Molecular Immunology 44 (2007) 2737–2748

Molecular and biochemical analysis of rainbow trout LCK suggests a conserved mechanism for T- signaling in gnathostomes Kerry J. Laing a,1, Stacey Dutton b, John D. Hansen a,c,∗ a Department of Pathobiology, University of Washington, Seattle, WA 98195, USA b Department of Neuroscience, Emory University, Atlanta, GA 30322, USA c USGS-Western Fisheries Research Center, Seattle, WA 98115, USA Received 28 September 2006; received in revised form 16 November 2006; accepted 18 November 2006 Available online 18 December 2006

Abstract Two genes were identified in rainbow trout that display high sequence identity to vertebrate Lck. Both of the trout Lck transcripts are associated with lymphoid tissues and were found to be highly expressed in IgM-negative . In vitro analysis of trout lymphocytes indicates that trout Lck mRNA is up-regulated by T-cell mitogens, supporting an evolutionarily conserved function for Lck in the signaling pathways of T-lymphocytes. Here, we describe the generation and characterization of a specific monoclonal antibody raised against the N-terminal domains of recombinant trout Lck that can recognize Lck (s) from trout thymocyte lysates that are similar in size (∼57 kDa) to mammalian Lck. This antibody also reacted with permeabilized lymphocytes during FACS analysis, indicating its potential usage for cellular analyses of trout lymphocytes, thus representing an important tool for investigations of salmonid T-cell function. Published by Elsevier Ltd.

Keywords: Lck; T ; Rainbow trout; Salmonid

1. Introduction lates immunoreceptor -based activation motifs (ITAMs) in the tails of the CD3 and TCR-␨ chains of the TCR complex. Cell mediated immunity is initiated when foreign peptides These phosphorylated ITAMsbecome high affinity docking sites () are presented to specific receptors (TCRs) for another , ZAP-70, that is also phosphorylated and acti- on T-lymphocytes by major histocompatibility complex (MHC) vated by Lck (Chan et al., 1995; Wange et al., 1995). Activated molecules on presenting cells (APCs). This process is ZAP-70 molecules autophosphorylate and provide docking sites well defined for mammalian lymphocytes. The TCR interacts for other signaling molecules such as the membrane adaptor pro- also with other molecules, such as CD3 and TCR-␨ and the tein LAT (linker for activation of T-cells), which in turn, recruits co-receptors CD4 and CD8, to co-ordinate the recruitment and and activates additional to initiate C and activation of various involved in the signaling pathways the MAP-kinase pathways that are fundamental to T-cell differ- of T-cells. Among these kinases is the lymphocyte cell kinase entiation, proliferation and effector functions (Sommers et al., (Lck) that belongs to the Src family of protein tyrosine kinases 2004). (Williams et al., 1998; Robinson et al., 2000). Lck phosphory- Lck is composed of four major subunits; an N-terminal unique domain, two Src homology domains SH3 and SH2, and a kinase domain (SH1); followed by a short regulatory tail. ∗ Corresponding author at: USGS-Western Fisheries Research Center, 6505 Activation and regulation of Lck occurs as a result of differen- NE, 65th Street, Seattle, WA 98115, USA. Tel.: +1 206 526 6588; tial of two important . One (Tyr505 in fax: +1 206 526 6654. human Lck) is located in the short C-terminal tail that in its phos- E-mail addresses: [email protected], [email protected] phorylated form, binds to the SH2 domain of Lck to maintain (J.D. Hansen). 505 1 Present address: Fred Hutchinson Cancer Research Center, 1100 Fairview Lck in an inactive state. Tyr is phosphorylated and dephos- Avenue North, D3-100, PO Box 19024, Seattle, WA 98109, USA. phorylated by the activities of the C-terminal Src kinase (Csk)

0161-5890/$ – see front matter. Published by Elsevier Ltd. doi:10.1016/j.molimm.2006.11.021 2738 K.J. Laing et al. / Molecular Immunology 44 (2007) 2737–2748 and CD45, respectively (Hermiston et al., 2002). Dephosphory- Lck, suggesting they will have similar protein:protein interac- lation of Tyr505 results in dissociation of the SH2 domain and tions. To support this, chicken Lck associates with both avian activation of Lck. The other essential tyrosine (Tyr394 in human CD4 and CD8 (Veillette and Ratcliffe, 1991; Chow et al., Lck) resides in an activation loop between the two lobes of the 1992). kinase domain. Autophosphorylation of Tyr394 causes a confor- T-cell populations of teleost fish are relatively undefined mational change in Lck that is essential for its kinase activity compared to mammals. Although increasing evidence shows (Holdorf et al., 2002; Giannini and Bijlmakers, 2004). In addi- the conservation of molecules associated with T-lymphocytes, tion, the SH3 domain interacts with the linker region between such as the T cell receptor (TCR), CD3, CD8 and CD4, in the SH2 and kinase domains to maintain the inactive state of the these lower vertebrates, cellular reagents capable of defining Lck (Hofmann et al., 2005). these critical molecules have yet to be developed. One of the The N-terminal unique sequence of Lck distinguishes it from major teleost models of immunity is rainbow trout, a commer- other Src family members and is responsible for its functional cially important species of the Salmonidae family. In order to specificity (Marth et al., 1985; Turner et al., 1990). This unique improve the availability of molecular reagents for lymphocyte region allows Lck to associate with the cytoplasmic tails of CD4 analyses in salmonids, it is important to identify and function- or CD8 via non-disulphide related protein-protein interactions ally characterize relevant lymphocyte specific markers for trout. between four cysteine residues (CXXC in Lck; CXC in CD4 and Our previous studies described CD8␣ and CD4-like molecules CK8␣)inaZn2+-dependent manner (Huse et al., 1998; Lin et al., from rainbow trout (Hansen and Strassburger, 2000; Laing et 1998). This in turn results in recruitment of Lck to the TCR com- al., 2006) that appear to contain important amino acids for plex, and the initiation of T cell signaling. Localization of Lck interacting with Lck. Prior to this study, however, no sequence to the cell membrane occurs via myristylation (Gly2, Ser6) and information was available for salmonid Lck, thus, we describe palmytilation sites (Cys3) within the N-terminal unique domain here the identification and analysis of two Lck genes from (Yasuda et al., 2000). rainbow trout. Lck-GST fusion proteins were also expressed Lck is well conserved throughout jawed vertebrates, with for the generation of monoclonal antibodies against trout Lck highly conserved gene sequences reported for chicken Gallus that were used in biochemical and cellular analyses of trout gallus (Chow et al., 1992), pufferfish Fugu rubripes (Brenner et lymphocytes. al., 2002) and zebrafish Danio rerio (Langenau et al., 2004) that are abundantly expressed in T-cell rich tissues such as thymus 2. Materials and methods and spleen. Important cysteine and tyrosine residues identified for Lck function in mammals are retained in non-mammalian 2.1. Identification of Lck transcripts in a rainbow trout cDNA library

Table 1 A cDNA probe containing nucleotides 30–980 of human Lck Primers and probes transcript (Perlmutter et al., 1988) was used to screen a rainbow Primer name Primer sequence 5–3 trout thymus cDNA library as previously described (Hansen LCK-F1 GARCCIGARCCITGGTTYTT et al., 1997) using moderate stringency. From 57 hybridizing ◦ LCK-R4 TTDATIGTRAAIGTNCCRTA clones, 18 were randomly selected for further 2 screening by LCK-23387 GAAATACAACAGGCTCCTTCA PCR using primers LCK-23387 and LCK-23388. These primers LCK-23388 AGGACGACTTGAGAAAATCTAC had been designed against sequences from a trout Lck cDNA LCK2-F1 GGCTCCTTCTCCTTATCTGTC fragment that was amplified by degenerate PCR using LCKF1 LCK2-R1 TCAGGGCAGTTCTCAGGTC Lck1-115F CTTAGTTGCTGTTTTTGATAGGATTACTAAA and LCKR4 (Table 1) with thymic cDNA resulting in a 928 bp Lck1-231R CTTCAACAGGGTGTCAGCGA amplicon. The fragment was cloned and sequence verified. Lck2-67F TTGCTGTTTTTGATAAGATTTATATTTGATGA PCR parameters were as described for RT-PCR below. Three Lck2-188R AAGCCTAATATAATGCTCTTCAACAGGG clones were identified from the cDNA library that contained LCK-23489 AGAGAGAATTCATGGGATGCAACTATAGT- Lck. One clone containing the full-length cDNA for trout Lck1 TCAGATTATGAT LCK-23491 AGAGACTCGAGTTAGCAAGGCTGTTCTTG- was sequenced to completion using universal (T3/T7) vector GTACTGCCCCTCTGT primers and one gene specific primer, LCK-F1. Subsequently, ARP-For GAAAATCATCCAATTGCTGGATG routine random sequencing of a thymus cDNA library identi- ARP-Rev CTTCCCACGCAAGGACAGA fied a second distinct Lck clone designated as Lck2, which was IgM-MEM-F1 AAAGCCTACAAGAGGGAGACCGAT sequenced to completion using universal primers with LCK2-F1 IgM-MEM-R1 AGAGTTATGAGGAAGAGTATGATGAAGGTG TCR-23743 CAGCTTGAAGTCAAGAAATAC and LCK2-R1 (Table 1). TCR-23744 TATCAGCACGTTGAAAACGAT 2.2. Animals and RNA extraction Taqman probes Probe sequence 5–3

LCK1-177T ATGGATAAAGTTGCTGCTCTGGCTGTAGACCAGTGAA Outbred rainbow trout (Clear Springs Foods Inc., Buhl, LCK2-104T AGGTCTCTTAGTCATGGATAAAGCTGCTCTGGCTG Idaho) were used unless otherwise stated and maintained at a ARP-T CTATCCCAAATGTTTCATTGTCGGCGC constant temperature of 15 ◦C in sand-filtered and UV-treated Note:R=AorG,N=G,A,CorT,Y=CorT,D=A,GorT. freshwater at the WesternFisheries Research Center, Seattle WA. K.J. Laing et al. / Molecular Immunology 44 (2007) 2737–2748 2739

Trout weighing 150–300 g were euthanized in buffered MS-222 2.6. Reverse transcription-polymerase chain reaction and various tissues removed immediately, snap frozen in liquid (RT-PCR) nitrogen and stored at –80 ◦C until required for RNA extrac- tion. Frozen spleen tissues for homozygous OSU-142 (OSU), First strand cDNAs were generated as previously described Hotcreek (HC) and Arlee (ARL) trout were kindly supplied by (Hansen et al., 1997). Reaction volumes were made up to 100 ␮L Dr. Gary Thorgaard (Washington State University). Total RNA by the addition of 80 ␮L sterile water. PCR was performed was isolated from tissues using the Trizol (Gibco-BRL) method using 1 U Amplitaq DNA polymerase (Applied Biosystems) in as previously described for Northern analyses (Hansen et al., 25 ␮L reactions containing 12 pmol each of forward and reverse 1997) or using the RNeasy (Qiagen) with in column DNase primers, 400 ␮M dNTP mix (Promega, UK), 1× PCR buffer treatment as per manufacturer’s instructions for PCR analyses containing MgCl2 (1.5 mM) (Applied Biosystems) with 1 ␮L (Laing et al., 2006). cDNA template. The cycling protocol was 1 cycle of 94 ◦C for 2 min, 30–38 cycles of 94 ◦C for 30 s, 50–60 ◦C for 30 s 2.3. Isolation and stimulation of trout peripheral blood and 72 ◦C for 1 min, with a final extension step of 5 min at leucocytes 72 ◦C.

Lymphocytes were isolated as previously described (Laing 2.7. Quantitative RT-PCR analysis of Lck1 and Lck2 et al., 2006). 5 × 106 cells were seeded in 24-well plates expression with 500 ␮L L15 medium (Gibco), containing 10% FCS and 50 ␮g/mL kanamycin and incubated at 15 ◦C. Following a Quantitative PCR (qPCR) was used to estimate levels of Lck1 14 h recovery period, PHA (20 ␮g/mL), PMA (10 ng/mL) or and Lck2 mRNA expression in different tissues of na¨ıve rainbow PBS, in another 500 ␮L L15 medium, were added to the trout and isolated peripheral blood leucocytes (PBL) following cultures. At various time points (0, 4, 8, 12 and 24 h) the in vitro stimulation. Tissues and PBL cDNAs were generated media was removed and the cells were harvested by direct from total RNA isolated using the RNeasy kit (Qiagen) as previ- addition of 350 ␮L RLT buffer (Qiagen RNeasy kit) contain- ously described (Laing et al., 2006). Primers and Taqman probes ing 1% ␤-mercaptoethanol. Of the resulting total RNA, 1 ␮g sets for qPCR were designed using the Primer Express Soft- was used to synthesize first strand cDNA in 20 ␮L volumes ware (ABI v2.0; Applied Biosystems). See Table 1 for primer as described previously (Hansen et al., 1997). First strand and probe combinations. Primer sets were confirmed for speci- cDNAs were diluted 1:10 for all subsequent quantitative PCRs ficity by cloning and sequencing amplified products as described (qPCR). above. Acid ribosomal phosphoprotein (ARP) was used for nor- malization as described previously (Purcell et al., 2004). Each 2.4. Northern blot analysis probe was synthesized, dual-labeled (5-FAM/3-TAMRA) and QPCR assays were performed using the ABI PRISM 7900HT For Northern analysis, mRNA was isolated from total RNA sequence detection system as previously described (Laing et using the mRNA isolation kit (Qiagen). The purified mRNA al., 2006). Treatment groups were compared to control groups (0.5 ␮g) was separated by electrophoresis under denaturing from the same time point using the Wilcoxon signed rank (non- conditions and blotted onto a nylon membrane. Overnight parametric) statistical test on the normalized hybridizations with a trout specific radiolabeled Lck probe was levels. performed under stringent conditions as previously described (Hansen et al., 1997). After exposure, the blots were stripped in 2.8. Lck expression in separated sIgM+ and sIgM− 1% glycerol (95 ◦C) and re-probed for the rainbow trout EfTu- lymphocytes 1 housekeeping gene as previously described (Hansen et al., 1997). Lymphocytes isolated from pronephros and PBL were sep- arated into surface IgM-positive (sIgM+) or IgM negative 2.5. Probes (sIgM−) cells by flow cytometry and subjected to RNA extrac- tion and cDNA synthesis as described previously (Laing et cDNA probes used during library screening and Northern al., 2006). RT-PCR analyses were performed for several genes, analyses were generated by PCR from cloned and sequence ver- including LCK1 (Lck1-115F/231R), LCK2 (Lck2-67F/188R), ified templates. Probe template (1 pg) was amplified in 1× PCR TCR-␣ chain (TCR-23743/TCR-23744), and the membrane- buffer (Applied Biosystems) supplemented with 1.5 mM MgCl2, bound form of IgM (IgM-MEM-F1/R1) in addition to the 200 ␮M dNTPs and 1 U of Amplitaq (Applied Biosystems) housekeeping gene ARP (ARP-For/ARP-Rev) (see Table 1 for using 30 cycles consisting of 94 ◦C for 10 s, 60 ◦C for 30 s and all primer sequences). Standard RT-PCR conditions were used 72 ◦C for 30 s. Amplified products were purified over Qiaquick as described above. columns (Qiagen), prior to random labeling with (32P)-dCTP (Amersham) using the Random Primers DNA Labeling System 2.9. Cloning, plasmid DNA isolation and sequence analysis (Gibco-BRL). Unincorporated nucleotides were removed from the labeled probes using sephadex G-50 spin columns (Roche Products generated by PCR were ligated into the pCR2.1- Applied Science). TOPO vector (Invitrogen) and three randomly selected clones 2740 K.J. Laing et al. / Molecular Immunology 44 (2007) 2737–2748 representing each product were sequenced using vector or gene 2.12. Generation of cell lysates specific primers. Comparisons of nucleotide and amino acid sequences with EMBL and SWISSPROT databases were per- Single cell suspensions of trout thymocytes or splenocytes formed using BLAST (Altschul et al., 1997). Direct comparisons were obtained by passing homogenized tissues through 40 ␮m between two sequences were performed using the NEEDLE cell strainers (BD Falcon) with PBS. Single cell suspensions global alignment program (Needleman and Wunsch, 1970)of were lysed at 2 × 107 cells/mL in 1% Digitonin (Sigma), 50 mM EMBOSS (Rice et al., 2000). Multiple sequence alignments Tris–HCl (pH 8.0) (Sigma) and 150 mM NaCl (Sigma), sup- were generated using CLUSTALW (v1.82) (Thompson et al., plemented with Complete protease inhibitor cocktail (Roche 1994). Phylogenetic trees were constructed from CLUSTALW Applied Science). A similar procedure was performed to obtain generated alignments using the Neighbor-joining method lysates of the human Jurkat T-cell line (ATCC) that had been (Saitou and Nei, 1987), within the MEGA 3.1 program (Kumar grown in Dulbecco’s modified Eagle medium (DMEM; Invitro- et al., 2004) with Poisson correction, deletion of gaps and boot- gen) supplemented with kanamycin and 10% FBS. Following strapped 1000 times. lysis for 30 min at 4 ◦C, the lysates were frozen at −80◦C until use. 2.10. Expression and purification of recombinant trout Lck To determine the effect of mitogen stimulation on trout Lck expression, single cell suspensions of thymocytes (1 × 107) Trout Lck1 cDNA was amplified using LCK-23489 and were incubated for 8 h at 15 ◦C in the presence of either PHA LCK-23491 primers that annealed at the 5 end of the SH4 (20 ␮g/mL), PMA (10 ng/mL) or with an equivalent volume of and near the 3 end of the SH2 domains, respectively. These PBS (Con) prior to lysis as described. primers incorporated EcoRI and XhoI restriction sites to facilitate directional cloning into pGEX-4T1 (Amersham Bio- 2.13. Western blot analysis sciences/GE Healthcare) for generating a fusion protein with a GST tag at the N-terminus (GST-LCK). The pGEX-4T:LCK Recombinant trout Lck or cell lysates (∼5 × 105) were expression vector was transformed into TOP 10 E. coli. Recom- electrophoresed through 10% SDS-polyacrylamide gels under binant protein expression of GST-LCK was induced for 4 h either reducing or non-reducing conditions. Separated proteins in 1 L cultures in LB broth using IPTG (1 ␮g/mL) follow- were electrotransferred to Hybond-P membrane (Amersham) ing an initial 2-h incubation at 37 ◦C with shaking. GST-LCK and after blocking overnight in PBS-Tween (PBS-T, 0.1% was purified using glutathione sepharose-4B (Amersham Bio- and 5% blocking milk (Biorad), membrane strips were rinsed sciences/GE Healthcare) using the GST microspin purification (PBS-T) and incubated with gentle agitation for 1 h at room method. temperature in the presence of the appropriate primary anti- body. Detection of recombinant GST-LCK was performed with 2.11. Monoclonal antibody production hybridoma supernatants diluted 1:1 with PBS-T or with a rabbit ␣-GST monoclonal antibody (Sigma) diluted 1:4000 in PBS-T. Briefly, purified GST-LCK was refolded by dialysis twice Hybridoma supernatants were also diluted 1:1 during detection each in the following buffers: (A) 20 mM Tris–HCl (pH of trout Lck in cell lysates. For positive controls during Western 8.0) containing DTT (0.1 mM), (B) 20 mM Tris–HCl (pH analyses, a commercially available ␣-mouse Lck monoclonal 8.0), and (C) 20 mM Tris–HCl (pH 8.0) containing 1 mM antibody (3A5; Biocarta) was used. Membranes were washed reduced and 0.2 mM oxidized glutathione in 1× phosphate and incubated for 1 h with HRP-conjugated ␣-mouse-IgG, ␣- buffered saline. Proteins were mixed with Freund’s com- rabbit-IgG or ␣-rat-IgG secondary antibodies (Amersham) in plete adjuvant and approximately 100 ␮g of GST-LCK was PBS-T (1:15,000) for detection of the 3A5, ␣-GST or hybridoma used for 1◦ injection (subcutaneous). Rats were then boosted derived antibodies, respectively. Bound antibodies were detected 2× using Freund’s incomplete adjuvant at the Basel Insti- by ECL (Amersham). tute for Immunology prior to harvesting their spleens. Single cell suspensions were obtained and splenocytes were washed 2.14. Immunoprecipitation three times in serum-free DMEM and resuspended in 20 mL serum-free medium. A 3:1 ratio of splenocytes to fusion part- Cell lysates (2 × 106 cell equivalents) were pre-cleared with ners (P3X63Ag8.653 myeloma cells) was used for fusion 20 ␮L Gammabind Plus sepharose (50% slurry in PBS) (Amer- (PEG 1500, Sigma) following standard procedures. Cells sham) for 2 h and then incubated at 4oC overnight with 100 ␮L were grown in selective media (HAT) and after approx- ␣-trout Lck or Lck-negative hybridoma supernatants, or with imately 10–14 days, supernatants from growing cultures 10 ␮L 3A5 antibody (Biocarta). Gammabind Plus sepharose were tested for reactivity against GST-LCK by ELISA using (10 ␮L; Amersham) was added to each sample and the reac- goat-antiRATHRP (Amersham) as the 2◦ reagent. Clones pos- tions incubated on ice for 1 h and pelleted by centrifugation. itive for the GST-LCK fusion protein were subcloned 3× After removal of the supernatants, the sepharose pellets were each in HT media by limiting dilution in media containing washed three times with lysis buffer and resuspended in either HT and scaled up. Supernatants were isolated and charac- 50 ␮L non-reducing or reducing PAGE loading buffer. Samples terized by Western blot analyses, immunoprecipitation and were heated to 100 ◦C for 5 min and 25 ␮L was then loaded FACS. onto a 10% SDS-PAGE gel for separation and detection as K.J. Laing et al. / Molecular Immunology 44 (2007) 2737–2748 2741 described. As a control, Jurkat cell lysates were subjected to Table 2 immunoprecipitation using the ␣-mouse Lck (3A5) antibody. Global alignment values from comparisons of trout LCK1 and LCK2 with peptide sequences of the human SRC family 2.15. FACS analysis Subfamily Human Amino acid identity to Amino acid identity to peptide trout LCK1 (%) trout LCK2 (%)

Single cell suspensions were obtained by disrupting trout SRC-A FGR 51.9 49.5 thymus through nylon sieves as described above. The cell FYN 52.8 53.0 concentration was adjusted to 1 × 106 cells/mL in 1× PBS SRC 47.7 51.5 containing 2% FBS, pelleted by centrifugation at 500 × g YES 52.4 51.9 and resuspended in PBS containing freshly prepared 2% Paraformaldehyde (Sigma). After incubation at room temper- SRC-B BLK 57.5 55.5 HCK 58.7 59.7 ature for 20 min, cells were washed twice in PBS. The cells LCK 70.2 70.9 5 (1 × 10 ) were then seeded into V-bottomed 96-well plates LYN 59.7 59.7 (Costar), pelleted by centrifugation, and resuspended (50 ␮L) in permeabilization buffer (0.2% FBS, 0.5% Saponin, 10 mM HEPES pH 7.5 and 0.1% Na-azide in PBS) and 50 ␮L of either homology to all members of the Src kinase family (ranging from 30A8 (␣-trout Lck) or 27F5 (an IgG isotype matched irrelevant ∼47 to 60% identity for non-Lck Src kinases), the trout Lcks are rat mAb) hybridoma supernatant. For negative (no antibody) clearly most related to mammalian Lck, with >70% amino acid controls, 50 ␮L permeabilization buffer were used in the place identity (Table 2) and conservation of Lck-specific amino acids. of the primary antibody. The cells were incubated at room tem- In addition, phylogenetic comparison of Src peptides clearly perature for 30 min, washed four times with permeabilization groups trout Lck1 and Lck2 with other vertebrate Lck peptides buffer and resuspended with PBS containing a goat anti-rat-IgG- with high bootstrap support (100%)(Fig. 2). Both trout Lck1 and FITC conjugated secondary antibody (1:200; Southern Biotech). Lck2 are highly conserved in SH1, SH2 and SH3 domains with Following a 20 min incubation at room temperature, cells were respect to other vertebrate Lck proteins and within the entire Src washed four times, resuspended in 200 ␮L PBS with 2% FBS kinase family. Although more variation does occur in the unique and analyzed using a BD FACSCaliber (Becton Dickson). domain, there are still a high number of conserved residues in the N-terminal segment between members of the Lck subfamily. Of 3. Results importance, the CXXC motif essential for Lck interaction with CD4 and CD8␣ in mammals is present in the unique domain 3.1. Isolation and sequence analysis of trout Lck for both forms of trout Lck, and represents a sequence specific motif for Lck that is absent from other Src kinases (data not While screening a rainbow trout thymus cDNA library, two shown). A second motif (GCXCS) that represents a myristoy- distinct sequences (designated Lck1 and Lck2) were identified lation and site in mammalian Lck, is conserved that are homologous to mammalian Lck. These two sequences in trout Lck2 (GCNCS), while trout Lck1 contains a tyrosine share 84.5% nucleotide and 94.6% amino acid identities to residue in place of the second cysteine of this motif (GCNYS). each other and were submitted to Genbank (AY973032 and AY973033, respectively). Lck1 was identified as a cDNA clone 3.2. Tissue-specific expression of trout Lck that hybridized to a human Lck probe. The full-length sequence comprises 2135 nucleotides with an open reading frame (ORF) Expression of trout Lck mRNA was detected in tissues asso- of 1515 bp (nucleotides 261–1775) that potentially encodes a ciated with lymphocyte function and development. Northern peptide containing 504 amino acids (∼57.9 kDa predicted pro- analysis detected high levels of Lck message in trout thymus tein). Lck2 was identified by random screening of cDNA clones with lower levels observed in spleen, pronephros and intestine and has 2246 nucleotides, coding for a potential peptide with using a probe for trout Lck1 (Fig. 3A). This finding is supported 503 amino acids (∼57.5 kDa) between nucleotides 207 and by real-time PCR (Fig. 3B) that shows high Lck expression in 1718. Both trout Lck sequences possess long 5 untranslated thymus with moderate expression in spleen, pronephros and PBL regions (UTRs) containing several upstream ATGs. Within the using primers and probes that can distinguish Lck1 and Lck2. 3 UTRs, both Lck1 and Lck2 display two mRNA instability Real time PCR also indicated that basal Lck2 expression was (ATTTA) motifs, Lck1 has one potential polyadenylation signal higher in thymus and spleen than Lck1 since the Lck1 transcript sequence (2257–2267) whereas Lck2 possesses two (2159–2164 was consistently detectable with 1–2 Ct higher than Lck2 in and 2198–2203). each na¨ıve fish tested. This correlated with observations using Comparison of trout Lck1 and Lck2 with Lck-related proteins conventional RT-PCR to detect Lck expression in the spleens highlights the conserved regions associated with Lck and other of three isogenic trout strains (OSU-142, Hot Creek and Arlee) Src family kinases (Table 2 and Fig. 1). Src kinases can gener- using a different set of primers (Fig. 4), with very weak expres- ally be divided into 4 Src homology (SH) domains, where SH1 sion observed for Lck1 in the Arlee strain. In this instance, is the kinase domain at the C-terminal end and SH4 is the unique Lck1 detection required two rounds of PCR compared to a sin- domain at the N-terminus associated with the specificity of the gle round of amplification for Lck2, although both genes were different family members. Although the trout sequences display clearly detectable in all three trout strains. The amplification of 2742 K.J. Laing et al. / Molecular Immunology 44 (2007) 2737–2748

Fig. 1. Multiple alignment of the amino acid sequences for selected Lck molecules, generated using the CLUSTALW program. Amino acids that are identical relative to human Lck are represented with a dot (·) while gaps in the alignment are shown with a dash (−). The positions of domains SH1-4 are shown. Residues at the N-terminus ‘GCXCS’ associated with and palmitoylation and the conserved CD4/CD8 interaction site ‘CXXC’ are highlighted. Conserved tyrosines Tyr505 and Tyr394, which are dephosphorylated and phosphorylated, respectively during Lck activation, are shaded. Hs = Homo sapiens,gg=Gallus gallus,fr=Fugu rubripes,dr=Danio rerio,om=Oncorhynchus mykiss.

Lck1 and 2 from the homozygous clones of trout also demon- confirming the efficiency of RT-PCR. Expression of both Lck strates that Lck1 and Lck2 represent individual genes (paralogs), isoforms in trout was restricted to the sIgM− fraction of head and not allelic variants. kidney (pronephros) lymphocytes (Fig. 5).

3.3. Expression of trout Lck in isolated lymphocytes 3.4. Modulation of trout Lck by mitogens

Lymphocytes from the pronephros (bone-marrow equivalent) Real time RTPCR was used to determine changes in trout and PBL were separated into populations expressing cell surface Lck1 and Lck2 transcript levels in response to T-cell mitogens. IgM (sIgM+) or lacking cell surface IgM (sIgM−) using FACS- The lymphocytes were cultured in vitro with either PHA or PMA based cell sorting. The separated lymphocytes were assessed for 4, 8, 12 or 24 h. Lck1 expression was up-regulated four- for expression of both Lck1 and Lck2 by RT-PCR. Levels of fold and seven-fold in PBL exposed to PHA for 8 and 12 h, IgM and TCR-␣ were also determined in these populations to respectively compared to controls (Fig. 6). The changes at both confirm the separation of sIgM cells from other lymphocytes, of these time points were significant (p ≤ 0.05). Similarly, Lck1 and ARP was amplified (as a housekeeping gene) in all samples levels increased four-fold and nine-fold above control levels in K.J. Laing et al. / Molecular Immunology 44 (2007) 2737–2748 2743

Fig. 3. Tissue-specific expression of trout Lck transcripts. (A) Total Lck mRNA Fig. 2. Phylogenetic relationships between members of the SRC kinase fam- was estimated by Northern Analysis to be most abundant in thymus, with mod- ily. The tree was constructed from CLUSTAL generated alignments of amino erate expression observed in the spleen, and low expression in the kidney. (B) acid sequences using the Neighbor-joining method, with Poisson correction Lck1 and Lck2 mRNA levels were analyzed individually using qPCR, con- and complete deletion of gaps. Tree topography was evaluated by bootstrap- firming high expression of both these transcripts in thymus, with expression ping 1000 times (shown as a percentage). The accession numbers and species also observed in spleen, pronephros, and PBL. The expression levels for qPCR abbreviations used are as follows. Human: SRC (P12931), YES (P07947), FYN are presented as fold expression levels (with fold standard deviations) com- (P06241), FGR (P09769), BLK (P51451), HCK (P08631), LCK (P06239) and pared to levels in the liver for each transcript. The levels between transcripts are LYN (P07948). Mouse: SRC (P05480), YES (Q04736), FYN (P39688), FGR not compared. Th = thymus, PN = pronephros, MN = mesonephros, Sp = spleen, (P14234), BLK (P16277), HCK (P08103), LCK (P06240) and LYN (P25911). Lv = liver, In = intestine, Mu = muscle, Hr = heart, Ts = testis, PBL = peripheral Chicken (CHICK): SRC (P00523), YES (P09324), FYN (Q05876) and LCK blood leukocytes. (P42683). Xenopus laevis (XENLA): SRC1 (P13115), SRC2 (P13116), YES (P10936), and FYN (P13406). Zebrafish (DANIO): SRC (Q6EWH0), YRK (yes-related kinase)(Q8AWF1) and LCK (Q6TPQ4). Fugu: LCK (Q8QGJ9). 3.6. Biochemical analysis of Lck from trout lymphocytes Swordtail Xiphophorus helleri (XIPHE): YES (P27447) and FYN (P27446). Atlantic salmon: SCK (Q9DDK6). The 22C9 and 30A8 mAbs (that bound exclusively to the PBL exposed to PMA, although due to high variation between Lck portion of GST-Lck) could detect two distinct protein bands individual fish, these increases were not statistically significant. between 55 and 60 kDa in size in trout thymus lysates under non- For both mitogens, levels of Lck decreased towards basal levels reducing conditions that were similar in size to Lck detected at the 24-hour time-point. Significant changes in Lck2 levels in Jurkat cell lysates using a mammalian anti-Lck mAb, 3A5 were not observed with either mitogen over this time span. (Fig. 8). The 30A8 rat-anti-trout Lck mAb was used to immuno- precipitate and detect Lck from trout thymus or spleen lysates 3.5. Generation of monoclonal antibodies against (data not shown). When analyzed under reducing conditions, recombinant Lck the trout Lck protein was the same size as human Lck immuno-

Rainbow trout Lck1 and Lck2 were generated as partial GST fusion proteins by inserting cDNA encoding the SH4-SH3-SH2 regions into the pGEX-4T vector and transfecting TOP 10 E. coli. Proteins of ∼60 kDa were generated using this system that could be detected during Western analysis using a com- mercial anti-GST monoclonal antibody (Fig. 7A). Monoclonal antibodies (mAbs) against trout Lck were generated against GST-Lck1 fusion protein. Four of these hybridoma supernatants contained mAbs (called 22C9, 30A8, 27A5 and 27F5) that bound to the GST-Lck1 fusion peptide during Western analysis Fig. 4. Expression of trout Lck1 and Lck2 mRNA from the spleens of homozy- (Fig. 7A) although 27A5 and 27F5 could also bind a different gous trout. O = Oregon State University strain, A = Arlee strain, H = Hot Creek strain, C = water control (negative). Lck2 was easily detectable after 38 cycles GST-fusion peptide (i.e. these antibodies were not specific for in one round of amplification (A), but expression of Lck1 was low/undetectable the Lck portion)(Fig. 7B). These antibodies also reacted with a by standard RTPCR unless a second round of amplification (20 cycles) was GST-Lck2 fusion peptide (data not shown). performed (B). 2744 K.J. Laing et al. / Molecular Immunology 44 (2007) 2737–2748

Fig. 5. Analysis of trout Lck mRNA expression in FACS-sorted lymphocyte populations. Leukocytes from head kidney and PBL were separated under strin- Fig. 7. Specificity of monoclonal antibodies for trout Lck. Monoclonal antibod- gent conditions to give two lymphocyte populations, those expressing surface ies (mAbs) were generated against recombinant a trout LCK GST-fusion protein IgM and those lacking surface IgM. Total RNA was extracted from separated (GST-LCK) and were tested for their ability to detect GST-LCK during Western lymphocytes and subjected to RT-PCR analysis at 38 cycles to determine the analysis under denaturing conditions (A). A commercially available anti-GST presence of transcripts for ARP (housekeeping control), surface IgM, TCR-␣, antibody (␣GST) was used as a positive control that recognizes the GST portion Lck1 and Lck2. Negative controls for RT-PCR included a no template control of the fusion protein and an irrelevant isotype-matched mAb supernatant (H7) (−) and a control lacking reverse transcriptase during cDNA synthesis (RT-). was used as a negative control. To distinguish between mAb recognition of the Note: the band present in the PBL IgM+ lane for LCK2 is the primer set and not GST portion and LCK portions of the GST-LCK, the same mAbs were tested an LCK2 band. for affinity to an irrelevant GST-fusion protein (not containing the Lck moiety) (B). precipitated from Jurkat cell lysates using the anti-Lck antibody 3A5. The 30A8 antibody was used in further immunoprecipi- mAb or the secondary antibody alone. The 27F5 antibody also tation experiments using lysates from isolated thymocytes that failed to precipitate any proteins that could be detected using had been stimulated with T-cell mitogens (Fig. 9). A visible the 30A8 mAb, proving specificity of the immunoprecipitation increase in the level of Lck protein was observed from trout reaction between 30A8 and Lck. thymocytes exposed to either PHA or PMA for 8 h. Once again, Lymphocyte populations from trout thymus bind the 30A8 these proteins were specifically detected with the 30A8 mAb and antibody during FACS analysis following permeabilization could not be detected with the 27F5 (isotype matched, anti-GST) using Saponin (Fig. 10). The IgG isotype matched negative control antibody (27F5), failed to bind with the trout lympho- cytes, confirming the reaction was specific for the anti-Lck mAb (30–A8) and not the secondary antibody. Neither mAb could bind to non-permeabilized lymphocytes, confirming the speci- ficity of 30A8 for a membrane-associated intracellular protein and not a protein expressed on the cell surface.

Fig. 6. Regulation of trout Lck mRNA upon T-cell mitogen stimulation. Trout PBL were isolated from four individual fish, resuspended in L15 medium and incubated for 0, 4, 8, 12 or 24 h with either PHA (20 ␮g/mL), PMA (10 ng/mL) or PBS (control). Total RNA was extracted from the cultured cells and subjected to qPCR to quantify changes in transcript levels. Transcript levels are presented Fig. 8. Immunoprecipitation of trout Lck. The anti-trout Lck mAb (30A8) was as fold increases of Lck1 or Lck2 in response to either PHA or PMA above that used to detect native Lck in the thymic cell lysates by non-reducing SDS-PAGE. of control cultures from the same time-point. Non-parametric statistical analysis An IgG-matched mAb (27F5), which is specific for GST, was used as a negative (Wilcoxon signed rank test) was performed to compare normalized expression control primary antibody on the thymus lysates. The anti-mouse Lck (3A5) mAb levels of each Lck transcript in treatment groups vs. controls. Statistically sig- was simultaneously used to detect human Lck in Jurkat cell lystates. Secondary nificant changes of transcript expression (p < 0.05) are shown with an asterisk. antibodies was also used alone as controls for all reactions. ␣-mIgG = anti-mouse Error bars represent fold standard error. IgG antibody, ␣-rIgG = anti-rat IgG antibody. K.J. Laing et al. / Molecular Immunology 44 (2007) 2737–2748 2745

3.7. Discussion

The presence of the lymphocyte cell specific kinase, Lck, was confirmed in salmonids by the isolation and sequence analysis of two distinct cDNAs (Lck1 and Lck2) from rainbow trout that are homologous to mammalian and avian Lck. Mammalian Lck is transcribed as two distinct mRNAs from two distinct promoters; a downstream promoter generates type I Lck primarily in imma- ture thymocytes and an upstream promoter generates type II Lck in both immature and mature T-cells (Reynolds et al., 1990). The two mammalian Lck transcripts differ only in their 5 UTR sequences. In chicken, however, only one Lck transcript has been identified that is expressed in both thymus and spleen (i.e. Fig. 9. Immunoprecipitation of trout Lck post mitogenic stimulation. Thy- immature and mature T-cells) (Chow et al., 1992). Salmonids mocytes (1 × 107) were incubated at 15 ◦C in L15 medium with either PHA possess many duplicated genes that arose from a genome dupli- (20 ␮g/mL), PMA (10 ng/mL) or with an equivalent volume of PBS (Con). cation event that occurred in the lineage leading to salmonids After 8 h, the cells were lysed in a digitonin-based lysis buffer. Lysates were (Moghadam et al., 2005). This likely explains the occurrence of pre-cleared and then incubated separately with 30A8, 27F5 or no antibody (IP two distinct trout Lck genes in trout while only one Lck gene Ab), followed by precipitation using Gammabind plus sepharose. Western anal- ysis was performed on immunoprecipitated proteins using these same mAbs as appears in other teleost fish genomes (i.e. zebrafish, tetraodon primary antibodies (1◦ Ab). and fugu) implying that trout Lck1 and Lck2 do not correspond to the type I and type II Lck transcripts seen in mammals. Again, Lck1 and Lck2 are distinct genes and not allelic variants as verified by the amplification of both genes from three homozy- gous trout lines (OSU, Hot Creek and Arlee). The functional implications of two distinct Lck genes in salmonid fish are not known. Trout Lck1 and Lck2 contain upstream ATGs in their 5 UTR sequences (Table 3) suggesting some level of post- transcriptional regulation. Trout Lck1 has six upstream ATGs while three upstream ATGs occur in the 5 UTR obtained for Lck2. At least one of these upstream ATGs (occurring at nucleotide 177 of Lck1) is flanked by favorable nucleotides for the initiation of translation (Kozak, 1981)(Table 3). Chicken, mouse and human Lck genes also possess upstream ATGs Fig. 10. FACS analysis of Lck+ trout thymocytes. Thymocytes were isolated (Takadera et al., 1989; Chow et al., 1992). Some of these from 6 trout, pooled, fixed (paraformaldehyde) and permeabilized (saponin). The upstream ATGs in the murine type I Lck transcript have been remaining cells were resuspended in PBS-FBS and not permeabilized. 30A8 and shown to reduce the efficiency of Lck translation confirming × 6 27F5 mAbs were used to stain 1 10 non-permeabilized (A) or permeabilized their importance in post-transcriptional regulation of Lck (Marth (B) thymocytes and the stained cells analyzed by FACS. Histograms represent the gated population of lymphocytes from 10,000 total events. et al., 1985). Whether trout Lck is regulated in the same way remains to be determined.

Table 3 Upstream ATGs of trout Lck 2746 K.J. Laing et al. / Molecular Immunology 44 (2007) 2737–2748

Conserved amino acids can be observed in the predicted trout sion observed in the thymus for both genes and lower expression Lck proteins suggesting that they share the same functions as levels in other lymphoid tissues such as pronephros, spleen their mammalian orthologs. For example, they both possess four and PBL. Little detectable expression of either transcript was highly conserved functional domains (SH1, SH2, SH3 and SH4) observed in liver, intestine, muscle, heart or testis of na¨ıve trout, with an especially conserved kinase domain (SH1) involved in although these tissues (with the exception of liver) were not stud- the enzyme activity of mammalian Lck. Searching the Con- ied by qPCR. This differs from the expression pattern of Lck served Domain Database (CDD; http://www.ncbi.nlm.nih.gov/ described for fugu, which showed a more widespread expres- Structure/cdd) with the trout Lck molecules recognizes these sion pattern, occurring in intestine, spleen, gill, heart, liver (in predicted SH1, SH2 and SH3 domains of the trout Lck addition to kidney) and low expression in muscle and ovary molecules. Both trout Lck molecules possess conserved tyrosine (Brenner et al., 2002), although the fugu study utilized a qual- residues in the kinase and C-terminal domains that correspond itative form of RT-PCR to detect expression. No analyses of to Tyr394 and Tyr505 of human Lck (Giannini and Bijlmakers, thymus or PBL were shown for fugu (Brenner et al., 2002). In 2004). In mammals, these tyrosines are implicated in controlling contrast, the trout Lck expression patterns are similar to those of Lck activation; therefore, conservation in trout Lck further high- mammalian and chicken Lck, that are restricted to tissues asso- lights their importance in regulating the kinase activity of fish ciated with T-lymphocytes such as peripheral blood, thymus and Lck. A conserved CXXC motif can be observed in the unique spleen (Takadera et al., 1989; Chow et al., 1992). domain (SH4) of both trout Lck molecules which likely interacts Examination of isolated lymphocytes confirmed expression with the intracellular tails of CD4 and CD8␣ T-cell co-receptors of trout Lck in the lymphocyte population of pronephros and during T-cell signaling (Turner et al., 1990; Campbell et al., PBL, and infers an association of salmonid Lck with lympho- 1995). Of interest, both trout Lck protein sequences possess a cyte function. When these lymphocytes are further separated into third cysteine in the CD4/CD8 binding region to give the motif populations expressing surface IgM or lacking surface IgM (i.e. CXXCXC. The additional cysteine is not present in either the into B-cell and non-B-cell lymphocytes) the expression of both human or chicken Lck molecules (these both have a tyrosine isoforms is restricted to the IgM-negative population in the lym- residue in this position, an amino acid that is also capable of phocytes derived from the kidney, but in both populations from binding metal ions) but can be observed in Fugu and zebrafish PBL. Mammalian Lck is predominantly expressed in T cells, Lck (Fig. 1). Trout Lck1, but not Lck2, also has a histidine but has also been shown in NK cells, associated with a role in residue immediately preceding the second cysteine of this motif KIR/SHP-1 signaling (Marti et al., 1998) by interacting with a that is conserved in the chicken and fugu Lck molecules. Cys- CXC motif in the cytoplasmic tail of the NK-receptor protein-1 teine and histidine residues can both bind metal ions such as (CD161) (Ljutic et al., 2005). It is, therefore, conceivable that Zn2+, a process known to be involved during Lck interactions expression of Lck1 and Lck2 in the sIgM− fraction of trout lym- with CD4 and CD8 (Huse et al., 1998). The presence of these phocytes may be associated with both T and NK cells (or their residues in fish Lck may be significant for the protein:protein fish equivalents). The presence of Lck transcripts in the IgM+ interaction of this kinase. Interestingly, trout CD4 and CD4REL fraction of PBL may reflect a role for Lck in signaling within (Dijkstra et al., 2006; Laing et al., 2006) both possess the canon- peripheral B-cells in fish. Certainly, this is plausible since Lck ical Lck association motif (CXC) in their cytoplasmic tails but has been shown to play a role in certain subsets of mammalian all teleost CD8␣ sequences lack the cononical motif and instead, B-cells (Ulivieri et al., 2003; Dal Porto et al., 2004; Frances et al., encode a CXH motif (Hansen and Strassburger, 2000; Moore et 2005; Paterson et al., 2006). However, the tissue-specific expres- al., 2005; Somamoto et al., 2005). Whether the unique CXXCXC sion levels measured by real-time PCR and Northern analyses motif found in teleost Lck can associate with CD8␣ has yet to as well as the restricted expression in the IgM-negative lym- determined. In addition, both molecules possess features simi- phocyte population of trout head kidney remain consistent with lar to myristoylation (Gly2 and Ser6) and palmitoylation (Cys3 a predominance of Lck transcript in fish T-cells and possibly and Cys5) sites (although Cys5 is replaced with Tyr5 in trout within fish NK cells. Lck1) suggesting trout Lck1 and Lck2 will associate with the cell Although basal Lck2 expression levels appear higher than membrane in a CD4/CD8 independent manner, as observed for Lck1 expression in the secondary lymphoid tissues, Lck1 mammalian Lck (Yasuda et al., 2000). Both palmitoylation sites expression is more responsive to in vitro mitogenic stimula- have been implicated in the localization of Lck into lipid rafts tion than Lck2 expression. Lck1 levels increase in isolated trout of mammalian CD4 + T cells, with Cys3 having a higher impor- PBL following 8 h exposure to both PHA and PMA, reach- tance in the processes of localization and activation (Kosugi ing maximal expression after 12 h. This is consistent with a et al., 2001), suggesting the substitution of the equivalent Cys role for trout Lck1 in T-cell activation and/or signaling, yet with Tyr5 in trout Lck1 will not impact its membrane associating differs from the regulation of Lck in mammalian peripheral potential. lymphocytes. Human blood mononuclear cells activated with The tissue and cell specific expression patterns of the trout an anti-CD3 antibody and PMA showed a transient down- Lck transcripts were determined using Northern analysis for regulation of Lck between 2 and 10 h after incubation with total Lck levels (Lck1 and Lck2), and isoform-specific qPCR the mitogen, which then returned to basal levels although no analysis to show the mRNA expression levels of the two dis- changes in Lck levels occurred with PMA alone (Paillard and tinct trout lck genes. Both trout Lck1 and Lck2 share a similar Vaquero, 1991). However, in a different study using two dis- mRNA tissue-specific expression pattern, with highest expres- tinct human leukemic T cell lines (Jurkat and P30/OKUBO), K.J. Laing et al. / Molecular Immunology 44 (2007) 2737–2748 2747 other phorbol esters such as tetradecanoylphorbol acetate (TPA) that were previously not available for salmonid fish and com- and phorbol dibutyrate (PDB) induced the type II Lck transcript plement ongoing research for generation of similar reagents for (i.e. the transcript derived from the upstream/distal promoter) trout CD4 and CD8. while down-regulating Lck type I (the transcript derived from the proximal promoter) (Leung and Miyamoto, 1991). Perhaps Acknowledgements in trout PBL, a similar promoter to that of mammalian type II Lck may control expression of the trout Lck1 gene. Although We thank Scott La Patra for providing fish, Gary Thorgaard we have no data to confirm this for trout, the fugu Lck gene for supplying splenic tissue from homozygous trout lines, Mau- was found to possess two promoters; the distal promoter was reen Purcell for assisting with experimental design for qPCR responsible for controlling Lck expression in lymphoid tissues and Steve Kaattari for providing biotinylated mAb1-14 used for while the proximal promoter generated an alternately spliced lymphocyte separation. This project was supported by National transcript that localized to testis (Brenner et al., 2002). Science Foundation Molecular and Cellular Biosciences Grant Trout Lck1 was generated as a recombinant GST fusion pro- 0453924 to JDH and the National Research Initiative of the tein using a bacterial expression system and used to immunize USDA Cooperative State Research, Education and Extension rats in order to generate monoclonal antibodies. From the array Service, grant number 0206006. SD was supported by the of antibodies that reacted with the GST-LCK fusion protein, NOAA-EPP funded LMRCSC. Mention of trade names does one was identified as a good candidate monoclonal antibody not imply U.S. Government endorsement. for use in detecting and analyzing trout Lck. This antibody (called 30A8) reacted with recombinant forms of both Lck1 References and Lck2 due to their high amino acid identity and bound two distinct proteins (56–60 kDa) during Western analysis on thy- Altschul, S.F., Madden, T.L., Schaffer, A.A., Zhang, J.H., Zhang, Z., Miller, W., mus cell lysates. These two proteins most likely represent the Lipman, D.J., 1997. Gapped BLAST and PSI-BLAST: a new generation of two distinct Lck isoforms of trout, but may reflect Lck under protein database search programs. Nucl. Acids Res. 25, 3389–3402. different levels of phosphorylation as observed for mammalian Brenner, S., Venkatesh,B., Yap,W.H.,Chou, C.F., Tay, A., Ponniah, S., Wang, Y., Tan, Y.H., 2002. Conserved regulation of the lymphocyte-specific expres- Lck (Watts et al., 1993). 30A8 could also immunoprecipitate sion of Lck in the Fugu and mammals. Proc. Natl. Acad. Sci. U.S.A. 99, native Lck from trout thymus lysates, and was used in immuno- 2936–2941. precipitation experiments that showed an up-regulation of trout Campbell, K.S., Buder, A., Deuschle, U., 1995. Interactions between the amino- Lck protein in isolated thymocytes stimulated for 8 h with PHA terminal domain of p56 (Lck) and cytoplasmic domains of CD4 and CD8- and PMA. This increase in protein expression is consistent with alpha in yeast. Eur. J. Immunol. 25, 2408–2412. Chan, A.C., , M., Johnson, R., Kong, G.H., Wang, T., Thoma, R., increases in transcription of trout Lck1 shown during qPCR in Kurosaki, T., 1995. Activation of ZAP-70 kinase-activity by phosphory- this study. When utilized in flow cytometric analysis, the 30A8 lation of tyrosine-493 is required for lymphocyte antigen receptor function. antibody could also detect proteins in trout lymphocytes that Embo J. 14, 2499–2508. had been permeabilized with saponin. The process of perme- Chow, L.M.L., Ratcliffe, M.J.H., Veillette, A., 1992. Tkl is the avian homolog abilization includes a cell-fixing step with paraformaldehyde of the mammalian lck tyrosine protein-kinase gene. Mol. Cell. Biol. 12, 1226–1233. that likely denatures some proteins, consistent with the abil- Dal Porto, J.M., Burke, K., Cambier, J.C., 2004. Regulation of BCR signal ity of 30A8 to recognize denatured Lck and reflects suitability transduction in B-1 cells requires the expression of the of 30A8 for use in single cell analyses of Lck in this manner. Lck. Immunity 21, 443–453. Thus, the 30A8 anti-trout Lck monoclonal antibody represents a Dijkstra, J.M., Somamoto, T., Moore, L., Hordvik, I., Ototake, M., Fischer, U., good tool that can be used to investigate the biochemical proper- 2006. Identification and characterization of a second CD4-like gene in teleost fish. Mol. Immunol. 43, 410–419. ties and protein-protein interactions of salmonid Lck and T-cell Frances, R., Tumang, J.R., Rothstein, T.L., 2005. Cutting edge: B-1 cells are signaling. deficient in Lck: defective receptor signal transduction in B-1 cells occurs in the absence of elevated Lck expression. J. Immunol. 175, 27–31. 4. Summary Giannini, A., Bijlmakers, M.J., 2004. Regulation of the Src family kinase Lck by Hsp90 and ubiquitination. Mol. Cell. Biol. 24, 5667–5676. Hansen, J.D., Strassburger, P., 2000. Description of an ectothermic TCR core- This report identified two distinct Lck gene products (par- ceptor, CD8 alpha, in rainbow trout. J. Immunol. 164, 3132–3139. alogs) in rainbow trout that are highly conserved relative to Lck Hansen, J.D., Strassburger, P., DuPasquier, L., 1997. Conservation of a master in other vertebrates, both in sequence similarity and in cell and hematopoietic switch gene during vertebrate evolution: isolation and char- tissue-specific expression suggesting that they are orthologous acterization of Ikaros from teleost and amphibian species. Eur. J. Immunol. to mammalian Lck. Since T-cell specific mitogens can regulate 27, 3049–3058. Hermiston, M.L., Xu, Z., Majeti, R., Weiss, A., 2002. Reciprocal regulation of Lck gene and protein expression in trout lymphocytes, it is highly lymphocyte activation by tyrosine kinases and phosphatases. J. Clin. Invest. likely that salmonid Lck also plays an important role in T-cell 109, 9–14. signaling in fish. Future studies utilizing the recombinant pro- Hofmann, G., Schweimer, K., Kiessling, A., Hofinger, E., Bauer, F., Hoffmann, teins and mAbs generated and characterized against trout Lck S., Rosch, P., Campbell, I.D., Werner, J.M., Sticht, H., 2005. Binding, domain during this investigation will allow investigations of the individ- orientation, and dynamics of the Lck SH3-SH2 domain pair and comparison with other Src-family kinases. Biochemistry 44, 13043–13050. ual cell types expressing Lck thereby facilitating experiments Holdorf, A.D., Lee, K.H., Burack, W.R., Allen, P.M., Shaw, A.S., 2002. Regu- to uncover the function of Lck proteins in fish. These reagents lation of Lck activity by CD4 and CD28 in the . Nat. provide essential tools for understanding lymphocyte function Immunol. 3, 259–264. 2748 K.J. Laing et al. / Molecular Immunology 44 (2007) 2737–2748

Huse, M., Eck, M.J., Harrison, S.C., 1998. A Zn2+ ion links the cytoplas- Perlmutter, R.M., Marth, J.D., Lewis, D.B., Peet, R., Ziegler, S.F., Wilson, C.B., mic tail of CD4 and the N-terminal region of Lck. J. Biol. Chem. 273, 1988. Structure and expression of Lck—transcripts in human lymphoid- 18729–18733. cells. J. Cell. Biochem. 38, 117–126. Kosugi, A., Hayashi, F., Liddicoat, D.R., Yasuda, K., Saitoh, S., Hamaoka, T., Purcell, M.K., Kurath, G., Garver, K.A., Herwig, R.P.,Winton, J.R., 2004. Quan- 2001. A pivotal role of cysteine 3 of Lck for localization to titative experssion profiling of immune response genes in rainbow trout glycolipid-enriched microdomains and T cell activation. Immunol. Lett. 76, following infectious haematopoietic necrosis virus (IHNV) infection or DNA 133–138. vaccination. Fish Shellfish Immunol. 17, 447–462. Kozak, M., 1981. Possible role of flanking nucleotides in recognition of the AUG Reynolds, P.J., Lesley, J., Trotter, J., Schulte, R., Hyman, R., Sefton, B.M., 1990. initiator codon by eukaryotic ribosomes. Nucl. Acids Res. 9, 5233–5252. Changes in the relative abundance of type-I and type-II Lck messenger-RNA Kumar, S., Tamura, K., Nei, M., 2004. MEGA3: integrated software for molecu- transcripts suggest differential promoter usage during T-cell development. lar evolutionary genetics analysis and sequence alignment. Brief. Bioinform. Mol. Cell. Biol. 10, 4266–4270. 5, 150–163. Rice, P., Longden, I., Bleasby, A., 2000. EMBOSS: the European molecular Laing, K.J., Zou, J.J., Purcell, M.K., Phillips, R., Secombes, C.J., Hansen, biology open software suite. Trends Gen. 16, 276–277. J.D., 2006. Evolution of the CD4 family: teleost fish possess two diver- Robinson, D.R., Wu, Y.M., Lin, S.F., 2000. The protein tyrosine kinase family gent forms of CD4 in addition to lymphocyte activation gene-3. J. Immunol. of the . Oncogene 19, 5548–5557. 177, 3939–3951. Saitou, N., Nei, M., 1987. The Neighbor-Joining method—a new method for Langenau, D.M., Ferrando, A.A., Traver, D., Kutok, J.L., Hezel, J.P.D., Kanki, reconstructing phylogenetic trees. Mol. Biol. Evol. 4, 406–425. J.P., Zon, L.I., Look, A.T., Trede, N.S., 2004. In vivo tracking of T cell Somamoto, T., Yoshiura,Y., Nakanishi, T., Ototake, M., 2005. Molecular cloning development, ablation, and engraftment in transgenic zebrafish. Proc. Natl. and characterization of two types of CD8alpha from ginbuna crucian carp, Acad. Sci. U.S.A. 101, 7369–7374. Carassius auratus langsdorfii. Dev. Comp. Immunol. 29, 693–702. Leung, S., Miyamoto, N.G., 1991. Differential expression of 2 classes of Lck Sommers, C.L., Samelson, L.E., Love, P.E., 2004. LAT: a T lymphocyte adapter transcripts upon phorbol ester treatment of human leukemic T-cells. J. Cell. protein that couples the antigen receptor to downstream signaling pathways. Physiol. 148, 344–352. Bioessays 26, 61–67. Lin, R.S., Rodriguez, C., Veillette, A., Lodish, H.F., 1998. Zinc is essential Takadera, T., Leung, S., Gernone, A., Koga, Y., Takihara, Y., Miyamoto, for binding of p56 (Lck) to CD4 and CD8 alpha. J. Biol. Chem. 273, N.G., Mak, T.W., 1989. Structure of the 2 promoters of the human lck 32878–32882. gene—differential accumulation of 2 classes of Lck transcripts in T-cells. Ljutic, B., Carlyle, J.R., Filipp, D., Nakagawa, R., Julius, M., Zuniga-Pflucker, Mol. Cell. Biol. 9, 2173–2180. J.C., 2005. Functional requirements for signaling through the stimulatory Thompson, J.D., Higgins, D.G., Gibson, T.J., 1994. Clustal-W—improving the and inhibitory mouse NKR-P1 (CD161) NK cell receptors. J. Immunol. sensitivity of progressive multiple sequence alignment through sequence 174, 4789–4796. weighting, position-specific gap penalties and weight matrix choice. Nucl. Marth, J.D., Peet, R., Krebs, E.G., Perlmutter, R.M., 1985. A lymphocyte- Acids Res. 22, 4673–4680. specific protein-tyrosine kinase gene is rearranged and overexpressed in the Turner, J.M., Brodsky, M.H., Irving, B.A., Levin, S.D., Perlmutter, R.M., murine T-cell lymphoma LSTRA. Cell 43, 393–404. Littman, D.R., 1990. Interaction of the unique N-terminal region of tyro- Marti, F., Xu, C.W., Selvakumar, A., Brent, R., Dupont, B., King, P.D., sine kinase p56lck with cytoplasmic domains of CD4 and CD8 is mediated 1998. LCK-phosphorylated human killer cell-inhibitory receptors recruit by cysteine motifs. Cell 60, 755–765. and activate phosphatidylinositol 3-kinase. Proc. Natl. Acad. Sci. U.S.A. Ulivieri, C., Valensin, S., Majolini, M.B., Matthews, R.J., Baldari, C.T., 2003. 95, 11810–11815. Normal B-1 cell development but defective BCR signaling in LCK−/− mice. Moghadam, H.K., Ferguson, M.M., Danzmann, R.G., 2005. Evolution of Hox Eur. J. Immunol. 33, 441–445. clusters in Salmonidae: a comparative analysis between Atlantic salmon Veillette, A., Ratcliffe, M.J.H., 1991. Avian CD4 and CD8 interact with a cel- (Salmo salar) and rainbow trout (Oncorhynchus mykiss). J. Mol. Evol. 61, lular tyrosine protein-kinase homologous to mammalian p56Lck. Eur. J. 636–649. Immunol. 21, 397–401. Moore, L.J., Somamoto, T., Lie, K.K., Dijkstra, J.M., Hordvik, I., 2005. Char- Wange, R.L., Guitian, R., Isakov, N., Watts, J.D., Aebersold, R., Samelson, L.E., acterisation of salmon and trout CD8alpha and CD8beta. Mol. Immunol. 42, 1995. Activating and inhibitory mutations in adjacent tyrosines in the kinase 1225–1234. domain of ZAP-70. J. Biol. Chem. 270, 18730–18733. Needleman, S.B., Wunsch, C.D., 1970. A general method applicable to search for Watts, J.D., Welham, M.J., Kalt, L., Schrader, J.W., Aebersold, R., 1993. IL- similarities in amino acid sequence of 2 proteins. J. Mol. Biol. 48, 443–453. 2 stimulation of T-lymphocytes induces sequential activation of mitogen- Paillard, F., Vaquero, C., 1991. Down-regulation of Lck messenger-RNA by T- activated protein-kinases and phosphorylation of p56 (Lck) at serine-59. J. cell activation involves transcriptional and posttranscriptional mechanisms. Immunol. 151, 6862–6871. Nucl. Acids Res. 19, 4655–4661. Williams, J.C., Wierenga, R.K., Saraste, M., 1998. Insights into Src kinase Paterson, J.C., Tedoldi, S., Craxton, A., Jones, M., Hansmann, M.L., Collins, G., functions: structural comparisons. Trends Biochem. Sci. 23, 179–184. Roberton, H., Natkunam, Y., Pileri, S., Campo, E., Clark, E.A., Mason, D.Y., Yasuda, K., Kosugi, A., Hayashi, F., Saitoh, S., Nagafuku, M., Mori, Y., Ogata, Marafioti, T., 2006. The differential expression of LCK and BAFF-receptor M., Hamaoka, T., 2000. Serine 6 of Lck tyrosine kinase: a critical site for and their role in apoptosis in human lymphomas. Haematolog. Hematol. J. Lck myristoylation, membrane localization, and function in T lymphocytes. 91, 772–780. J. Immunol. 165, 3226–3231.