Expression and Temperature-Dependent Regulation of the Beta2-Microglobulin (Cyca-B2m) Gene in a Cold-Blooded Vertebrate, the Common Carp (Cyprinus carpio L.)

Expression of beta2-microglobulin (β 2m) in the common carp was studied using a polyclonal antibody raised against a recombinant protein obtained from eukaryotic expression of the Cyca-B2m gene. β 2m is expressed on peripheral blood Ig+ and Ig- lymphocytes, but not on erythrocytes and thrombocytes. In spleen and pronephros, dull- and bright-positive populations could be identified correlating with the presence of erythrocytes, thrombocytes, and mature leucocytes or immature and mature cells from the lympho-myeloid lineage, respectively. Thymocytes were shown to be comprised of a single bright-positive population. The Cyca-B2m polyclonal antiserum was used in conjunction with a similarly produced polyclonal antiserum to an MHC class I (Cyca-UA) α chain to investigate the expression of class I molecules on peripheral blood leucocytes (PBL) at different permissive temperatures. At 12℃, a temporary downregulation of class I molecules was demonstrated, which recovered to normal levels within 3 days. However, at 6℃, a lasting absence of class I cell-surface expression was observed, which could be restored slowly by transfer to 12C. The expression of immunoglobulin molecules on B cells was unaffected by temperature changes. The absence of the class cell-surface expression was shown to be the result of a lack of sufficient Cyca-B2m gene transcription, although Cyca-UA mRNA was present at comparable levels at all temperatures. This suggests that class I expression is regulated by a temperature-sensitive transcription of the Cyca-B2m gene.


er is tempera
ure-dependent, as shown in experiments, which uses the RMA-S cell line (Ljunggren et al., 1990), nothing is known of such requirements for class I trimolecular stability in poikilothermic animals such as teleostean fish.

The immune system of fish, although with some differences, seems to be comprised of the same basic features found in other vertebrates (Turner, 1994).For some time, functional assays (e.g., skin transplantation, MLR, in vitro antibody production) were considered to be evidence for the presence of MHC in teleostean fish (Stet and Egberts, 1991).Recently, however, a large number of MHC (class Ice, /32m, class II ce and/3) and MHC-related genes of different teleost species have been identified (Dix n et al.,  1995)).As yet, no data are available concerning the functional properties of the molecules encoded by these MHC genes (Stet et al., 1996).Studies on function requires appropriate tools, such as anti- bodies.Due to availability of carp/32m (Cyca-B2m) and class I ce-chain (Cyca-UA) full-length cDNA, it became possible to produce recombinant proteins and raise polyclonal antibodies to such proteins, as has recently been demonstrated for the Cyca-UA class I molecule (Van Erp et al., 1996a).

Experiments have shown that the immune response of fish kept at the lower limit of their physiological temperature was severely impaired (Bly and Clem,  1992).These findings were investigated further by studies carried out in the channel catfish (Ictalurus punctatus), where it was demonstrated that in vitro teleost antigen-presenting cells were able to take up, process, and present exogenous antigen and trigger antibody production at low permissive temperatures (Vallejo et al., 1992).It was concluded that the previously observed temperature-dependent suppres- sion of primary T-cell responses in fish was not attributed to impaired class II expression.However, the effect on MHC class I molecule expression has never been investigated in fish.

In this study, the expression of class I molecules was investigated with the aid of polyclonal antibodies raised against recombinant carp class I ce-chain (Cyca-UA) and/32m (Cyca-B2m) molecules in carp that were kept at different permissive temperatures.


RESULTS


Characterization of the Carp/32-Microglobulin


Antiserum

We have previously reported the cloning of a full- length cDNA encoding carp/32-microglobulin (Cyca- B2m) (Dixon et al., 1993).Part (73 codons) of the Cyca-B2m cDNA was cloned in frame into the pRSET vector, and the resultant recombinant protein was used to immunize rabbits.The antiserum obtained was tested against the purified recombinant protein, membrane lysates of erythrocytes, and leuco- cytes in a western blot.The serum reacted strongly with a band of relative molecular mass Mr) of 11.0 kD from the recombinant protein extraction (Figure 1), and only very weakly with a band of Mr of 13.4 kD from the leucocyte membrane lysate.No reactivity was detected with membrane lysates of erythrocytes, A B 2 3 1 2 3 18.9 -6.8


FIGURE

Cell membrane lysates and recombinant Cyca-B2m protein were transferred to a nitrocellulose filter and the western blot was incubated with the Cyca-B2m antiserum.(A) Coomassie stained polyacrylamide gel; PBL membrane lysate (lane 1), erythrocyte membrane lysate (lane 2), and recombinant Cyca-B2m protein (lane 3).(B) Western blot of the same samples showing reactivity of the Cyca-B2m antiserum as indicated by triangles.Arrows indicate positions of molecular weight markers in kD.

although a large amount of protein was present in the molecular weight range studied.

The antiserum was also tested on live cells by FACS analyses (Figure 2).The preimmune serum staining showed only weak reactivity, being close to the range of the c

jugate o
ly.This preimmune reactivity could be reduced by absorption to erythro- cytes.The absorped antiserum was used in all subsequent experiments.The specific antiserum rec- ognized an antigen on the cell surface of carp peripheral blood leucocytes (PBL) labeling strongly the majority (66.5%) of the leucocytes (Figure 2A).Detailed FACS analyses, using double staining of the Cyca-B2m antiserum combined with monoclonal antibodies to Ig B cells (WCI12) and to thrombo- cytes (WCL6), was carried out on carp PBL (Figures 2B and 2C).The different percentages of the labeled PBL were determined from the FACS experiments.For double staining with the monoclonal against surface Ig, the percentages were Cyca-B2m+/WCI12 (42%), Cyca-B2m+/WCI12 (17%), Cyca-B2m-/ WCI12 (2%), and Cyca-B2m- WCI12-(37%).In the case of the thrombocyte marker, the following proportions were found: Cyca-B2 +/WCL6 (7%), Cyca-B2m+/WCL6 (57%), Cyca-B2m-/WCL6 (17%), and Cyca-B2m-/WCL6- (19%).

In order to study the expression of Cyca-B2m in major lymphoid organs (namely, thymus, spleen, and pronephros), leucocytes were isolated, labeled with the polyclonal anticarp/32m, and analyzed by FACS.No expression was detected on erythrocytes, as indicated by the fact that the staining of these cells was similar to the negative control, that is, conjugate only (Figure 3).The antiserum recognized leucocytes isolated from the thymus, pronephros, and spleen, but the level of expression differed among these cell populations.Pronephrocytes and splenocytes could be divided into three populations: a negative, dull-, and bright-positive population.In the pronephros, equal representation of negative and bright-positive populations was observed, whereas with splenocytes, the majority were found in the negative and dull-positive populations.Thymocytes comprised only bright- positive cells (Figure 3B).,: :i.'::.:"":" WCI 12 WCL 6 lla 4

FIGURE 2 Peripheral blood leucocytes (PBL) were labeled with the anti-/32m recombinant protein serum and the b nding of the polyclonal was detected with goat anti-rabbit IgG conjugated with PE (GAM-PE).(A) The fluorescence intensity histogram depicted represents: (1) conjugate only, (2) preimmune serum, and (3) immune serum.(B and C) Fluorescence intensity contour graphs of double-labeled PBL with the Cyca-B2m antiserum, combined with the monoclonal antibody to B cells (WCI12) and to thrombocytes (WCL6).For the detection of the Cyca-B2m antiserum and the monoclonal antibodies, GAM-PE and goat anti-mouse Ig conjugated to FITC (GAM-FITC) were used, respectively.

A B

.7, : ..! :1

q.

1 I lla 113 2 113 3

Log Fluorescence Intensity Log Fluorescence Intensity FIGURE 3 Leucocytes were isolated from cell suspensions of thymus, pronephros, and spleen.Erythrocytes were isolated from peripheral blood.Cells were labeled with the Cyca-B2m antiserum and the binding of the polyclonal was detected with GAM-PE.Fluorescence intensity histograms of cells labeled with the Cyca-B2m antiserum are shown. (A) (1) Erythrocytes control, (2) splenocytes control, (3) erythrocytes, and (4) splenocytes.(B) (1) Pronephrocytes control, (2) thymocytes control, (3) pronephrocytes, and (4) thymocytes.

Temperature Effect on/32-Microglobulin and MHC Class I Expression on Carp Leucocytes MHC class I expression has been shown to be critically dependent on association with /32m and peptide, and in RMA-S cell line, this expression can be modulated by temperature (Ljunggren et al., 1990).In this study, the effect of three different temperatures (24C, 12C, and 6C) on in vivo cell-surface expression of MHC molecules was investigated.For this purpose, the Cyca-UA and Cyca-B2m polyclonal antiserum were used in conjunction with the WCI12 monoclonal antibody as a control (Figure 4).The antiserum raised against recombinant Cyca-UA, a locus known to be expressed, may not recognize all carp class I molecules, whereas the Cyca-B2m polyclonal antiserum is expected to give an indirect estimation of the levels of all class I cell-su face expression.

Carp PBL were isolated at different time points from the different experimental groups and labeled with the polyclonal against Cyca-B2m and Cyca-UA.

In the first experiment, carp were transferred from their standard rearing temperature of 24C to 12C.

Subsequently, they were exposed to this temperature for 2 weeks and then moved back to 24C.Fish living at standard conditions of 24C show the typical histograms for the three antibodies used as depicted in Figures 4 and 5 under day 0. After 3 days of exposure to 12C, the levels of expression of Cyca-B2m and Cyca-UA were reduced, but recovered to control levels after 6 days.The expression of surface Ig during this period was constant.Transfer of carp from 12C back to 24C did not change all of the expression patterns studied to any significant degree (Figure 4).In the second experiment, the carp were moved from 24C to 12C, and finally to 6C.The fish remained at this temperature for 2 weeks bef re being transferred back to 12C.Fish kept at 6C showed, after 3 days at this temperature, reduced levels of Cyca-B2m and Cyca-UA expression.However, after 6 days, all cell-surface expression studied is los and remained absent for prolonged periods, except Ig.

After transferring the carp to 12C, t

y recovered to normal levels within 6 days at this temperature.Cell- surface expression
f surface Ig again remained constant during the experimental period Figure 5).However, absence of Cyca-B2m and Cyca-UA at 6C has been observed for a period up to 4 weeks.

Temperature Effect on/32-Microglobulin and MHC Class I Transcription in Carp Leucocytes

In the previous experiment, a number of animals were kept at the three different temperatures (24C, 12C and 6C) for 4 weeks and samples for transcription analysis were taken.For each temperature group, PBL from two animals was isolated, RNA was extracted, and DNA prepared.A PCR using specific primers for Cyca-B2m and Cyca-UA genes was carried out.The concentrati ns of the Cyca-B2m and Cyca-UA transcripts in PBL were compared between groups by evaluating the yield obtained at different cycle numbers (Figure 6).

Within temperature groups, no clear difference in amplification was found both for Cyca-B2m and


24C 12C

Cyca-UA.However, between groups, a clear differ- ence in yield of amplification was found for Cyca-B2m, but not for Cyca-UA.In the case of Cyca-B2m (Figure 6A), the PCR products generated from the cDNA of the 6C temperature group were not detectable within 20 cycles, and lower yields were seen with 25 and 30 cycles when compared to the products generated in both the 24C and 12C temperature groups.No visible difference was fo

d between t
e amount of PCR products from PBL of fish living at 24C and 12C.In contrast, in all temperature groups, the yield of Cyca-UA (

gure 6B)
enerated by the PCR amplification increased propor- tionally with the number of cycles (20, 25, and 30).


DISCUSSION

The expression of MHC molecules and their tissue distribution cannot be dissociated from age and 24 C day 0 day 3 day 6 day 3


Cyca-IgM

Cyca-B2m

i,Cyca-UA FIGURE 4 A group of carp was kept at to two different temperatures.Fish maintained at the normal temperature of 24C were transferred to 12C and were kept at this temperature for 2 weeks.After this period, the fish were transferred back to 24C.Carp were bled at regular intervals, PBL were isolated and labeled wi h the Cyca-B2m and the Cyca-UA antisera.As a control, the WCI12, a monoclonal antibody against carp Ig was used.Labeled PBL from carp kept at different temperatures were analyzed by FACS and the relevant fluorescence intensity histograms are depicted.

immunological state of the species considered; how- ever, it is generally accepted that class I molecules are present on the cell surface of most somatic cells, whereas class II expression occurs mainly on the cells of the immune system (Klein, 1986).With the recent isolation of carp B2m gene (Cyca-B2m) from a spleen/pronephros cDNA library (Dixon et al., 1993), it became possible to study the expression levels and tissue distribution of this gene.PCR amplification of cDNA obtained from leucocytes from spleen, prone- phros, blood, and thymus confirmed the presence of transcripts in these organs (Rodrigues et al., in preparation).However, due to the limitations of this type of analysis, a polyclonal antiserum was produced against Cyca-B2m recombinant protein.

The Cyca-B2m antiserum showed substantial reac- tivity against the recombinant protein, representing only part of the native/32m molecule.However, less reactivity could be detected with cell-surface /3_m, 24C 6C

which as expected ran at a higher Mr. The weak reactivity might be explained because only approxi- mately 50% of PBL is Cyca-B2M+.

Because it is was hypothesized that Cyca-B2m is associated with all carp class I gene products, it was expected that all nucleated cells would be positive with the serum against carp/32m.However, although nucleated erythrocytes and thrombocytes were anti- Cyca-B2m negative, and thus also class I negative.This has been confirmed by using antibodies to Cyca-UA class I molecules (Van Erp et al., 1996a).These findings are in disagreement with the distribution of Xenopus class I molecules, which appear at the time of metamorphosis to adult stage on leucocytes and erythrocytes (Flaj

987).Also
in chicken, the B-F encoded molecules are expressed on nucleated erythrocytes (Mller et al., 1991).Pre- viously, a serologically defined major histocompatibility locus K in carp has been described (Kaastrup et 12C Cyca-IgM


Cyca-B2m

Cyca-UA day 0 day 3 day 6 day 6

FIGURE 5 A group of carp was subjected to three diflcvcnt tcmpcratures.Fish reared at the normal temperature of 24C were transferred to 12C and subsequently to 6C, and were kept at this temperature for 2 weeks.After this period, the fish were transferred back to 12C.

Carp were bled at regular intervals, PBL were isolated and labeled with the Cyca-B2m and the Cyca-UA antisera.As a control, the WCI12, a monoclonal antibody against carp Ig was used.Labeled PBL from carp kept at different temperatures were analyzed by FACS and the relevant fluorescence intensity histograms are depicted.

A B 1 2 3 4 5 6 3O 1 2 3 4 5 6 25 2O FIGURE 6 Carp were kept for 4 weeks at three different temperatures (24C, 12C, and 6C).After this period, PBL were isolated from the three temperature groups.RNA was extracted, cDNA prepared, and PCR with specific primers for Cyca-B2m and the Cyca-UA was carried out and aliquots visualized by agarose gel electrophoresis.(A) PCR products for the Cyca-B2m amplification after different cycle numbers.(B) PCR products for the Cyca-UA amplification.Lanes to 6 depict PCR yields from individual carp kept at different temperatures: and 2 (24C); 3 and 4 (12C); and 5 and 6 (6C).

al., 1989), which alleles correlated with transplant rejection, and thus were thought to ressemble class I alleles.It is clear, from the lack of/32m expression on erythrocytes as described in this study, that the K locus is not a class I locus, but a locus that codes for an alloantigen perhaps closely linked to one of the class I major histocompatibility complex loci described in carp (Van Erp et al., 1996b).

The fluorescence histogram of labeled carp PBL showed dulland a bright-positive populations.With the help of the only two monoclonal antibodies against carp leucocytes available, the B cell (Secombes et al., 1983) and thrombocyte markers (Rombout et al., 1996), the nature of the different populations was further investigated.The double labeling of PBL revealed that the WCL6 cells stained only weakly with the Cyca-B2m antiserum (Figure 2C).In this experiment, a population is seen that is Cyca-B2M-/WCL6-, which represents contaminat- ing erythrocytes.This lack of Cyca-B2m expression on carp thrombocytes is in contrast to what is known of expression of class I for the mammalian counter- parts of thrombocytes, the platelets (Klein, 1986) and the chicken thrombocytes (Pink et al., 1985).It was also apparent from the PBL double labeling that almost all Ig lymphocytes (WCI12/) are Cyca-B2m bright-positive.There was also an Igpopulation observed that was also bright-positive, which accounts for 17% of the PBL, and are most probably putative peripheral T cells.

To investigate the tissue distribution of Cyca-B2m molecules, cells from lymphoid organs were labeled with the Cyca-B2m antiserum.Erythropoiesis and trombopoiesis occur mainly in the spleen (Rombout et  al., 1996), so the proportion of thrombocytes and several stages of erythrocytes in the density-isolated cell fraction is probably high.The large proportion of dull-positive cells isolated from the spleen are most likely accounted for by the presence of thrombocytes and erythrocytes.A similar staining pattern was found with an antiserum that recognizes Cyca-UA class I molecules (Van Erp et al., 1996a).The pronephros is thought to be the hemopoietic organ in fish and therefore equivalent to the mammalian bone marrow, as well as a peripheral immunological organ contain- ing numerous antibody-producing cells after antigenic stimulation.In such tissue, a large proportion of undifferentiated cells are present, so the dull-positive cells most likely represent the developmental stages of different leucocyte lineages, and the bright-positive cells are the mature Ig and Ig-leucocytes.In chicken bone marrow and spleen, a similar distribution of class I (B-F) dull-and bright-positive cell populations has been reported (Dunon et al., 1990).The fluores- cence histogram of the thymocytes revealed in carp just one type of cells, all of which were brightpositive.It has been demonstrated that in frogs, birds, and mammals, mainly mature thymocytes are bright- positive for class I, whereas cortical are dull-positive (Klein, 1986).A similar situation where two popula- tions are detected in the thymus has been observed with Cyca-UA antiserum Van Erp et al., 1996a).This discrepancy may be explained because subpopulations of thymocytes use different class I ce chains from different lineages (Van Erp et al., 1996b), which are all probably noncovalently associated with /32m (Rodrigues et al., in preparation).

In the next set of experiments, we addressed the effect of temperature on the expression of the /32m and Cyca-UA class I molecules.It has been demon- strated that both/32m and peptide are instrumental in the proper folding of the trimolecular structure of class I molecules (Elliot, 1991).There are, however, instances in which eithe of the two requirements can be circumvented.One of the best known example is the RMA-S cell line, in which in the absence of peptides, class I expression can be rescued by lowering the temperature (Ljunggren t al., 1990).

The/32m molecule seems to be more essential for a proper class I expression (Zijlstra et al., 1990), although low levels of refolding with exogenous/32m and peptide have been observed (Vitiello et al.,  1990).

The experiments described in this study clearly showed that lowering the ambient temperature of the carp resulted in a decreased level of/32m and class I expression, which remained undectable when carp were kept at 6C. Recovery to normal expression levels, after transfer to a higher temperature, was achieved in all groups, albeit with a slower rate in he 6C group.In contrast, no changes were observed in the expression of the Ig on the B cells.The latter observation is consistent with the fact that temperature effects have been noted for T-cell but not B-cell functions (Bly and Clem, 1992).

The question that arises is whether the observed downregulation of class I molecule expression is due to (1) a failure to transport the molecules to the cell surface, and (2) the inability to fold the trimolecular structure due to the absence of one or more of the constituents.The transport system of macromolecules does not seem to be impaired at low permissive temperatures because normal levels of Ig molecules have been observed on the B cells in the different temperature groups.Thus, the absence of either peptide,/2m, or class I molecules seems to be a likely cause.Although peptide is required to fold a func- tional class I molecule, low temperature would allow for the expression of empty class I molecules (Ljunggren et al., 1990).In addition, the fact that antigen uptake and processing have been observed in fish at low permissive temperatures, although at a slower rate, would argue against an absence of protein-processing activities (Vallejo et al., 1992).The latter study also showed that exogenous antigens are being presented, which seems to suggest in fish a different effect of temperature on expression of class II molecules.Although peptides bound to class I or class II molecules follow differ nt processing routes (Brodsky et al., 1996), it seems unlikely that the class I route has a different susceptibility to temperature changes compared to class II, which in fish seems to function only at a slower rate at lower temperatures.

The remaining possibility of absence of either the /32m or the class I a-chain molecule was investigated by semiquantative PCR.This experiment clearly indicated a downregulation of Cyca-B2m gene tran- scription only at 6C. Cyca-UA gene transcription is unaffected at all temperatures.Therefore, the absence of normal transcription levels of the Cyca-B2m gene accounts for the downregu ation of class I molecules on the cell surface.This observation is reminiscent of /32m knockout mice, where also normal levels of class I transcripts can be found in the absence of class I cell-surface expression (Zijlstra et al., 1990).

In conclusion, temperature changes within permis- sive ranges can result in temporal or longlasting changes in class I expression, which is determined by a temperature-sensitive transcription regulation mechanism of the/32m gene.This observation could account, in part in aquacultural practices, for the observed "immunological disasters" after severe and sudden temperature changes (Bly and Clem, 1992).This applies especially to those cases where CTL responses are required for a proper immunological response to the causative agent, such as might be the case in "winter saprolegniosis" (Bly et al., 1992) or perhaps more importantly in viral infections.gen, Madison, WI) and recombinant protein produc- tion was induced using 0.8 mM IPTG.The recombi- nant B2m protein was purified by nickel-affinity chromatography using the Xpress system (Invi- trogen).The recombinant B2m protein was further purified over Sephadex G-100 to remove contaminat- ing bacterial proteins.Under denaturating conditions, SDS-PAGE was used to determine the purity and size of the recombinant protein.Four hundred micrograms of the protein were conjugated to 2 mg of KLH by glutaraldehyde and w

e injected i.m. into a rabbit at two si
es in combination with complete Freund's adjuvant.The animal was boosted with antigen and incomplete Freund's adjuvant after month and bled at 2 months postinjection.Serum was collected after clotting an stored at -80C for future use.


MATERIALS AND METHODS

Western Blotting


Animals

Common carp (Cyprinus carpio L.) were reared at 23C in recirculating UV-sterilized water, and fed pelleted dry food (Provimi, Rotterdam, The Ne

er- lands) at a ratio of 2% of body weight per day.

Anima
s from a single F1 hybrid family R3 R8, weighing between 150-250 g (10-16 months), were used.R3 and R8 are partly inbred strains of common carp originating from Poland and Hungary, respec- tively (Wiegertjes et al., 1994).


Production of Recombinant Protein and Polyclonal Antisera

A PCR fragment was produced by anchored PCR using the previously described Cyca-B2m cDNA clone in conjunction with the specific primer OL-85  (Dixon et al., 1993).This fragment, which contained the last 73 codons and the 3' untranslated region of the clone, was cloned into the vector pCRII and sequenced using the Sequenase version 2.0 kit (U.S.

Biochemical).Following sequence confirmation, this fragment was excised from the vector using EcoRI and ligated in frame into the vector pRSET (Invi- trogen, Leek, The Netherlands).The plasmid was transformed into bacterial strain BL21-DE3 (Nova-Membrane lysates were prepared from both peri- pheral blood leucocytes and erythrocytes.Briefly, cells (108) were subjected to three consecutive freeze- thaw cycles in PBS, adjusted to 270 mOsm, contain- ing mM MgCI2.After each cycle, membrane fragments and nuclei were collected by centifugation at 13,000 g.After the last cycle, membranes were solubilized in lysisbuffer (10 mM Tris/HC1, pH 8, mM MgC12, 150 mM NaC1, 0.1 mM protease inhibitors, and 1% CHAPS).Nuclei were removed by centrifugation at 13,000 g and lysates stored at -80C for futher use.The recombinant protein was isolated on a preperative SDS polyacrylamide gel (Prep-gel, Biorad).The protein samples, obtained after this producere, were separated on a 10-20% SDS poly- acrylamide gel.The protein was transferred to a nitrocellulose f

ter by electrob
otting and the filter was incubated with the polyclonal antisera at an appropriate dilution.Reactivity of the rabbit serum was determined using an alkaline phosphatase conju- gated goat anti-rabbit immunoglobulin and visualized using and NBT and BCIP substrate buffer.


Cell Isolation

Fish were anesthetized in tricaine methane sulphonate (TMS, Crescent Research Chemicals, Phoenix, AZ) at 3 g/10 l, and heparinized blood was collected from the dorsal aorta.This was diluted 1:1 in cRPMI (RPMI 1640 adjusted to 270 mOsm), and peripheral blood leucocytes were separated on Lymphoprep (Nycomed, Oslo, Norway) by centrifugation at 840 g for 30 min at 4C.The cells were harvested from the interface, washed twice, and resuspended in cRPMI.

The peripheral blood leucocytes were used for further analyses.Pronephros, spleen, and thymus-cell suspen- sions were prepared by forcing the tissues through a 50-mesh nylon gauze filter w

le adding cRPMI.After
eing washed in cRPMI (680 g for 10 min at 4C) and resuspended, the cell suspensions were separated on Lymphoprep by centrifugation at 840 g for 30 min at 4C.The cells were harvested, washed twice, and resuspended at a concentration of 10 7 cells/ ml in cRPMI.


Flow Cytometry (FACS)

For the flowcytometry studies, the following anti- bodies were used: the polyclonal described in this paper raised against recombinant Cyca-B2m protein; an anti-class I (Cyca-UA) polyclonal serum (Van Erp et al., 1996a); WCI12, a monoclonal antibody that detects carp surface immunoglobulin (Secombes et  al., 1983); WCL6, a monoclonal antibody specifically recognizing carp thrombocytes (Rombout et al.,  1996).Periphera blood leucocytes (PBL), isolated as described earlier, were incubated for 30 min on ice in 0.5 ml of appropriately diluted (usually 1:100) Cyca-UA or Cyca-B2m polyclonal antibodies.For all the incubation and washing steps, FACS medium con- taining cRPMI, 1% BSA, and 0.1% NaN3 was used.

After washing, binding of the polyclonal antibodies was detected by incubating the cells for 15 min on ice with phycoerythrine (PE) conjugated goat anti-rabbit immunoglobulin antibody (GAR-PE; Southern, Bir- mingham, AL), diluted 1:100 in FACS medium, containing 1% of pooled carp serum.For the PBL double labeling, leucocytes were incubated with the Cyca-B2m polyclonal antiserum together with WCI12 or WCL6, followed by an incubation with GAR-PE and fluorescein isothiocyanate (

TC) con- jugated goat an
i-mouse immunoglobulin serum (Dakopatts, Denmark).Cells were analyzed using a FACStar (Becton Dickinson Immunocytometry Sys- tems, Mountain View, CA) with an argon laser tuned at 488 nm.For analyzing the FACS results, the Consort 30 data analysis package was used.

Temperature Experiments A group of eighteen carp from the same F1 generation (R3 R8) were kept at three different temperatures.The control group of six individuals was maintained at normal rearing conditions, that is, at a temperature of 24C throughout the experiment.The remaining twelve individuals were kept at 12C.After 3 days, six animals from the 12C group were transferred to 6C.The resulting three groups of six animals were kept at these temperatures for at least 2 weeks.After this period, the carp at 12C were transferred to 24C, and the animals at 6C were placed at 12C.Experimental fish were bled at regular intervals, and isolated PBL were labeled with different sera and analyzed by FACS.Small numbers of carp were kept at the three different temperatures for up to month, after which they were sacrificed for further transcrip- tion analyses using RNA isolation procedures.

RNA and cDNA Preparation RNA and cDNA preparation for the use in semi- quantitative PCR analyses have been performed as described by Rodrigues et al. (1995).Briefly, cells  were thawed out in lysis buffer (4 M guanidium thiocyanide, 25 mM sodium citrate, pH 7.0, 0.5% sarcosyl, 0.1 M 2/3-ME) followed by phenol/chloro- form extractions.Total RNA was precipitated in ethanol, washed and dissolved in water, and stored at -80C.Samples containing 10 mg of total RNA were converted into cDNA using the Riboclone cDNA Synthesis System (Promega, Madison, WI).Effi- ciency of cDNA synthesis was traced by determining the incorporation of fluorescein-dUTP in a parallel reaction.The labeled cDNA samples were serially diluted and blotted onto a nylo