Fish & Shellfish Immunology 18 (2005) 49e60 www.elsevier.com/locate/fsi
The ontogeny of MHC class I expression in rainbow trout (Oncorhynchus mykiss) Uwe Fischera,), Johannes Martinus Dijkstrab, Bernd Ko¨llnera, Ikunari Kiryuc, Erling Olav Koppangd, Ivar Hordvike, Yoshihiro Sawamotof, Mitsuru Ototakec a
Friedrich-Loeffler-Institutes, Federal Research Centre for Virus Diseases of Animals, D-17493 Greifswald-Insel Riems, Germany b Institute for Comprehensive Medical Science, Fujita Health University, Toyoake, Aichi, Japan c Inland Station, National Research Institute of Aquaculture, Fisheries Research Agency, Tamaki, Mie, Japan d Department of Basic Sciences and Aquatic Medicine, Norwegian School of Veterinary Science, Oslo, Norway e Institute for Fisheries and Marine Biology, The University of Bergen, Bergen, Norway f Nagano Prefectural Experimental Station of Fisheries, Akashina, Nagano, Japan Received 2 February 2004; revised 30 April 2004; accepted 19 May 2004
Abstract In the present study, clonal rainbow trout (Oncorhynchus mykiss) embryos and larvae were assayed for the expression of key molecules involved in specific cell-mediated cytotoxicity using an anti-MHC class I monoclonal Ab and by RT-PCR using specific primers derived from classical MHC class I (class Ia), TCR and CD8. Whereas RT-PCR revealed that MHC class Ia and CD8 were expressed from at least 1 week after fertilisation (p.f.) on, TCR expression was detectable from 2 weeks p.f. Immunohistochemistry indicated an early and distinct expression of MHC class I protein in the thymus. Positive lymphoid, epithelial and endothelial cells were found in the pronephros, in the spleen and in the inner and outer epithelia at later stages. Whereas in older rainbow trout the intestine is counted among the organs of the highest class I expression, during ontogeny it was the last site (39 days after hatching) where such expression was detectable. Knowledge on the appearance of the assayed key molecules during fish development is relevant for the pathogenesis of infections as well as for early vaccine delivery. Besides such information regarding the development of the adaptive immune system, immunohistochemistry revealed that in early larvae MHC class I was expressed in neurons whereas in older rainbow trout this was not observed. Ó 2004 Elsevier Ltd. All rights reserved. Keywords: Rainbow trout (Oncorhynchus mykiss); Ontogeny; MHC; T cell receptor; CD8
1. Introduction Immunocompetence is a feature acquired by the organism during ontogeny. The development of the adaptive immune system is of interest because memory responses are possible targets for vaccines ) Corresponding author. Tel.: +49-3-8351-7105; fax: +49-3-8351-7219. E-mail address: uwe.fi
[email protected] (U. Fischer). 1050-4648/$ - see front matter Ó 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.fsi.2004.05.006
50
U. Fischer et al. / Fish & Shellfish Immunology 18 (2005) 49e60
protecting juvenile fish from disease. Considerable progress in the understanding of the development of the adaptive immune system in fish has been made. In rainbow trout, specific antibodies are transmitted from mother to embryo whereas first cytoplasmic and surface IgM positive cells can be observed 12 and 8 days (respectively) before hatching [1]. However, in the same species a first adaptive humoral immune response to a bacterial antigen can only be induced 1 month after hatching in the same species [2]. Functional studies in rainbow trout juveniles indicative of adaptive cell-mediated cytotoxicity are restricted to allograft rejection: fry as young as 14 days post-hatching can destroy skin allografts, a process accompanied by lymphocyte infiltration [3]. The thymus in fish is probably essential for the education of T lymphocytes and thus for immunocompetence. The thymus anlage was detected in rainbow trout embryos in concert with the transcriptional start of the terminal deoxynucleotidyl transferase (TdT), which has a function in T cell receptor (TCR) recombination [4]. Recombination-activating gene (RAG)1 transcription as well as that of the TCR was found in the thymus of early zebrafish (Danio rerio) embryos [5,6]. Other genes important in T cell function, i.e. major histocompatibility complex (MHC) class Ia and class II, are also transcribed early in fish ontogeny. However, they appear to be transcribed even before the onset of thymus development, since common carp genes of both classes are transcribed from day 1 after fertilisation [7]. The present study focuses on MHC class I expression in rainbow trout embryos and larvae, since for these early fish stages no reports on MHC class I histology are available yet. Mammalian MHC class Ia molecules are involved in offering antigenic peptides derived from endogenous proteins at the cell surface for recognition by the TCR/CD8 complex of cytotoxic T lymphocytes [8,9]. MHC class Ia molecules mediate the acquisition of a T cell repertoire by participating in the positive and negative selection of CD8 positive T cells in the thymus. The finding of rainbow trout sequence homologues for MHC class Ia [10e15], b2 microglobulin (b2m) [16], low molecular mass protein (LMP), transporter associated with antigen processing (TAP), MHC class II [13], TCR [17,18] and CD8 [19] suggests that antigen presentation in fish is similar to that in higher vertebrates. Therefore, it is not surprising that in fish cell-mediated cytotoxicity against allogeneic cells is executed by CD8 expressing lymphocytes [20] and that fish cytotoxic cells distinguish between virus-infected syngeneic and allogeneic target cells [21]. Cell-mediated cytotoxicity against virus-infected cells is probably MHC class I restricted [22]. Salmonid fish only express a single MHC class Ia locus designated Onmy-UBA [14,15]. Moreover, rainbow trout MHC class I molecules are expressed in similar cell types as their mammalian homologues as shown by means of a monoclonal Ab (mAb), designated H9, raised against recombinant rainbow trout MHC class Ia [23]. Thus TCR, CD8 and MHC class I seem to represent key molecules in specific cell-mediated cytotoxicity as in mammals. The present study applied the H9 antibody for investigation of class I ontogeny. Earliest MHC class I expression was detected in the primary and secondary lymphoid tissues, in the epithelia, and in the nervous tissues. Thymic expression suggests a function in T cell selection, and expression in the secondary lymphoid tissues and epithelia should be associated with immune surveillance. The expression in nervous tissues is puzzling, but may be associated with a function of MHC class I in neuron development, as discussed for mammals [24]. In rainbow trout no antibodies are available against TCR and CD8. Therefore, in this study RT-PCR was applied to collect data on their expression at the transcriptional level.
2. Materials and methods 2.1. Fish Eyed eggs of homozygous isogeneic rainbow trout (clone C25) were derived from the Nagano Prefectural Experimental Station of Fisheries, Nagano, Japan. The clone was produced by gynogenesis over two generations by suppression of mitosis and meiosis in the first and second generations, respectively. Clonality was confirmed by DNA fingerprinting (data not shown). For convenient propagation of the
U. Fischer et al. / Fish & Shellfish Immunology 18 (2005) 49e60
51
homozygous strains some of the gynogenetic animals were subject to a treatment with methyltestosterone and developed as homozygous neomales. These males then were crossed with genetically identical females. Eggs and sperm were transported to the National Research Institute of Aquaculture (NRIA), Mie, Japan, where the eggs were fertilised. Eyed eggs were sent to the Friedrich-Loeffler-Institutes, Insel Riems, Germany. Fish were hatched over three consecutive years (2001e2003) in commercial incubators at 8e12 (C and then were kept in aerated water tanks at 12 (C. At the NRIA, the 2003 eggs hatched 25 days after fertilisation, while at Insel Riems the hatching was delayed due to transport. Regression of the yolksac was completed 2e4 weeks after hatching, and automated feeding started. Two months after hatching the fish weighed approximately 0.7 g. One-year-old and 2-year-old (adult) rainbow trout were used for comparative studies. These fish were kept in aerated water tanks at 12 (C. 2.2. Expression analysis of Onmy-UBA*501, TCRb and CD8a by RT-PCR The embryos were removed from the yolk-sac (until 2 weeks after hatching) and were transferred as a whole to microtubes with 1 ml ‘TRIzol’ (Gibco BRL, Life Technologies, Grand Island, NY, USA). Then they were fragmented using a plastic grinder, frozen and thawed, followed by vigorous shaking for 10 min. Subsequently, total RNA was isolated following the TRIzol protocol with the exception that the supernatant of the initial TRIzol/chloroform separation was added to another ml of TRIzol for additional removal of DNA and proteins. RNA purification was performed for three individual fish at each time point as well as for pooled samples. The numbers of pooled individuals were: day 7 p.f. ( post fertilisation), n ¼ 10; day 14 p.f., n ¼ 10; day 21 p.f., n ¼ 10; day 25 p.f., n ¼ 10; day 7 p.h. ( post hatching), n ¼ 5; day 14 p.h., n ¼ 3; day 28 p.h., n ¼ 2. The RT-PCR reaction schedules using a ‘RT-PCR high-PLUS’ kit (Toyobo, Osaka, Japan) were as follows. The 25 ml RT-PCR reaction mixtures were prepared as suggested by the manufacturer, with 2.5 mM Mn(OAc)2, 1 mM of each primer and 500 ng total RNA. The RT-PCR schedules were: 60 (C for 30 min, then 94 (C for 2 min, followed by 35 or 40 cycles of 94 (C for 1 min, 60 (C for 1.5 min, and finally 60 (C for 7 min. For amplification of a 446 bp MHC class Ia fragment from Onmy-UBA*501 (GenBank AF287488) the primers pI-a1f, 5#-CTACACCGSATCTTCTGAAGTTCCCA (forward) and p5-a2r, 5#AATGTTTATCCCGCTCAGTCAT (reverse) were used, applying 35 cycle repeats. For amplification of a 395 bp fragment from the TCR b chain (GenBank AF329700) the primers pTCRb-1, 5#-CAAAGTCACTGAACCCACAG (forward) and pTCRb-2, 5#-TCTGGGTGCTCTTCACATAG (reverse) were used, applying 40 cycle repeats. For amplification of a 357 bp fragment of the CD8a chain (GenBank AF178054), the primers CD8a-f, 5#-GTGGAGATCACTTGTGCACC (forward) and CD8a-r3, 5#GGCAGTTGTAGAGCTGGTTAGC (reverse) were used, applying 35 cycle repeats. None of the primer sets used could amplify a similar fragment from genomic DNA (data not shown). 2.3. Immunocytochemistry analysis of MHC class I expression Distribution of MHC class I expression was monitored by immunocytochemistry using the murine mAb H9 raised against recombinant Onmy-UBA*501 protein [23]. MHC class II expression was monitored using a rabbit serum ( pAb Ø127) produced against recombinant salmon MHC class II beta chain [25]. C25 eggs and larvae were sampled at different times before and after hatching, then shock frozen in liquid nitrogen and stored at 70 (C until use. Serial cryosections were made from whole eggs and larvae at a thickness of 20 mm using a Leica CM3050 cryostat (Carl Zeiss, Jena, Germany). Sections were placed on glass slides coated with poly-L-lysine, fixed with 4% paraformaldehyde for 20 min at 4 (C and permeabilised with 0.1% Triton X-100 (Sigma-Aldrich, St. Louis, MO, USA). Then, sections were labelled with mAb H9 and an anti-mouse-IgG AlexaÔ488 conjugate (Molecular Probes, Leiden, The Netherlands)
52
U. Fischer et al. / Fish & Shellfish Immunology 18 (2005) 49e60
or with pAb Ø127 and an anti-rabbit-Ig-FITC (Sigma-Aldrich). For double labelling, a combination of the two primary antibodies and a mixture of anti-rabbit-Ig-FITC (Medac) and anti-mouse-Ig-Cy3 (Dianova, Hamburg, Germany) conjugates as secondary antibodies were applied. Control sections were labelled with the conjugate only. Sections which were single stained with FITC conjugates were counterstained with PI (Sigma-Aldrich; contained in the mounting medium at a concentration of 2 mg ml 1). Serial sections comprising the whole sample were examined using a confocal laser scanning microscope (Zeiss LSM510). Suitable optical sections were scanned using the 488 nm line of an Argon laser and a pinhole diameter allowing optical slides of approximately 1 mm thickness. The confocal laser scanning microscope was set up using conjugate control sections leaving no background of unspecific fluorescence or improper optical or electronic signal amplification. Labelled sections were scanned under the same conditions for all parameters applied for the conjugate control samples.
3. Results 3.1. RT-PCR analysis of MHC class I, TCRb and CD8a At the NRIA total RNA was isolated from the embryos without yolk-sac on days 7, 14, 21 and 25 p.f. (day 25 was the day of hatching), and on days 7, 14 and 28 p.h. RT-PCR data for Onmy-UBA*501 (the MHC class Ia allele found in C25 trout), TCRb and CD8a from pooled samples are shown in Fig. 1. Onmy-UBA*501 and CD8a are detectable already from the first week p.f. while TCRb was first detected 2 weeks after fertilisation, although not all individual fish expressed TCRb 2 weeks after fertilisation (data for individual fish not shown). Increasing mRNA levels were observed for Onmy-UBA*501 and TCRb during the time of sampling while low CD8a levels were detected at the day of hatching. After hatching CD8a expression increased to higher levels than before hatching (Fig. 1). RT-PCR data for individual whole-embryo samples (including yolk-sac) taken at the Federal Research Centre in Germany were also comparable with Fig. 1 (data not shown). Most important for the present study is that the RT-PCR data for Onmy-UBA*501 are in agreement with the H9 immunohistochemistry (see below): the amounts of
Fig. 1. Ontogeny of MHC class Ia and its probable receptors. Total RNA (RNA) was isolated from pooled C25 individuals 1, 2 and 3 weeks after fertilisation (1w, 2w, 3w, left side), at the day of hatching (ha), and 1, 2 and 4 weeks after hatching (1w, 2w, 4w, right side) and compared with samples from the spleen (sp) and muscle (mu) from a 1-year-old C25 trout. RT-PCR was performed analysing Onmy-UBA*501 (UBA), TCRb and CD8a expression.
53
U. Fischer et al. / Fish & Shellfish Immunology 18 (2005) 49e60
Onmy-UBA are low during early ontogeny (lower than in muscle of adult fish) and an increase can be observed starting around hatching (when the first H9 positive cells are observed). 3.2. Expression of MHC class I protein Expression of MHC class I protein was examined by immunohistochemistry using mAb H9 raised against the recombinant rainbow trout MHC class Ia protein Onmy-UBA*501. Previous studies [23] had proven the specificity of this antibody for rainbow trout MHC class I by comparison of rainbow trout and goldfish tissues in immunohistochemistry, and immunocytochemistry analysis of Onmy-UBA*501 transfected mammalian cells. For each time point three C25 trout were analysed. The subjective semiquantitative H9 binding intensities for the various organs are summarised in Table 1. Generally, polymorphic dendritic-like cells, including those forming the thymus meshwork, and epithelial cells of the gills, the skin and the digestive tract stained most intensively. Figs. 2ae3m exemplify the findings. The earliest fluorescence signals of MHC class I expression in cryosections were observed in the thymus and in the nervous system. During ontogeny MHC class I expression increased in all organs except in the nervous system where the expression intensity reached its peak on day 12, then disappeared and was again observed from day 73 p.h. on. In elder fish, the strongest expression was observed in various epithelial and lymphoid tissues [23] (see below). Table 1 Expression of MHC class I protein in different organs (semi-quantitative immunohistochemistry dataa)
days p.h.
-3 0
3
6
9 12 15 18 21 24 29 31 39 46 53 63 73 80 365 *
thymus pronephros spleen gill skin intestine brain a
semiquantitative staining intensity of MHC class I in
immunohistochemistry:
At least two embryos/larvae/fish per day were examined. ) Start of feeding.
absent low to medium
high
54
U. Fischer et al. / Fish & Shellfish Immunology 18 (2005) 49e60
U. Fischer et al. / Fish & Shellfish Immunology 18 (2005) 49e60
55
3.2.1. MHC class I in thymus Since the medulla and cortex assignments are uncertain (see Section 4) the below refrains from such designations and only describes the thymus topography by distinguishing the inner and outer zones. The thymus was the first organ among the lymphoid tissues in which MHC class I protein was observed and it showed the highest expression levels in 1-year-old fish. Positive cells first were concentrated in foci (Fig. 3a), and later became confluent in the inner zone (Fig. 2b). The early clusters could not clearly be attributed to the inner or outer zone of the thymus because cutting directions in small embryos varied and the topography of the sections was hard to interpret. The cell types found to express MHC class I in the early thymus were similar to those in the adult thymus. Intensive staining was limited to large polymorphic (Fig. 3b) and epithelial cells, whereas most of the round lymphocyte-like cells were unstained. In transversal sections of the thymus of adult fish we observed structures similar to the Hassal’s corpuscles of higher vertebrates (see Section 4). However, in vertical sections these structures appeared as MHC class II expressing epithelial cords reaching from the subepithelial zone through the outer zone into the inner zone. These epithelial cords did not penetrate the inner zone with higher MHC class I and II expression. Although fewer, there are cells apart from the epithelial cords expressing MHC class II protein in the outer zone too. Interestingly, some express either class I or II, whereas others express both (Fig. 2a,b). Both class I and II expressing subepithelial cells and cells of the inner zones consist of confluent tissue formations (Fig. 2b). 3.2.2. MHC class I in kidney and spleen In the haematopoietic part of the kidney (the head kidney), large polymorphic cells irregularly spread over the organ showed positive staining with mAb H9 (Fig. 3d). In the diuretic part epithelia of renal tubules expressed MHC class I (Fig. 3c). Melanomacrophages were not stained by mAb H9 (Fig. 3d). In both parts of the kidney MHC class I positive cells sporadically appeared from the 15th day p.h. on, but were stably detected from the 46th day p.h. on in all samples assayed. MHC class I positive splenocytes were first detected on day 39 p.h. Positive cells were polymorphic and randomly distributed over the organ section (data not shown). Fig. 2. Confocal laser scanning micrographs of thymus sections from adult rainbow trout (2 years of age) double stained for MHC class I (mAb H9+a-mouse-Ig-Cy3, red) and II (a-MHC II serum, a-rabbit-Ig-FITC, green). (a) Mucosal epithelial cells (ep) are double stained, whereas most other cells are single stained. Epithelial cords (ec) lining from the subepithelial layer into the deeper thymus layers mostly are single MHC class II positive. MHC class I positive cells are polymorphic reticular cells accumulating in small clusters in the outer zone of the thymus (image size 325.7!325.7 mm). (b) In the inner zone, double positive ( yellow-brownish) confluent cells are present. Cells positive for MHC class I and II mainly are polymorphic, and MHC class II single positive cells are somewhat larger than MHC class I single positive cells. The figure consists of 3!5 single images (total image size approximately 1500!900 mm). Fig. 3. Confocal laser scanning micrographs of tissue sections from rainbow trout larvae at different stages of development stained for MHC class I (mAb H9+a-mouse-Ig-FITC, green; counterstaining with PI, red). (a) Thymusd21 days p.h. Foci of MHC class I positive cells are shown (image size 325.7!325.7 mm). (b) Thymusd31 days p.h. Larger magnification of an MHC class I positive polymorphic cell (image size 41.8!41.8 mm). (c) Kidneyd31 days p.h. Renal tubulus epithelium is positive for MHC class I (image size 325.7!325.7 mm). (d) Kidneyd1-year-old rainbow trout. Polymorphic cells in the haematopoietic tissue are MHC class I positive. Melanomacrophages (mm) are negative for MHC class I (image size 183.2!183.2 mm). (e) Gilld18 days p.h. Only few cells are positive for MHC class I (image size 202.3!202.3 mm). pl, primary lamella; sl, secondary lamella. ( f) Gilld73 days p.h. Confluent MHC class I positive epithelial cells can be seen (image size 325.7!325.7 mm). ( g) Skind31 days p.h. Foci of MHC class I positive cells are found in the epidermis becoming confluent later on (data not shown) (image size 325.7!325.7 mm). (h) Intestine ( proximal part)d39 days p.h. A focus of MHC class I positive cells in the mucosa is shown (image size 233.9!233.9 mm). (i) Upper jawd46 days p.h. Superficial and deeper epithelial cells are strongly positive for MHC class I (image size 325.7!325.7 mm). ( j) Oesophagusd46 days p.h. Epithelial and submucosal cells are H9 positive (image size 208.6!208.6 mm). (k) Braindday of hatching. MHC class I positive neurons (image size 143.9!143.9 mm). (l) Spinal cordd73 days p.h. MHC class I positive paravertebral ganglion-like structures can be seen (image size 162.9!162.9 mm). (m) Braind2-year-old rainbow trout. Endothelial cells of a capillary are stained. Note the weak staining in the neuropile (image size 140.4!140.4 mm).
56
U. Fischer et al. / Fish & Shellfish Immunology 18 (2005) 49e60
3.2.3. MHC class I in the gills The gill epithelia stained positive with mAb H9 from day 18 p.h. on in some larvae (Fig. 3e) with stable detection in all sampled larvae from day 24 p.h. on. In larger larvae (Fig. 3f) and 1-year-old fish, the gill epithelium mostly showed very intensive H9 staining. In the pseudobranchia, only the epithelium separating the organ from the operculum was H9 positive whereas the lamellae remained negative in all individuals (data not shown). 3.2.4. MHC class I in the skin MHC class I expression in the skin started with epithelial foci on day 24 p.h. These foci, predominantly consisting of squamous epithelial cells of the upper epidermal layers (Fig. 3g), became confluent from day 46 p.h. on. In addition to epithelial cells, several Langerhans-like cells and endothelial cells of capillaries in the dermis were positive from day 53 p.h. on. 3.2.5. MHC class I in the digestive tract The first signals of H9 staining in the intestine started on day 39 p.h. Expression first was limited to foci of mucosal cells (Fig. 3h) and polymorphic submucosal cells (data not shown). The exact location of these foci in serial sections of whole larvae was difficult to attribute to a particular part of the intestine at that early developmental stage. Later, these foci were localised in the posterior part of the intestine, i.e. in the rectum, but also in the upper part of the oesophagus close to the mucosa of the oral cavity, which was intensely stained from day 46 p.h. on (Fig. 3i). Few polymorphic cells in the mucosa of the oesophagus were stained with mAb H9 on day 9 p.h. (Fig. 3j). No further positive cells were detected in the entire digestive tract prior to day 39 p.h. Generally, in later stages of the ontogeny, particularly from day 80 p.h. on, the expression levels of MHC class I molecules varied in the different parts of the digestive tract with low to moderate levels in the anterior parts (oesophagus, stomach) and with high to very high levels in the posterior (rectum) and middle parts. At later stages of the development of the fish, endothelia of blood vessels in the intestinal wall and foci of mononuclear cells in the submucosa also stained positive for mAb H9. 3.2.6. MHC class I in the nervous system MHC class I protein expression in the nervous system was observed early in the development of rainbow trout when compared to other organs. Fluorescence signals of H9 staining were first detected on the day of hatching. This was the only day when H9 positive neurons could be clearly distinguished from other cells of the brain, although only in a single section from one fish (Fig. 3k). Other positive signals detected in the brain neuropile during the first month were mostly impossible to attribute to a certain cell type. A peak of MHC class I expression in the central nervous system was observed around day 12 p.h. (Table 1). Whereas considerable levels of MHC class I expression were detected in the upper brain stem of younger larvae, staining intensity shifted to the medulla oblongata and the spinal cord of older larvae. H9 positive foci were then localised in paravertebral ganglion-like structures close to the spinal cord (Fig. 3l). In older fish, endothelial cells of smaller and larger blood vessels of the brain (Fig. 3m) were stained clearly whereas some neuropile regions only showed very faint H9 reactivity.
4. Discussion MHC class I expression during rainbow trout ontogeny was studied by means of RT-PCR and immunohistochemistry. RT-PCR assays showed that MHC class Ia and CD8a in rainbow trout were expressed from 1 week p.f. and that TCRb could be found from 2 weeks p.f. (Fig. 1). One week after fertilisation TCRb transcription was not detectable (Fig. 1), suggesting that the earliest MHC class Ia
U. Fischer et al. / Fish & Shellfish Immunology 18 (2005) 49e60
57
expression in rainbow trout has different functions from the interaction with the TCRab/CD3 complex expressed by its T lymphocytes. In zebrafish larvae, which develop much faster than rainbow trout, TCRa transcription was first found in the thymus at least 8 days after fertilisation [6]. Although they did not show the data, Hansen and Zapata [26] mentioned that rainbow trout TCRb was transcribed from around 15 days after fertilisation on (similar to Fig. 1), and that this was in concert with early thymus development. In fish, as in all vertebrates, the thymus is the first lymphoid organ to develop and the first one to become lymphoid. In rainbow trout lymphoid colonisation of the thymus has been found to occur 5 days before hatching [27]. The thymus is followed by the kidney, with the spleen developing later [2]. This implies that at least the MHC class Ia and the CD8 molecules would have a function in early rainbow trout ontogeny prior to the development of any lymphoid organ. There are no studies showing CD8 transcription in any species prior to the onset of thymus development; studies in mammals report a lack of CD8 expression before the thymus appears [28]. In mice, CD8 positive lymphocytes were detected from day 15 of foetal development (approximately 35 days of gestation) [29]. However, very early MHC class Ia expression has been described. In common carp MHC class Ia protein was detectable already on day 1 after fertilisation [7]. In chicken, class Ia transcripts were found from day 6.5 after fertilisation [30], 1 week before chicken TCRab T cells were detectable [31,32]. In mice MHC class Ia expression was even reported from the one cell stage [33]. For such early class Ia expression, a role in the development of the embryo independent of antigen presentation was suggested [33]. In order to allow conclusions on the function of very early CD8 and class Ia expression in rainbow trout, the cell types expressing these genes should be identified in future experiments. Previous research [23] showed that the H9 antibody is rainbow trout specific and recognises OnmyUBA*501, the single class Ia molecule in the clonal C25 trout. However, in outbred rainbow trout lacking Onmy-UBA*501 allele, the H9 antibody also showed reactivity, suggesting that it can bind either to a conserved epitope in different Onmy-UBA alleles or to related Ib molecules [23]. So far, we have found three expressed class Ib loci sharing significant homology with Onmy-UBA (unpublished results). However, since the H9 data obtained in the present study agree with findings for mammalian class Ia expression, it is likely that most H9 binding indicates Onmy-UBA expression. H9 analysis indicated that the thymus was the first lymphoid organ in which MHC class I was expressed. Whereas in the young larvae class I expression was focal, in older rainbow trout expression was found in confluent layers, especially in the epithelium separating the organ from the gill cavity and the inner zone probably representing the medulla. Appearance of MHC class I molecules in the thymus early in ontogeny is not unique to rainbow trout. In chicken, MHC class I protein was first detectable in this organ [30,34] and, like in rainbow trout, the expression changed from cellular foci of thymic reticular cells into a confluent medulla/ inner zone [34]. In mammals, medulla formation also starts with MHC expressing epithelial islets, later forming the confluent medulla [35] with a significantly higher MHC expression than in the surrounding cortex (e.g. Ref. [36]). During migration from the subcapsular zone to the medulla, double positive thymocytes are induced by MHC class I- and by MHC class II-presenting cells to differentiate into single CD8 or single CD4 positive cells, respectively. These MHC class I and II presenting cells form a meshwork of fibroblasts, macrophages, and epithelial and dendritic cells. Whereas in the subcapsular zone and in the cortex TCR gene rearrangement and positive selection of MHC binding thymocytes are performed, there is a negative selection against thymocytes with high MHC binding activity in the medulla [37]. The findings on MHC class I and II antigen expression in the rainbow trout thymus suggest similar functions in fish as in mammals. A similar functional organisation of the fish thymus was also suggested since RAG1 (involved in TCR rearrangement) is expressed in the cortex but not in the medulla of the zebrafish thymus [38]. The fact that rainbow trout (Fig. 2a,b) has a thymic MHC class I and II distribution similar to that in mammals further supports the idea that the outer and inner zones of the rainbow trout thymus correspond with their cortex and medulla. However, since morphological differences in the thymus organisation between fish and mammals have been described (reviewed by Ref. [39]), further investigations are needed with regard to these designations in fish. For example, Hassal’s corpuscles which can be used to locate the thymus medulla in mammals can not be
58
U. Fischer et al. / Fish & Shellfish Immunology 18 (2005) 49e60
found in rainbow trout. In a previous study [23] we apparently mistook cross-sections of epithelial cords in the outer zone (Fig. 2a) for Hassal’s corpuscles. Furthermore, nurse cells have been described to be located in the inner zone of fish thymus whereas in mammals they are situated in the cortex [40]. Early MHC class I expression was also observed in the central nervous system, i.e. in the neuropile and in some neurons. Cells other than neurons positively stained for MHC class I with mAb H9 were not attributable to a certain cell type due to the lack of specific cell markers for neuronal cells in fish. In mice, MHC class I was found to be expressed in endothelial, microglial, and oligodendrocyte cell types [41] as well as in neurons [24,42]. The latter was doubted by most of the scientific world until recently, probably because most neurons do not express MHC class I [43]. However, it has been found recently that class Ia is expressed in rearranging neurons and a function of class Ia in the rearrangement of these cells was hypothesised [42]. Although they originally found class Ia expression in neurons of neonatal cats and mice, they later detected class Ia as well as Ib expression in the neurons of adult mice [24,42]. Very recent studies by other groups showed class Ib expression in neuron subsets in mice [44]. The detection of rainbow trout class I in neurons during ontogeny suggests that these molecules may play a role in neuron rearrangement, since the chance occurrence of finding such neurons should be higher than in adult rainbow trout, where we never found neurons clearly stained with the H9 antibody [23]. However, prior to further conclusions, the nature of the neuronally expressed class I molecules (class Ia or Ib) and the detailed distribution in the nervous system should be studied. Anyhow, the finding seems to fit into a general pattern of common characteristics of thymus and nervous tissues. For example, in the zebrafish larvae RAG1, Ikaros and GATA3 genes are all expressed in nervous tissues as well as in the thymus [6]. The onset of class I expression in the mucosal epithelia of skin, gill and intestine of the rainbow trout only follows that in the lymphoid and nervous tissues. And in the intestine, the first class I expression rather seems to be associated with small epithelial foci than with a general epithelial class I expression, despite the fact that in adult rainbow trout, mucosal epithelia count among the sites of the highest class I expression [15,23]. An interesting fact is that MHC class I is expressed in the intestinal epithelia later (day 39) than in the gill epithelia (day 18). A possible reason for this observation may be that water borne antigens are met directly after hatching, when the protection provided by the egg shell is lost. Food borne antigens are met at a later stage from a few days on after feeding has started (day 31 p.h.). In addition, there may be qualitative differences in early class I function in epithelia among vertebrates: Xenopus class Ia expression is high in the tadpole gill whereas that in the tadpole intestine and thymus is virtually absent until after metamorphosis [45]. The delay between food uptake and intestinal class I expression in rainbow trout may be related to a need for toleration of new food. That would agree with data in mammals, since in pigs a significant decrease in intestinal class I expression is observed for a period of a few days just after weaning and the start of self-feeding [46]. Data shown in the present study are important for considering vaccine strategies, as they show that the MHC class I expression patterns in young rainbow trout are comparable to those of adults already a few weeks after hatching. This implies that before this age MHC class I restricted memory responses probably cannot be fully stimulated. In short, the ontogeny of class I expression in rainbow trout is very similar to that in higher vertebrates. The expression patterns suggest a function in immunity as well as in development. Future studies should distinguish between class Ia and Ib expression and should correlate the data to the possibility of inducing CTL memory by vaccines at an early age.
Acknowledgements This work was partly supported by a grant from the Deutsche Forschungsgemeinschaft (FI 604/3-1) and by ‘‘The Promotion of Basic Research Activities for Innovative Biosciences’’ funded by the Bio-oriented
U. Fischer et al. / Fish & Shellfish Immunology 18 (2005) 49e60
59
Technology Research Advancement Institution (BRAIN), Japan. The authors wish to thank Mrs Helga Noack and Mrs Anja Schulz for excellent technical assistance and Mrs Anette Beidler for critical reading of the manuscript.
References [1] Castillo A, Sanchez C, Dominguez J, Kaattari SL, Villena AJ. Ontogeny of IgM and IgM-bearing cells in rainbow trout. Dev Comp Immunol 1993;17:419e24. [2] Tatner MF. The ontogeny of humoral immunity in rainbow trout, Salmo gairdneri. Vet Pathol Immunopathol 1986;12:93e105. [3] Tatner MF, Manning MJ. The ontogeny of cellular immunity in the rainbow trout, Salmo gairdneri Richardson, in relation to the stage of development of the lymphoid organs. Dev Comp Immunol 1983;7:69e75. [4] Hansen JD. Characterization of rainbow trout terminal deoxynucleotidyl transferase structure and expression. TdT and RAG1 co-expression define the trout primary lymphoid tissues. Immunogenetics 1997;46:367e75. [5] Willett CE, Zapata AG, Hopkins N, Steiner LA. Expression of zebrafish rag genes during early development identifies the thymus. Dev Biol 1997;182:331e41. [6] Trede NS, Zapata A, Zon LI. Fishing for lymphoid genes. Trends Immunol 2001;22:302e7. [7] Rodrigues PN, Hermsen TT, van Maanen A, Taverne-Thiele AJ, Rombout JH, Dixon B, et al. Expression of MhcCyca class I and class II molecules in the early life history of the common carp (Cyprinus carpio L.). Dev Comp Immunol 1998;22:493e506. [8] Townsend A, Bodmer H. Antigen recognition by class I-restricted T lymphocytes. Annu Rev Immunol 1989;7:601e24. [9] Germain RN. MHC-dependent antigen processing and peptide presentation: providing ligands for T lymphocyte activation. Cell 1994;76:287e99. [10] Hansen JD, Strassburger P, Du Pasquier L. Conservation of an alpha 2 domain within the teleostean world, MHC class I from the rainbow trout Oncorhynchus mykiss. Dev Comp Immunol 1996;20:417e25. [11] Klein J, Sato A. Birth of the major histocompatibility complex. Scand J Immunol 1998;47:199e209. [12] Flajnik MF, Ohta Y, Namikawa-Yamada C, Nonaka M. Insight into the primordial MHC from studies in ectothermic vertebrates. Immunol Rev 1999;167:59e67. [13] Hansen JD, Strassburger P, Thorgaard GH, Young WP, Du Pasquier L. Expression, linkage, and polymorphism of MHC-related genes in rainbow trout, Oncorhynchus mykiss. J Immunol 1999;163:774e86. [14] Shum BP, Guethlein L, Flodin LR, Adkison MA, Hedrick RP, Nehring RB, et al. Modes of salmonid MHC class I and II evolution differ from the primate paradigm. J Immunol 2001;166:3297e308. [15] Aoyagi K, Dijkstra JM, Xia C, Denda I, Ototake M, Hashimoto K, et al. Classical MHC class I genes composed of highly divergent sequence lineages share a single locus in rainbow trout (Oncorhynchus mykiss). J Immunol 2002;168:260e73. [16] Shum BP, Azumi K, Zhang S, Kehrer SR, Raison RL, Detrich HW, et al. Unexpected beta2-microglobulin sequence diversity in individual rainbow trout. Proc Nat Acad Sci U S A 1996;93:2779e84. [17] Partula S, de Guerra A, Fellah JS, Charlemagne J. Structure and diversity of the T cell antigen receptor beta-chain in a teleost fish. J Immunol 1995;155:699e706. [18] Partula S, de Guerra A, Fellah JS, Charlemagne J. Structure and diversity of the TCR alpha-chain in a teleost fish. J Immunol 1996;157:207e12. [19] Hansen JD, Strassburger P. Description of an ectothermic TCR coreceptor, CD8 alpha, in rainbow trout. J Immunol 2000;164: 3132e9. [20] Fischer U, Utke K, Ototake M, Dijkstra JM, Ko¨llner B. Adaptive cell-mediated cytotoxicity against allogeneic targets by CD8-positive lymphocytes of rainbow trout (Oncorhynchus mykiss). Dev Comp Immunol 2003;7:323e37. [21] Somamoto T, Nakanishi T, Okamoto N. Role of specific cell-mediated cytotoxicity in protecting fish from viral infections. Virology 2002;297:120e7. [22] Fischer U, Dijkstra JM, Aoyagi K, Xia C, Ototake M, Nakanishi T. MHC class I restricted cytotoxicity in rainbow trout, Oncorhynchus mykiss. Dev Comp Immunol 2000;24:76e7. [23] Dijkstra JM, Ko¨llner B, Aoyagi K, Sawamoto Y, Kuroda A, Ototake M, et al. The rainbow trout classical MHC class I molecule Onmy-UBA*501 is expressed in similar cell types as mammalian classical MHC class I molecules. Fish Shellfish Immunol 2003;14: 1e23. [24] Huh GS, Boulanger LM, Du H, Riquelme PA, Brotz TM, Shatz CJ. Functional requirement for class I MHC in CNS development and plasticity. Science 2000;290:2155e9. [25] Koppang EO, Hordvik I, Bjerkas I, Torvund J, Aune L, Thevarajan J, et al. Production of rabbit antisera against recombinant MHC class II beta chain and identification of immunoreactive cells in Atlantic salmon (Salmo salar). Fish Shellfish Immunol 2003; 14:115e32. [26] Hansen JD, Zapata AG. Lymphocyte development in fish and amphibians. Immunol Rev 1998;166:199e220.
60
U. Fischer et al. / Fish & Shellfish Immunology 18 (2005) 49e60
[27] Manning MJ. Fishes. In: Turner RJ, editor. Immunology. A comparative approach. Chichester: Wiley; 1994. p. 69e100. [28] Maddox JF, Mackay CR, Brandon MR. Ontogeny of ovine lymphocytes. I. An immunohistological study on the development of T lymphocytes in the sheep embryo and fetal thymus. Immunology 1987;62:97e105. [29] Andjelic S, Jain N, Nikolic-Zugic J. Ontogeny of fetal CD8lo4lo thymocytes: expression of CD44, CD25 and early expression of TcR alpha mRNA. Euro J Immunol 1993;23:2109e15. [30] Dunon D, Salomonsen J, Skjodt K, Kaufman J, Imhof BA. Ontogenic appearance of MHC class I (B-F) antigens during chicken embryogenesis. Dev Immunol 1990;1:127e35. [31] Chen CH, Cihak J, Lo¨sch U, Cooper MD. Differential expression of two T cell receptors, TcR1 and TcR2, on chicken lymphocytes. Euro J Immunol 1988;18:539e43. [32] Cooper M, Chen C, Bucy R, Thompson C. Avian T cell ontogeny. Adv Immunol 1991;50:187e217. [33] Cooper JC, Fernandez N, Joly E, Dealtry GB. Regulation of major histocompatibility complex and TAP gene products in preimplantation mouse stage embryos. Am J Reprod Immunol 1998;40:165e71. [34] Sgonc R, Hala K, Wick G. Relationship between the expression of class I antigen and reactivity of chicken thymocytes. Immunogenetics 1987;26:150e4. [35] Rodewald HR, Paul S, Haller C, Bluethmann H, Blum C. Thymus medulla consisting of epithelial islets each derived from a single progenitor. Nature 2001;414:763e8. [36] Vicente A, Varas A, Sacedon R, Zapata AG. Histogenesis of the epithelial component of rat thymus: an ultrastructural and immunohistological analysis. Anatomical Record 1996;244:506e19. [37] Janeway CA, Travers P, Walport M, Shlomchik M. 7. Entwicklung und U¨berleben von Lymphozyten. In: Janeway CA, Travers P, Walport M, Shlomchik M, editors. Immunologie. 5th ed. Translated into German by Beginnen K, Seidler L, Haußer-Siller I. Berlin: Spektrum Akdaemischer Verlag, p. 239. Original title: Immunobiology. The immune system in health and disease. 5th ed. New York: Garland Science; 2002. [38] Lam SH, Chua HL, Gong Z, Wen Z, Lam TJ, Sin YM. Morphologic transformation of the thymus in developing zebrafish. Dev Dynamics 2002;225:87e94. [39] Zapata AG, Chiba A, Varas A. Cells and tissues of the immune system of fish. In: Iwama G, Nakanishi T, editors. The fish immune system. San Diego, London: Academic Press; 1996. p. 22. [40] Flano E, Alvarez F, Lopez-Fierro P, Razquin BE, Villena AJ, Zapata AG. In vitro and in situ characterization of fish thymic nurse cells. Dev Immunol 1996;5:17e24. [41] Horwitz MS, Evans CF, Klier FG, Oldstone MB. Detailed in vivo analysis of interferon-gamma induced major histocompatibility complex expression in the central nervous system: astrocytes fail to express major histocompatibility complex class I and II molecules. Lab Invest 1999;79:235e42. [42] Corriveau RA, Huh GS, Shatz CJ. Regulation of class I MHC gene expression in the developing and mature CNS by neural activity. Neurone 1998;21:505e20. [43] Joly E, Oldstone MB. Neuronal cells are deficient in loading peptides onto MHC class I molecules. Neurone 1992;8:1185e90. [44] Ishii T, Hirota J, Mombaerts P. Combinatorial coexpression of neural and immune multigene families in mouse vomeronasal sensory neurons. Curr Biol 2003;13:394e400. [45] Salter-Cid L, Nonaka M, Flajnik MF. Expression of MHC class Ia and class Ib during ontogeny: high expression in epithelia and coregulation of class Ia and lmp7 genes. J Immunol 1998;160:2853e61. [46] McCracken BA, Spurlock ME, Roos MA, Zuckermann FA, Gaskins HR. Weaning anorexia may contribute to local inflammation in the piglet small intestine. J Nutrition 1999;129:613e9.