The CD8αβ heterodimer functions as a coreceptor with the TCR, influencing the outcome of CD8+ T cell responses to pathogen-infected and tumor cells. In contrast to the murine CD8B gene, the human gene encodes alternatively spliced variants with different cytoplasmic tails (M-1, M-2, M-3, and M-4). At present, little is known about the expression patterns and functional significance of such variants. We used quantitative RT-PCR to demonstrate differential mRNA expression patterns of these splice variants in thymocytes and in resting, memory, and activated primary human CD8+ T cells. In total CD8+ T cells, mRNA levels of the M-1 variant were the most predominant and levels of M-3 were the least detected. The M-4 isoform was predominant in effector memory CD8+ T cells. Upon stimulation of CD8+ T cells, the M-2 variant mRNA levels were elevated 10–20-fold relative to resting cells in contrast to the other isoforms. Curiously, the M-2 isoform was not expressed on the cell surface in transfected cell lines. Using fluorescent chimeras of the extracellular domain of mouse CD8β fused to the cytoplasmic tails of each isoform, the M-2 isoform was localized in a lysosomal compartment regulated by ubiquitination of a lysine residue (K215) in its cytoplasmic tail. In contrast, upon short-term stimulation, the M-2 protein localized to the cell surface with the TCR complex. The relatively recent evolution of CD8B gene splice variants in the chimpanzee/human lineage is most likely important for fine-tuning the CD8+ T cell responses.

Mature human cytotoxic MHC class I (MHC-I)3-restricted TCRαβ T lymphocytes are characterized by the membrane expression of CD8 molecules as αα and αβ (1, 2). Human CD8β can also be expressed as a ββ homodimer in transfected COS cells and transgenic mice (3). A critical step in activation of the cytotoxic CD8+ T cells is the recognition and binding of the TCR/CD3 complex to specific peptide-MHC-I (pMHC-I) complexes on the surface of APCs. The CD8 coreceptor forms a molecular complex with TCR to the p-MHC-I complex, thereby localizing p56lck, which is associated with CD8, to the TCR/CD3 complex to facilitate early signaling events (4, 5). The kinase phosphorylates the CD3ζ-chain leading to activation of the adaptor protein ZAP70 that then initiates the MAP/ERK signaling pathway. Once the TCR engages a pMHC-I complex, the TCR undergoes down-modulation, leading to receptor translocation from the plasma membrane to early endosomes, which are enriched for signaling molecules and attenuation of TCR signaling by targeting TCR from endosomes to a lysosomal compartment for degradation (6, 7, 8). This is regulated initially through movement of TCR into clathrin-coated vesicles (9). The p56lck kinase phosphorylates the clathrin H chain, a regulatory step in clathrin-mediated endocytosis. In addition to phosphorylation, the CD3ζ-chain, which has multiple lysines, is also ubiquitinated upon T cell activation, and p56lck activity is required for this modification to occur (10). Recycling of the TCR in thymocytes is thought to regulate levels of expression of the complex, which is much lower on CD4+CD8+ thymocytes than on mature T cells (11). Recent experimental evidence suggests that regulation of CTL activity is mediated by modifications of CD8 coreceptor functions in vivo, including down-regulation of CD8 expression on the cell surface (12, 13, 14, 15) and switching to expression of the CD8αα homodimer (16).

CD8αβ is a better coreceptor with the TCR because of the properties associated with the CD8β-chain. Although the cytoplasmic domain of the CD8α-chain contains the binding site for p56lck required for the initiation of early T cell signaling (17, 18, 19), it is palmitoylation of the CD8β-chain that facilitates the partitioning of CD8 into lipid rafts (20, 21). Studies in mice indicated a role for the CD8β cytoplasmic tail in thymic development and activation of CD8+ T cells although it contains no known protein binding motifs (22, 23, 24, 25). A direct association of CD8 with the TCR-CD3 membrane complex by interaction of CD8β with the CD3δ-chain contributes to recruitment of the TCR into lipid rafts, and it has been suggested that the CD8-pMHC interaction increases the probability of an encounter of pMHC with the appropriate TCR (26). Changes in glycosylation of CD8β during development and after activation have been proposed to modulate interaction with MHC-I (27, 28).

Unlike the mouse, the human CD8B gene encodes additional exons that give rise to CD8β protein isoforms (Fig. 1,A). The two additional exons VIII and IX are also present in Pan troglodytes (chimpanzee) but not in Macaca mulatta (rhesus macaque), indicating their relatively recent evolution. Multiple alternatively spliced transcript variants of human CD8β encoding distinct membrane associated or secreted isoforms were described that may have distinct roles in T cell function and development (29, 30). The initial studies sequenced cDNA clones in the early 1990s and found multiple spliced variants that differed at the 3′ end of the gene as well as two secretory forms. S1 nuclease studies on thymus, peripheral blood, and a T cell thymoma line reported the presence of membrane-spanning isoforms, whereas the mRNA-encoding secretory forms were difficult to detect (30). The membrane-associated isoforms share similar extracellular, transmembrane, and membrane proximal cytoplasmic domains but differ in the rest of their cytoplasmic tails (Fig. 1,B). The cytoplasmic tail of isoform M-1 (also known as transcript variant 5) shows high sequence homology with that of murine CD8β, whereas the M-2, M-3, and M-4 tails have no reported counterparts. The isoform M-1 was shown to be a spliced product of exons I, II, III, V, VI, and VII. Isoform M-2 (also known as transcript variant 1) arose from the same set of exons as M-1 except that it derives from the use of an “aberrant” splice acceptor site located 58 bp upstream of the regular splice acceptor signal of exon VII (31). This insertion changes the reading frame, thereby altering the predicted cytoplasmic tail amino acid sequence and termination codon, and extending the length of the predicted protein sequence by 36 amino acids. Isoform M-3 (also known as transcript variant 3) lacks exon VII and includes two alternate exons, VIII and IX, encoding the carboxyl terminus and 3′ UTR. Another variant described by DiSanto et al. (29) that we have designated as M-4 (also known as transcript variant 2) possesses the transmembrane domain and lacks exons VII and VIII but carries exon IX, giving rise to a unique 36 amino acid sequence in the 3′ region. All isoforms contain the membrane proximal cytoplasmic domain of 8 amino acids encoded by the C1 exon that has a cysteine required for palmitoylation (20, 32), but the rest of the tails range in size from 3 to 39 amino acids (Fig. 1 B).

FIGURE 1.

Schematic representation of the alternatively spliced forms of human CD8β cDNA. A, Organization of the CD8B1 gene showing exon-intron boundaries (30 ). All the splice variants share a common leader (L), extracellular (VJ), transmembrane (TM) and membrane proximal intracellular region encoded by CY1 exon. S indicates secretory domain. B, CY2, CY2′, CY3, and CY4 depict the 3′ terminal exons that are added as a result of alternative splicing to confer unique amino acid sequence to the cytoplasmic tails of M-1, M-2, M-3, and M-4 splice variants, respectively. C, Schematic representation of the location of primers and probe used for quantitative real-time PCR. To distinguish between the splice variants, the sequences of reverse primers were designed to correspond to exon/exon borders of the complementary strand. The gray arrow represents the common forward primer, and black arrows represent the unique reverse primers. Dotted box indicates the FAM-labeled TaqMan probe that is common to all the splice variants.

FIGURE 1.

Schematic representation of the alternatively spliced forms of human CD8β cDNA. A, Organization of the CD8B1 gene showing exon-intron boundaries (30 ). All the splice variants share a common leader (L), extracellular (VJ), transmembrane (TM) and membrane proximal intracellular region encoded by CY1 exon. S indicates secretory domain. B, CY2, CY2′, CY3, and CY4 depict the 3′ terminal exons that are added as a result of alternative splicing to confer unique amino acid sequence to the cytoplasmic tails of M-1, M-2, M-3, and M-4 splice variants, respectively. C, Schematic representation of the location of primers and probe used for quantitative real-time PCR. To distinguish between the splice variants, the sequences of reverse primers were designed to correspond to exon/exon borders of the complementary strand. The gray arrow represents the common forward primer, and black arrows represent the unique reverse primers. Dotted box indicates the FAM-labeled TaqMan probe that is common to all the splice variants.

Close modal

The purpose of the present work was to examine the expression pattern of the membrane-associated human CD8β splice variants by quantitative RT-PCR analysis of mRNA. If the isoforms had different functions, as suggested by the very different sequences of their cytoplasmic tail, we hypothesized that they would likely exhibit differences in expression patterns. We did find differences in the expression patterns between naive, stimulated, and effector memory CD8+ T cell populations. In thymocytes and PBLs, multiple isoforms were expressed. Strikingly, in effector memory CD8+ T cell populations the M-4 isoform was predominant and, upon initial stimulation, the M-2 isoform was significantly up-regulated.

We also found that the expression of the M-2 isoform was regulated by ubiquitination. Thus, the regulated expression of the M-2 isoform is controlled by both changes in splicing and posttranslational modification. This allows for a more rapid method of controlling expression of a protein on the cell surface where it can function. These findings support the hypothesis that changing isoform expression is a mechanism for regulating CD8 function.

First overlap extension PCR was used to generate chimeras of mouse CD8β (Lyt-3) extracellular and transmembrane domains with human CD8β cytoplasmic tails of each variant. In the second overlap extension PCR, the chimeras generated in the first step were used as a template to fuse with enhanced yellow fluorescent protein (EYFP) through a linker (33). The sequences of the forward and reverse primers used to generate all chimeric genes are shown in Table I. The cDNA of each chimeric gene was cloned into a pBMN-I retroviral expression vector (from Gary Nolan, Stanford University, Stanford, CA) using an EcoRI site. Mutagenesis of lysine residues was performed using QuickChange mutagenesis kit from Stratagene. Oligos were designed to change the designated amino acid and to create a restriction site to identify mutants (Table I). Mutagenesis was conducted in the expression vector pBMN-I carrying chimeric M-2 (K215G) Y, and each construct was sequenced to confirm the presence of mutation. Another arginine to valine (R196V) mutation was introduced in the CD8β chimeras (in the human C1 exon encoding cytoplasmic tail) to generate a palmitoylation motif identical to that of a mouse. Mutants are represented using amino acid single letter code representing the original amino acid, residue number, and substituted amino acid; for example, K215G is lysine at position 215 changed to a glycine. pCDNA-3 mammalian expression vector carrying CD8α or each CD8β isoform or hemagglutinin-tagged ubiquitin (HA-Ub) was generated by standard cloning procedure.

Table I.

Sequences of primers and probes used for PCR amplification and cloning

Primers
CD8β: real-time PCR  
 FAM-labeled probe CAGCCCACCAAGAAGTCCACCCTCA 
 Common forward GTGTGGTTGATTTCCTTCCCA 
 Specific reverse M-1 GGCTCTGCTTATTTGTAAAATTGTTTC 
 Specific reverse M-2 GGCAAAGCATATTGAATTTCTGTTTC 
 Specific reverse M-3 TTCTGGAACATTTCTCCAGTGG 
 Specific reverse M-4 ATACCTTCCCCTTGAGGCTGTT 
β2 microglobulin: real-time PCR  
 FAM-labeled probe TGATGCTGCTTACATGTCTCGATCCCA 
 Forward TGACTTTGTCACAGCCCAAGATA 
 Reverse AATCCAAATGCGGCATCTTC 
First step overlap extension PCR  
 Lyt-3 forward CAAAAGCGCCAAGATGCAGC 
 Lyt-3 (TM) reverse GGCTCTCCTCCGCCGACAGTAAAAGTAGAC 
 CD8β Cyt forward GGCCGTCTACTTTTACTGTCGGCGGAGGAGAGC 
 M-1 Cyt reverse GGCTCTGCTTATTTGTAAAATTGTTTC 
 M-2 Cyt reverse CTCATTACTGACCGATGTCTTTTTG 
 M-3 Cyt reverse CCTGTATATTCAGTAGTCCATTC 
 M-4 Cyt reverse GTGCTTCTTGCCTATGTTTTCAGGATC 
Second step overlap extension PCR  
 Lyt-3 forward TCGAATTCGCCAAGATGCAGCCATG 
 EYFP forward GGTGGAGGCGGTAGCGGTGG 
 EYFP reverse CGGAATTCCTTTACTTGTACAGCTCGTC 
 M-1 Cyt reverse CCGCTACCGCCTCCACCTTTGTAAAATTGTTTCATG 
 M-2 Cyt reverse CCGCTACCGCCTCCACCCTGACCGATGTCTTTTTG 
 M-3 Cyt reverse CCGCTACCGCCTCCACCGTAGTCCATTCTGGAAC 
 M-4 Cyt reverse CCGCTACCGCCTCCACCTGTTTTCAGGATCCATG 
Mutants  
 K206/208A CGGCTTCGTTTCATGGCACAAGCTTTCAATATCGTTTGCC 
 K215G TCAATATCGTTTGCCCTCGGGTAAGTGGTTTCACAA 
 K242G GGTGTCCTGCTACAAGGAGATATCGGTCAGTAACGA 
 R196V CCGTCTACTTTTACTGCGTACGGAGGAGAGCCCGGC 
Primers
CD8β: real-time PCR  
 FAM-labeled probe CAGCCCACCAAGAAGTCCACCCTCA 
 Common forward GTGTGGTTGATTTCCTTCCCA 
 Specific reverse M-1 GGCTCTGCTTATTTGTAAAATTGTTTC 
 Specific reverse M-2 GGCAAAGCATATTGAATTTCTGTTTC 
 Specific reverse M-3 TTCTGGAACATTTCTCCAGTGG 
 Specific reverse M-4 ATACCTTCCCCTTGAGGCTGTT 
β2 microglobulin: real-time PCR  
 FAM-labeled probe TGATGCTGCTTACATGTCTCGATCCCA 
 Forward TGACTTTGTCACAGCCCAAGATA 
 Reverse AATCCAAATGCGGCATCTTC 
First step overlap extension PCR  
 Lyt-3 forward CAAAAGCGCCAAGATGCAGC 
 Lyt-3 (TM) reverse GGCTCTCCTCCGCCGACAGTAAAAGTAGAC 
 CD8β Cyt forward GGCCGTCTACTTTTACTGTCGGCGGAGGAGAGC 
 M-1 Cyt reverse GGCTCTGCTTATTTGTAAAATTGTTTC 
 M-2 Cyt reverse CTCATTACTGACCGATGTCTTTTTG 
 M-3 Cyt reverse CCTGTATATTCAGTAGTCCATTC 
 M-4 Cyt reverse GTGCTTCTTGCCTATGTTTTCAGGATC 
Second step overlap extension PCR  
 Lyt-3 forward TCGAATTCGCCAAGATGCAGCCATG 
 EYFP forward GGTGGAGGCGGTAGCGGTGG 
 EYFP reverse CGGAATTCCTTTACTTGTACAGCTCGTC 
 M-1 Cyt reverse CCGCTACCGCCTCCACCTTTGTAAAATTGTTTCATG 
 M-2 Cyt reverse CCGCTACCGCCTCCACCCTGACCGATGTCTTTTTG 
 M-3 Cyt reverse CCGCTACCGCCTCCACCGTAGTCCATTCTGGAAC 
 M-4 Cyt reverse CCGCTACCGCCTCCACCTGTTTTCAGGATCCATG 
Mutants  
 K206/208A CGGCTTCGTTTCATGGCACAAGCTTTCAATATCGTTTGCC 
 K215G TCAATATCGTTTGCCCTCGGGTAAGTGGTTTCACAA 
 K242G GGTGTCCTGCTACAAGGAGATATCGGTCAGTAACGA 
 R196V CCGTCTACTTTTACTGCGTACGGAGGAGAGCCCGGC 

For flow cytometry studies the following Abs were used: 5F2 (CD8β, unconjugated, Santa Cruz Biotechnology), 2ST8.5H7 (CD8β, unconjugated, Beckman Coulter), OKT8 (CD8α, FITC, hybridoma), CD8α (PE-Cy7, eBioscience), CD8α (RPA-T8, PE, BioLegend), CD4 (RPA-T4, FITC, or APC, eBioscience), CD28 (APC, eBioscience), CD69 (FN50, PE, Biolegend), CD3 (UCHT1, PE, BioLegend), CD45RA (HI100, Pacific Blue, BioLegend), CD45RO (UCHL1, Alexa Fluor 700, BioLegend), CD62L (DREG-56, APC, BioLegend), and CCR7 (FITC, R&D Systems); and anti-mouse Abs for lysosomal-associated membrane protein-1 (LAMP-1) (1D4B, BD Pharmingen), CD8α (hybridoma supernatant), CD8β (hybridoma supernatant), and anti-ubiquitin (Covance and Upstate Biotechnology). Secondary Abs include goat anti-mouse PE and anti-rat Cy5 (Jackson ImmunoResearch Laboratories) and goat anti-mouse HRP (eBioscience).

M-4 peptide-specific Ab was generated in rabbits against the 3′ terminal 19 amino acids (YYSNTTTSQKLLNPWILKT) of the M-4 cytoplasmic tail (Proteintech Group). Briefly, the peptide was conjugated to KLH through a cysteine residue that was added at the 5′ end of the peptide. Two rabbits were immunized with the KLH-conjugated peptide using a standard 102-day protocol.

OT-1 TCR hybridoma cells expressing CD8α were provided by N. Gascoigne (Scripps Research Institute, La Jolla, CA). Retroviral vector pBMN-I containing the chimeric gene was expressed in phoenix packaging cells as described earlier (33). Supernantants of these cells were used to transduce OT-1 hybridomas, and positive clones were selected using G418 for selection. Cells were analyzed by flow cytometry for EYFP expression. M-4 isoform encoding cDNA was cloned into pBMN-I retroviral vector. JM human thymoma was retrovirally transduced as described above.

All human cells and tissues were obtained under protocols approved by the Yale Human Investigations Committee. PBLs were obtained from blood donors 18–60 years of age after informed consent. Thymocytes were obtained from thymus that was removed from infants during a cardiothoracic procedure. PBMCs were isolated by centrifugation at 900 × g for 30 min at 20–25°C over a Ficoll-Hypaque Plus (GE Healthcare) gradient. Cells were washed and incubated in RPMI 1640 medium to remove adherent cells. The nonadherent cells were then seeded into a 12-well plate at a concentration of 1 × 106/ml of T cell growth medium. If indicated, 12-well flat-bottom plates were coated with unconjugated OKT3 (CD3) monoclonal antibodies (10 μg/ml) and anti-CD28 (1 μg/ml) for stimulation.

Two-, three-, four-, or six-color immunostaining was conducted using standard protocol. Sorting of cells is described earlier. For sorting resting or stimulated CD8 and CD4, cells were gated for CD3 using anti-CD3 FITC and further separated using either anti-CD4 APC or anti-CD8α PE.

RNA was isolated from purified cells by an RNeasy spin kit (Qiagen). One microgram of total RNA was subjected to first strand cDNA synthesis. The oligo(dT)12–18-primed reverse transcriptase (RT) reaction was conducted in a total volume of 20 μl either with SuperScript II RT or without the enzyme (−RT control) according to the manufacturer’s protocol (Sprint Powerscript, Clontech). A total of 1 μl of cDNA was used for each TaqMan measurement. Quantitative PCR amplifications were performed with a Stratagene MxP3000 real-time system. As an internal control, the expression of β2 microglobulin was measured. cDNA was subjected to amplification in a 15-μl reaction mixture containing 2× TaqMan master mix (Applied Biosystems), sense primer, antisense primer, TaqMan probe, and sterile distilled water. The PCR conditions were as follows: initial denaturation at 95°C for 10 min followed by 40 cycles of denaturation at 95°C for 10 s, annealing at 62°C for 45 s, and extension at 72°C for 45 s. The sequences of the common forward primer, TaqMan probe, and unique reverse primers for all the CD8β isoforms are in Table I. A standard curve was generated using fluorescence data from the serial dilutions of either the plasmid or cDNA pool from donors, and the relative amount of each splice variant was determined and expressed relative to β2 microglobulin. Probability values <0.05 were considered to indicate statistical significance.

COS-7 fibroblasts and HEK-293 T cell line were cultured and maintained in RPMI 1640 or DMEM, respectively, each supplemented with 10% FCS (HyClone). Cells were transfected using Lipofectamine 2000 (Invitrogen) as described previously (34). Human CD8αβ expression was assessed using the CD8α Ab OKT8 and CD8β Abs 5F2 and 2ST8. Immunostaining and flow cytometry were performed using standard techniques and are presented as histograms (lin/log). The mean fluorescence intensity (MFI) was used as an index of fluorochrome staining intensity.

Cells (1 × 107) were solubilized in Triton-lysis buffer (1% Triton X-100, 10 mM Tris-HCl (pH 7.6), 100 mM NaCl, 5 mM MgCl2) or RIPA buffer (10 mM Tris-HCl (pH 7.6), 100 mM NaCl, 0.5% SDS, 1% sodium deoxycholate, 1% Nonidet P-40) that were freshly supplemented with protease inhibitor mixtures. The soluble cell lysates were mixed with 5 μg of anti-mouse CD8α Ab. Protein G-sepharose beads were added for 2 h, washed three times, and run on SDS-PAGE. Western blot was conducted with primary Ab followed by secondary (mouse True Blot; eBioscience) and developed using ECL Plus Western blot detection system (Amersham Biosciences).

Cells were seeded on 12-mm-diameter glass coverslip in 24-well plates and allowed to adhere for 1 h at 37°C and fixed with 2.5% paraformaldehyde followed by permeabilization. Incubations with primary and Cy5 conjugated secondary Abs were in FBS in PBS for 1 h at room temperature. Coverslips were mounted in ProLong Gold antifade reagent (Invitrogen) and observed with a Leica TCP SP2 confocal microscope. Quantitation of fluorescence intensities was done with Image J software (NIH).

Quantitative RT-PCR was performed to determine the expression patterns of CD8β splice variants based on the relative mRNA levels in various subsets of CD8+ T cells, including thymocytes and naive, activated, and memory cells. A common forward primer and probe and unique reverse primer specific for the cytoplasmic region of each isoform were developed (Fig. 1,C). Relative standard curves were generated to measure the expression levels of each isoform, and their amounts were normalized to β2 microglobulin (Table II). Values for the expression of each CD8β splice variant were expressed as a percentage of total CD8β. Variable patterns of all four splice variants were detected in thymocytes (Fig. 2,A). In two of the three samples, the mRNA pattern of CD8β variants was Μ-1 > Μ-4 > Μ-2, and in the third sample these were similarly expressed. M-3 mRNA was barely detectable. The mRNA for the same three isoforms was present in total CD8+ T cells Μ-1 > Μ-2/Μ-4 with an exception of one sample where M-4 levels were almost 2-fold higher than M-1 isoform. Again, M-3 mRNA was barely detectable (Fig. 2 B).

Table II.

Relative amount of each CD8β splice variant normalized to β2 microglobulin

Relative Amount of:
M-1M-2M-3M-4
Thymus 1 0.94 ± 0.075 0.9 ± 0.05 0.28 ± 0.034 0.85 ± 0.082 
Thymus 2 0.45 ± 0.12 0.12 ± 0.036 0.007 ± 0.001 0.25 ± 0.072 
Thymus 3 0.84 ± 0.274 0.35 ± 0.132 0.125 ± 0.2 0.46 ± 0.063 
Donor 1 0.27 ± 0.025 0.114 ± 0.02 0.038 ± 0.015 0.166 ± 0.04 
Donor 2 0.6 ± 0.151 0.274 ± 0.13 0.002 ± 0.001 0.177 ± 0.047 
Donor 3 1.217 ± 0.227 0.49 ± 0.02 0.004 ± 0.001 0.198 ± 0.013 
Donor 4 1.574 ± 0.39 1.034 ± 0.11 0.108 ± 0.02 0.427 ± 0.086 
Donor 5 0.76 ± 0.131 0.275 ± 0.084 0.18 ± 0.034 1.496 ± 0.15 
CD45RA+CD62L+CCR7+ 1.21 ± 0.27 0.613 ± 0.128 0.015 0.794 ± 0.297 
CD45RO+CD62L+CCR7+ 0.318 ± 0.113 0.104 ± 0.033 0.0005 0.4 ± 0.141 
CD45RO+CD62LCCR7 0.014 ± 0.002 0.013 ± 0.001 0.00002 0.125 ± 0.039 
CD8β+CD28+ 1.573 ± 0.39 1.034 ± 0.004 0.108 ± 0.021 0.427 ± 0.086 
CD8βlowCD28 0.027 ± 0.003 0.023 ± 0.004 0.017 ± 0.005 0.126 ± 0.01 
Relative Amount of:
M-1M-2M-3M-4
Thymus 1 0.94 ± 0.075 0.9 ± 0.05 0.28 ± 0.034 0.85 ± 0.082 
Thymus 2 0.45 ± 0.12 0.12 ± 0.036 0.007 ± 0.001 0.25 ± 0.072 
Thymus 3 0.84 ± 0.274 0.35 ± 0.132 0.125 ± 0.2 0.46 ± 0.063 
Donor 1 0.27 ± 0.025 0.114 ± 0.02 0.038 ± 0.015 0.166 ± 0.04 
Donor 2 0.6 ± 0.151 0.274 ± 0.13 0.002 ± 0.001 0.177 ± 0.047 
Donor 3 1.217 ± 0.227 0.49 ± 0.02 0.004 ± 0.001 0.198 ± 0.013 
Donor 4 1.574 ± 0.39 1.034 ± 0.11 0.108 ± 0.02 0.427 ± 0.086 
Donor 5 0.76 ± 0.131 0.275 ± 0.084 0.18 ± 0.034 1.496 ± 0.15 
CD45RA+CD62L+CCR7+ 1.21 ± 0.27 0.613 ± 0.128 0.015 0.794 ± 0.297 
CD45RO+CD62L+CCR7+ 0.318 ± 0.113 0.104 ± 0.033 0.0005 0.4 ± 0.141 
CD45RO+CD62LCCR7 0.014 ± 0.002 0.013 ± 0.001 0.00002 0.125 ± 0.039 
CD8β+CD28+ 1.573 ± 0.39 1.034 ± 0.004 0.108 ± 0.021 0.427 ± 0.086 
CD8βlowCD28 0.027 ± 0.003 0.023 ± 0.004 0.017 ± 0.005 0.126 ± 0.01 
FIGURE 2.

Differential expression pattern of human CD8β splice variants in thymocytes and PBLs. A, Flow cytometric analysis of human thymocytes showing CD8α and CD8β surface levels. Right panel, Relative mRNA abundance of each CD8β splice variant in thymocytes quantitated by real-time RT-PCR (thymus 1 represents thymocytes derived from a 7-mo-old infant, thymus 2 from a newly born infant, and thymus 3 was obtained from a 4-day-old infant). B, mRNA expression pattern of human CD8β splice variants (M-1, M-2, M-3, and M-4) in primary peripheral CD8+ T cells. For all the samples, β2 microglobulin PCR served as a control for integrity of cDNA samples and to normalize mRNA expression. Normalized values for each splice variant were represented as percentage of CD8β. Each bar represents the mean and SD of assays from duplicate samples. C, A peptide Ab was generated in rabbits against the cytoplasmic tail of the M-4 isoform. The specificity of this antiserum for M-4 isoform was first tested in a transient assay in HEK-293T cells. Cells were transfected with either plasmid containing CD8α alone (negative control; lane 1) or cotransfected with CD8α and different CD8β isoforms (M-1, lane 2; M-3, lane 3; M-4, lane 4; and M-2, lane 5). Forty-eight hours posttransfection cells were detergent solubilized by radioimmunoprecipitation assay (RIPA) buffer, and resulting lysates were immunoprecipitated with anti-CD8α mAb (OKT8) and protein G-coupled sepharose beads. Immunoprecipitates were resolved by SDS-PAGE, immunoblotted with either CD8β-specific mAb (5F2) to detect total CD8β (upper panel) or with M-4 peptide-specific Ab (lower panel), and visualized by ECL. To assay for expression in primary T cells, thymocytes obtained from thymus 3 (A) were treated as described above and similarly analyzed for M-4 isoform relative to total CD8β protein (lane 8). Human PBLs were stimulated in vitro using plate-bound Abs (anti-CD3 and anti-CD28) for 14 days and cells were either unstimulated (lane 6) or restimulated for 24 h (lane 7) before lysis and analyzed. JM human thymoma, retrovirally transduced with M-4 isoform, was used as a positive control (lane 9) and untransduced cells served as a negative control (lane 10).

FIGURE 2.

Differential expression pattern of human CD8β splice variants in thymocytes and PBLs. A, Flow cytometric analysis of human thymocytes showing CD8α and CD8β surface levels. Right panel, Relative mRNA abundance of each CD8β splice variant in thymocytes quantitated by real-time RT-PCR (thymus 1 represents thymocytes derived from a 7-mo-old infant, thymus 2 from a newly born infant, and thymus 3 was obtained from a 4-day-old infant). B, mRNA expression pattern of human CD8β splice variants (M-1, M-2, M-3, and M-4) in primary peripheral CD8+ T cells. For all the samples, β2 microglobulin PCR served as a control for integrity of cDNA samples and to normalize mRNA expression. Normalized values for each splice variant were represented as percentage of CD8β. Each bar represents the mean and SD of assays from duplicate samples. C, A peptide Ab was generated in rabbits against the cytoplasmic tail of the M-4 isoform. The specificity of this antiserum for M-4 isoform was first tested in a transient assay in HEK-293T cells. Cells were transfected with either plasmid containing CD8α alone (negative control; lane 1) or cotransfected with CD8α and different CD8β isoforms (M-1, lane 2; M-3, lane 3; M-4, lane 4; and M-2, lane 5). Forty-eight hours posttransfection cells were detergent solubilized by radioimmunoprecipitation assay (RIPA) buffer, and resulting lysates were immunoprecipitated with anti-CD8α mAb (OKT8) and protein G-coupled sepharose beads. Immunoprecipitates were resolved by SDS-PAGE, immunoblotted with either CD8β-specific mAb (5F2) to detect total CD8β (upper panel) or with M-4 peptide-specific Ab (lower panel), and visualized by ECL. To assay for expression in primary T cells, thymocytes obtained from thymus 3 (A) were treated as described above and similarly analyzed for M-4 isoform relative to total CD8β protein (lane 8). Human PBLs were stimulated in vitro using plate-bound Abs (anti-CD3 and anti-CD28) for 14 days and cells were either unstimulated (lane 6) or restimulated for 24 h (lane 7) before lysis and analyzed. JM human thymoma, retrovirally transduced with M-4 isoform, was used as a positive control (lane 9) and untransduced cells served as a negative control (lane 10).

Close modal

We were able to corroborate the presence of the M-4 protein by generating an antisera specific for its unique cytoplasmic tail. The difference between these isoforms is subtle and limited to their cytoplasmic tails. In addition to the C1 domain encoded amino acids, M-1 isoform has only 3 aa and M-3 has 14 unique aa residues, whereas M-2 (39 aa) and M-4 (36 aa) carry relatively longer unique amino acid sequences. We were able to generate a peptide-specific Ab for M-4 isoform cytoplasmic tail (terminal 19 aa residues) and detected its expression relative to the total CD8β protein. Initially we determined the specificity of this peptide Ab for M-4 isoform in a transient assay system using HEK-293 T cells transfected with mammalian expression vector expressing each isoform. Only the cells transfected with M-4 isoform showed a band corresponding to ∼32 kDa, slightly larger than the expected size of M-4 (28 kDa based on the primary amino acid sequence) largely due to the glycosylation of the protein. Cells transfected with the other three isoforms did not show any cross-reactivity with this Ab (Fig. 2,C, lower panel, lanes 1–5). All the isoforms were detected by immunoblotting with CD8β-specific Ab (5F2) that recognizes an epitope in the extracellular region of CD8β that is shared by all the isoforms (Fig. 2,C, upper panel, lanes 1–5). These isoforms could be distinguished by the difference in their sizes, with M-1 being the smallest (Fig. 2C, lane 2) and M-2 being the largest (Fig. 2 C, lane 5).

We were able to detect an M-4 isoform-specific band in both thymocytes (Fig. 2,C, lower panel, lane 8) and total PBLs (Fig. 2,C, lower panel, lanes 6–7) with the peptide Ab. Immunoprecipitation with IgG control Ab from the PBLs did not show any band corresponding to M-4 (data not shown). A quantitative analysis of M-4 isoform relative to total CD8β protein is not possible because the peptide Ab is a polyclonal antisera and anti-CD8β is a mAb. The intensity of the M-4-specific band as seen in primary T cells is less (Fig. 2,C, lower panel) as compared with the total CD8β-specific band (Fig. 2,C, upper panel), suggesting that M-4 constitutes a small fraction of total CD8β. In contrast, the JM thymoma line (human CD4+CD8α+CD8β), retrovirally transduced with M-4 isoform, showed a specific band (Fig. 2 C, lane 9) of higher intensity than total CD8β, suggesting that the avidity of the peptide Ab is not an issue contributing to weak binding in thymocytes or PBLs.

We analyzed two subsets of memory CD8+ T cells (CD45RO+ positive), the central and effector memory cells for CD8β isoform expression. CD3+ cells were gated for CD8+ T lymphocytes. Memory CD8+ T cells were further distinguished based on CD45RA or CD45RO staining, respectively (35). CD45RO gated cells were separated as central memory based on higher CD62L and CCR7 expression levels, whereas effector memory cells were characterized by lower levels of the same phenotypic markers (Fig. 3,A, left panel). The mRNA pattern was strikingly different in the effector memory population, which showed a predominance of the M-4 isoform (Fig. 3,A, lower right panel). In the representative sample shown, the M-4 variant was ∼7-fold higher than in M-1 in the effector memory subset, whereas in the central memory M-4 was comparable to the M-1 variant. Sample variation ranged from 2- to 7-fold. The naive CD8+ T cells showed ∼10-fold higher CD8β mRNA levels than in the memory subset, and the M-1 isoform mRNA was higher, as observed previously (Fig. 3 A, right panel, and 2B).

FIGURE 3.

Higher mRNA levels of M-4 splice variant in memory CD8+ T lymphocytes. Purified PBMCs were isolated, labeled with Abs, and subpopulations were sorted by FACS. A, For sorting, cells were gated on CD3 using anti-CD3 (PE) and CD8α populations using PE-Cy7-conjugated Ab. The memory CD8+ T cell population was distinguished based on CD45RA (Pacific Blue) and CD45RO (Alexa Fluor 700). Central memory CD8+ T cells were separated from effector memory CD8+ T cells based on CD62L (APC) and CCR7 (FITC) phenotypic markers. The human PBLs in this donor contained 33% CD3+, 8% CD4CD8+, 6% CD45RA, and 1.2% CD45RO. RNA was extracted from the defined subpopulations (naive, central, and effector memory), real time RT-PCR was conducted, and data were expressed as described previously. B, Purified PBMCs were stained with anti-CD4 (FITC), anti-CD8β Ab (5F2 followed by PE-conjugated goat anti-mouse IgG), and anti-CD28 (APC). PBLs in this donor containing 20% CD4CD8+ were sorted by gating on naive (CD8βhighCD28+) and TEMRA (CD8βlowCD28, 9.2%) surface expression levels. The purified T cell populations were stained with CD45RA Ab. Relative mRNA levels of CD8β variants were determined and data are expressed as described above, and one representative experiment of three is shown.

FIGURE 3.

Higher mRNA levels of M-4 splice variant in memory CD8+ T lymphocytes. Purified PBMCs were isolated, labeled with Abs, and subpopulations were sorted by FACS. A, For sorting, cells were gated on CD3 using anti-CD3 (PE) and CD8α populations using PE-Cy7-conjugated Ab. The memory CD8+ T cell population was distinguished based on CD45RA (Pacific Blue) and CD45RO (Alexa Fluor 700). Central memory CD8+ T cells were separated from effector memory CD8+ T cells based on CD62L (APC) and CCR7 (FITC) phenotypic markers. The human PBLs in this donor contained 33% CD3+, 8% CD4CD8+, 6% CD45RA, and 1.2% CD45RO. RNA was extracted from the defined subpopulations (naive, central, and effector memory), real time RT-PCR was conducted, and data were expressed as described previously. B, Purified PBMCs were stained with anti-CD4 (FITC), anti-CD8β Ab (5F2 followed by PE-conjugated goat anti-mouse IgG), and anti-CD28 (APC). PBLs in this donor containing 20% CD4CD8+ were sorted by gating on naive (CD8βhighCD28+) and TEMRA (CD8βlowCD28, 9.2%) surface expression levels. The purified T cell populations were stained with CD45RA Ab. Relative mRNA levels of CD8β variants were determined and data are expressed as described above, and one representative experiment of three is shown.

Close modal

We also used an alternative strategy for distinguishing memory population based on a previous report (36) where T cells from peripheral blood were gated on CD8β and CD28 high or low surface expression levels and sorted by flow cytometry. Higher surface expression of CD8β and CD28 on T cells is characteristic of naive cells (CD8βhigh, CD28high, CD45RA+), but after encountering Ag the terminally differentiated effector CD8+ T cells lose surface expression of CD28 and down-regulate CD8β (CD8βlow, CD28, CD45RA+). Both naive and effector memory RA CD8+ T cell (TEMRA) subsets stained positive with CD45RA Ab. The expression pattern in naive CD8+ T cells was M-1 > M-2 > M-4, whereas terminally differentiated effector memory cells or TEMRA expressed high levels of M4 mRNA relative to the other isoforms (Fig. 3 B).

To determine the effect of in vitro stimulation of T cells on the expression pattern of CD8β splice variants, T lymphocytes were activated using plate-bound Abs to cross-link the TCR/CD3 complex and the costimulatory molecule CD28. Surface expression of CD8β, CD8α, and CD69 (T cell activation marker, gated on CD8+ T cells) was determined by flow cytometry for both resting and activated cells (Fig. 4, upper panel). CD8β surface expression was higher after 24 h of stimulation and decreased at 72 h in activated cells relative to the resting population. CD8+ T cells were sorted after different time points (24 and 72 h) and RNA was extracted. By quantitative RT-PCR analysis, higher expression levels were demonstrated for M-1 and M-2 in activated relative to resting CD8+ T cells, whereas no significant change was observed in the mRNA levels of either M-3 (data not shown) or M-4 splice variants (Fig. 4, lower panel). The change in the expression pattern was distinct for both M-1 and M-2 whereas M-2 mRNA levels showed an ∼10–20-fold increase after 24 h of stimulation followed by a gradual decline through day 7 (data not shown). There was a constant increase in the mRNA levels of M-1 isoform for the same time points reaching a 2- to 7-fold increase by 72 h (Fig. 4, lower panel). The mRNA expression data were normalized to β2 microglobulin mRNA, and therefore the increase in mRNA results from an effect of stimulation on T lymphocytes.

FIGURE 4.

In vitro stimulation of human T lymphocytes alters mRNA expression of CD8β variants with significant increase in M-2 mRNA levels. Human T lymphocytes were either resting or stimulated with plate-bound OKT3 (10 μg/ml) and anti-CD28 (1 μg/ml) for 1 and 3 days. Upper panel, FACS plots showing CD8β, CD8α, and CD69 (T cell activation marker) surface levels in stimulated cells (black line) relative to resting (gray line) (mRNA expression pattern of total CD8+ T cells from donor 1 is shown in Fig. 1 B). Lower panel, Real-time PCR was used to quantify relative mRNA levels of M-1, M-2, and M-4 variants. β2 microglobulin PCR served as a control for integrity of cDNA samples and to normalize mRNA expression. The graphical analysis represents the fold induction or reduction of mRNA expression in the treated samples compared with the untreated sample. The fold induction of the untreated sample equals 1 in all experiments. The human PBLs from donor 1 contained 34% CD3+, 12.6% CD4+CD8, and 14.8% CD4CD8+. *, p < 0.05 as determined by Student’s t test. Each bar represents the mean and SD of assays from duplicate samples, and one representative of five independent samples is shown.

FIGURE 4.

In vitro stimulation of human T lymphocytes alters mRNA expression of CD8β variants with significant increase in M-2 mRNA levels. Human T lymphocytes were either resting or stimulated with plate-bound OKT3 (10 μg/ml) and anti-CD28 (1 μg/ml) for 1 and 3 days. Upper panel, FACS plots showing CD8β, CD8α, and CD69 (T cell activation marker) surface levels in stimulated cells (black line) relative to resting (gray line) (mRNA expression pattern of total CD8+ T cells from donor 1 is shown in Fig. 1 B). Lower panel, Real-time PCR was used to quantify relative mRNA levels of M-1, M-2, and M-4 variants. β2 microglobulin PCR served as a control for integrity of cDNA samples and to normalize mRNA expression. The graphical analysis represents the fold induction or reduction of mRNA expression in the treated samples compared with the untreated sample. The fold induction of the untreated sample equals 1 in all experiments. The human PBLs from donor 1 contained 34% CD3+, 12.6% CD4+CD8, and 14.8% CD4CD8+. *, p < 0.05 as determined by Student’s t test. Each bar represents the mean and SD of assays from duplicate samples, and one representative of five independent samples is shown.

Close modal

Fluorescently tagged chimeric CD8β genes were generated to study each isoform in an OT-1 murine hybridoma system (33). The advantage of this system is that it lacks endogenous CD8β, expresses peptide-specific OT-1-TCR, and carries fluorescently labeled CD3ζ-enhanced cyan fluorescent protein (ECFP) and CD8β-EYFP. The system was established for studying CD8-TCR interaction (37). A murine CD8β (Lyt-3) cDNA encoding the extracellular and transmembrane domains was fused to the cytoplasmic tails of each isoform and linked at the carboxyl terminus to EYFP (Fig. 5,A). These genes were retrovirally transduced into CD8β-negative OT-1 hybridomas previously transfected with wild-type CD8α and CD3ζ-ECFP, generating stable cell lines (OT-1.βM1Y, OT-1.βM2Y, OT-1.βM3Y, OT-1.βM4Y). All of the chimeric isoforms localized predominantly to the plasma membrane with the exception of CD8β M-2 isoform, which exhibited a unique distribution. Most of the CD8β M-2 protein was localized in endocytic compartments (Fig. 5,B). Flow cytometric analysis of these transductants further confirmed the reduced expression levels of chimeric M-2-EYFP (Fig. 5 C). Because the only difference between these chimeric proteins is in their cytoplasmic tails, it is unlikely that this was due to differences in the translational efficiency or misfolding of the protein. Barely detectable surface expression of nonchimeric wild-type CD8β M-2 isoform was consistently observed in stable transfectants of the human JM T cell thymoma (CD8α+CD4+) despite adequate levels of mRNA determined by RT-PCR (data not shown).

FIGURE 5.

Unique intracellular distribution of chimeric murine-human CD8β isoform with M-2 cytoplasmic tail in OT-1 T cell hybridoma. A, Schematic representation of EYFP-tagged chimeric murine-human CD8β isoforms. OT-1 hybridoma expressing CD3ζ-ECFP was retrovirally transduced with the retrovirus carrying an individual chimeric CD8β isoform (28 ). B, Confocal microscopy of OT-1.βM1Y, OT-1.βM2Y, OT-1.βM3Y, and OT-1.βM4Y hybridomas showing EYFP expression (first column). One representative of each cell is shown at a higher resolution with CD8β-ΕYFP expression (second column), surface marker CD3ζ-ECFP (third column) and merged (fourth column). C, Flow cytometric analysis of the same OT-1 hybridomas showing MFI for EYFP expression (filled gray line). CD3 expression levels served as internal control (empty black line).

FIGURE 5.

Unique intracellular distribution of chimeric murine-human CD8β isoform with M-2 cytoplasmic tail in OT-1 T cell hybridoma. A, Schematic representation of EYFP-tagged chimeric murine-human CD8β isoforms. OT-1 hybridoma expressing CD3ζ-ECFP was retrovirally transduced with the retrovirus carrying an individual chimeric CD8β isoform (28 ). B, Confocal microscopy of OT-1.βM1Y, OT-1.βM2Y, OT-1.βM3Y, and OT-1.βM4Y hybridomas showing EYFP expression (first column). One representative of each cell is shown at a higher resolution with CD8β-ΕYFP expression (second column), surface marker CD3ζ-ECFP (third column) and merged (fourth column). C, Flow cytometric analysis of the same OT-1 hybridomas showing MFI for EYFP expression (filled gray line). CD3 expression levels served as internal control (empty black line).

Close modal

One possible explanation for the intracellular localization of CD8β M-2 isoform was the possibility of ubiquitination of lysine residues in the M-2 cytoplasmic tail leading to internalization and degradation. Membrane trafficking of proteins can be controlled by ubiquitination of lysine residues. We therefore used site-directed mutagenesis to replace each lysine (K) residue of wild-type CD8β M-2 cytoplasmic domain with either glycine (G) or alanine (A). Of the four lysine residues, K206 is shared by all other isoforms while the remaining three are unique to the M-2 cytoplasmic tail. Lysine 206 and 208 were mutated to alanine, whereas 215 and 242 were changed to glycine (Fig. 6,A). COS-7 fibroblasts were transiently cotransfected with constructs carrying each mutant along with human CD8α. Vector alone transfectant was used as a negative control and M-1 construct as a positive control for CD8β expression. Transfectants were stained with anti-CD8α Ab OKT8, anti-CD8β Ab 5F2 (recognizes epitope on CD8β-chain alone), or anti-CD8αβ Ab 2ST85H7 (recognizes epitope only present on the CD8αβ heterodimer) to determine the surface expression of each mutant relative to the wild type (Fig. 6,B). The resulting transfectants with wild-type M-2 isoform showed <10% surface staining with either hCD8β-specific Ab 5F2 or anti-CD8αβ Ab 2ST85H7 relative to M-1 transfectant (Fig. 6,B). Neither the double mutant K206/208A nor the single mutant K242G altered cell surface expression of the M-2 isoform. In contrast, mutation of K215 that lies in between the other two unique lysine residues resulted in >50% surface expression of CD8β relative to the wild-type M-2 (Fig. 6 B). We further generated a quadruplet mutant in which all four lysine residues were mutated to either glycine or alanine to determine whether there was a synergistic effect of all the residues on the surface expression of M-2. The M-2 expression of this mutant was no better than K215G alone (data not shown). Therefore, the instability of M-2 was most likely due to a posttranslational modification of this residue that was targeting it for degradation.

FIGURE 6.

Expression of M-2 isoform is regulated by modification of a single lysine residue. A, The amino acid sequence of the cytoplasmic tail of M-2 variant highlighting the lysine residue in C1 exon shared by all variants (K206) and three lysine residues unique to M-2 tail (K208, K215, and K242). B, COS-7 cells were transiently cotransfected with expression vector carrying human CD8α and CD8β as indicated. Forty-eight hours posttransfection cells were stained with anti-CD8α (FITC) and anti-CD8β (2ST8 followed by PE-conjugated goat anti-mouse IgG) Abs. Left panel, Cells cotransfected with wild-type M-2, M-1 (positive control), or vector alone (negative control). Right panel, Cells cotransfected with M-2 lysine mutants. C, HEK-293 cells were transfected with vector alone (upper panel), M-2 wild-type (middle panel), or M-2 mutant K215G (lower panel). Transfected cells were stained with anti-CD8β (5F2) Ab. These experiments were performed at least three times, and representative data are shown. D, Upper panel, Confocal microscopy of OT-1.βM2 (K215G) Y mutant; lower panel, FACS analysis of total CD8β-EYFP expression in wild-type OT-1.βM2Y and mutant OT-1.βM2 (K215G) Y.

FIGURE 6.

Expression of M-2 isoform is regulated by modification of a single lysine residue. A, The amino acid sequence of the cytoplasmic tail of M-2 variant highlighting the lysine residue in C1 exon shared by all variants (K206) and three lysine residues unique to M-2 tail (K208, K215, and K242). B, COS-7 cells were transiently cotransfected with expression vector carrying human CD8α and CD8β as indicated. Forty-eight hours posttransfection cells were stained with anti-CD8α (FITC) and anti-CD8β (2ST8 followed by PE-conjugated goat anti-mouse IgG) Abs. Left panel, Cells cotransfected with wild-type M-2, M-1 (positive control), or vector alone (negative control). Right panel, Cells cotransfected with M-2 lysine mutants. C, HEK-293 cells were transfected with vector alone (upper panel), M-2 wild-type (middle panel), or M-2 mutant K215G (lower panel). Transfected cells were stained with anti-CD8β (5F2) Ab. These experiments were performed at least three times, and representative data are shown. D, Upper panel, Confocal microscopy of OT-1.βM2 (K215G) Y mutant; lower panel, FACS analysis of total CD8β-EYFP expression in wild-type OT-1.βM2Y and mutant OT-1.βM2 (K215G) Y.

Close modal

Similar results were observed using expression studies in the human cell line HEK-293. Transfection of the expression vector containing the CD8β M-2 mutant K215G in the absence of CD8α was sufficient for cell surface expression of CD8β in HEK-293 cells, as reported earlier for the M-1 isoform (3). The cells transfected with M-2 mutant K215G stained positive with hCD8β-specific Ab 5F2 but showed a weak staining for M-2 wild-type transfectant (Fig. 6 C). Moreover, the cell surface expression levels were improved as a result of heterodimerization in the presence of CD8α (data not shown).

We further examined retrovirally transduced OT-1.βM2Y, in which the lysine K215 was mutated to glycine, to investigate whether ubiquitination affected the cellular localization of the chimeric M-2 protein. Indeed, the M-2 (K215G) mutant was found primarily at the plasma membrane as opposed to the wild-type M-2, which localized in lysosomal compartment (Fig. 6,D, upper panel). These results were quantified by flow cytometric analysis of wild type OT-1.βM2Y and mutant OT-1.βM2 (K215G) Y. CD8β M-2 (K215G) mutant-expressing cells exhibited a 10-fold increase in the total levels of CD8β-EYFP relative to the wild type (Fig. 6 D, lower panel). Expression of wild-type M-2 fused to EYFP, and the CD8β M-2 mutant (K215G) in human JM thymoma showed a similar distribution of M-2 protein (data not shown).

Monoubiquitination by the addition of a single ubiquitin molecule to a lysine affects protein trafficking, whereas the addition of multiple ubiquitins to a lysine (polyubiquitination) leads to degradation by the 26S proteasome or other proteolytic pathways depending on how the ubiquitin chain is formed (38, 39). Chimeric M-2 indeed showed a distribution pattern similar to LAMP-1, a lysosomal marker, as can be observed in the overlay of the EYFP and LAMP-1, whereas the mutant K215G localized to the cell surface similar to CD3ζ (Fig. 7,A). Thus, degradation of M-2 appeared to reflect lysosomal proteolysis. Treatment of OT-1.βM2Y hybridoma with the lysosomotropic agent chloroquine resulted in a significant increase in chimeric M-2 (∼57 kDa) protein. Tagging CD8β with EYFP increased its size relative to CD8α and facilitated the detection of CD8β. A dose-dependent increase of high-molecular mass polyubiquitinated proteins was observed (Fig. 7,B) after chloroquine treatment. Treatment with higher concentrations of chloroquine was toxic to the cells (Fig. 7,B, lane 5). A mild increase in high-molecular mass proteins was also observed when the cells were treated with the proteasomal inhibitor MG132 (Calbiochem) (Fig. 7 B, lane 6).

FIGURE 7.

M-2 isoform localizes in the lysosomal compartment and is modified by ubiquitination. A, Immunofluorescence microscopy of OT-1.βM2Y (upper panel) and OT-1.βM2 (K215G) Y (lower panel) cells stained with anti-LAMP-1 Ab. CD3ζ surface expression for both of the cells is shown. B, Immunoprecipitation of chimeric CD8β from OT-1.βM2Y: untreated (lane 2), treated with increasing doses of chloroquine (0.5, 5, and 50 μM; lanes 3–5, respectively) and treatment with MG132 (100 μM, lane 6). Nontransduced cells were loaded as a control (lane 1). Shown are Western blot with anti-CD8β Ab (upper panel) and with anti-ubiquitin (lower panel). C, Ubiquitination of wild-type CD8β M-2 isoform. HEK-293 cells were transiently cotransfected with plasmid containing HA-tagged ubiquitin and vector alone (lane 1) or plasmid containing CD8α (lane 2) or CD8α and CD8β M-1 (lane 3) or CD8α and CD8β M-2 (lane 4). Forty-eight hours posttransfection cells were lysed and immunoprecipitated with anti-CD8α. Western blot was conducted with anti-CD8β (5F2) Ab (upper panel) and anti-HA Ab (lower panel).

FIGURE 7.

M-2 isoform localizes in the lysosomal compartment and is modified by ubiquitination. A, Immunofluorescence microscopy of OT-1.βM2Y (upper panel) and OT-1.βM2 (K215G) Y (lower panel) cells stained with anti-LAMP-1 Ab. CD3ζ surface expression for both of the cells is shown. B, Immunoprecipitation of chimeric CD8β from OT-1.βM2Y: untreated (lane 2), treated with increasing doses of chloroquine (0.5, 5, and 50 μM; lanes 3–5, respectively) and treatment with MG132 (100 μM, lane 6). Nontransduced cells were loaded as a control (lane 1). Shown are Western blot with anti-CD8β Ab (upper panel) and with anti-ubiquitin (lower panel). C, Ubiquitination of wild-type CD8β M-2 isoform. HEK-293 cells were transiently cotransfected with plasmid containing HA-tagged ubiquitin and vector alone (lane 1) or plasmid containing CD8α (lane 2) or CD8α and CD8β M-1 (lane 3) or CD8α and CD8β M-2 (lane 4). Forty-eight hours posttransfection cells were lysed and immunoprecipitated with anti-CD8α. Western blot was conducted with anti-CD8β (5F2) Ab (upper panel) and anti-HA Ab (lower panel).

Close modal

Polyubiquitination of M-2 isoform was further confirmed in HEK-293 cells that were transiently cotransfected with the expression vectors carrying HA-tagged ubiquitin and M-2 wild type. M-2 protein could be detected in Western blot with anti-CD8β Ab even with weak surface expression (Fig. 7,C, upper panel). Cells transfected with HA-Ub and vector alone/CD8α or M-1 isoform served as controls. Probing the same membrane with anti-HA Ab showed a predominant band at ∼70 kDa corresponding to the size of M-2 conjugated with three to four ubiquitin moieties indicating its polyubiquitination (Fig. 7 C, lower panel). Thus, surface expression of the M-2 isoform is regulated by ubiquitination of K215.

Because we had earlier observed increased levels of M-2 variant in activated primary human CD8+ T cells after stimulation (Fig. 4), we investigated whether M-2 protein localization would change upon stimulation. OT-1.βM2Y cells were stimulated by either antibody cross-linking (anti-CD3 and anti-CD28) or incubation with RMA-S cells pulsed with Kb-peptide (data not shown) for various time points at 37°C. After fixing, the cells were analyzed for EYFP distribution. An increase in EYFP signal at the cell surface was observed at 1 h poststimulation but gradually decreased at later time points (3 and 6 h, respectively; Fig. 8,A). The distribution of M-2 protein in activated cells was polarized similar to CD3ζ-CFP at the cell surface as indicated by arrows in Fig. 8,A. Moreover, increased colocalization was observed between the two fluorochromes (CD8β M-2-EYFP and CD3ζ-ECFP) as observed in the merged image (Fig. 8,A, lower panel). In some cells we did not observe a similar pattern of M-2 distribution on the cell surface, probably due to lack of synchronization between cells. Quantitation of immunofluorescence for both EYFP and ECFP showed that in unstimulated cells EYFP expression was higher intracellularly, and upon stimulation a significant increase in surface expression was observed in contrast to ECFP surface expression, which hardly changed (Fig. 8,B). The increase in surface expression of CD8β M-2 by immunofluorescence of EYFP correlated with the surface staining as determined using flow cytometry (Fig. 8 C). Increased surface expression of M-2 isoform after activation suggests a physiological role for this isoform in activated CD8+ T cells.

FIGURE 8.

T cell stimulation increases surface expression of M-2 isoform and its colocalization with CD3ζ. A, OT-1.βM2Y hybridoma were resting or stimulated in the presence of plate-bound OKT3 and anti-CD28 for different time points as indicated. Upper panel, Immunofluorescence microscopy of OT-1.βM2Y for CD8β M-2-EYFP; middle panel, CD3ζ-ECFP; lower panel, the two fluorochromes are merged. Arrows indicate the cells where colocalization of EYFP and ECFP is evident. B, Quantification of CD8β M-2-EYFP and CD3ζ-ECFP fluorescence intensities from A. The interior and exterior of the z-axis sections at 2 μm from the coverslip were manually marked (not shown), and the mean EYFP (upper panel) or ECFP (lower panel) fluorescence intensity was quantified in each area; the relative intensity is shown in the bar graph where data are expressed as the mean and SEM (n = 20 for all the time points shown; *, p < 0.001, Student’s t test). C, Flow cytometric analysis of OT-1.βM2Y hybridoma shown in A. Gray line (filled) represents isotype control, black line (dotted) shows unstimulated cells, and black line (empty) represents stimulated cells.

FIGURE 8.

T cell stimulation increases surface expression of M-2 isoform and its colocalization with CD3ζ. A, OT-1.βM2Y hybridoma were resting or stimulated in the presence of plate-bound OKT3 and anti-CD28 for different time points as indicated. Upper panel, Immunofluorescence microscopy of OT-1.βM2Y for CD8β M-2-EYFP; middle panel, CD3ζ-ECFP; lower panel, the two fluorochromes are merged. Arrows indicate the cells where colocalization of EYFP and ECFP is evident. B, Quantification of CD8β M-2-EYFP and CD3ζ-ECFP fluorescence intensities from A. The interior and exterior of the z-axis sections at 2 μm from the coverslip were manually marked (not shown), and the mean EYFP (upper panel) or ECFP (lower panel) fluorescence intensity was quantified in each area; the relative intensity is shown in the bar graph where data are expressed as the mean and SEM (n = 20 for all the time points shown; *, p < 0.001, Student’s t test). C, Flow cytometric analysis of OT-1.βM2Y hybridoma shown in A. Gray line (filled) represents isotype control, black line (dotted) shows unstimulated cells, and black line (empty) represents stimulated cells.

Close modal

Alternative pre-mRNA splicing is a key molecular event that generates a highly dynamic human proteome through a network of coordinated splicing events. At least 70% of human genes express multiple mRNAs through alternative splicing of exons or exon segments (40). This process allows a single gene to increase its coding capacity by expressing several related proteins with diverse and even antagonist functions, which is thought to be a major source for accelerated, lineage-specific evolution and eventual speciation, especially in mammals. Recently, Sorek (41) described the implication of new exons being added to evolving genomes that they are frequently associated with alternative splicing, with the new exon-containing variant typically being the rare one. This allows the new variant to be evolutionarily tested without compromising the original protein, and provides an evolutionary strategy for generation of novel functions with minimum damage to the existing functional repertoire, and, over time, if the new variant has a selective advantage it is positively selected (42). A prototypical example is the Homo sapiens and P. troglodytes CD8B gene that carries two additional exons (VIII and IX) that are missing in rhesus macaques and mice, which generate multiple alternatively spliced variants.

Using quantitative real-time PCR to determine the relative mRNA expression levels of the membrane-associated CD8β variants (M-1, M-2, M-3, and M-4), we found differential expression in thymocytes and in naive, memory, and activated CD8+ T lymphocytes. The CD8β M-1 variant was predominant in most thymocytes and peripheral blood CD8+ T cells. An increase in mRNA levels of the M-1 variant upon in vitro T cell stimulation supports the importance of this CD8β isoform in CD8+ T cell function. This is consistent with the fact that the only form of CD8β in the mouse is similar to the M-1 isoform with the last exon encoding FHK and FYK for mice and humans, respectively.

In contrast, the predominant isoform in the effector memory CD8+ T cell subset was the M-4 isoform. In two independent approaches to distinguish the memory subset from total CD8+ T cells, we observed much higher mRNA levels of M-4 variant than the other isoforms. The relatively higher levels of M-4 in the effector memory population may explain the one of five samples of PBLs in which mRNA for M-1 isoform was not the highest, but rather M-4 mRNA levels were almost 2-fold higher than M-1. This could be due to the higher percentage of memory CD8+ T cell population in PBMCs obtained from this donor. Alternatively, differences in the relative levels of the M-1, M-2, and M-4 isoforms could result from individual differences in splicing frequency. The presence of the M-4 isoform in effector memory population might confer unique properties to this cell type. Unlike the other isoforms, M-4 carries unique potential motifs in its cytoplasmic tail, including a Grb-2 binding motif (YSN) (43), a protein kinase C binding motif (SQK), and a potential di-leucine-based receptor internalization motif (LL) that can possibly influence signaling events downstream of TCR and modulate T cell function (44).

Activation-induced change in CD8B gene splicing transiently favored splicing toward M-2 mRNA, causing a significant up-regulation of M-2 RNA levels. The M-2 variant was described as an “aberrant” form in an earlier report, as no product of its cDNA was detected on the cell surface of transfected L cells, suggesting that the encoded protein either failed to bind to CD8α or resulted in a complex that was excluded from the cell surface (31). We obtained similar results in that the M-2 protein failed to show cell surface expression in the COS-7 monkey kidney cell line and in two human cell lines, the JM T cell thymoma and the HEK-293 cells. To address this paradox of why a cell would up-regulate a message for this isoform that is not expressed on the cell surface, we generated fluorescently tagged chimeras of these variants to determine their cellular distribution. Unlike the other variants that localized to the cell surface, the M-2 isoform showed a unique distribution in intracellular compartments. Our work establishes that exclusion from the cell surface of CD8β M-2 isoform in resting cells is due to posttranslational modification of K215 in the cytoplasmic tail by the addition of ubiquitin, which targets the protein for endocytosis as well as lysosomal/proteasomal degradation (Fig. 7). The large increase in the mRNA observed after activation of the human T cells most likely results in surface expression of this isoform, as we observed that after stimulation in the OT-1 hybridoma cells, M-2 was now expressed on the cell surface. The M-2 protein might play a unique role in stimulated CD8+ T cells.

Ubiquitination of proteins is a posttranslational modification that affects protein interaction and trafficking. A recent example that is related to changes in ubiquitination regulating cell surface expression is MHC class II (45). In murine immature dendritic cells, MHC class II is located in multivesicular bodies, a late stage endosome. In the multivesicular bodies, deubiquitinating enzymes remove ubiquitin before fusion with lysosomes for degradation. Dendritic cell maturation leads to expression of MHC class II on the cell surface, which is no longer ubiquitinated. The CD8β M-2 variant appears to be regulated similarly to MHC class II in that it undergoes constitutive ubiquitination in the absence of T cell stimulation and after stimulation appears on the cell surface, presumably as a result of changes in ubiquitination. Increased colocalization with CD3 presumably leads to increased down-regulation of the TCR complex. The di-leucine motif present in the M-2 cytoplasmic tail might contribute to the down-regulation.

The sequence diversity of the cytoplasmic tails of these isoforms in conjunction with the major differences in their phenotypic expression patterns support the hypothesis that these isoforms have specialized roles that fine tune or uniquely contribute to the function of the CD8+ T cells. An interesting question concerns how the CD8β isoforms contribute to the T cell function during differentiation, in memory subsets, or upon activation. These isoforms most likely influence coreceptor-mediated signaling pathways downstream of TCR engagement. This might not only affect the Ag sensitivity of CTL activation, but also effector functions both quantitatively and qualitatively. This could shape the intensity and/or duration of signaling and hence influence the production of particular cytokines/chemokines upon T cell activation. Additionally, CD8 internalization has been postulated to be a mechanism that modulates higher (chemokine secretion and cytolysis) or lower order (IFN-γ, TNF-α, and IL-2) CTL functions (46).

According to our data, the M-1 isoform is predominant in naive CD8+ T cells while the M-4 isoform is predominant in effector memory CD8+ T cells. Compared with naive CD8+ T cells located primarily in lymph nodes, Ag-specific memory CD8+ T cells have an increased ability to survey nonlymphoid (peripheral) sites for the presence of infection and respond quickly and effectively (47). They have a lower capacity to proliferate and secrete homeostatic cytokines. Hence, the cytokines/chemokines profile secreted by effector memory CD8+ T cells is different from activated naive T cells. We have preliminary data using JM T cell thymoma line expressing M-1 or M-4 isoforms that show differences in cytokine/chemokine secretion upon stimulation (unpublished data). Therefore, the two isoforms most likely contribute to functional differences between naive and effector memory CD8+ T cells. We are in the process of analyzing the functional differences between these isoforms, which most likely play a critical role in T cell development and function.

We thank the Yale University FACS core facility staff for their help with cell sorting, Professor Nicholas Gascoigne (Scripps Research Institute, La Jolla, CA) for OT-1 hybridomas and expression vectors with mouse CD8β-EYFP, Dr. Gary Kopf for providing the human thymus samples, and Professor Susan Kaech (Yale University) for critical reading of the manuscript. We also thank Shanta Nag for technical support and Farhana Dewan, a summer research student.

The authors have no financial conflicts of interest.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1

This work is supported by National Institutes of Health Grant RO1-CA48115 (to P.B.K.) and a Trudeau postdoctoral research fellowship (to D.T.).

3

Abbreviations used in this paper: MHC-I, MHC class I; ECFP, enhanced cyan fluorescent protein; EYFP, enhanced yellow fluorescent protein; HA-Ub, hemagglutinin-ubiquitin; LAMP-1, lysosomal-associated membrane protein-1; MFI, mean fluorescence intensity; pMHC-I, peptide-MHC class I; RT, reverse transcriptase; TEMRA, effector memory RA CD8+ T cell.

1
Moebius, U., G. Kober, A. L. Griscelli, T. Hercend, S. C. Meuer.
1991
. Expression of different CD8 isoforms on distinct human lymphocyte subpopulations.
Eur. J. Immunol.
21
:
1793
-1800.
2
Norment, A. M., D. R. Littman.
1988
. A second subunit of CD8 is expressed in human T cells.
EMBO J.
7
:
3433
-3439.
3
Devine, L., L. J. Kieffer, V. Aitken, P. B. Kavathas.
2000
. Human CD8β, but not mouse CD8β, can be expressed in the absence of CD8α as a ββ homodimer.
J. Immunol.
164
:
833
-838.
4
Janeway, C. A., Jr.
1992
. The T cell receptor as a multicomponent signalling machine: CD4/CD8 coreceptors and CD45 in T cell activation.
Annu. Rev. Immunol.
10
:
645
-674.
5
Straus, D. B., A. Weiss.
1992
. Genetic evidence for the involvement of the lck tyrosine kinase in signal transduction through the T cell antigen receptor.
Cell
70
:
585
-593.
6
Alcover, A., B. Alarcon.
2000
. Internalization and intracellular fate of TCR-CD3 complexes.
Crit. Rev. Immunol.
20
:
325
-346.
7
Luton, F., V. Legendre, J. P. Gorvel, A. M. Schmitt-Verhulst, C. Boyer.
1997
. Tyrosine and serine protein kinase activities associated with ligand-induced internalized TCR/CD3 complexes.
J. Immunol.
158
:
3140
-3147.
8
Valitutti, S., S. Muller, M. Salio, A. Lanzavecchia.
1997
. Degradation of T cell receptor (TCR)-CD3-ζ complexes after antigenic stimulation.
J. Exp. Med.
185
:
1859
-1864.
9
Crotzer, V. L., A. S. Mabardy, A. Weiss, F. M. Brodsky.
2004
. T cell receptor engagement leads to phosphorylation of clathrin heavy chain during receptor internalization.
J. Exp. Med.
199
:
981
-991.
10
Cenciarelli, C., D. Hou, K. C. Hsu, B. L. Rellahan, D. L. Wiest, H. T. Smith, V. A. Fried, A. M. Weissman.
1992
. Activation-induced ubiquitination of the T cell antigen receptor.
Science
257
:
795
-797.
11
Myers, M. D., T. Sosinowski, L. L. Dragone, C. White, H. Band, H. Gu, A. Weiss.
2006
. Src-like adaptor protein regulates TCR expression on thymocytes by linking the ubiquitin ligase c-Cbl to the TCR complex.
Nat. Immunol.
7
:
57
-66.
12
Couedel, C., M. Bodinier, M. A. Peyrat, M. Bonneville, F. Davodeau, F. Lang.
1999
. Selection and long-term persistence of reactive CTL clones during an EBV chronic response are determined by avidity, CD8 variable contribution compensating for differences in TCR affinities.
J. Immunol.
162
:
6351
-6358.
13
Demotte, N., D. Colau, S. Ottaviani, D. Godelaine, A. Van Pel, T. Boon, P. van der Bruggen.
2002
. A reversible functional defect of CD8+ T lymphocytes involving loss of tetramer labeling.
Eur. J. Immunol.
32
:
1688
-1697.
14
Kao, C., M. A. Daniels, S. C. Jameson.
2005
. Loss of CD8 and TCR binding to class I MHC ligands following T cell activation.
Int. Immunol.
17
:
1607
-1617.
15
Maile, R., C. A. Siler, S. E. Kerry, K. E. Midkiff, E. J. Collins, J. A. Frelinger.
2005
. Peripheral “CD8 tuning” dynamically modulates the size and responsiveness of an antigen-specific T cell pool in vivo.
J. Immunol.
174
:
619
-627.
16
Cawthon, A. G., H. Lu, M. A. Alexander-Miller.
2001
. Peptide requirement for CTL activation reflects the sensitivity to CD3 engagement: correlation with CD8αβ versus CD8αα expression.
J. Immunol.
167
:
2577
-2584.
17
Hoeveler, A., B. Malissen.
1993
. The cysteine residues in the cytoplasmic tail of CD8α are required for its coreceptor function.
Mol. Immunol.
30
:
755
-764.
18
Shaw, A. S., J. Chalupny, J. A. Whitney, C. Hammond, K. E. Amrein, P. Kavathas, B. M. Sefton, J. K. Rose.
1990
. Short related sequences in the cytoplasmic domains of CD4 and CD8 mediate binding to the amino-terminal domain of the p56lck tyrosine protein kinase.
Mol. Cell. Biol.
10
:
1853
-1862.
19
Veillette, A., M. A. Bookman, E. M. Horak, J. B. Bolen.
1988
. The CD4 and CD8 T cell surface antigens are associated with the internal membrane tyrosine-protein kinase p56lck.
Cell
55
:
301
-308.
20
Arcaro, A., C. Gregoire, N. Boucheron, S. Stotz, E. Palmer, B. Malissen, I. F. Luescher.
2000
. Essential role of CD8 palmitoylation in CD8 coreceptor function.
J. Immunol.
165
:
2068
-2076.
21
Irie, H. Y., K. S. Ravichandran, S. J. Burakoff.
1995
. CD8 beta chain influences CD8 alpha chain-associated Lck kinase activity.
J. Exp. Med.
181
:
1267
-1273.
22
Bosselut, R., S. Kubo, T. Guinter, J. L. Kopacz, J. D. Altman, L. Feigenbaum, A. Singer.
2000
. Role of CD8β domains in CD8 coreceptor function: importance for MHC I binding, signaling, and positive selection of CD8+ T cells in the thymus.
Immunity
12
:
409
-418.
23
Crooks, M. E., D. R. Littman.
1994
. Disruption of T lymphocyte positive and negative selection in mice lacking the CD8 beta chain.
Immunity
1
:
277
-285.
24
Itano, A., D. Cado, F. K. Chan, E. Robey.
1994
. A role for the cytoplasmic tail of the beta chain of CD8 in thymic selection.
Immunity
1
:
287
-290.
25
Nakayama, K., K. Nakayama, I. Negishi, K. Kuida, M. C. Louie, O. Kanagawa, H. Nakauchi, D. Y. Loh.
1994
. Requirement for CD8 beta chain in positive selection of CD8-lineage T cells.
Science
263
:
1131
-1133.
26
Doucey, M. A., L. Goffin, D. Naeher, O. Michielin, P. Baumgartner, P. Guillaume, E. Palmer, I. F. Luescher.
2003
. CD3δ establishes a functional link between the T cell receptor and CD8.
J. Biol. Chem.
278
:
3257
-3264.
27
Daniels, M. A., L. Devine, J. D. Miller, J. M. Moser, A. E. Lukacher, J. D. Altman, P. Kavathas, K. A. Hogquist, S. C. Jameson.
2001
. CD8 binding to MHC class I molecules is influenced by T cell maturation and glycosylation.
Immunity
15
:
1051
-1061.
28
Moody, A. M., D. Chui, P. A. Reche, J. J. Priatel, J. D. Marth, E. L. Reinherz.
2001
. Developmentally regulated glycosylation of the CD8αβ coreceptor stalk modulates ligand binding.
Cell
107
:
501
-512.
29
DiSanto, J. P., D. Smith, D. de Bruin, E. Lacy, N. Flomenberg.
1993
. Transcriptional diversity at the duplicated human CD8 beta loci.
Eur. J. Immunol.
23
:
320
-326.
30
Nakayama, K., Y. Kawachi, S. Tokito, N. Minami, R. Yamamoto, T. Imai, G. Gachelin, H. Nakauchi.
1992
. Recent duplication of the two human CD8 beta-chain genes.
J. Immunol.
148
:
1919
-1927.
31
Shiue, L., S. D. Gorman, J. R. Parnes.
1988
. A second chain of human CD8 is expressed on peripheral blood lymphocytes.
J. Exp. Med.
168
:
1993
-2005.
32
Pang, D. J., A. C. Hayday, M. J. Bijlmakers.
2007
. CD8 raft localization is induced by its assembly into CD8αβ heterodimers, not CD8αα homodimers.
J. Biol. Chem.
282
:
13884
-13894.
33
Yachi, P. P., J. Ampudia, N. R. Gascoigne, T. Zal.
2005
. Nonstimulatory peptides contribute to antigen-induced CD8-T cell receptor interaction at the immunological synapse.
Nat. Immunol.
6
:
785
-792.
34
Devine, L., D. Thakral, S. Nag, J. Dobbins, M. E. Hodsdon, P. B. Kavathas.
2006
. Mapping the binding site on CD8β for MHC class I reveals mutants with enhanced binding.
J. Immunol.
177
:
3930
-3938.
35
Sallusto, F., J. Geginat, A. Lanzavecchia.
2004
. Central memory and effector memory T cell subsets: function, generation, and maintenance.
Annu. Rev. Immunol.
22
:
745
-763.
36
Werwitzke, S., A. Tiede, B. E. Drescher, R. E. Schmidt, T. Witte.
2003
. CD8β/CD28 expression defines functionally distinct populations of peripheral blood T lymphocytes.
Clin. Exp. Immunol.
133
:
334
-343.
37
Yachi, P. P., J. Ampudia, T. Zal, N. R. Gascoigne.
2006
. Altered peptide ligands induce delayed CD8-T cell receptor interaction: a role for CD8 in distinguishing antigen quality.
Immunity
25
:
203
-211.
38
Haglund, K., P. P. Di Fiore, I. Dikic.
2003
. Distinct monoubiquitin signals in receptor endocytosis.
Trends Biochem. Sci.
28
:
598
-603.
39
Hicke, L., R. Dunn.
2003
. Regulation of membrane protein transport by ubiquitin and ubiquitin-binding proteins.
Annu. Rev. Cell. Dev. Biol.
19
:
141
-172.
40
Wang, G. S., T. A. Cooper.
2007
. Splicing in disease: disruption of the splicing code and the decoding machinery.
Nat. Rev. Genet.
8
:
749
-761.
41
Sorek, R..
2007
. The birth of new exons: mechanisms and evolutionary consequences.
RNA
13
:
1603
-1608.
42
Xing, Y., C. Lee.
2006
. Alternative splicing and RNA selection pressure: evolutionary consequences for eukaryotic genomes.
Nat. Rev. Genet.
7
:
499
-509.
43
Songyang, Z., S. E. Shoelson, J. McGlade, P. Olivier, T. Pawson, X. R. Bustelo, M. Barbacid, H. Sabe, H. Hanafusa, T. Yi, et al
1994
. Specific motifs recognized by the SH2 domains of Csk, 3BP2, fps/fes, GRB-2, HCP, SHC, Syk, and Vav.
Mol. Cell. Biol.
14
:
2777
-2785.
44
Dietrich, J., X. Hou, A. M. Wegener, C. Geisler.
1994
. CD3γ contains a phosphoserine-dependent di-leucine motif involved in down-regulation of the T cell receptor.
EMBO J.
13
:
2156
-2166.
45
Shin, J. S., M. Ebersold, M. Pypaert, L. Delamarre, A. Hartley, I. Mellman.
2006
. Surface expression of MHC class II in dendritic cells is controlled by regulated ubiquitination.
Nature
444
:
115
-118.
46
Laugel, B., D. A. Price, A. Milicic, A. K. Sewell.
2007
. CD8 exerts differential effects on the deployment of cytotoxic T lymphocyte effector functions.
Eur. J. Immunol.
37
:
905
-913.
47
Kaech, S. M., E. J. Wherry.
2007
. Heterogeneity and cell-fate decisions in effector and memory CD8+ T cell differentiation during viral infection.
Immunity
27
:
393
-405.