HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Ramiro, A. R. Articles by Toribio, M. L. Search for Related Content PubMed PubMed Citation Articles by Ramiro, A. R. Articles by Toribio, M. L.

Differential Developmental Regulation and Functional Effects on Pre-TCR Surface Expression of Human pT^a and pT^b Spliced Isoforms¹

Almudena R. Ramiro^*, María N. Navarro^*, Aura Carreira^*, Yolanda R. Carrasco^*, Virginia G. de Yébenes^*, Graciela Carrillo^*, José L. San Millán, Bent Rubin and María L. Toribio^2,*

^* Centro de Biología Molecular "Severo Ochoa," Consejo Superior de Investigaciones Científicas, Universidad Autónoma de Madrid, Cantoblanco, Madrid, Spain; Genetics Unit, Hospital Ramón y Cajal, Madrid, Spain; and Centre National de la Recherche Scientifique, Centre Hospitalier Universitaire Purpan, Toulouse, France

	Abstract

Top Abstract Introduction Materials and Methods Results Discussion References

 
Functional rearrangement at the TCR locus leads to surface expression on developing pre-T cells of a pre-TCR complex composed of the TCR-chain paired with the invariant pre-TCR (pT) chain and associated with CD3 components. Pre-TCR signaling triggers the expansion and further differentiation of pre-T cells into TCR mature T cells, a process known as selection. Besides the conventional pT transcript (termed pT^a), a second, alternative spliced, isoform of the pT gene (pT^b) has been described, whose developmental relevance remains unknown. In this study, phenotypic, biochemical, and functional evidence is provided that only pT^a is capable of inducing surface expression of a CD3-associated pre-TCR complex, which seems spontaneously recruited into lipid rafts, while pT^b pairs with and retains TCR intracellularly. In addition, by using real-time quantitative RT-PCR approaches, we show that expression of pT^a and pT^b mRNA spliced products is differentially regulated along human intrathymic development, so that pT^b transcriptional onset is developmentally delayed, but selection results in simultaneous shutdown of both isoforms, with a relative increase of pT^b transcripts in -selected vs nonselected pre-T cells in vivo. Relative increase of pT^b is also shown to occur upon pre-T cell activation in vitro. Taken together, our data illustrate that transcriptional regulation of pT limits developmental expression of human pre-TCR to intrathymic stages surrounding selection, and are compatible with a role for pT^b in forming an intracellular TCR-pT^b complex that may be responsible for limiting surface expression of a pT^a-containing pre-TCR and/or may be competent to signal from a subcellular compartment.

	Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

 
Progenitors of T lymphocytes that differentiate inside the thymus have to face at least two selection steps that involve signaling through two molecular sensors, the pre-TCR and the TCR, which are sequentially expressed during intrathymic development (reviewed in Refs. 1, 2, 3). Early thymocytes that succeed in a functional rearrangement at the TCR locus express first a pre-TCR complex composed of the TCR chain paired with a nonrearranging pre-TCR (pT)³ chain and associated with CD3 components (4, 5, 6, 7), which triggers the developmental checkpoint generally referred to as selection. Pre-TCR-mediated selection involves the survival and proliferation of those thymocytes carrying a productive TCR gene rearrangement, and the feedback inhibition of further rearrangements (allelic exclusion) at this locus (8, 9, 10, 11, 12, 13). Following selection, pre-TCR-mediated cellular expansion ends up abruptly, and TCR rearrangements are then induced in the resulting population of small nondividing pre-T cells (14, 15). On productive TCR gene rearrangements and substitution of pT by TCR, the TCR is expressed associated with CD3 on the membrane of developing thymocytes, which are then allowed to undergo a second step of selection (positive or TCR selection), which results in rescue from programmed cell death and final differentiation into mature T cells (2).

Most of the information regarding the functional expression of the pre-TCR in thymocyte development arises from approaches involving targeted disruption of components critical for the generation and assembly of the murine pre-TCR (16, 17, 18). Available data support that activity of the pre-TCR begins in CD44^+/-CD25⁺ thymocytes, when pT transcription is maximal (5) and functional TCR chain rearrangements are achieved (8, 19). However, actual detection of the pre-TCR on normal murine thymocytes has been elusive, mostly because the pre-TCR is expressed at very low levels (reviewed in Ref. 13), and up-to-date unambiguous biochemical evidence of a surface pre-TCR has only been obtained from TCR-deficient mice (4, 7). This barely detectable level of surface pre-TCR led us to wonder whether expression on the membrane of thymocytes was indeed essential for the selection process. Supporting the requirement of a surface pre-TCR, it was shown that the TCR chain has to exit the endoplasmic reticulum (ER) compartment to induce differentiation, cellular proliferation, and TCR allelic exclusion (20). However, pre-TCR ligand binding seems dispensable for its function, as T cell development was shown to be independent of pre-TCR extracellular Ig domains (21, 22). In this regard, it has recently been suggested that the pre-TCR could be endowed with constitutive activation properties, as it was found to be spontaneously recruited into lipid rafts enriched in activated signaling transducing molecules (23).

In contrast to mice, most of the information regarding pre-TCR expression in humans must be derived from direct analysis of ex vivo isolated normal thymocytes. Following such approaches, we have recently succeeded in the detection of surface pT expression on primary human thymocytes (15), and showed that it was confined in vivo to a limited fraction of CD4⁺CD8⁺ double-positive (DP) cycling thymocytes that coexpress low surface CD3. As CD3^low cycling DP thymocytes are the immediate progeny of CD4⁺CD8⁺ pre-T cells (24), in which TCR rearrangements are first induced in humans (25), they represent the most immature -selected pre-T cells. Pre-TCR surface expression is then rapidly down-regulated from the surface of -selected thymocytes that stop cycling and eventually become small resting cells, this coinciding with the onset of TCR rearrangements (15).

Despite the restricted developmental window of cell surface pre-TCR expression, pT transcription has been shown to span, both in mice and humans, all pre-T cell stages and to diminish only immediately before the single-positive stage (5, 26). Regulation of surface pre-TCR expression is therefore still an open question. In this regard, a second, alternatively spliced, isoform of pT, termed pT^b, has been described recently; this isoform lacks the second pT exon, coding for most of the extracellular Ig-like domain, but retains the cysteine residue presumably involved in TCR binding (27). pT^b is coexpressed in the murine thymus with the conventional pT^a isoform, although at significantly (10-fold) lower levels (27). Conservation of this pT^b alternatively processed transcript in humans (28) suggests that it may play an important role in thymocyte development. At present, the only available data on the putative function of pT^b come from studies in mice showing that each pT isoform has functionally distinct effects on TCR-chain expression on transfected cells (27). These data led to the proposal that pT^b may be more efficient than pT^a in bringing TCR to the cell surface, allowing higher expression of TCR as part of an alternative CD3-associated functional pre-TCR. However, the regulation of pT^b expression throughout T cell development has not been directly addressed as yet and, therefore, the physiological relevance of such an alternative pre-TCR remains to be determined.

In this study, real-time quantitative RT-PCR analyses provide evidence that developmental expression of pT^a and pT^b RNA spliced products is differentially regulated in the human thymus, so that maximal pT^b transcription is delayed as compared with pT^a, although both pT isoforms drop simultaneously immediately following selection. Moreover, our biochemical and functional studies support that, in contrast to mice, only pT^a can promote the expression of a functional surface pre-TCR in humans. The possibility that pT^b plays a role in regulating surface pre-TCR expression, and/or forms a pre-TCR complex competent to signal from an inner compartment is discussed.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

 
cDNA constructs, cell lines, and transfections

The human pT^a and pT^b full-length cDNAs were generated by amplification of total thymus cDNA with the sense 5'-GGG CCC GGA TCC ATA TGG CCG GTA CAT GGC TG-3' and antisense 5'-GGG GGA TCC CCG GCA GCT CCA GCC TGC AG-3' primers, followed by digestion and ligation into the BamHI site of the pEGFP-N1 vector (Clontech Laboratories, Palo Alto, CA). The TCR (V12.1) full-length cDNA was obtained by amplification from the TCR-pCDNA3 plasmid (29) with the sense 5'-ATG GAT CCT CTA GAT GAT TTT TGC CAG CCT GTT G-3' and antisense 5'-GGG GGA TCC CCG CTG GAC CAC AGC CGC AG-3' primers and further cloning into the BamHI site of pEGFP-N1 (Fig. 1A). The pT^a, pT^b, and TCR cDNAs were also subcloned, as previously described (29), into the BamHI site of the pcDNA3-derived bicistronic expression vector pCIGFP, which contains an internal ribosomal entry site (IRES) sequence, followed by the green fluorescent protein (GFP) cDNA (Fig. 1B). The human mature T cell line Jurkat (ATCC, American Type Culture Collection, Manassas, VA), the pre-T cell line SupT1 (ATCC), and the TCR-deficient JR3.11 mutant (30), derived from Jurkat, were grown in RPMI 1640 medium (BioWhittaker, Walkersville, MD) supplemented with 10% FCS (Life Technologies, Paisley, U.K.). Transfections were conducted by electroporation, as previously described (15). Briefly, 50 µg plasmid DNA were transfected at 264 V and 975 µF in a Gene Pulser II (Bio-Rad, Richmond, CA). Transient transfectants were analyzed 14–24 h after electroporation. Stable transfectants were generated by G418 selection, as described (15). Cotransfections were performed by electroporation with a fixed amount (10 or 20 µg) of pT^a-GFP plasmid cDNA plus increasing amounts of either pT^b-GFP plasmid cDNA or empty pCIGFP plasmid cDNA (29) as control, plus up to 70 µg carrier pcDNA3 plasmid (Invitrogen, Carlsbad, CA) per transfection point. Then, 6–8 h later, electronically gated GFP⁺-transfected cells were examined for surface CD3 levels by flow cytometry. The percentage of CD3 expression per transfection point was determined from the mean fluorescence intensity values using the corresponding pCIGFP controls as reference.

View larger version (26K): [in this window] [in a new window]   FIGURE 1. Pre-TCR surface expression is detected in pTa, but not pTb transfectants. A, Schematic representation of pTa-GFP, pTb-GFP, and TCR-GFP constructs used in transfection experiments, and CD3 surface expression in pTa, pTb, and TCR transient transfectants. EC, TM, and CYT stand for extracellular, transmembrane, and cytoplasmic regions, respectively. B, Schematic representation of bicistronic expression vectors containing pTa, pTb, or TCR cDNAs, followed by an IRES sequence and the EGFP cDNA, and CD3 surface expression in pTa, pTb, and TCR transient transfectants. JR3.11 cells were transfected with the constructs shown in A and B and analyzed 14–24 h later by flow cytometry after staining with Leu-4 PE anti-CD3 mAb. CD3 expression in GFP+ electronically gated cells (bold histograms) is overlaid for comparison with CD3 expression in GFP- cells (thin histograms). C, Surface expression of pre-TCR components in representative pTa-GFP (a13.2) and pTb-GFP (b71.2) stable transfectant cell lines. Anti-CD3 (Leu-4 PE), anti-TCR (anti-V8 mAb plus goat anti-mouse Ig PE), and anti-pT (K5G3 mAb plus anti-mouse Ig PE) stainings vs GFP expression are shown. Results are representative of >30 different pTa and pTb stable transfectants, and 10 independent experiments.

Directly PE-labeled anti-CD3 and unlabeled anti-CD3 mAbs were obtained from BD Biosciences (Franklin Lakes, NJ), and PE-labeled anti-CD69 from Caltag Laboratories (Burlingame, CA). The 1734-14 anti-V8 mAb was a generous gift from A. Alcover (Institute Pasteur, Paris, France). PE-labeled anti-mouse Ig Ab was purchased from Caltag Laboratories. Staining and flow cytometric analysis were performed as described (15) in an Epics XL cytometer (Coulter, Hialeah, FL). Analysis of intracellular Ca²⁺ fluxes was conducted as previously reported (31). Briefly, 10⁷ cells/ml were loaded with fura 2-acetoxymethyl ester (Sigma, St. Louis, MO). Fluorometric responses were recorded in a PerkinElmer (Norwalk, CT) fluorometer after stimulation with 10 µg/ml anti-CD3 OKT3 (ATCC) or anti-pT K5G3 (see below) mAbs. CD69 expression was analyzed by flow cytometry after a 12-h incubation of cells in plates previously coated with 20 µg/ml anti-CD3 OKT3 or in the presence of 20 ng/ml PMA (Sigma) plus 1 µM ionophore (Sigma).

Generation of the anti-pT K5G3 mAb

A cDNA encoding the extracellular region of the human pT-chain was generated by amplification with the sense 5'-CCG GAT CCA TAT GCT ACC CAC AGG TGT GGG C-3' and antisense 5'-CCC CGG ATC CTCA CAG CGC CCC ACC CGG TGT-3' oligonucleotides. The amplification was digested with BamHI and ligated into the BamHI site of the isopropyl -D-thiogalactoside-inducible, His-tag-containing pQE32 vector (Qiagen, Hilden, Germany). A 15-kDa recombinant protein was purified from transformed M15 Escherichia coli after isopropyl -D-thiogalactoside induction under 8 M urea-denaturalizing conditions following manufacturer’s instructions. A total of 50 µg keyhole limpet hemocyanin-coupled Ag was used to immunize BALB/c mice i.p. in combination with CFA (Difco, Detroit, MI). Second and third immunizations were conducted 30 days later with 50 µg Ag and incomplete Freund’s adjuvant (Difco), and 50 days later with 70 µg Ag alone, respectively. Lymph node cells from immunized mice were fused with the murine myeloma cell line Ag8653, according to standard procedures. Ag specificity of hybridoma supernatants was assessed by ELISA and confirmed by immunofluorescence staining of transfected COSA7 cells (not shown).

Confocal microscopy

JR3.11 transfectant lines were adhered to poly(L-Lys) (Sigma)-precoated coverslips (5 x 10⁵ cells/coverslip) by incubation at 37°C for 2 h. Coverslips were then washed in PBS, fixed with 2% paraformaldehyde in PBS for 10 min, and blocked with 2% BSA/PBS. For intracellular staining, cells were permeabilized for 5 min with 0.05% Triton X-100 (Sigma) before blocking. Cells were stained by incubation for 30 min with 4 µg/ml of either anti-protein disulfide isomerase (PDI) Ab (Stressgen, Victoria, Canada) or Leu-4 anti-CD3 mAb (BD Biosciences) plus E43 anti-CD59 mAb (generously provided by V. Horejsi, Institute of Molecular Genetics, Prague, Czech Republic). Secondary reagents include anti-rabbit Ig rhodamine (tetramethylrhodamine isothiocyanate (TRITC); Southern Biotechnology, Birmingham, AL), anti-mouse IgG1 TRITC, and anti-mouse Ig Cy5 (Jackson ImmunoResearch, West Grove, PA). Preparations were viewed using a Radiance 2000 system (Bio-Rad) coupled to an Axiovert S100TV inverted microscope (Zeiss, Oberkochen, Germany). Enhanced GFP (EGFP), TRITC, and Cy5 fluorescences were detected using bandpass filter HQ515/30, longpass filter HQ600/50, and longpass filter HQ660/LP, respectively.

Metabolic labeling and immunoprecipitation

Twenty-four hours after transfection, cells were washed twice in PBS and incubated for 30 min at 37°C in Met/Cys-free DMEM medium at 10⁷ cells/ml. Labeling was performed for 30 min by adding 300 µCi/ml of a combination of [³⁵S]Met/Cys (Amersham Pharmacia Biotech, Uppsala, Sweden). After an additional hour of incubation with 10% FCS-supplemented RPMI 1640 medium (BioWhittaker), cells were washed twice in PBS and lysed in 1% Brij 96 containing lysis buffer, as described (29). After preclearings with preimmune rabbit serum plus protein G-Sepharose beads (Amersham), lysates were immunoprecipitated with a rabbit antiserum against the pT cytoplasmic region (CT-1) (29) coupled to protein G-Sepharose beads. Immunoprecipitates were extensively washed with lysis buffer and resolved in two-dimensional SDS-PAGE.

Isolation of thymocyte samples

Postnatal thymocyte subsets were isolated from thymus samples removed during corrective cardiac surgery of patients aged 1 mo to 3 years, as previously described (24). Briefly, large- and small-sized thymocyte subsets were recovered from the 1.068 and 1.08 density layers, respectively, after centrifugation on stepwise Percoll density gradients (LKB, Uppsala, Sweden), and depletion of mature T, B, NK, and myeloid cells, as described (15). CD34⁺ and CD4⁺CD8⁺ cells were immunomagnetically sorted using anti-CD34- and anti-CD8-coated magnetic beads (Dynabeads; Dynal, Oslo, Norway), respectively. CD4⁺CD8^- were isolated from the CD8^- fraction with anti-CD4-coated beads (Dynal). Large DP CD3^- thymocytes were further fractionated into CD8⁺ and CD8⁺ cells by sorting in an EPICS Elite Cell Sorter (Coulter) after labeling with anti-CD8 (2ST8-5H7, kindly provided by E. L. Reinherz, Dana-Farber Cancer Institute, Boston, MA) plus PE-labeled goat anti-mouse IgG2a (Caltag Laboratories).

Real-time quantitative RT-PCR

Total RNA from thymocyte subsets was isolated using standard procedures, as described (26), and reverse transcribed into cDNA following manufacturer’s instructions (Life Technologies) using an oligo(dT) primer. A total of 10–50 ng retrotranscribed RNA was used per PCR reaction. The primers and TaqMan probes for RT-PCR were designed using the Primer Express software (Applied Biosystems, Foster City, CA). TaqMan probes were labeled with 6-carboxy fluorescein (Applied Biosystems). For pT^a the sense 5'-GTG TCC AGC CCT ACC CAC-3' and antisense 5'-ATC CAC CAG CAG CAT GAT TG-3' primers were used in combination with the 5'-TGT GGG CGG CAC ACC CTT TC-3' TaqMan probe. pT^b isoform was amplified separately using the sense 5'-GCC GGT ACA TGG CTG CTA CT-3' and antisense 5'-CTG TAG AAG CCT CTC CTG TG-3' primers together with the 5'-CCT GGC CCT TGG GTG TCC AGC-3' TaqMan probe. Primers and probes were used at a final concentration of 300 nM and 200 nM, respectively. GAPDH amplifications were conducted with the Pre-Developed TaqMan Assay Reagent specific for human GAPDH gene expression quantification (Applied Biosystems), according to manufacturer’s indications. All PCR reactions were set in triplicates using the TaqMan Universal PCR Master Mix (Applied Biosystems). Amplifications, detection, and analysis were performed in an ABI PRISM 7700 system (Applied Biosystems).

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

 
Impaired surface expression of human pT^b as part of a functional pre-TCR complex

RT-PCR analysis of RNA isolated from unfractionated human thymocytes, using pT-specific primers in exons 1 and 4, allowed us to amplify two different fragments, one of the expected size corresponding to pT^a, and a smaller one lacking 320 bp (data not shown). Cloning and sequencing of the smaller PCR product confirmed that it corresponded to the previously described human pT^b spliced isoform, which lacks exon 2 coding for most of the extracellular Ig-like domain (28). To gain some insights into the functional expression of a putative alternative pT^b-containing pre-TCR complex, transfection experiments were conducted using a TCR-deficient human T cell line, JR3.11, derived from Jurkat, in which TCR-chain transfection was shown to reconstitute surface expression of a functional CD3-associated TCR heterodimer (30). To detect expression in transfected cells, pT^a and pT^b cDNAs were fused to GFP (Fig. 1A) and then they were independently transfected into JR3.11 cells for comparison. Expression of surface CD3 was analyzed 24 h later by flow cytometry on transfected JR3.11 cells traced by their GFP expression. As shown in Fig. 1A, CD3 was expressed on GFP⁺ pT^a transfectants, but at very low levels, such as found on primary pre-T cells (15), while no CD3 was found on GFP^- cells. The situation was markedly different in pT^b-GFP transfectants, as no CD3 expression could be detected on GFP⁺ cells (Fig. 1A). Therefore, our data indicate that, unlike pT^a, transfection with pT^b does not result in surface expression of a pre-TCR complex. To confirm that the different effects of pT^a and pT^b on pre-TCR expression were not due to the GFP tag, the experiments were repeated with nontagged pT isoforms, using the bicistronic expression vector pCIGFP containing either pT^a or pT^b, followed by an IRES sequence and the GFP cDNA (Fig. 1B). Again, surface CD3 expression was induced on GFP⁺ cells transfected with pT^a, but not pT^b. In addition, control transfection experiments using either a TCR-GFP fusion construct or a bicistronic pCIGFP-TCR construct revealed normal surface expression levels of a CD3-associated TCR complex on GFP⁺, but not GFP^-, transfectants (Fig. 1, A and B).

Our results were further confirmed in JR3.11 pT^a-GFP and pT^b-GFP stable transfectants. As shown in Fig. 1C, the heterogeneous expression of GFP within each transfectant cell line served to trace the specificity of surface CD3 expression by two-color flow cytometry. Again, pT^a-GFP⁺ cells expressed CD3, but no CD3 expression was found on GFP⁺ cells from >30 pT^b transfectant cell lines analyzed. Of notice, all pT^a transfectants derived in this study (35 independent cell lines) displayed low levels of CD3, similar to those found on transient transfectants. More importantly, surface labeling with an anti-V8 mAb revealed that such GFP⁺ pT^a, but not pT^b, transfectants coexpressed CD3 and the endogenous TCR chain in stoichiometric amounts (Fig. 1C), suggesting that CD3 and TCR were expressed on pT^a transfectants as part of a conventional pT^a-TCR pre-TCR complex.

To seek direct evidence that pT^a was in fact expressed together with CD3 and TCR, different pT^a-GFP transfectant cell lines were then assayed for their reactivity with a mAb (K5G3) that was produced against the extracellular region of the human pT chain, as described in Materials and Methods. Immunofluorescence microscopy of COS transfectants provided evidence of the specificity of the anti-pT reagent, as it was reactive against all GFP⁺ COS cells transfected with pT^a-GFP, but not against pT^b-GFP transfectants (data not shown). As shown in Fig. 1C, flow cytometric analyses revealed that surface reactivity with K5G3 was exclusively detected in pT^a-GFP⁺ (but not in pT^b-GFP⁺) stable transfectants. These results provide formal proof that surface expression of all TCR, pT^a, and CD3 pre-TCR components is specifically induced after transfection with pT^a. Taken together, these data support the notion that, in contrast to mouse, human pT^a is unique, or at least more efficient than pT^b, in bringing TCR to the cell surface as part of a CD3-associated pre-TCR complex.

Still, it was possible that an alternative pT^b-containing pre-TCR could be expressed at the cell surface, although at levels below the detection limit of the flow cytometry technique. To rule out this possibility, more sensitive assays were performed that allow detection of surface expression of a functional pre-TCR. First, intracytosolic Ca²⁺ increases were analyzed in distinct pT^a and pT^b stable transfectants after CD3/pre-TCR engaging with mAbs against distinct surface epitopes. As shown in Fig. 2A, no Ca²⁺ fluxes were recorded in a representative pT^b transfectant in response to CD3 stimulation. In contrast, pT^a transfectants were all responsive to CD3 ligation, although Ca²⁺ signals were consistently lower than those induced in the parental Jurkat cells upon stimulation of the mature CD3-TCR. Importantly, Ca²⁺ mobilization was also induced by specific pT^a ligation with K5G3 mAb, further supporting surface expression of a functional pre-TCR.

View larger version (39K): [in this window] [in a new window]   FIGURE 2. pTa, but not pTb, can promote surface expression of a signaling-competent pre-TCR. A, Analysis of intracellular Ca2+ increases in pTa and pTb transfectants after stimulation. Jurkat, pTa-GFP (a10.4), and pTb-GFP (b1.4) cell lines were loaded with fura 2-acetoxymethyl ester, stimulated with anti-CD3 (OKT3) and anti-pT (K5G3) mAbs, and fluorometrically measured for Ca2+ mobilization. B, Analysis of CD69 expression after stimulation of pTa and pTb transfectants. pTa and pTb cell lines were stimulated with either anti-CD3 mAb, anti-pT mAb, or PMA for 12 h and analyzed for CD69 expression by flow cytometry. CD69-PE vs GFP dot plots are shown. Results are representative of five different experiments and four distinct pTa and pTb stable transfectants.

pT^b retains the TCR chain intracellularly

To determine the subcellular localization of the chimeric pT^b-GFP protein, several stable pT^b-GFP transfectants were analyzed by confocal microscopy and compared with pT^a-GFP transfectants. Green fluorescence examination of permeabilized cells (Fig. 3A) revealed that GFP was mostly found scattered throughout the cytosol in both pT transfectants (left panels), indicating that both pT^a and pT^b isoforms were preferentially located intracellularly, most likely in the ER. This possibility was confirmed after staining with an Ab recognizing the protein disulfide isomerase ER-resident protein. Most of the pT^a-GFP and pT^b-GFP chimeric proteins displayed a localization pattern similar to PDI staining (middle panels). Overlay analysis (right panels) showed that both proteins do in fact colocalize, and therefore, that the vast majority of both pT^a and pT^b chains remains retained within the ER in transfected cell lines.

View larger version (58K): [in this window] [in a new window]   FIGURE 3. Confocal microscopy analysis of pTa-GFP and pTb-GFP stable transfectants. pTa-GFP (a13.2) and pTb-GFP (b1.4) transfectants were adhered to coverslips and fixed, as described in Materials and Methods. A, Permeabilized cells were stained with anti-PDI Ab plus goat anti-rabbit Ig TRITC. Representative GFP, PDI, and overlaid GFP/PDI expression images are shown for both pTa and pTb transfectants. B, Nonpermeabilized cells were stained with anti-CD3 (Leu-4) mAb plus goat anti-mouse IgG1 TRITC and anti-CD59 (E43) mAb plus goat anti-mouse Ig Cy5. Representative images for CD3, and overlaid GFP/CD3, CD3/CD59, and GFP/CD3/CD59 expression patterns are shown.

To next investigate whether impaired surface expression of pT^b was due to a defective association to TCR, pT^a and pT^b transient transfectants were metabolically labeled with [³⁵S]methionine, and their lysates immunoprecipitated with a rabbit antiserum (CT-1) raised against a synthetic peptide contained in the cytoplasmic region of the human pT molecule, and hence shared by both pT isoforms (29), and resolved in two-dimensional gels. As expected, the CT-1 antiserum immunoprecipitated a pT^a-TCR dimer from pT^a transfectants that migrated out of the diagonal (Fig. 4, left). More importantly, pT^b-TCR dimers were also reduced from pT^b precipitates (Fig. 4, right), showing that the hampered expression of a pT^b-containing pre-TCR on the cell surface is not due to a deficient association of TCR.

View larger version (70K): [in this window] [in a new window]   FIGURE 4. The pTb isoform retains the TCR chain intracellularly. JR3.11 cells transiently transfected with either pTa-GFP (left) or pTb-GFP (right) constructs were metabolically labeled, immunoprecipitated with an anti-pT (CT) rabbit antiserum, and resolved by two-dimensional (nonreducing (NR)/reducing (R)) SDS-PAGE. Molecular mass markers in kilodaltons are shown on the right.

View larger version (19K): [in this window] [in a new window]   FIGURE 5. The pTb isoform is able to reduce surface pre-TCR expression levels induced on pTa transfectants. A, CD3 surface expression on SupT1 pre-T cells transiently transfected with pTa-GFP or pTb-GFP constructs (upper histograms), or with a bicistronic vector encoding untagged pTa or pTb isoforms plus EGFP. Twenty-four hours after transfection, cells were stained with Leu-4 PE anti-CD3 mAb and analyzed by flow cytometry. CD3 expression in GFP--untransfected cells (unshaded histograms) and GFP+-transfected cells (shaded histograms) is shown. B, Surface CD3 expression levels induced on TCR-deficient JR3.11 cells transfected with pTa are reduced in a dose-dependent manner upon cotransfection with pTb. Electronically gated GFP+ JR3.11 cells cotransfected with a fixed amount of pTa-GFP cDNA, plus increasing amounts of either pTb-GFP cDNA or empty pCIGFP plasmid as control, were examined for surface CD3 levels 8 h after transfection, as described in A. Expression changes were recorded as the percentage of anti-CD3 mAb binding (estimated as mean fluorescence intensity (MFI)) in samples cotransfected with pTa and pTb, relative to the corresponding negative controls cotransfected with pTa and empty pCIGFP. Data from two representative independent experiments are shown.

Independent regulation of pT^a and pT^b expression during human intrathymic development

To seek a possible developmental role for pT^b in vivo, we aimed at establishing the expression patterns of both pT^a and pT^b isoforms during human intrathymic development. This issue was approached by real-time quantitative PCR, which provides a unique means to quantify transcription accurately, because detection of the amplified product proceeds along the whole reaction, by using a combined thermocycler detector system. This allows performance of the quantification from the earliest phases of amplification, when the amount of product is proportional to the starting target DNA quantity. Primers and TaqMan probes were designed to independently amplify and detect pT^a and pT^b in different thymus subsets. The primer-probe combinations for both isoforms are represented schematically in Fig. 6A. Primers for pT^b amplification were designed in the first exon and in the junction between exons 1 and 3, respectively. The TaqMan probe for pT^b annealed with the first exon. As the specificity of this amplicon relied on only in the exon 1-to-exon 3 junction, which is partially conserved among splice sites, different 3' primers had to be tested to fulfill the required specificity. The specificity of the chosen pair is depicted in Fig. 6B, which shows that amplification of a plasmid containing pT^b cDNA was detected from very early cycles of the reaction. On the contrary, when a pT^a-containing plasmid was used as template, amplification curves overlaid with the control reactions where no template was added. This result ensured that pT^a would not be cross-detected with the pT^b primers in the thymus samples. Specificity of the pT^a primers and probe was tested as well (not shown).

View larger version (60K): [in this window] [in a new window]   FIGURE 6. Specificity of pTb detection by quantitative RT-PCR. A, Design of pTa and pTb amplicons. Arrows, Primers; , TaqMan probes. Exon composition is also depicted. B, pTb primers and TaqMan probe were used to amplify pTb and pTa plasmid templates. PCR reactions were performed in triplicates. No-template amplification curves are shown as a negative control. Rn, change in normalized reporter signal.

cDNA was synthesized from RNA obtained from thymic subsets representing different developmental stages, and subjected to amplification. The results obtained for one representative experiment covering all thymocyte subsets from the very same individual are represented in Fig. 7A. It was found that the most immature CD34⁺CD4^- thymocytes display moderate levels of pT^a expression, which are increased to maximal levels in the downstream CD4⁺CD8^-CD3^- pre-T cell stage. Such a high transcriptional rate is maintained in CD4⁺CD8⁺ cells, immediately preceding the -selection checkpoint (24). However, pT^a transcription drops abruptly after this critical step, such that pT^a mRNA levels are 10-fold reduced in both cycling (large) and resting (small) CD4⁺CD8⁺ thymocytes, when functional expression of the pre-TCR has already taken place (24), but before TCR expression. A distinct transcription pattern was observed for pT^b. Up-regulation of pT^b transcription is developmentally delayed as compared with pT^a, such that pT^b expression levels are increased first at the CD34⁺CD4^- to CD4⁺CD8^- transition, but peak at the next CD4⁺CD8⁺ stage, immediately upstream of -selection (Fig. 7A, upper graph). Transcription level of pT^b decreases afterward, although somewhat less dramatically than pT^a, in -selected CD4⁺CD8⁺ thymocytes. As a consequence of this apparently independent regulation of pT^a and pT^b expression, pT^b:pT^a ratio increases along intrathymic development, indicating a relative accumulation of pT^b isoform at later pre-T cell stages (Fig. 7A, lower graph). This relative increase of pT^b:pT^a ratio in -selected vs -unselected (CD4⁺CD8^-) thymocytes is shown in Table I for four independent experiments.

View larger version (28K): [in this window] [in a new window]   FIGURE 7. Differential regulation of pTa and pTb isoforms during human intrathymic development. A, Intrathymic subsets representing sequential maturational stages were isolated, as described in Materials and Methods, and subset identification is depicted at the bottom. RNAs were isolated, reverse transcribed, and subjected to quantitative RT-PCR. pTa, pTb (see Fig. 6A), and GAPDH-specific amplifications were conducted in parallel in triplicate reactions for each of the cDNA samples. Top, pTa and pTb values were obtained after normalization to GAPDH values. Bottom, pTb:pTa ratio. Relative transcriptional amounts are given in arbitrary units. B, Susceptibility of pTa and pTb mRNAs to cellular activation. SupT1 pre-T cells were activated by treatment with PMA plus ionophore. Activation was assessed by staining with anti-CD69 PE Ab and flow cytometry analysis. Histograms on the left show CD69 expression on SupT1 cells before (unshaded histogram) and after treatment (shaded histogram). Expression of pTa and pTb mRNAs measured by quantitative RT-PCR before () and after () SupT1 activation is shown on the right for a representative experiment.

View this table: [in this window] [in a new window]   Table I. pTb/pT1 transcriptional ratio in pre--selected vs post--selected thymocytes in four independent experiments1

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

 
In both mice and humans, a second, alternatively spliced, mRNA isoform of pT, pT^b, has been described, whose developmental relevance remains unknown. The aim of the present study has been to approach this issue in the human system. By focusing on the biochemical properties and regulation of expression of pT^b in vivo, we show in this study that, in contrast to pT^a, pT^b is unable to promote the expression of a human pre-TCR complex on the surface of transfected cells, but retains the ability to bind to TCR intracellularly. Moreover, both isoforms are shown to display a different pattern of mRNA expression along human intrathymic development, which provides evidence of their independent regulation in vivo, and supports their individual roles during the T cell developmental process.

Experimental evidence for the different capacities of pT isoforms to regulate surface TCR expression was originally provided by Barber and coworkers (27) in mice. However, the murine pT isoforms were shown to behave differently from the human isoforms. Indeed, in contrast to its human counterpart, the murine pT^b isoform was proposed to be more efficient than pT^a in bringing a pre-TCR to the cell surface. This proposal was based on the finding that pT^a, but not pT^b, was capable of reducing surface levels of a transfected TCR chain on a TCR-deficient T cell line. Although species-specific functional differences cannot be ruled out, it is also likely that discrepancies between mice and humans rely on the particular experimental systems used. In this regard, it is worth noting that Barber and coworkers (27) showed that transfection of TCR-deficient murine cells with TCR alone resulted in measurable levels of surface TCR, which was thus expressed independently of pT; however, surface expression pT or CD3 was difficult to detect upon cotransfection with pT^a or pT^b isoforms. The situation was markedly different in the TCR-deficient human T cell line JR3.11 used in this study, since no surface TCR expression occurred in the absence of a TCR-pairing subunit. Therefore, we could directly assess the contribution of pT^a and pT^b isoforms in bringing the endogenous TCR chain to the cell surface.

Transfection of JR3.11 cells with pT^a and pT^b, either untagged or tagged to GFP, has yielded important clues on the differential expression, biochemical properties, and possible functions of both isoforms in humans. In the first place, GFP tracing of transfected cells, together with the availability of appropriate specific reagents against individual pre-TCR components, has allowed us to conclude that pT^a, but not pT^b, is able to support surface expression of a CD3-associated TCR-pT heterodimer proved to be fully functional in terms of CD69 induction and Ca²⁺ mobilization. This is a somehow unexpected result regarding previous experimental arguments on the requirement of an additional surrogate VpreT chain, whose identification has remained elusive as yet, to promote pre-TCR expression on the membrane (5). Therefore, our findings indicate that, unless expressed on the mature JR3.11 line, the hypothetical VpreT chain either does not exist in humans, or is not absolutely required for the assembly and expression of a surface pre-TCR.

In addition to phenotypic and functional data, confocal microscopy examination confirmed the expression of a surface CD3-associated complex in pT^a, but not in pT^b transfectants, and also revealed that, in line with previous reports (1, 12, 15, 33), extremely low pre-TCR levels are able to reach the plasma membrane of our pT^a transfectants regardless of the amount of protein found in the cytoplasm, as traced by GFP expression (data not shown). Interestingly, such low surface pre-TCR expression levels displayed a raft-associated distribution that closely resembles that recently reported for murine pre-T cells (23), and further reinforces the idea that the pre-TCR is spontaneously recruited into rafts and reaches the plasma membrane as a constitutively active complex also in humans. This in turn has been related to the apparently dispensable (or nonexistent) pre-TCR ligand (22, 23). More interestingly, this analysis showed that, regardless of their differential surface expression patterns, both pT^a and pT^b human isoforms, similarly to their murine counterparts (27), are predominantly found in the ER within the cell. This fact may reflect the existence of specific retention mechanisms that, as proposed recently (29), could map within the pT cytoplasmic domain shared by both pT isoforms.

The absence of pre-TCR complexes from the surface of our pT^b transfectants raised the possibility that, despite retaining the cysteine residue involved in TCR binding, pT^b was unable to form dimers with the TCR chain and, hence, could not promote the expression of a pre-TCR complex on the cell surface. However, we provide evidence that this is not the case, as TCR-pT^b dimers could be precipitated from the cytoplasm of transfected cells. Therefore, other reasons must account for the impaired expression of pT^b-containing complexes on the plasma membrane. One attractive possibility is that one or more CD3 chains may be displaced or absent from such pT^b complexes, therefore preventing their release from the ER and their transport to the cell surface. Experiments are currently in progress to determine the exact composition of the cytoplasmic pT^b-TCR heterodimer in terms of CD3 association, and preliminary results suggest that CD3 can associate with the pT^a-TCR, but not with the pT^b-TCR heterodimer (unpublished results). Regardless of structural constraints, our finding that overexpression of pT^b inhibits or reduces the expression of endogenous pre-TCR from the surface of SupT1 pre-T cells could suggest a possible competitive or regulatory role for pT^b, which may be unable to reach the plasma membrane, but instead is capable of binding and retaining TCR and/or CD3 chains intracellularly. This possibility was further supported by the finding that surface pre-TCR expression levels induced on TCR-deficient JR3.11 cells transfected with pT^a can be reduced in a dose-dependent manner upon cotransfection with pT^b.

Although data regarding the regulation or stability of the pT^b protein in vivo are still lacking, our observation that, as in mice (27), pT^a mRNA is on average about 10-fold more abundant than pT^b in the whole thymocyte population in humans (unpublished results) may argue against a competitive role for pT^b in thymic differentiation. However, as pT^b expression is expected to be time and stage specific in developmental terms, no definitive conclusions on its role could be drawn based on quantitative data obtained from total thymus samples. With this view in mind, we sought a possible function for pT^b by studying its expression throughout T cell development. In this respect, quantitative RT-PCR has allowed us to compare accurately the shifts in pT^a and pT^b expression from small intrathymic subsets representative of independent maturation stages. This analysis has revealed that while transcription of both pT isoforms is maximal at the selection checkpoint, pT^b peak of expression is delayed with respect to pT^a and consistently decreases less dramatically (about 5-fold on average) than pT^a following selection, which together entails an increasing pT^b:pT^a ratio during intrathymic development. On the one hand, this differential transcriptional pattern supports that expression of both isoforms (either at the mRNA processing itself or at the stability level) is tightly controlled throughout T cell development, most likely reflecting functional differences between them. A similar finding has been reported for the alternatively spliced products of Pax-5 during B cell development (34). On the other hand, our data suggest that pT^b function may be restricted to a narrow developmental window either coincident or immediately following TCR chain expression and selection. Finally, although previous data indicate that anti-CD3-induced pre-TCR signaling could silence the pT locus (35), this is the first demonstration that normal ex vivo cells have lost expression of both pT^a as well as pT^b following selection. As selection simultaneously results in transcriptional activation of the TCR locus, as measured by TEA transcription (3, 26), such a transcriptional regulatory mechanism would, in turn, limit developmental expression of the pre-TCR to stages preceding TCRchain expression. Therefore, competitive displacement of pT by TCR during TCR assembly, although proved functionally relevant in vitro in mice (36), does not seem to be required to control surface pre-TCR expression in vivo in humans. According to this view, both pre-TCR and mature TCR can be coexpressed on SupT1 transfectants (29). This proposal concurs with our finding that pT^a transcription drops abruptly in pre-T cells lacking TCR immediately following PMA-induced activation. In addition, pT^b is only slightly reduced under such experimental conditions, most likely reflecting the relative increase of pT^b transcripts observed in -selected cells in vivo.

Although definitive conclusions on the physiological relevance of the relative pT^b transcriptional increase observed in -selected pre-T cells demand protein expression analysis at the single cell level, our data are compatible with a competitive function for pT^b in thymocyte differentiation responsible for limiting or regulating surface expression of the TCR-pT^a pre-TCR following -selection, but before TCR expression. In addition, there still remains the possibility that pT^b plays an independent signaling function from a subcellular compartment. It is currently assumed that the pre-TCR has to exit the ER to trigger selection (20). However, as it has been discussed elsewhere (33), folding constraints could at least in part be responsible for that effect and, therefore, signaling from a post-ER compartment could also be considered. While rigorous testing of this attractive possibility awaits functional approaches such as those recently reported in mice (37), it seems reasonable to think that access to the surface could be dispensable provided that a competent complex is formed inside the cell, as could be the case for human pT^b.

	Acknowledgments

 
We thank Drs. A. Alcover, V. Horejsi, and E. L. Reinherz for the generous gift of Abs; Drs. C. Hernández and M. Viribay for invaluable help with quantitative RT-PCR; Drs. R. Bragado, P. Fuentes, and E. Teixeiro for technical advice on functional assays; Dr. C. Sánchez for advice on confocal microscopy; and Drs. B. Alarcón, J. Millán and J. M. Redondo for helpful discussions.

	Footnotes

 
¹ This work was supported by grants from the Comisión Interministerial de Ciencia y Tecnología (SAF 97-0161), Dirección General de Enseñanza Superior (PB97-1194), Comunidad Autónoma de Madrid (08.3/0015.1/99 and 08.3/0021/2000), and Fondo de Investigación Sanitaria (FIS 00/1044), and by an institutional grant from the Fundación Ramón Areces. A.R.R. is supported by a postdoctoral fellowship from the Consejo Superior de Investigaciones Científicas.

² Address correspondence and reprint requests to Dr. María L. Toribio, Centro de Biología Molecular "Severo Ochoa." Universidad Autónoma de Madrid, Cantoblanco, 28049 Madrid, Spain. E-mail address: mtoribio{at}cbm.uam.es