Immunogenetics

, Volume 56, Issue 1, pp 47–55

Characterization and expression of the immunoglobulin light chain in the fugu: evidence of a solitaire type

Authors

  • Nil Ratan  Saha
    • Fisheries Laboratory, Graduate School of Agricultural and Life SciencesThe University of Tokyo
  • Hiroaki Suetake
    • Fisheries Laboratory, Graduate School of Agricultural and Life SciencesThe University of Tokyo
    • Fisheries Laboratory, Graduate School of Agricultural and Life SciencesThe University of Tokyo
Original Paper

DOI: 10.1007/s00251-004-0662-5

Cite this article as:
Saha, N.R., Suetake, H. & Suzuki, Y. Immunogenetics (2004) 56: 47. doi:10.1007/s00251-004-0662-5

Abstract

In this study, we characterized the immunoglobulin light (IgL) chain gene and examined its expression in the fugu (Takifugu rubripes). The cDNA fragment that partially encodes the IgL chain was isolated by RACE and used as a probe for screening for IgL in a fugu splenic cDNA library. The IgL cDNA sequence that we found consisted of a variable (VL) and a constant (CL) segment. Its structural features were similar to the IgL isotype commonly found in teleosts. Genomic sequence analysis revealed that the IgL gene was organized as two VL gene segments (designed VL1 and VL2) followed by single joining (JL) and CL segment. In addition, an unusual duplicate VL1 gene segment was found downstream of the CL segment. The transcriptional orientation of the VL exons was found to be opposite to that of the JL and CL segments. Genomic blot hybridizations with VL and CL probes gave multibands, supporting the contention that the teleost IgL forms a multicluster. Both genomic and cDNA sequences analyses showed that all of the constant segments found in the fugu are identical, suggesting that no other isotypes could be found in this species. Comparison of the deduced amino acid sequence of the fugu CL domain with those of other species showed a high degree of identity (from 40 to 77%). IgL mRNAs were found to be expressed primarily in the lymphoid tissues. In situ hybridization revealed the presence of IgL-positive cells widely distributed throughout the spleen, head kidney, kidney, and thymus. These results support the contention that the lymphoid tissues are the major sites of antibody production in fish. Since IgL mRNA was also expressed in the skin and gill that are exposed to external antigens, it is likely that mucosal Ig plays an important role in immune protection.

Keywords

FuguImmunoglobulin light chainGene structureGene expressionLymphoid organs

Introduction

Immunoglobulins (Igs) play a major role in mediating vertebrate humoral immunity. In fish, IgM appears to be the major Ig isotype, though a putative homologue of the IgD gene has also been found in some species (Wilson et al. 1997; Hordvik et al. 1999; Hirono et al. 2003). Fish IgM is a tetrameric molecule with each monomer consisting of a pair of covalently linked identical heavy-light (H-L) chain dimers. Its Ig heavy (IgH) chain gene exists in two forms; i.e., secretory and membranous. On the other hand, its Ig light (IgL) chain lacks a membrane domain and is joined to both forms of IgH to form a H-L chain dimmer, putting it in a position to contribute to the variability of the antigen-binding site. Our efforts have focused on defining the role of Igs in the development of the immune system in the fugu, which is a good model species to study because of its short genome and nearly equal number of genes as mammals (Aparicio et al. 2002). The structure of the fugu IgH gene has already been elucidated (manuscript submitted). The purpose of this study is to extend those findings and examine the structure of the IgL gene in order to better understand the factors that contribute to the variability of the antigen-binding site.

The IgL chain has only one constant region (CL), whereas the H chain has three to four constant domains. The structure of the IgL chain gene is organized such that the VL segment is followed by a joining (JL) and CL segment. In contrast to IgH-chain genes, L-chain genes lack a diversity segment (D). Within the variable region, variability occurs in three clustered hypervariable regions, which are brought together by peptide folding in association with heavy chain hypervariable regions to form the antigen-binding site (Okamoto et al. 2003). IgL-chain genes in mammals exist in two distinct forms called κ and λ. In the κ type, the locus is arranged with multiple VL and JL segments and a single copy of a CL segment (Zachau 1989; Kirschbaum et al. 1996); on the other hand, the λ type contains several CL domains. IgL genes in teleosts and elasmobranches appear to be organized in a multicluster arrangement (Daggfeldt et al. 1993; Ghaffari and Lobb 1993), though many fish species have not as yet been examined. We were interested in investigating whether this pattern was also found in the fugu whose genomic size is comparatively smaller than in the human, mouse, and zebrafish, in spite of containing a similar number of genes.

Fish possess two to three IgL isotypes. Two isotypes of IgL have been identified in rainbow trout (Oncorhynchus mykiss) (Partula et al. 1996) and channel catfish (Ictalurus punctatus) (Ghaffari and Lobb 1993, 1997), while three have been reported in the common carp (Cyprinus carpio) (Tomana et al. 2002), zebrafish (Danio rerio) (Haire et al. 2000), and Atlantic salmon (Salmo salar) (Solem and Jørgensen 2002), all of which are bony fish. The sturgeon, a chondrostean fish (Acipenser baeri), was reported to have three IgL isotypes (Lundqvist and Pilström 1999). Molecular genetic analyses have shown that most, if not all, cartilaginous fish have at least three IgL isotypes, one of which has a V region orthologous to κ (Greenberg et al. 1993; Sitnikova and Nei 1998), while the other two have a structure more similar to λ (Rast et al. 1994). A similar finding was reported in Xenopus laevis (Haire et al. 1996; Schwager 1991; Zezza et al. 1991). A phylogenetic tree has previously been used to examine the evolutionary relationships of various fish species vis-à-vis their immunoglobulin gene structure (Partula et al. 1996; Espelid et al. 2001; Pilström 2002; Tomana et al. 2002; Okamoto et al. 2003). We herein examined and characterized the structure and expression of the IgL gene in the fugu and compared it with its phylogenetic neighbors. We previously found that, in addition to being expressed in lymphoid tissues, the IgM H-chain gene is expressed in the mucosal organs of the fugu. In order to determine that the expression of IgM is typical, we analyzed the expression of IgL mRNA in different tissues of the fugu.

Materials and methods

Animals and harvesting of tissues

Fugu (Takifugu rubripes) specimens were kindly provided by the Fukui Prefecture Fisheries Experimental Station and were reared in a tank of running seawater that was kept at 20 °C. The fish were fed once a day with commercial fish pellets. For tissue harvesting, the fish were first anesthetized with 2-phenoxyetanol, after which they were dissected. The spleen, head kidney, kidney, thymus, skin, gill, intestine, liver, gonad, and muscle were harvested and immediately fixed in RNAlater (Ambion), to be used later for RNA extraction. The spleen, head kidney, kidney, and thymus from a different cohort of animals were similarly harvested and immediately fixed in 4% paraformaldehyde (PFA) in phosphate buffer solution for later use for in situ hybridization.

RNA isolation and cDNA preparation

Total RNA was extracted from the spleen using the RNA extraction reagent (ISOGEN; Nippon Gene) according to the manufacture’s instructions, and the first strand cDNA was synthesized from the purified poly (A)+ RNA using the SMART RACE cDNA amplification kit (Clontech).

Molecular cloning and sequencing

Specific primers (IgL-F3: ggtgtgtctggccagcgggggctcc and IgL-R1: gtcgctccatcaagatgttgaggacag) were prepared from a partial nucleotide sequence homologue to the IgL gene from the fugu genomic database (http://fugu.hgmp.mrc.ac.uk) and the first-strand cDNA was amplified by thermal cycling (i-Cycler, Bio-Rad) in 20 μl of the reaction mixture containing 5 units of Taq DNA polymerase (Takara, Japan) and 500 nM of each primer. Amplification was performed for 30 cycles at 94 °C for 5 s, 60 °C for 10 s, and 72 °C for 2 min, and the final elongation was conducted at 72 °C for 10 min. The PCR product was ligated into the pCR-II TOPO vector (Invitrogen). The plasmid DNA was purified and sequenced using 310 Genetic Analyzer (Applied Biosystem). Sequences were analyzed using the SeqED version 1.0.3 sofware (Applied Biosystem). Analyses were repeated on at least three independent PCR amplifications to avoid PCR errors.

cDNA library construction and screening

Purified poly (A)+ RNA from the fugu spleen was used for the construction of a cDNA library in the λZipLox vector (Gibco BRL) by directional cloning into SalI and NotI sites. The library was screened by DNA hybridization on nitrocellulose membranes with plaque replicas as recommended by the manufacturer.

A partial cDNA fragment that was isolated by RACE was used to prepare a probe for the screening of the IgL chain. The probe was labeled using an AlkPhos Direct labeling kit (Amersham). The membrane was pre-incubated with AlkPhos Direct hybridization buffer at 55°C for 1 h, after which it was hybridized using hybridization buffer containing an AlkPhos-labeled probe for 4 h at 55°C. The membrane was washed at high stringency after which it was treated with CDP-Star chemiluminescent detection reagent (Amersham) for 5 min; signals were detected using a camera LAS-1000 (Fuji Film) and LAS-1000 Lite software.

Southern blot analysis

Genomic DNA from fugu erythrocytes was isolated using standard methodologies, after which it was digested with BamHI, EcoRI and XbaI, PstI and HindIII. The DNA (7.5 μg/lane) was separated on a 0.8% agarose gel, after which the gel was treated with 0.5 M NaOH containing 1.5 M NaCl for 30 min and the DNA transferred onto a Biodyne B membrane (PALL). Probe labeling, hybridization, and signal detection procedures were the same as those already described above except that the hybridization period was 16 h.

Northern blot analysis

Total RNA was isolated from the spleen, head kidney, trunk kidney, thymus, skin, gill, intestine, liver, gonad and muscle of three animals. Specific primers for the CL region (IgCL-F: GTGGTGCGTCCCACCCTGACCGTCC and IgCL-R: CGTGTCGCTCCATCAAGATGTTGAGGAC) were used to prepare the probe. Ten micrograms of total RNA from each tissue were separated on a 1.2% agarose-formamide denatured gel and transferred onto a Biodyne B membrane (PALL) using 20×SSC as the transfer buffer. The remainder of the protocol was identical to that described above.

In situ hybridization

In order to prepare RNA probes, cDNA fragments encoding CL and VL regions were ligated in pCRII-TOPO vector (Invitrogen). Large-scale quantities of DNA were purified from positive clones using a Geno Pure Plasmid kit (Roche Molecular Biochemicals, Indianapolis, Ind.). The DNA was linearized with XhoI and BamHI. RNAs labeled with digoxigenin (DIG) were then transcribed as recommended by the manufacturer (DIG RNA Labeling Kit, Roche).

Tissues fixed in PFA were transferred to methanol after an overnight incubation at 4 °C and were kept at −20 °C until use. The tissues were then dehydrated sequentially in a series of methanol baths, cleared with xylene, and embedded in paraffin. Sections of 3-μm thickness were prepared and mounted onto coated slides. Before hybridization, tissues sections were deparaffinized, rehydrated, and incubated for 20 min at 65 °C to suppress endogenous peroxidase activity. Hybridization was carried out using DIG-labeled probes in hybridization solution (SSC, 25% formamide, 0.5 mg/ml yeast tRNA, 10 μg/ml heparin, 9.2 mM citric acid, 0.1% Tween-20) for 16 h at 45 °C in a humidified chamber, after which the slides were washed using standard methods. The sections were then incubated with blocking solution (2 mg BSA/ ml PBS-T, 2% sheep serum, 0.8% goat serum) for 1 h, after which they were incubated overnight at 4 °C with anti-digoxigenin-AP conjugate antibody (Roche) at a dilution (in blocking solution) of 1:5,000. Finally, the color reaction was developed over 6 h in the dark at room temperature using Nitro Blue Tetrazolium chloride and 5-bromo-4 chloro-3inkdolyl-phosphate.

Results

Characterization of the fugu IgL chain

A partial cDNA sequence isolated from the fugu spleen by RACE showed high homology to the IgL sequences of other known species. This cDNA fragment was used to prepare a probe for screening of IgL from the cDNA library. A total of 20 positive clones were obtained and sequenced. The complete nucleotide and deduced amino acid sequences of the IgL (clone F-L311) are shown in Fig. 1. The cDNA sequence of fugu IgL was found to comprise 950 nucleotides, including a short 5′-untranslated region (35 bp), an open reading frame of 693 bp that encoded 231 amino acids, and a 222-bp 3′-untranslated sequence. A putative polyadenylation signal (AATAAA) sequence was found 19 bp upstream from the poly(A) tail.
Fig. 1

Nucleotide and deduced amino acid sequence of the fugu IgL-chain cDNA. An asterisk indicates the stop codon. Cysteine residues are bolded. The sequence in the box is the polyadenylation signal. The positions of leader peptide, VL, JL and CL are denoted by arrows above the nucleotide sequence

Genomic organization of the IgL chain

Fugu database mining (Fugu rubripes v3.0) based on the fugu IgL cDNA sequence showed that there are only two full (Fig. 2a, b) and two short (Fig. 2c, d) constant exons in the fugu genome, but all of them were located in different Scaffolds. Since all of those constant exons were identical, it could be explained that no other IgL isotypes could be found in the fugu. On the other hand, those sequences were failed to assemble in a contig and that makes it difficult to say how many light chain genes could be found in the fugu. Two VL genes (designed VL1 and VL2) were found in the fugu. The VL2 gene was predicted by GENSCAN (http://genes.mit.edu/GENSCAN.html) and confirmed by homology searching (http://www.ncbi.nlm.nih.gov/BLAST/). The identities between two VL genes were 50 and 37% for nucleotide and amino acid sequences, respectively. In addition, an unusual duplication of the VL1 genes (partial, C′-terminal end) was found downstream of the CL segment (Fig. 2a). All of these VL genes were located in reverse direction.
Fig. 2a–d

Genomic organization of the fugu IgL chain. The size of the IgL gene was calculated and the values represented as scale bars. The boxes represent exons and the horizontal lines adjacent to the exons represent introns. The arrows indicate relative transcriptional orientation based on the cDNA sequence. The box with doted line represents 99% identity of nucleotide sequences. The sites for restriction enzymes: B BamHI, H HindIII, E EcoRI, P PstI, X XbaI. The ID numbers of genomic sequences are Scaffold_4242 (a), Scaffold_5366 (b), Scaffold_11871 (c) and Scaffold_3715 (d)

Homology analysis of the IgL chain

Multiple alignment of the deduced amino acid sequence of the fugu IgCL with the known IgCL sequences of other fish species showed that the cysteine residues (boxed) involved in the formation of disulfide bonds were completely conserved (Fig. 3), as were the tryptophan residues (shaded) that are important for the formation and stabilization of the tertiary structure of protein. Alignment results (Table 1) indicated that the deduced amino acid sequence of the fugu IgL chain constant region was very similar to that seen in the sea bass (77% identity), wolffish (74%), cod (68%), trout (53%), salmon (51%) and carp (49%). The fugu CL sequence showed 37% and 27% identity with λ chain sequences of humans and mice, respectively, and 28% identity with their κ chain sequences. Thus, the fugu IgL could not be classified as either a mammalian λ or κ isotype.
Fig. 3

Alignment of the deduced amino acid sequence of the fugu CL region with that of other fish species. Asterisks indicate amino acid residues that are identical to those in the fugu sequence. Periods and semicolons indicate amino acid residues that are similar to those in the fugu sequence. Dashes indicate gaps introduced to facilitate maximal alignment. Cysteine residues involved in interdomain (Cys-26 and Cys-85) and intradomain (Cys-104) interactions are indicated by @ and # symbols, respectively. The shaded areas indicate the location of tryptophan residues. The accession numbers for each species are as follows: sea bass (AJ400216), wolffish-L2 (AF137398), trout-L1 (X68521), salmon-1 (AF273019), cod-1 (X68515), carp-L1A (AB015902) and zebrafish-1 (AF246185)

Table 1

Percent identity of the amino acids of the fugu IgCL with those of other representative fish species. Alignment was made between constant regions. The highest identity of fugu IgCL was found with the IgCL of the sea bass (77%)

Seabass

Wolffish L2

Cod L1

Trout L1

Salmon 1

Carp L1A

Zebra fish 1

Fugu

77

74

67

53

52

49

40

Seabass

89

77

55

54

55

43

Wolffish L2

78

55

54

52

42

Cod L1

55

55

50

49

Trout L1

92

57

53

Salmon 1

56

53

Carp L1A

76

Phylogenetic analysis of the fugu IgL chain

Phylogenetic analysis was carried out to clarify the relationship between the fugu IgCL and the IgCL of other vertebrates (Fig. 4). A neighbor-joining tree using the CLUSTAL W and TreeView (2) packages was constructed based on the nucleotide sequences found in the EMBL/DDBJ/GenBank databases. The results showed that the fugu CL region shared the same cluster as those of the cod L1, wolffish-L2, and sea bass, which were different from the cluster associated with mammalian IgL isotypes. The estimated genetic distance indicated that the fugu IgL comprised its own group.
Fig. 4

Phylogenetic analysis of the vertebrate IgCL region. A neighbor-joining tree was constructed based on the nucleotide sequence of the fugu and other vertebrate IgCL domains. The fugu is marked in bold print. Node values represent a bootstrap analysis of 1,000, but only the values more than 500 are shown in the figure. The scale bar corresponds to the estimated evolutionary distance units. The sequences were obtained from the EMBL/DDBJ/GenBank databases and had the following accession numbers: carp L1A (AB015902), carp L1B (AB035728), carp L3 (AB035730), catfish G (L25531), catfish F (U25705), chicken (M33049), cod L1 (X68515), cod L2 (AJ293807), duck (M25726), horse λ (L07562), horse κ (X75612), human λ (X51755), human κ (S83373), mouse κ (M21795), mouse λ (J00582), raja 1 (L25568), raja 2 (L25566), rat (J02574), ratfish (L25549), salmon 1 (AF273019), salmon 2 (AF297518), salmon 3 (AF462234), sandbar shark (M81314), sea bass (AJ400216), shark 1 (M64307), shark 2 (L25561), sheep κ (X54110), sturgeon (AJ133189), trout L1 (X68521), trout L2 (U69987), wolffish L1 (AF137397), wolffish L2 (AF137398), yellowtail 1 (AB062664), yellowtail 2 (AB062653), yellowtail 3 (AB062667), Xenopus 1 (M94392), zebrafish 1 (AF246185), zebrafish 2 (AF246162), zebrafish 3 (AF246193)

Southern blot analysis

Genomic DNA blots from an individual fugu were hybridized with probes specific for the VL and CL regions (Fig. 5) and one to several restriction fragments were found, depending on which restriction enzyme was used. Specifically, only one fragment was detected when DNA was digested with PstI, while two bands were detected when HindIII was used. On the other hand, several bands (4–7) were detected while BamHI, EcoRI and XbaI were used.
Fig. 5

Southern blot analysis of the fugu genomic DNA using probes specific for the VL and CL regions. The restriction enzymes used were BamHI, EcoRI, XbaI, PstI and HindIII. Numbers on the left indicate molecular size (kb)

Northern blot analysis

Hybridization with a specific CL probe yielded a single band of the expected size (Fig. 6). Signal intensity was high in the spleen, head kidney, kidney, and gill (standardized by densitometry). A faint signal was also detected in the skin, while no bands were seen in the other tissues examined.
Fig. 6

Northern blot analysis of mRNA from various fugu tissues using a CL-specific probe. The 28S line represents the control (quality and quantity of RNA). Number on the left represents molecular size. The signal was detected in the spleen, head kidney, kidney, thymus, skin, and gill

In situ hybridization

IgL-positive cells were detected in various lymphoid tissues (Fig. 7). Cells that expressed IgCL were scattered in the hemopoietic tissue of the spleen (a), head kidney (b), and kidney (c). Though the distribution pattern of IgCL-positive cells was similar in these three lymphoid tissues, they were found to be present in greatest number in the head kidney. Positive cells were also detected in the inner zone of the thymus (Fig. 7d). Hybridization using VL region probe yielded similar results (Fig. 7e–h).
Fig. 7a–h

Cells expressing IgCL and IgVL mRNA transcripts were detected in the spleen, head kidney, kidney, and thymus using in situ hybridization. Cells positive for CL transcripts were scattered throughout the spleen (a), head kidney (b), kidney (c) and thymus (d). Similar distribution patterns were also detected when VL-specific probes were used (e–h) (bv blood vessels, gl glomerulus, mmc melanomacrophage center, iz inner zone, oz outer zone) (magnification ×40)

Discussion

The complementary cDNA sequence that encoded the fugu IgL gene was obtained from a fugu splenic cDNA library. The fugu IgL gene contains VJ and CL domains, indicating that this IgL gene is similar to the major IgL isotypes isolated from other fish species. Analysis of the IgL genomic sequences revealed that CL segments located in the fugu genome are identical. These findings are supported by the fact that the CL segment sequences of all other clones picked up from the cDNA library were identical (data not shown). These results suggest that the no other IgL isotypes could be found in the fugu. Two VL segments and a single JL segment were found upstream of the CL segment. Like fugu, two VL genes were also reported in the channel catfish (G-type IgL) (Rast et al. 1994). In addition, an unusual duplication of the VL1 gene (partial, C′-terminal end) was found downstream of the CL segment. In elasmobranches, VL, CL, and JL were all found to be located in the same transcriptional orientation (Rast et al. 1994), while we found that the VL genes were opposite to that of the JL and CL segments in the fugu IgL gene. A similar orientation has been reported in the channel catfish (Rast et al. 1994) and rainbow trout (Timmusk et al. 2000) and may be a common feature of teleosts. The genomic sequences reported in this study failed to assemble into a contig. Further studies are thus required to clarify how many light chains could be found in the fugu.

Multiple alignment comparison of the deduced amino acid sequence of the IgL showed its cysteine residues form disulphide bonds (Cys-26 and Cys-85) or an interdomain disulfide linkage (Cys-104) and that tryptophan residues contribute to the tertiary structure of the protein (Williams and Barclay 1988); the CL region was conserved in a host of different fish species. The highest identity of deduced amino acids in the fugu CL region was found with the CL sequence of the sea bass (77%), followed by the wolffish L2 (74%) and cod L1 (51%). Phylogenetic analysis of teleost, cartilaginous fish, and other vertebrate immunoglobulin CL sequences also demonstrated that the fugu CL forms a cluster with the above species, and is quite close to the group comprising the wolffish L1, yellowtail (1–3), carp L1B and catfish G in the phylogenetic tree. On the other hand, vertebrate λ and κ IgL isotypes belong to a separate branch of the tree. Nucleotide sequence comparison suggests that the fugu CL could not be readily classified as a mammalian κ or λ chain isotype and might be considered as common fact in teleosts.

Catilagenous fish, considered as primitive vertebrates, possess three IgL isotypes (Rast et al. 1994), the same number reported in bony fish like the zebrafish (Haire et al. 2000), common carp (Tomana et al. 2002), salmon (Solem and Jørgensen 2002), and yellowtail (Okamoto et al. 2003); the trout (Partula et al. 1996), catfish (Ghaffari and Lobb 1997) and wolffish (Espelid et al. 2001), on the other hand, each possess two isotypes. A single isotype has been reported in the sea bass (dos Santos et al. 2001). In fugu, only four exons represented either the full or part of the constant region and showed 99% identity. The data obtained from cDNA and genomic sequence strongly suggest that the fugu has no other isotypes. In teleosts, the IgL chains that have been identified thus far can be largely classified into four groups: L1/G, which is found in various species, L3/F which is found in the channel catfish, and different versions of L2 in rainbow trout/zebrafish and Atlantic cod (Tomana et al. 2002). According to CL region phylogenetic analysis, the fugu IgL falls into the L1/G group. Southern blot analyses using VL and CL probes indicated that the fugu IgL might form a multicluster in its locus, as found in other fish species (Ghaffari and Lobb 1993; Draggfeldt et al. 1993; Partula et al. 1996; Tomana et al. 2002).

Northern blot analyses revealed a single band only in the spleen, head kidney, kidney, thymus, skin, and gill; the transcripts were expressed as messages of the expected size. Although an additional, smaller band was reportedly detected in the common carp (Tomana et al. 2002) and cod (Solem and Jørgensen 2002), we did not detect such a band in the fugu. The expression of the single band in our animals was intense in the lymphoid tissues. Furthermore, IgL-positive cells were widely scattered in the spleen, head kidney, kidney, and thymus in a pattern that we also found for IgH-chain-positive cells (manuscript submitted). These results support the notion that these lymphoid organs are the main sites of antibody production. Notably, the head kidney, which is known to be a site of B-cell differentiation in fish (Razquin et al. 1990; Koumans-van Diepen et al. 1994; Breuil et al. 1997; Romano et al. 1997; Schrøder et al. 1998) was found to contain very high numbers of positive cells, on a par with those seen in the salmon (Press et al. 1994) and turbot (Fournier-Betz et al. 2000). In some instances, IgL-positive cells were found to be densely packed around the renal tubules; this was confirmed using sIgM probes. Unexpectedly, a large number of IgL-positive cells were detected in the inner lymphoid zones of the thymus, an organ thought to be primarily a site of T-cell maturation. Both sIgM- and mIgM-positive cells were also detected in this tissue (manuscript submitted). Interestingly, plasma cells were also reported to be present in the thymus of juvenile spotted wolffish (Grøntvedt and Espelid 2003). These results suggest that the thymus might be a source of B cells. However, the distribution of B cells in thymus is still in dispute and is complicated by the fact that mature B cells can migrate there from the peripheral blood.

IgL expression was also observed in the mucous membranes of the skin and gill; we also detected IgM-secreting plasma cells in these organs (data not shown). Furthermore, the skin mucus of fish was reported to contain IgM (Rombout et al. 1993; Lumsden et al. 1995; Hatten et al. 2001; Grøntvedt and Espelid 2003). The skin and gills form the first line of defense against external pathogenic antigens in the aquatic environment. Thus, the expression of Ig in these mucosal tissues suggests that mucosal IgM might play a role in the immune protection of the body surface.

In conclusion, our data suggest that no other IgL isotype could be found in the fugu. The IgL is intensely expressed in the lymphoid organs. Moreover, the expression of the IgL mRNA was also observed in the mucosal tissues, suggesting that these tissues contribute not only to innate immunity, but humoral, adaptive immunity as well.

Acknowledgements

This work was supported by a Grant-in-Aid for Creative Basic Research 12NP0201 from the Ministry of Education, Science, Sports, and Culture of Japan. We thank the staff at the Fukui Prefecture Fisheries Experimental Station for providing us with pufferfish.

Copyright information

© Springer-Verlag 2004