Tuberculosis-induced variant IL-4 mRNA encodes a cytokine functioning as growth factor for (E)-4-hydroxy-3-methyl-but-2-enyl pyrophosphate-specific Vgamma2Vdelta2 T cells

Abstract

The possibility that mycobacterial infections induce variant cytokine mRNA encoding a functionally distinct protein for immune regulation has not been addressed. In this study, we reported that Mycobacterium tuberculosis and bacillus Calmette-Guérin infections of macaques induced expression of variant IL-4 (VIL-4) mRNA encoding a protein comprised of N-terminal 97 aa identical with IL-4, and unique C-terminal 96 aa including a signaling-related proline-rich motif. While VIL-4 could be stably produced as intact protein, the purified VIL-4 induced apparent expansion of phosphoantigen (E)-4-hydroxy-3-methyl-but-2-enyl pyrophosphate (HMBPP)-specific Vgamma2Vdelta2 T cells in dose- and time-dependent manners. The unique C-terminal 96 aa bearing the proline-rich motif (PPPCPP) of VIL-4 appeared to confer the ability to expand Vgamma2Vdelta2 T cells, since simultaneously produced IL-4 had only a subtle effect on these gammadelta T cells. Moreover, VIL-4 seemed to use IL-4R alpha for signaling and activation, as the VIL-4-induced expansion of Vgamma2Vdelta2 T cells was blocked by anti-IL-4R alpha mAb but not anti-IL-4 mAb. Surprisingly, VIL-4-expanded Vgamma2Vdelta2 T cells after HMBPP stimulation appeared to be heterologous effector cells capable of producing IL-4, IFN-gamma, and TNF-alpha. Thus, mycobacterial infections of macaques induced variant mRNA encoding VIL-4 that functions as growth factor promoting expansion of HMBPP-specific Vgamma2Vdelta2 T effector cells.

FIGURE 1

BCG and M. tuberculosis infection induced expression of variant IL-4 mRNA (VIL-4), which encoded a cytokine protein comprised of 97 aa identical with IL-4 at N-terminal and unique 96 aa including a signaling-related proline-rich motif at C-terminal. a, The agarose gel showed PCR products of VIL-4 (marked by an arrow) and IL-4 cDNA derived from cells isolated from spleens of two BCG-infected monkeys. Numbers on top indicate PCR samples ID. Numbers 1–3 were PCR products of cDNA derived from the monkey 278; number 3 was the sample from lymphocytes of the spleen with TB-like granulomatous lesions 2 years after BCG/SIVmac coinfection, whereas numbers 1–2 were the samples from PBMC and BAL obtained on 4 wk, respectively, after BCG infection. Numbers 4–6 were PCR products from cDNA derived from the monkey 276, with the same sample orders as Numbers 1–3. Number 7 was negative control for PCR. No VIL-4 were amplified by PCR in samples from the other four BCG-infected monkeys. b, The agarose gels showed first-round (left) and second-round (right) PCR products of VIL-4 (marked by an arrow) and IL-4 cDNA from spleen lymphocytes of eight monkeys infected 2–5 mo with M. tuberculosis. Numbers 1–8 were samples from monkeys FB32, 33, 35, 36, 37, 38, 43, and 45, respectively. The first-round PCR primers yielded 380-bp VIL-4 (marked by an arrow) and 229-bp IL-4 fragments after amplification of samples Number 6 and 8; the second-round PCR primers amplified 337-bp VIL-4 (marked by an arrow) and 186-bp IL-4 fragments. NC, negative control for PCR. Sequencing of the 380-bp or 337-bp fragment from difference monkeys showed the same VIL-4 sequence. The lower band next to the VIL-4 fragment was an untranslatable IL-4 variant. The PCR-based cloning approach using a new 5′ primer for IL-4 leader sequence and a poly-A 3′ primer confirmed the existence of the variant sequence of IL-4, as shown in c. c, VIL-4 DNA sequence had two insertions: the 151 bp insertion was introduced right between the third and fourth exons of IL-4 gene, and positioned from 292 through 443; the second insertion containing 38 bp joined from the downstream of the fourth exon through the 3′ terminal. Both inserted parts shared a homology of 94% with Homo sapiens chromosome 5 genomic sequence of IL-4 intron. d, VIL-4 shared identical 97 residues with IL-4 at N-terminal, but displayed 96 different amino acid sequences thereafter in the downstream due to the open reading-frame shift caused by the 151 bp insertion. Note that there is a proline-rich motif (PPPCPP) as underlined at the position from 137 through 142 at the C-terminal of VIL-4.

FIGURE 2

Both VIL-4 and IL-4 were expressed and purified. VIL-4 or IL-4 protein was first purified by Ni-NTA agarose, and then further purified by an anion exchange column of Biologic Duoflow Chromatography system. a, The unbound Peak 1 fraction collected from the column was either VIL-4 or IL-4 protein. Peaks 2 and 3 fractions represented unknown proteins that were bound to the column and later eluted by elution buffer after the protein in Peak 1 passed through. b, SDS-PAGE analysis showed that both IL-4 (Lane 1) and VIL-4 (Lane 2) were apparently pure after the two-round purification, with each protein exhibiting a single dominant band on the gel. The molecular weights of expressed fusion protein IL-4 and VIL-4 were ∼20 and ∼25 kDa on SDS-PAGE gel. c, Purified IL-4 and VIL-4 were quantified by a BCA Protein Quantitation Assay kit. The result showed that purified VIL-4 was 161 μg/ml; IL-4 was186 μg/ml (control protein was 219 μg/ml). d, 2 ng IL-4 and 3 ng VIL-4 were loaded in the dot blot membrane, and both IL-4 and VIL-4 were recognized by the anti-human IL-4 mAb. e, 3 ng IL-4 and 4.5 ng VIL-4 were run on the SDS-PAGE gel in the Western blot; only IL-4, not VIL-4, was recognized by the anti-human IL-4 mAb. f, Different concentration of IL-4 and VIL-4 were run in the SDS-PAGE gel in a repeated Western blot analysis, and results were same as described above. Loading 1 was negative control, loading 2 was 1 ng IL-4 sample, loading 3 was 1.5 ng VIL-4, loading 4 was 5 ng IL-4, and loading 5 was 7.5 ng VIL-4.

FIGURE 3

VIL-4 induced proliferation and expansion of HMBPP-stimulated Vγ2Vδ2 T cells at a dose- and time-dependent manner. a, Flow cytometry histograms from one representative monkey (7289) showed that VIL-4 induced apparent expansion of HMBPP-stimulated Vγ2Vδ2 T cells. PBMC from the monkeys were stimulated with HMBPP for 3 days, cultured with VIL-4 or IL-4 for 7 days, and then assessed for changes in numbers of Vγ2Vδ2 T cells. CD3⁺ T cells were gated for analysis of HMBPP-stimulated Vγ2Vδ2 T cells. IL-4 induced only minor expansion of HMBPP-stimulated Vγ2Vδ2 T cells. The stimulation with HMBPP only, IL-4 only, or VIL-4 only did not induce detectable expansion of Vγ2Vδ2 T cells. b, VIL-4-induced expansion of HMBPP-stimulated Vγ2Vδ2 T cells was dose-dependent, with an optimal concentration of VIL-4 being 15 ng/ml (10 ng/ml for IL-4). Data were percentage numbers of cells in CD3⁺ T cells. Shown were representative data from one (7289) of two monkeys in two independent experiments. c, VIL-4-induced expansion of HMBPP-stimulated Vγ2Vδ2 T cells was time-dependent. After a 3-day stimulation with HMBPP, Vγ2Vδ2 T cells were cultured with VIL-4 for 3 or 7 days. Expansion of HMBPP-stimulated Vγ2Vδ2 T cells on day 7 was higher than that on day 3. Shown were representative data from one (7289) of three monkeys in two independent experiments. d, VIL-4-induced expansion of HMBPP-stimulated Vγ2Vδ2 T cells in PBMC form other monkeys. Shown were the data from PBMC stimulated for 3 days with HMBPP and cultured for 7 days with 10 ng IL-4/ml or 15 ng VIL-4/ml. No expansion for the cultures using HMBPP alone, IL-4 alone, or VIL-4 alone (data not shown). e, VIL-4 induced much greater expansion of HMBPP-stimulated Vγ2Vδ2 T cells than IL-4 (p < 0.05). Data were means and SEM values derived from PBMC cultures of six monkeys.

FIGURE 4

Anti-IL-4 receptor α-chain mAb blocked VIL-4-induced expansion of HMBPP-stimulated Vγ2Vδ2 T cells. a, Anti-IL-4 receptor α-chain mAb blocked VIL-4-induced expansion of HMBPP-stimulated Vγ2Vδ2 T cells at a dose-dependent manner. b, Anti-IL-4 mAb was not able to block VIL-4-induced expansion of HMBPP-stimulated Vγ2Vδ2 T cells. IL-4-induced minor expansion appeared to be reduced by anti-IL-4 mAb. Data were derived from one of three monkeys in two independent experiments. Note that IL-4-induced minor expansion of Vγ2Vδ2 T cells were blocked by both anti-IL-4 mAb and anti-IL-4 receptor α-chain mAb.

FIGURE 5

HMBPP-stimulated VIL-4-expanded Vγ2Vδ2 T cells were heterologous effector cells capable of producing IFN-γ, TNF-α, and IL-4. a–c, showed IFN-γ, TNF-α, and IL-4, respectively, in culture supernatant of HMBPP-stimulated VIL-4-expanded PBMC. These cytokines were measured by double sandwich ELISA. Data were means and SEM values derived from PBMC of four monkeys. Difference in each of cytokines between VIL-4- and IL-4-treated cultures was statistically significant (*, p < 0.05; **, p < 0.01). Negative controls, HMBPP treated for 3 days and cultured without VIL-4 or IL-4 for additional 7 days. d, Representative intracellular cytokine staining data indicating that IFN-γ-producing Vδ2⁺ cells in HMBPP-stimulated VIL-4-expanded PBMC. Shown were the representative data from one of three experiments using PMBC from one of three monkeys.