Main

Reduced resistance to some congenital or perinatal infections in human neonates is well documented(1,2). Weakness in T cell response may account, at least partly, for this vulnerability. Indeed, it is well established that neonatal T cells release little or no interferon (IFN)-gamma(38), interleukin (IL) 4, IL5, and IL10(5,9) in primary responses to various stimuli. In addition, defective neonatal IL2 secretion has been reported by several laboratories including our laboratory. In contrast to other cytokines, this defect is selectively observed in response to monoclonal Ab that engages the T-cell receptor (TCR)-CD3 complex (anti-CD3)(10,11) or the CD2(11,12), but it is not found after stimulation with mitogens. Furthermore, the intensity of this defect clearly varies depending upon the monoclonals used for stimulation(11), and it must be noticed that this variability might account for discrepancies between laboratories with regard to reduction of IL2 secretion or not at birth(13).

Although many studies were undertaken to elucidate the functional properties of neonatal T cells very few studies have addressed the mechanisms involved in their defective responsiveness. A silencer for the IL2 gene(14) has been described in antigen stimulated naive adult T cells, which are highly similar to neonatal T cells regarding their high CD45 RA and low CD45 RO phenotype(15,16), and their poor responsiveness to CD3 monoclonal Ab(1719). Another mechanism that might be complementary to the former would include alteration of signaling pathways. Important progress has been made in the understanding of TCR-CD3 mediated signaling. It is now clear that activation of protein-tyrosine kinases (PTK) and phosphatases plays a crucial role(1922) and that subsequent tyrosine phosphorylation of intracellular substrates is an obligatory event for IL2 production(23). Indeed, these early signaling events are responsible for a complex cascade of downstream events that ultimately activate the AP-1 transcription factor which in turn activates the gene encoding IL2(24). Therefore, analysis of protein-tyrosine phosphorylation (Tyr-P) in CD3-stimulated cells from newborns may provide important insights into the basic mechanisms underlying defective IL2 response to CD3 in neonates. In the present study, we thus analysed comparatively Tyr-P in CD3-stimulated cells from newborns (cord) and adults.

METHODS

Antibodies. The murine MAb OKT3 (Orthodiagnostic, Roissy en France, France) specific for CD3 ε was used for cell stimulation; CD3 cross-linking was ensured by goat Fab'2 mouse IgG (Tebu, Le Perray en Yvelines, France). MAb used for immunoblotting of tyrosine phosphorylated proteins were mouse antiphosphotyrosine clone 4G10 (Upstate Biotechnology Inc, Lake Placid, NY).

Cell preparations. Peripheral blood mononuclear cells (PBMC) from cords and from healthy adults were isolated as previously described(11) and then depleted of adherent cells by adherence to plastic as described elsewhere(25). Cell preparations of nonadherent cells were referred to as "lymphocytes" in the text. Purification of CD4+ T cells was performed by negative selection. Briefly, lymphocytes were incubated with monoclonals specific for HLA-DR, CD8, CD56, CD19 (Dako, Trappes, France), CD16 (Immunotech, Marseille, France), and glycophorin A (Becton Dickinson, Grenoble, France) before adding magnetic beads coated with an antimouse IgG and subsequent passage over a column according to the manufacturer recommendations (Dynal, Compiegne, France). Cytometric analysis of this selected population demonstrated that more than 95% of cells were stained with anti-CD3 and anti-CD4 (data not shown).

RNA preparation and polymerase chain reaction amplification of cDNA. Cell stimulation by cross-linked CD3 (OKT3: 2 µg/mL) in the presence of goat Fab'2 mouse IgG (8 µg/mL) was performed for 3 h as previously described(11). Cytokine mRNA expression was analyzed by using the semiquantitative reverse-transcriptase polymerase-chain reaction (RT-PCR) method. Briefly, 2 µg of total cellular RNA were reverse-transcribed with oligo-p(dt)15 primer using the first strand cDNA synthesis kit for RT-PCR (Boehringer Mannheim, Meylan, France). Nonlooping and nonoverlapping oligonucleotide set of primers from separate exons were prepared by genset (Oligo-express, Montreuil, France). As controls, the constitutively expressed CD3 and β actin genes in T cells were evaluated in addition to IL2 in each sample. Sequence primers were: CCAGGCTGATAGTTCGGTGACC and TGTCTGAGAGCAGTGTTCCCAC (CD3); TCGTCGACAACGGCTCCGGCATGTGC and TTCTGCAGGGAGGAGTCGGAAGCAGC (βactin); and TGTCTAGAAGAAGAACTCAACCTCTGG and GTGAAG CTTTTTAGAGCCCTAGGGC (IL2). Amplification reactions were performed in a 100-µL mixture containing 20 units/mL Taq polymerase (Promega, Madison, WI), 200 µM deoxynucleotide triphosphate, 0.2 µM of the two primers, and 1.5 mM Mg+ in Promega buffer. PCR cycles were 30 sec at 94°C, 30 sec annealing at 55°C, and 45 sec extension at 72°C. The reaction was performed in a Perkin Elmer/Cetus DNA thermal cycler for various cycles as indicated in the figure legends, each ending with a step of 10 min at 72°C. The amplified products obtained at various cycles for each sample were visualized by electrophoresis on agarose gel using ethidium bromide.

Analysis of tyrosine phosphorylation of intracellular protein substrates. Cells (2×106/cells in 200 µL of RPMI culture medium) were stimulated with cross-linked CD3 (OKT3: 5 µg/mL) in the presence of goat Fab'2 mouse IgG (8 µg/mL). When indicated, pervanadate was added during the last 2 min of stimulation at concentrations determined as optimal to synergize with CD3 for detecting tyrosine phosphorylation (hydrogen peroxide: 3 mM + sodium orthovanadate: 0.1 mM). The procedures used to analyze Tyr-P were conducted as previously described(26) except that sample size was reduced to 2×106 cells. In brief, solubilized cells transferred to nitrocellulose after electrophoresis were incubated with 4G10 an antiphosphotyrosine Ab (1 µg/mL) and subsequently probed with horseradish peroxydase conjugated anti-mouse IgG Ab.

RESULTS

Defective IL2 mRNA accumulation in CD3-stimulated cord lymphocytes. We previously reported that lymphocyte stimulation through TCR-CD3 ligation could detect a defect in T cell function in neonates. This defect is evidenced by impaired T cell proliferation(27) that is associated with a defect in IL2 secretion and mRNA expression(11). In the present study, we confirmed defective IL2 expression in neonates using the RT-PCR cycle method. Indeed, as illustrated in Fig. 1, we found that signals for IL2 mRNA expression were clearly detected from 30 to 40 cycles of PCR in adult lymphocytes stimulated for 3 h by cross-linked CD3, whereas no signal was observed in cord cells after stimulation for 3 h (Fig. 1) and up to 24 h (Ref. 11 and data not shown). Of note, these results were highly reproducible because they were observed in three of three experiments using different blood samples. As a control, similar amounts of the constitutively expressed β actin and CD3 genes were found in both populations (Fig. 1).

Figure 1
figure 1

Study of IL2 mRNA accumulation in CD3-stimulated adult and cord lymphocytes. Adult or cord lymphocytes were stimulated for 3 h with cross-linked CD3. At that time total mRNAs were isolated and reverse-transcribed before amplification by PCR using primers specific for β actin, CD3 (controls) or IL2. For β actin and CD3, each PCR was stopped after 15(1), 20(2), 25(3), 30(4) cycles. For IL2 each PCR was stopped after 25(3), 30(4), 35(5) and 40(6) cycles as indicated.

Comparison of tyrosine containing protein substrates in newborn and adult lymphocytes or CD4 T cells after CD3-stimulation. In a first set of experiments, time course of tyrosine phosphorylation in CD3-stimulated adult and newborn cells was evaluated by stimulating cells with cross-linked CD3 for different times. In these kinetic experiments, total cellular extracts were immunoblotted with an antiphosphotyrosine Ab as previously described. As illustrated in Figure 2, cross-linked CD3 could induce detectable tyrosine phosphorylation of several intracellular proteins in lymphocytes (A) or purified CD4 T cells(B) from cord cells (lines 3 and 5 compared with line 1) and in adults (lines 4 and 6 compared with line 2). Tyrosine phosphorylation of protein substrates is an early event since it is observed as soon as 2′ after stimulation through CD3 and was still detectable up to 20′. These results are reminiscent with data previously found by us(26) or others(28) in adult cells. It must be noted that some variabilities between experiments were observed at the later time points. Indeed at 15′-20′ the signals either remained at plateau values (data not shown) or slightly decreased (Fig. 2). More importantly, and as illustrated in Figure 1, no difference in kinetics was observed between newborns and adults in samples treated the same way in the same experiment. Furthermore, similar CD3-induced profiles of Tyr-P were observed in both populations with two predominant bands around 110-120 and 65-75 kD, respectively. Additional bands could also be found but these bands were variable from one experiment to another in both adult and cord cells (compare results in Figs. 2 and 3). Of note, under these experimental conditions, inducibility of tyrosine phosphorylation in lymphocytes or CD4 T cells was always very low or even under the threshold of the sensibility of the assay. Furthermore, variabilities in the intensities of Tyr-P from one experiment to another in both adult and cord cells precluded clear comparison between both populations. Therefore, experimental conditions were designed to potentiate signal intensities.

Figure 2
figure 2

Kinetics of tyrosine phosphorylation after stimulation with CD3. Total cellular proteins from lymphocytes or purified CD4 T cells from cord (C) or adult (A) bloods were analyzed for phosphotyrosine abundance using the 4 G10 antibody in the absence or presence of CD3 stimulation for the indicated times. 2×106 cells were analyzed per line. Arrows denote positions of proteins with the indicated size in kD. Asterisk represents mobility of heavy chain antibody.

Figure 3
figure 3

Pervanadate at the concentration of 0.1 mM synergize with cross-linked CD3 to induce tyrosine phosphorylation. Cells were incubated for 10′ in the absence of stimulation (line 1); as in line 1 except that pervanadate was added during the last 2 min of incubation (line 2); cells were incubated with cross-linked CD3 for 10 min (line 3) or with CD3 for 10 min plus pervanadate during the last 2 min of incubation (line 4). Tyrosine phosphorylation in whole cell lysates was evaluated as in Figure 2. Asterisk represents the mobility of heavy chain antibody.

In T cells tyrosine phosphorylation is controlled by coordinate actions of PTKs and protein-tyrosine phosphatases (PTPs). Thus, the steady state level of phosphotyrosine on cellular proteins is a consequence of the relative catalytic activities of intracellular PTKs and PTPs. Consequently, tyrosine phosphorylated substrates might be poorly detectable in the experimental conditions described above because of their rapid dephosphorylation by intracellular PTPs. To circumvent this difficulty, in a second series of experiments, cells were stimulated with cross-linked CD3 for 10′, and during the last 2′ of stimulation, pervanadate, a potent inhibitor of PTPs(29), was added to prevent dephosphorylation. As illustrated in Figure 3, at the concentration of 0.1 mM per mL, pervanadate alone induced low levels of Tyr-P, (line 2 compared with line 1) but was markedly synergistic with CD3 for detection of phosphorylated substrates in both newborn and adult cells (line 4 compared with line 3). Interestingly, as shown in Figure 4, the small, although, detectable levels of Tyr-P induced by pervanadate seemed decreased in cord. More interestingly, when pervanadate was added during the last 2 min of CD3 stimulation, a clear defect of Tyr-P up-regulation in neonatal lymphocytes (experiments A) or purified CD4 T cells(experiment B) was observed.

Figure 4
figure 4

Impaired tyrosine phosphorylation in newborn cells after stimulation. Cells, consisting in adult and cord lymphocytes (experiments A) or purified CD4 T cells (experiment B), were treated as in Figure 2 before evaluation of tyrosine phosphorylation from whole cell lysates.

DISCUSSION

Human CD4 T cells isolated from umbilical cord blood display the phenotypic and functional features of the immature compartment of immunologically naive T cells in adults. Phenotypically, they express low levels or no CD45 RO and high levels of CD45 RA, 80% of them bearing the CD31 adhesion molecule antigen(15,16). Functionally, neonatal or adult naive CD4 cells have no helper activity, induce CD8 suppressor cells(30), and after stimulation with various mitogens, produce mainly IL2 but not IL4 and little or no IFN gamma(39). Furthermore, there is a general consensus that these cells are less responsive to stimulation via CD3 than CD4+ memory cells expressing CD45 RO (see a review inRef. 31). After stimulation with mitogens or immobilized anti-CD3, neonatal cells rapidly acquire the phenotypic and functional features of memory T cell(9,27). This indicates that acquisition of T cell memory is associated with fundamental cell changes.

Antigen-induced T cell changes converting naive T cells into memory T cells might include alteration in signaling pathways. This was initially suggested by two reports showing that memory cells have higher basal levels of diacylglycerol and protein kinase C catalytic activities than naive cells and that CD3 antigen-mediated calcium signals and protein kinase C activation are prevalent in the CD45 RO compared with the CD45 RA subset of adult T lymphocytes(32). More recently, the hypothesis of an alteration of signaling in less mature cells has been reinforced by the finding of differential activation of phospholipase C gamma 1, mitogen-activated protein kinase(33) and P21 ras(34) in CD45 RA compared with CD45 RO cells. In these studies, however, it is not clear whether the subset defined by the CD45 RA phenotype is truly antigenically naive, particularly since adults have been exposed repeatedly and over a long period of time to a myriad of antigens in the environment. To circumvent these uncertainties, in the present study, neonatal T cells cord blood representing a population of truly unprimed cells were compared with T adult cells widely exposed to various antigens. Our results show that although CD3 alone induces small, although significant, levels of tyrosine phosphorylation in both adult and cord cells, accumulation of tyrosine phosphorylation induced by cross-linked CD3 in the presence of pervanadate was clearly reduced in cord cells as compared with adult cells. This strongly suggests that the intracellular machinery required for optimal Tyr-P responses via the TCR-CD3 complex is either incomplete or suppressed in neonatal T cells.

The basic mechanisms underlying defective tyrosine phosphorylation in newborn T cells are unknown. One plausible explanation for defective tyrosine phosphorylation in newborn cells would be alteration of specific signal transduction apparatus controlled by CD45. Indeed, this family of phosphatases is essential for initiating TCR-dependent tyrosine phosphorylations(35) and IL2 production(36). Furthermore, IL2 response to TCR-CD3 is far more efficient in transfectants expressing the low molecular weight of CD45 isoforms compared with transfectants expressing the high molecular weight(37). Another possibility, not exclusive from the former, is alteration of PTK. Two families of PTKs play an integral and obligatory role in the activation of the T cell antigen receptor: the src (lck, fyn) and syk (zap, syk) families. These two families of PTKs function in a sequential manner and cooperate to result in the induction of a variety of cellular tyrosine phosphoproteins among which several, such as 2z 70 and CD3 zeta, play a key role in T cell activation. Therefore, a defect in PTK activation should be involved in the defect of phosphorylation observed in cord. Further studies either focused on phosphorylation of substrates such as Zap 70 and CD3 zeta on data on PTK activities remain to be done to provide new insights into alteration of signaling which accompanied defective IL2 response to CD3 at birth.

Equally important but less well understood are the mechanisms that negatively regulate signal transduction. Among those PTP3, the SHP-1 protein has been implicated in the negative regulation of TCR mediated signaling by down regulating PTK activities(38,39). Higher expression of SHP-1 in the thymic medullary area compared with the cortex suggests that this phosphase might differentially regulate T cell signaling depending on the state of T cell maturation. Further experiments would be also of great interest to examine whether this PTP might play a role in differential signaling observed in newborn and adult cells.

In summary, our data provide biochemical evidence that CD3 stimulation may not fully activate tyrosine phosphorylation in newborn T cells. Because tyrosine phosphorylation is an obligatory event for IL2 expression, these alterations might contribute to the defective IL2 response observed in neonates. This hypothesis, however, remains to be experimentally demonstrated. In addition, our data provide new perspective to elucidate the basic mechanisms of immune incompetence in neonates and their consequences in clinical application.

Acknowledgments. The authors thank I. Dorval for helpful technical assistance, G. Bismuth for helpful comments and discussions, and B. Dauthuille for excellent secretarial assistance.