A Dirofilaria immitis Polyprotein Up-Regulates Nitric Oxide Production

Abstract

We investigated the effect of recombinant Dirofilaria immitis polyprotein (rDiAg) on nitric oxide (NO) production by peritoneal macrophages. rDiAg induced NO production by macrophages from wild-type and lipopolysaccharide-hyporesponsive C3H/HeJ, but not CD40^−/−, mice. These results suggest that CD40 is involved in rDiAg-driven NO production by murine macrophages.

Nitric oxide (NO) has pleiotropic effects, including vasodilatation, neurotransmission, and immunomodulation (3). In immune responses, NO is known to be produced mainly by activated macrophages and to modulate Th1/Th2 balance, as well as to induce immunosuppression (1, 3, 18). In bacterium- or protozoon-infected hosts, NO also serves as a toxic molecule against these pathogens (3). In contrast, macrophage NO production driven by living helminth parasites or their products is considered to be involved in immunosuppression through the prevention of worm-specific T-cell responses (2, 4, 14, 15). CD40 is expressed on B cells, macrophages, dendritic cells, and endothelial cells and is involved in class switching (in B cells), cell proliferation, and cytokine production in these cells (21). CD40 has been recently shown to elicit NO synthesis in macrophages and dendritic cells (13, 17, 20).

Nematode polyprotein allergens (NPAs) found in numerous worms are generated via a unique biosynthetic process (7). The DNA encoding the NPA precursor is characterized by tandemly repeated sequences, which are composed of 10 to 50 repeating units, depending on the species. NPAs are first synthesized as a large precursor polyprotein that possesses a cleavage site at the C terminus of each repeating unit. The precursor is subsequently proteolytically processed into <15-kDa single repeating units, thereby yielding multiple copies of similar or identical proteins. NPAs are predominantly located in the pseudocoelomic fluid of ascarids (including the porcine intestinal roundworm Ascaris suum) or on the surface of filariae (including the canine heartworm Dirofilaria immitis) and are secreted as a major component of the excretory-secretory (ES) product (7). More recently, we have found that the D. immitis polyprotein (DiAg; a Dirofilaria NPA) can bind to human CD40 molecules and induces antigen-nonspecific polyclonal immunoglobulin E (IgE) production by highly purified human and murine B cells in the presence of interleukin-4 (6, 19). To verify that DiAg signaling occurs via CD40, we examined the effect of DiAg on NO synthesis by CD40-bearing and CD40-defective macrophages.

We first examined the effect of recombinant DiAg (rDiAg) on NO production by macrophages from lipopolysaccharide (LPS)-hyporesponsive C3H/HeJ mice (Clea Japan, Tokyo, Japan). rDiAg was prepared with Escherichia coli as described previously (19). The recombinant protein was then passed through a column of polymyxin B-immobilized beads and found to have negligible endotoxin levels (<2 pg of LPS/mg of protein) as measured by Endotoxin Test-D (Seikagaku, Tokyo, Japan) (19). Elicited peritoneal exudate cells were collected by washing the mouse peritoneal cavity with ice-cold RPMI 1640 medium 5 days after intraperitoneal injection with 1 ml of sterile 3% thioglycolate (Difco Laboratories, Detroit, Mich.). Adherent cell fractions were detached and resuspended in the culture medium as described previously (19). By nonspecific esterase staining, >95% of the cells exhibited macrophage characteristics. Macrophages (10⁶/ml) were stimulated with various concentrations of rDiAg for the indicated periods. Nitrite (a stable oxidation product of NO) accumulation in culture supernatants was determined with Griess reagent (14). Differences in NO levels between groups were evaluated with Student's t test. Nitrite accumulation was first detected at 24 h and reached a plateau at 48 to 72 h after rDiAg addition (Fig. 1A). rDiAg induced NO production in a dose-dependent manner, whereas ovalbumin (OVA; Sigma Chemical, St. Louis, Mo.), used as a control protein, failed to elicit NO production (Fig. 1B). Recently, CpG oligonucleotides from bacteria have been shown to induce NO synthesis by macrophages (9). To exclude the possibility that CpG contamination was responsible for the rDiAg activity, we next prepared a recombinant control protein (rCont; N-terminal half, amino acids [aa] 1 to 67, of ABA-1 [Ascaris NPA]) by the same procedure as rDiAg (19). rCont did not induce NO production at any of the concentrations used (Fig. 1B). These results suggest that DiAg acts as an inducer of NO synthesis and indicate that its activity is not due to contamination with bacterial components such as LPS and CpG oligonucleotides. NO production by macrophages is mediated by inducible NO synthase (iNOS) (3). To determine whether rDiAg-induced NO production is dependent on the activation of iNOS, C3H/HeJ macrophages were cultured with rDiAg in the presence of aminoguanidine (AMG) or dexamethasone (both from Sigma), iNOS-specific inhibitors, for 48 h. rDiAg-induced nitrite accumulation was completely inhibited by either inhibitor (Fig. 1C). We next examined whether rDiAg directly induces the expression of iNOS mRNA in macrophages. Reverse transcription (RT)-PCR was performed as described previously (19). The primer sequences were as follows: iNOS, 5′-CGTTGGATTTGGAGCAGAAGTG-3′ (sense) and 5′-CATGCAAAATCTCTCCACTGCC-3′ (antisense); glyceraldehyde 3-phosphate dehydrogenase (GAPDH), 5′-ACCACAAGTCCATGCCATCAC-3′ (sense) and 5′-TCCACCACCCTGTTGCTGTA-3′ (antisense). Only rDiAg induced the expression of iNOS mRNA (Fig. 1D). These results indicate that DiAg up-regulates the production of NO predominantly at the mRNA level.

FIG. 1.

Effect of rDiAg on NO production by macrophages from LPS-hyporesponsive C3H/HeJ mice. Peritoneal macrophages from C3H/HeJ mice were cultured with rDiAg (10 μg/ml) for the indicated periods (A) or with various concentrations of OVA, rCont, or rDiAg for 48 h (B). Macrophages were cultured with an iNOS inhibitor (1 mM AMG or 1 μM dexamethasone [DEX]) in the presence or absence of rDiAg (10 μg/ml) for 48 h (C). The data represent the mean values and standard deviations (error bars) from five independent experiments. ∗∗ and ∗, P < 0.01 and P < 0.05, respectively, compared with medium. Macrophages were cultured with OVA, rCont, or rDiAg at 10 μg/ml, and iNOS mRNA expression was detected by RT-PCR (D). The housekeeping gene for GAPDH was amplified as an internal control. The data represent one typical experiment out of five.

To define whether only DiAg in ES products from D. immitis has the ability to induce NO synthesis, native DiAg, whole ES products (including native DiAg), and DiAg-depleted ES products were prepared as described previously (19). Whole ES products elicited reduced, but significant, NO levels in C3H/HeJ macrophages, and native DiAg triggered the same amounts as rDiAg (Fig. 2A). AMG completely inhibited NO production. In contrast, DiAg-depleted ES products did not induce NO production. These responses also depended on the expression of iNOS mRNA (Fig. 2B). These results indicate that DiAg is an inducer of NO synthesis.

FIG. 2.

Effect of native DiAg on NO production by macrophages from LPS-hyporesponsive C3H/HeJ mice. Peritoneal macrophages were cultured with native DiAg, whole ES products, or DiAg-depleted ES products (DiAg⁻ ES) at 10 μg/ml in the presence or absence of AMG (1 mM) for 48 h (A). The data represent the mean values and standard deviations (error bars) from five independent experiments. ∗∗, P < 0.01 compared with AMG treatment. Macrophages were cultured with the same stimuli for 24 h, and iNOS mRNA expression was detected by RT-PCR (B). The housekeeping gene for GAPDH was amplified as an internal control. The data represent one typical experiment out of five.

NO production by macrophages and dendritic cells is triggered by CD40 signaling (13, 17, 20). rDiAg acts as an agonist to human CD40 (6). To assess whether CD40 is involved in rDiAg-driven NO synthesis, thioglycolate-elicited peritoneal macrophages from CD40^−/− (BALB/c background) and CD40 ligand-defective (CD40L^−/−; C57BL/6 background) mice (Jackson Laboratory, Bar Harbor, Maine) and their wild-type counterparts (Clea Japan) were stimulated with rDiAg, a hamster anti-mouse CD40 monoclonal antibody (HM40-3; PharMingen, San Diego, Calif.), or LPS (from E. coli O55:B5; Sigma) for 48 h. As shown in Table 1, rDiAg did not induce NO production by CD40-defective macrophages, whereas macrophages from wild-type and CD40L^−/− mice produced NO normally. Similar patterns were observed in cultures stimulated with the anti-CD40 monoclonal antibody (Table 1) and in cultures stimulated with native DiAg (data not shown). LPS significantly induced NO production by macrophages from all of the strains used. These results suggest that CD40 plays a central role in DiAg-driven NO production by murine macrophages.

TABLE 1.

Effect of rDiAg on macrophages from wild-type, CD40^−/−, and CD40L^−/− mice^a

Strain	Mean nitrite accumulation (μM) ± SD in:
	Medium	rDiAg	Anti-CD40 MAb	LPS
BALB/c	1.1 ± 0.15	9.9 ± 0.91b	10.7 ± 0.12b	20.7 ± 2.10b
CD40−/−	1.2 ± 0.18	1.2 ± 0.28	1.2 ± 0.15	24.0 ± 2.31b
C57BL/6	1.3 ± 0.16	12.8 ± 0.82b	13.9 ± 0.28b	23.0 ± 2.06b
CD40L−/−	1.2 ± 0.23	15.9 ± 1.88b	18.1 ± 0.37b	21.7 ± 2.44b

^aMacrophages from CD40^−/− (BALB/c background) and CD40L^−/− (C57BL/6 background) mice were cultured with rDiAg (10 μg/ml), an anti-CD40 MAb (2 μg/ml), or LPS (5 μg/ml) for 48 h. The data shown are from five independent experiments.

^bP < 0.01 compared with medium alone.

The role of NO in host-parasite interactions has been studied from two different perspectives, host and parasite protection. However, the possibility of the former appears to be excluded by recent studies that the killing of Brugia malayi third-stage larvae and Trichinella spiralis has been shown to be induced in iNOS^−/− mice (5, 10). The latter is supported by reports that spleen cells from B. pahangi microfilaria- or Echinococcus multilocularis-infected mice induce NO synthesis concomitant with antigen-specific proliferative suppression by restimulation with each antigen (4, 14). In addition, lacto-N-fucopentaose III, an oligosaccharide from Schistosoma mansoni, induces NO-dependent CD4⁺ T-cell proliferative suppression in vitro (2). Therefore, it is likely that NO production driven by living worms or their products is associated with parasite survival rather than host protection. We have recently demonstrated that DiAg preferentially induces polyclonal IgE synthesis by B cells (6, 19). These findings suggest that DiAg-driven NO production is involved in the down-regulation of parasite-specific responses.

Little is known about the immunological properties of NPAs, except that NPAs are recognized as antigens in mice immunized repetitively with the polyprotein plus adjuvant (7). We demonstrated here that DiAg is a potent inducer of NO synthesis by macrophages and that this response is not induced in CD40^−/− mice, suggesting that, like human CD40, murine CD40 also mediates DiAg signaling. This raises the hypothesis that all NPAs share CD40 signaling. The amino acid sequence identity between ABA-1 (133 aa), which was first characterized as an NPA, and DiAg (129 aa) can be as much as 42.9% (16); however, ABA-1, unlike DiAg, cannot trigger nonspecific IgE synthesis (11) and polyclonal B-cell expansion (12). In addition, full-length recombinant ABA-1 (133 aa) failed to induce NO synthesis by macrophages (data not shown). ABA-1 only prolongs ongoing IgE production (11). Therefore, it is unlikely that CD40-binding activity is a feature of all NPAs. Recently, somatic extracts from adult D. immitis worms caused relaxation of the canine aorta via NO production by endothelial cells, resulting in increased blood flow and decreased vascular resistance (8). Since adult D. immitis worms reside mainly in the pulmonary arteries of infected dogs, the vasodilatation mediated by NO may contribute to the provision of nutrients and the maintenance of habitat spaces for the invading parasite. Thus, D. immitis-derived factors, including DiAg, appear to facilitate parasitism of the worm via immunosuppression and arterial relaxation by NO. Together, these findings suggest that the role of NPAs in hosts differs among species and that these divergences may be based on the individual means of worm parasitism.

In conclusion, we have demonstrated that rDiAg-driven NO synthesis by macrophages is dependent on CD40, suggesting that DiAg mimics a host molecule.