Constant IFNγ mRNA to protein ratios in cord and adult blood T cells suggests regulation of IFNγ expression in cord blood T cells occurs at the transcriptional level

Abstract

Low levels of IFNγ produced by umbilical cord blood (UCB) T lymphocytes upon activation may be due to the need for a high threshold of activation or to intrinsic blocking transcription/translation. We examined IFNγ mRNA accumulation and protein expression in pharmacologically stimulated human UCB and adult blood (AB) T cells. Our data indicate that both IFNγ mRNA accumulation and protein synthesis were significantly lower in stimulated UCB T cells than the AB T cells. Since the RNA dependent kinase PKR, an inhibitor of translation, can be activated by low levels of IFNγ mRNA, we measured its involvement. Treatment with 2-amino-purine, an inhibitor of PKR, did not enhance IFNγ protein expression in UCB T cells. Furthermore, our studies indicated that IFNγ promoter hypermethylation does not appear to regulate IFNγ expression either, as treatment with the demethylating agent, 5-aza-2′-deoxycytidine, did not lead to a significant increase in IFNγ mRNA accumulation in UCB T cells. What is readily evident from our studies is that the IFNγ mRNA to protein ratio was similar in UCB and AB T cells and it was not altered by any of the treatments used. These results therefore suggests that IFNγ expression in UCB T cells is suppressed at the transcriptional level by an unknown mechanism(s).

Introduction

Neonatal T cells are generally considered to be immature due to their inability to mount an efficient immune response against pathogens [1–3]. Many reports have indeed shown that both cytotoxic and helper neonate T lymphocytes differ qualitatively and quantitatively from adult T lymphocytes in their response to produce effector cytokines, for example IFNγ, in vivo and in vitro[4–7]. The basis for this difference, however, remains unclear.

It has been proposed that the neonatal T cells obtained from cord blood require a high threshold of activation for producing mature effector functions. The proliferation of low cytokine producing neonatal cells can be enhanced with high concentrations of pharmacological agents PMA and ionomycin. In this regard, studies have shown that in the presence of strong adjuvants or costimulatory signals, neonatal CD4⁺ and CD8⁺ T cells are capable of mature-level responses [8–11]. It has also been suggested that UCB T lymphocytes have an intrinsic defect, the nature of which is unclear, for producing Th1 specific cytokines, and preferentially favour Th2 responses [12,13]. Previous findings from our laboratory have described that upon stimulation with the high concentrations of PMA and ionomycin both CD4⁺ and CD8⁺ UCB T lymphocytes expressed mature levels of the activation marker CD69 but did not produce Th1 cytokine such as IFNγ[14]. Therefore, activation of UCB T cells per se does not appear to be blocked, but rather production of Th1 cytokines, like IFNγ, is hindered.

The pleiotropic cytokine IFNγ is known to be principally produced by NK cells, cytotoxic CD8⁺ T cells and the Th1 subset of CD4⁺ T cells. IFNγ is essential for both innate and adaptive immunity, generally amplifying Th1 immune responses and acting through binding to the IFNγ receptors on the cell surface that trigger Jak-STAT signalling pathway [15,16]. Neonatal lymphocytes, in comparison to adult lymphocytes, produce 10–15 fold less IFNγ protein upon anti-CD3 crosslinking or stimulation with phorbol ester and calcium ionophore [17–21].

The poor IFNγ expression by neonatal T cells might involve mechanisms regulating gene transcription and/or translation. In this regard it has been proposed that differential methylation of the IFNγ promoter is closely associated with low IFNγ expression in the UCB T lymphocytes. In CD4⁺ UCB T cells IFNγ promoter is hypermethylated at CpG and non CpG sites within and adjacent to the promoter [22–24], which could reduce IFNγ transcription. In addition to the above reports describing transcriptional regulation of IFNγ expression, Ben Asouli et al. [25] have shown that the human IFNγ mRNA (at low concentration) auto regulates its own translation through a pseudoknot that activates the RNA dependent protein kinase PKR. When activated, PKR blocks translation through the phosphorylation of the eIF2 α subunit of the RNA polymerase II dependent elongation complex [26]. In addition to the mechanisms described above, reduced constitutive NFATc2 (nuclear factor of activated T cells) expression has been reported in UCB T lymphocytes during primary stimulation, which may be one underlying molecular mechanism for the low IFNγ expression in UCB T lymphocytes [27].

Therefore, in order to further explore the regulation of IFNγ expression in neonatal T cells, we analysed both umbilical cord blood (UCB) and adult blood (AB) T cells for IFNγ mRNA accumulation, using real time taqman® PCR, and IFNγ protein expression, either by FACS or ELISA. In order to obtain maximal T cell activation we used two pharmacological agents: PMA and ionomycin. In this study we show that the regulation of IFNγ expression in neonatal T cells occurs only at the transcriptional level, and an intrinsic program hinders the induction of IFNγ expression in these cells.

Materials and methods

The Cantonal Institutional Review Board of Basel, Switzerland approved this study. UCB samples were collected from healthy term babies at the University Women's Hospital, Basel and AB samples were obtained from healthy donors at the blood donation centre, University Hospital, Basel. Written informed consent was obatined in all instances.

Isolation and stimulation of T lymphocytes

PBMC and cord blood mononuclear cells were isolated from whole blood by centrifugation over a ficoll paque plus gradient (Amersham Biosciences, Uppsala, Sweden). Then CD4⁺ and CD8⁺ T cells were enriched with CD4⁺ and CD8⁺ T cell specific microbeads (Miltenyi Biotech, Gladbach, Germany) either together or separately using MACS mini columns (Miltenyi Biotech, Gladbach, Germany) according to the manufacturer's instructions. The purity of the enriched T lymphocytes routinely reached >90%, as confirmed by FACS analysis.

0·5 × 10⁶ cells/ml T cells were stimulated in 12 well plates (Nunc, Denmark) for varying time periods ranging from 3 to 48 h at 37°C in 5% CO₂ in RPMI 1640 medium (Gibco Invitrogen Life Technologies, Grand Island, NY, USA) supplemented with 5% FCS, 2 mm l-glutamine (Gibco Invitrogen Life Technologies), 100 U/ml penicillin, 100 µg/ml streptomycin (Gibco Invitrogen Life Technologies). 50 ng/ml PMA (Sigma Chemicals, St. Louis, MO, USA) and 1 µm ionomycin (Sigma Chemicals) were used as stimuli. To analyse intracellular IFNγ expression golgistop (BD Biosciences, Basel, Switzerland) was added to the cell culture 3 h prior to FACS analysis [28]. Further treatments included the use of 2-amino-purine (2-AP, Sigma Chemicals) and 5-aza-2′-deoxycytidine (Sigma Chemicals) as indicated in the figures or figures legends. Our analysis using the tryptan blue dye exclusion test indicated that at ≥ 90% of the T cells were viable after each drug treatment.

Intracellular cytokine staining and flow cytometry

Following stimulation, cultured T cells were washed twice in PBS containing 0·5% FCS and incubated with FITC labelled mAb (BD Biosciences) against CD3, CD4 or CD8 surface antigens for 15 min at 4°C. The cells were again washed twice with PBS containing 0·5% FCS then fixed and permeabilized using cytofix/cytoperm plus kit (BD Biosciences) according to the manufacturer's instructions. Cells were then incubated with PE labelled mAb (BD Biosciences) against IFNγ or IL-2 cytokines. Phenotypic analysis of naïve and mature T cells was performed by two colour flow cytometry using a Becton Dickinson FACScan flow cytometer (BD Biosciences). A minimum of 10 000 gated events was acquired and analysed with the Cell Quest Pro software (BD Biosciences).

IFNγ ELISA

IFNγ levels were measured using a commercial human IFNγ ELISA kit (Ebioscience Inc, San Diego, CA, USA) according to the manufacturer's instructions. ELISA plate was read at 450 nm in an ELISA reader (Molecular Devices, Sunnyvale, CA, USA) and the data was analysed using Soft Max Pro Software.

Total RNA isolation and real time RT-PCR

Total RNA was isolated using High Pure RNA isolation kit (Roche Diagnostics, Mannheim, Germany). cDNA was reverse transcribed from 1 µg of total RNA using a reverse transcription system (Promega, Madison, WI, USA) according to the manufacturer's instructions. For the real time RT-PCR analysis, 25 µl reaction  mixture consisted of 1 µl of cDNA, 1 × taqman universal master mix (Applied Biosystems, Foster City, CA, USA) and 1 × predeveloped taqman assay reagents (PDAR, Applied Biosystems), containing gene specific probes and primers for IFNγ or 18 s RNA. For these experiments the ABI Prism 7000 system (Applied Biosystems) was used. 18 s RNA was used as an internal control to correct unequal sample loading. It has previously been shown that the amount of 18sRNA per cell does not vary with cell activation [29]. Thermal cycler conditions comprised of 2 min at 50°C, 10 min at 95°C followed by 45 cycles for 15 s at 95°C and at 60°C for 1 min.

The change in IFNγ mRNA expression was assayed by normalization to the 18sRNA internal control. In order to obtain the fold difference the data was analysed using the ΔΔC_T method described previously by Livak and Schmittgen, where ΔΔC_T = (IFNγ C_T– 18sRNA C_T)_stimulated– (IFNγ C_T– 18sRNA C_T)_unstimulated[30].

Results

UCB T lymphocyte synthesize very low amount of IFNγ protein and mRNA

UCB and adult T lymphocytes were stimulated with the high concentrations of PMA and ionomycin and the frequency of IFNγ positive cells was measured by FACS. This analysis indicated that following 3 h of stimulation the frequencies of IFNγ positive UCB T cells were 100 fold less than the AB T cells (Fig. 1 a,b). In contrast the frequencies of the IL-2 positive T cells were similar in both UCB and AB (Fig. 1c,d). Although the frequencies of IFNγ positive UCB T cells further increased about 10 fold (3·39%) and 20 fold (5·17%) after 24 h and 48 h of stimulation, respectively (Fig. 1e,f), it never reached the equivalent levels attained by AB T cells.

Fig. 1.

IFNγ and IL-2 production by UCB and AB T cells. Enriched UCB and AB T cells were stimulated for 3 h to 48 h with PMA and ionomycin. The frequencies of the IFNγ and IL-2 positive cells were measured by FACS. 10 000 gated events were acquired. The frequencies of (a) the IFNγ positive UCB T cells were 100 fold less than (b) the IFNγ positive AB T cells, while the IL-2 positive T cell population was similar in both groups (c, d). Frequencies of IFNγ positive UCB T cells were increased approximately (e) 10-fold and (f) 20-fold upon 24 h and 48 h stimulation, respectively, in comparison to 3 h stimulation.

To examine whether low IFNγ expression by UCB T cells was a consequence of low IFNγ gene transcription, IFNγ mRNA accumulation was quantified by real time PCR. The amount of IFNγ mRNA was normalized against the quantity of 18 sRNA [29]. Following 24 h of stimulation with PMA and ionomycin, 12-fold less IFNγ mRNA was accumulated by UCB T lymphocytes than the AB T lymphocytes (Table 1). IFNγ mRNA accumulation, measured at 3 h after stimulation, by UCB T cells was 30 fold less than the AB T cells (data not shown).

Table 1.

Real time quantitative PCR analysis of IFNγ gene expression using 2^–ΔΔCT method. Cord and adult blood T lymphocytes were stimulated for 24 h using PMA and ionomycin. The fold change in IFNγ gene expression was normalized to an internal control (18sRNA gene). 10 samples of each, cord and adult blood, were analysed. The fold change in the IFNγ gene was calculated using 2^–ΔΔCT method as described in the materials and methods. Following 24 h stimulation UCB T lymphocytes accumulated 12 fold less IFNγ mRNA.

Samples	IFNγ CT	18 s RNA CT	Δ CT*	ΔΔ CT**	2–ΔΔ CT†
Unstimulated CB T cells	35·47 ± 0·78	15·93 ± 0·41	19·54 ± 0·82	0·0 ± 0·0	1
Stimulated CB T cells	25·98 ± 0·77	15·37 ± 0·34	10·61 ± 0·78	− 8·93 ± 0·15	489·10 ± 0·50
Unstimulated AB T cells	32·80 ± 0·94	15·78 ± 0·32	17·01 ± 0·76	0·0 ± 0·0	1
Stimulated AB T cells	20·00 ± 0·69	15·51 ± 0·40	4·49 ± 0·56	− 12·51 ± 0·80	5865·3 ± 0·57

^*Δ C_T (IFNγ C_T-18 s RNA C_T);

^**ΔΔ C_T (Δ C_T^stimulated–Δ C_T^unstimulated);

^†Normalized IFNγ amount relative to unstimulated T cells

IFNγ production in UCB T cells is not regulated by PKR

It has been reported earlier that when present in low amount, IFNγ mRNA negatively autoregulates its own translation by activating the IFNγ dependant protein kinase PKR [25]. Therefore, we hypothesized that this low IFNγ protein produced by UCB T cells could be due to PKR activation. To check this hypothesis, we stimulated UCB T cells in the presence of increasing concentrations of the PKR activation inhibitor, 2-amino-purine (2-AP), and measured frequencies of the IFNγ positive cells using FACS. FACS analysis indicated that 2-AP treatment did not enhance the frequency of IFNγ positive UCB T cells (Fig. 2).

Fig. 2.

Effect of 2-AP on IFNγ expression in UCB T cells. UCB T cells were stimulated for 3 h with PMA and ionomycin in the presence of varying concentrations of PKR activation inhibitor, 2-AP, which was added during the last 30 min of the incubation. Frequencies of IFNγ positive were measured using FACS. 10 000 gated events were acquired. The treatment of 2-AP did not show any increase in the frequency of IFNγ positive UCB T cells in comparison to non treated UCB T cells.

Promoter demethylation is not sufficient to induce IFNγ expression in UCB T lymphocytes

Recent findings have indicated that the hypermethylation of IFNγ promoter might contribute to the mechanisms preventing IFNγ gene expression in neonatal vs. adult blood T cells. We therefore examined whether treatment with 5-aza-2′-deoxycytidine (aza); a potent promoter demethylating agent [22,23,31] would relieve the transcriptional block, thereby resulting in optimal IFNγ mRNA accumulation in UCB T cell subsets. For this study, UCB CD4⁺ and CD8⁺ T cells were stimulated with PMA and ionomycin for 24 h in the presence of 5-aza-2′-deoxycytidine and IFNγ mRNA accumulation was quantified using real time- quantitative RT-PCR. It has been shown earlier that the treatment with 5-aza-2′-deoxycytidine cause demethylation of the IFNγ promoter without affecting T cell viability and their proliferation [22,23,31]. Our results show that neonate CD4⁺ and CD8⁺ T cells accumulated approximately 2 fold more IFNγ mRNA after 5-aza-2′-deoxycytidine treatment in comparison to nontreated controls (see Table 2). This increase in IFNγ mRNA also translated into a corresponding increase in IFNγ secretion (Fig. 3). Thus, demethylation of IFNγ gene promoter is not sufficient to establish mature-like levels of IFNγ expression. This observation indicates that the tight regulation of IFNγ expression in UCB T cells most likely involves several mechanisms.

Table 2.

Real time PCR analysis of the IFNγ mRNA accumulation UCB T cell after 5-aza-2′-deoxycytidine (Aza) treatment. Enriched UCB T cell subsets were stimulated with PMA and ionomycin in the presence of 5 µm aza for 24 h. The mean fold change in the IFNγ gene expression was calculated using equation ^ΔΔCT = (IFNγ C_T– 18sRNA C_T)_+Aza– (IFNγ C_T– 18sRNA C_T)_-Aza. The fold change was calculated by 2^–ΔΔCT. The aza treatment resulted in a two fold increase in IFNγ mRNA accumulation.

Samples	IFNγ CT	18 s RNA CT	Δ CT*	ΔΔ CT**	2–ΔΔ CT†
UCB CD4+T cells – Aza	26·35 ± 0·27	12·13 ± 0·11	14·22 ± 0·16	0·0 ± 0·0	1
UCB CD4+T cells + Aza	25·39 ± 0·25	12·23 ± 0·16	13·16 ± 0·09	− 1·06 ± 0·07	2·08 ± 0·18
UCB CD8+T cells – Aza	25·42 ± 0·49	12·58 ± 0·32	12·84 ± 0·17	0·0 ± 0·0	1
UCB CD8+T cells + Aza	24·48 ± 0·53	12·83 ± 0·38	11·65 ± 0·15	− 1·18 ± 0·02	2·26 ± 0·32

^*Δ C_T (IFNγ C_T-18 s RNA C_T);

^**ΔΔ C_T (Δ C_T^+Aza–Δ C_T^–Aza);

^†Normalized IFNγ amount relative to untreated T cells

Fig. 3.

Effect of 5-aza-2′-deoxycytidine on IFNγ protein production. Enriched UCB T cells were stimulated for 24 h in the presence of 5 µm 5-aza-2′-deoxycytidine (Aza). Following treatment supernatants from the cultures were collected and IFNγ levels were quantified using an commercial ELISA as described in the materials and methods. Each bar represent mean ± SD.

IFNγ mRNA to protein ratio is constant in UCB and AB T cells

We were able to precisely measure IFNγ mRNA and protein levels in the same samples, by the use of real-time RT-PCR and sensitive ELISA assays. This feature, enabled us to determine that the ratio between these two parameters. The IFNγ mRNA to protein ratio was constant in both UCB and AB T cell subsets (Table 3). Furthermore, we determined that this ratio was not altered by treatment with 5-aza-2′-deoxycytidine, which lead to small 2-fold increments in mRNA accumulation (see Table 2), as it was paralleled by a similar increase in IFNγ protein production (Fig. 3). Our results therefore imply that IFNγ expression in UCB T cells is not regulated at the translational level and that the main regulatory modes reside at the transcription level.

Table 3.

IFNγ mRNA to protein ratio in the UCB and AB T cell subsets. Enriched UCB and AB T cell subsets were stimulated for 24 h with PMA and ionomycin. IFNγ mRNA accumulation was measured using real-time PCR and calculated according to 2^–ΔΔCT method. IFNγ protein was measured by ELISA and the IFNγ mRNA to protein ratio was calculated. The ratio between these two parameters are fairly constant in the T cell subsets obtained from UCB and AB.

Samples	IFNγ mRNA	IFNγ protein (pg/ml)	mRNA/protein ratio	Mean ± SD (mRNA/protein ratio)
UCB CD4+T cells	″576	″883	0·652	0·72 ± 0·08
	″739	″982	0·752
	″3666	″4368	0·839
	″1370	″2068	0·662
AB CD4+T cells	31872	47945	0·664	0·68 ± 0·03
	99334	150347	0·660
	41189	55264	0·745
	33456	49264	0·679
UCB CD8+T cells	″8135	″7066	1·151	1·07 ± 0·08
	″2721	″2384	1·141
	″8481	″8159	1·039
	″4482	″4564	0·982
AB CD8+T cells	49324	40311	1·223	1·08 ± 0·13
	187682	94708	0·925
	66451	57811	1·149
	70728	68811	1.027

Discussion

The complex mechanism regulating IFNγ expression in neonatal T cells is still a puzzle, and it remains to be resolved whether this is regulated at the transcription or at translational level. In order to address this issue in more detail we have used real-time RT-PCR assays to accurately assess IFNγ mRNA levels, and FACS or ELISA analysis to examine IFNγ protein levels in the same samples. We used high doses of PMA and ionomycin to synergistically activate T cells in a receptor and APC-independent manner. PMA activates protein kinase C (PKC), but does not elevate Ca²⁺. Ionomycin elevates Ca²⁺, but Ca²⁺ alone does not activate PKC in the absence of PMA, addition of the Ca²⁺-mobilizing agent ionomycin along with PMA synergistically activates T cell proliferation [32,33]. Our data indicate that the induction of IFNγ mRNA accumulation in neonatal T cells stimulated with PMA and ionomycin is 30 fold less than similarly stimulated adult T cells. In a similar manner the level of IFNγ protein was found to be much lower in stimulated neonatal T cells than in adult T cells.

IFNγ mRNA has been shown to activate the RNA-dependent protein kinase PKR, a stress kinase that is also activated by double-stranded RNA. The cis-acting RNA elements within IFNγ transcripts function as sensors of intracellular PKR levels and regulate IFNγ mRNA splicing and translation [34,35]. Since low levels of IFNγ mRNA has been shown to activate PKR [25] and we could observe very low IFNγ mRNA accumulation by cord blood T cell therefore we examined the effect of PKR inhibitor drug 2-AP on stimulated UCB T cells. Our data indicate that treatment with 2-AP did not increase IFNγ protein expression in PMA and ionomycin stimulated UCB T cells, a finding that is in agreement with reports indicating low levels of PKR activation in T lymphocytes [36].

It is well established that the methylation of DNA is an epigenetic mechanism for the modulation of gene expression in mammalian cells [37]. DNA methylation changes chromatin structure and may help the recruitment of transcription factors to the target genes [38]. Many of the studies have reported that the differential CpG methylation of IFNγ promoter is responsible for the low IFNγ expression [22,23,31]. Here, we examined whether IFNγ promoter hypermethylation could be reversed by treatment with a demethylating agent, 5-aza-2′-deoxycytidine. Our results show that this treatment only lead to a 2 fold increase in IFNγ mRNA accumulation in PMA and ionomycin stimulated neonatal T cells. Of interest was that this 2 fold increase in the mRNA accumulation was paralleled by an equivalent increase in IFNγ protein production.

This finding lead us to assess the IFNγ mRNA: protein ratios in our various study groups. This analysis indicated that this ratio was fairly constant in all the groups examined, whether they be activated neonatal or adult T cells. This analysis strongly suggest that in UCB T cells IFNγ expression is not regulated on a post-transcriptional level but rather on transcriptional level, at least, partially involves promoter methylation.

The exact mechanism regulating efficient induction of IFNγ mRNA accumulation is still unclear but may involve several other mechanisms. The dynamic changes in the histone tail acetylation have been shown to play an important role in the effector functions of T cells [37,39]. Avni et al. [39] have suggested that TCR stimulation activates a histone tail modification mediated change in chromatin structure in the immature T cells, which allows the binding of TCR-induced transcription factors to the promoter regions of IFNγ gene. In the absence of polarizing cytokines, both early histone hyperacetylation and early cytokine gene expression are reduced to the low basal amounts as observed in UCB T cells. In UCB T cells certain cytokine genes, such as that for IFNγ, are positioned in such a manner that only limited gene transcription is possible upon TCR stimulation. The difference between neonatal and adult T cells is that inactive cytokine genes in neonatal T cells are located in euchromatin regions, whereas a large fraction of the silenced genes in differentiated T cells are repositioned to centromeric heterochromatin regions due to inhibitory modifications, previously established by the polarizing cytokines, that down-regulate the expression of inappropriate genes. It is likely therefore that in immature T cells, TCR stimulation initiates permissive chromatin modifications that facilitate early gene expression. These findings also provide a possible reason for the delayed response against stimulus that UCB T cells display before they are able to produce efficient effector functions upon TCR stimulation, as a period of time is required for this repositioning of key effector genes in maturation of naïve to memory T cells [40]. Our data provide some support for this hypothesis, in that longer periods of PMA and ionomycin stimulation (24 h and 48 h) enhanced the frequency of IFNγ positive cells in the UCB T cell population.

To summarize, we propose here that the regulation of IFNγ in UCB T lymphocytes occurs at the transcriptional level and not post-transcriptional or translational level, as we observed fairly constant mRNA to protein ratios in all of the T cell groups examined. The nature of this transcriptional block, however, still remains to be resolved or may involve several factors such as deficient regulatory proteins, RNA or transcription factors rendering IFNγ up-regulation slower and lower than that of adult T cells.