circulating haemocytes as a result of in vivo experimental infection with Fusarium was shown by flow cytometry [30] At 72 h post-stimulation, transcriptionally active, young or maturating haemocyte forms, . Expression and distribution of penaeidin antimicrobial peptides are regulated by haemocyte reactions in microbial challenged shrimp Marcelo. Polyne ´ sie Franc¸ aise Penaeidins are a family of antimicrobial peptides constitu- tively produced and stored in the haemocytes of penaeid shrimp. In