Identification of the most active interleukin-32 isoform

Abstract

Cytokines are crucial in host defence against pathogens such as bacteria, viruses, fungi and parasites. A newly described cytokine, interleukin-32 (IL-32), induces various proinflammatory cytokines (tumour necrosis factor-alpha, IL-1beta, IL-6) and chemokines in both human and mouse cells through the nuclear factor-kappaB and p38 mitogen-activated protein kinase inflammatory signal pathway. The IL-32 primarily acts on monocytic cells rather than T cells. In an attempt to isolate the IL-32 soluble receptor, we used an IL-32 ligand-affinity column to purify neutrophil proteinase 3, which is a serine proteinase involved in many inflammatory diseases. IL-32 has biological activity associated with Mycobacterium tuberculosis and chronic proinflammatory diseases such as rheumatoid arthritis. IL-32 is transcribed as six alternative splice variants and the biological activity of each individual isoform remains unknown. Here, we cloned the complementary DNA of the four IL-32 isoforms (alpha, beta, gamma and delta) that are the most representative IL-32 transcripts. To produce recombinant protein with a high yield, the amino acids of two cysteine residues were mutated to serine residues, because serine residues are not conserved among different species. The multi-step purified recombinant IL-32 isoform proteins were assessed for their biological activities with different cytokine assays. The gamma isoform of IL-32 was the most active, although all isoforms were biologically active. The present study will provide a specific target to neutralize endogenous IL-32, which may contribute to basic and clinical immunology.

Figure 1

The regulation of four different human interleukin-32 (IL-32) messenger RNA splice variants in human epithelial WISH cells and the diagram of four IL-32 isoforms with the indicated position for cysteine mutation. (a) WISH cells were stimulated with human interferon-γ (IFN-γ; 5 ng/ml) and the cells were harvested over time, as indicated at the top of the figure. Total RNA was isolated from the cells and semi-quantitative reverse transcription–polymerase chain reaction was performed. The bottom line, the GAPDH housekeeping gene, depicts equal amounts of total RNA template in each sample. (b) Four different IL-32 isoforms were shown with the amino acid (aa) positions, cysteine residues changed to serine residues. The hatched bar displays the additional amino acid sequence as a result of the absence of exon splicing in IL-32 isoforms β, γ, and δ. The N-terminal Cys/Ser-mutated residue is consistent as the second amino acid residue, however the C-terminal Cys/Ser-mutated residue is varied, and is listed by amino acid position because of distinct splicing in each isoform.

Figure 2

The expression of four recombinant interleukin-32 (rIL-32) isoform proteins and verification of their specificity by neutrophil proteinase 3 (PR3). (a) Four rIL-32 isoform proteins, indicated by letter at the top of the lane, were purified by a multi-step purification procedure and then each rIL-32 protein (50 ng/lane) was applied for sodium dodecyl sulphate–polyacrylamide gel electrophoresis/silver stain to evaluate their purity. (b) The verification of rIL-32 isoform protein by PR3, which is a specific binding protein for IL-32 and has the ability to digest IL-32 in a very specific manner.

Figure 3

The Western blot of four recombinant interleukin-32 (rIL-32) isoforms in the presence or absence of preincubation with proteinase 3 (PR3). (a) Four rIL-32 proteins (5 ng in each lane) were loaded and probed with rabbit anti-IL-32α polyclonal antibody. (b) The rIL-32 protein (5 ng in each lane) was preincubated with 10 ng of human PR3 for 30 min at 37°, loaded for Western blotting, and then probed with the same antibody as in (a).

Figure 4

The biological activity of four recombinant interleukin-32 (rIL-32) isoforms in human peripheral blood mononuclear cells (PBMC). Human PBMC were stimulated with four different rIL-32 isoforms (40 ng/ml) in the presence of polymyxin B (100 U/ml) for 18 hr, after which the cell culture supernatant was harvested for cytokine measurement. (a) IL-6 and (b) tumour necrosis factor-α (TNF-α), were measured by enhanced chemiluminescence. As shown in the second lane, lipopolysaccharide (LPS) induced a large amount of cytokine in both IL-6 and TNF-α, whereas polymyxin B (100 U/ml) in the third lane, abolished LPS activity.

Figure 5

The examination of recombinant interleukin-32 (rIL-32) isoform protein with the peritoneal macrophages of lipopolysaccharide (LPS)-resistant C3H/HeNj mice. The peritoneal macrophages of normal C57-Black/6 and LPS-resistant C3H/HeNj mice were stimulated four rIL-32 isoforms and then mouse (a) macrophage inflammatory protein 2 (MIP-2) and (b) tumour necrosis factor-α (TNF-α) were measured in the cell culture supernatant after 18 hr of stimulation. (a) and (b), Lane 6, staphylococcus stimulates a normal response in both normal C57-Black/6 and LPS-resistant C3H/HeNj peritoneal macrophages, whereas LPS, lane 7, shows a lack of response in LPS-resistant C3H/HeNj peritoneal macrophages.

Figure 6

Τhe biological activity of recombinant interleukin-32γ (rIL-32γ) in vivo with Toll-like receptor-2-deficient (TLR-2^−/−). The most active rIL-32γ isoform was administrated to five C57-Black/6 and TLR-2^−/− and then the mice were killed and cytokines were measured from sera. In both (a) C57-Black/6 and (b) TLR-2^−/− mice, tumour necrosis factor-α (TNF-α) and MIP-2 were sufficiently produced although TLR-2^−/− showed fewer cytokines.