Transforming Growth Factor β-Induced Failure of Resistance to Infection with Blood-Stage Plasmodium chabaudi in Mice

Abstract

The role of transforming growth factor β (TGF-β) in infection with Plasmodium chabaudi was investigated with resistant and susceptible mouse models. C57BL/10 mice produced gamma interferon (IFN-γ) and nitric oxide (NO) shortly after infection and cleared the parasite spontaneously. In contrast, BALB/c mice showed a transient enhancement of TGF-β production, followed by a relative lack of IFN-γ and NO production, and succumbed to the infection. However, there was no correlation between levels of serum TGF-β and splenic TGF-β mRNA in both mouse strains before and after infection. Administration of recombinant TGF-β (rTGF-β) rendered resistant mice susceptible because of suppression of subsequent production of IFN-γ and NO. Administration of anti-TGF-β antibody to the infected BALB/c mice resulted in remarkable increases in serum IFN-γ and NO, and the mice resisted the infection. Splenic CD4⁺ T and CD11b⁺ cells of C57BL/10 mice were significantly activated after infection, but this was completely abrogated by administration of rTGF-β. These results suggested that, in the P. chabaudi-susceptible but not resistant mice, production of TGF-β was promoted, and subsequent failure of IFN-γ- and NO-dependent resistance to the parasite was induced. This study is the first to indicate that TGF-β production was the key event in failure of resistance to mouse malaria.

Malaria remains a major cause of morbidity and mortality worldwide. Patients infected with malaria parasites show extremely diverse clinical manifestations, and the ability of a host to resist the infection depends on both immunological and inherent characteristics of the host (24). Elucidation of the mechanisms involved in protective host responses to the parasite is essential for the development of an effective vaccine or therapeutic method. Infection of inbred strains of mice with blood-stage murine malaria parasites is a recognized model of human malaria (6). Many different kinds of cytokines are produced after infection with malaria parasites of humans and mouse models, and they are believed to play major roles in determining the fate of the infected hosts (14, 21, 35, 36). Although, CD4⁺ T cells are solely responsible for the gamma interferon (IFN-γ)-dependent resistance of the mouse to acute infection with Plasmodium chabaudi subsp. chabaudi AS, mouse strains with competent functions of CD4⁺ T cells, such as BALB/c mice, often fail to resist the parasite (36). Thus, the presence of additional mechanisms which downregulate proper activation of CD4⁺ T cells of the infected mice is suggested.

Transforming growth factor-β (TGF-β) is secreted ubiquitously by many different cell types, including activated T cells and activated macrophages (25). The bioactivity of TGF-β is often bidirectional, depending on target cells or coexisting mediators (29). In vitro evidence indicated that TGF-β upregulated arginase activity of murine macrophages and limited macrophage-dependent cytostasis and cytolysis (2). Human monocyte functions such as H₂O₂ production and adherence were suppressed by TGF-β, but antimycobacterial activity and O₂⁻ release were unaffected (38). TGF-β also has important roles in generation of effector T cells (3) and Th2 cell development (1).

Levels of TGF-β in serum were found to be decreased in patients infected with Plasmodium falciparum, but returned to the normal range after initiation of treatment (39). In vitro, Vγ9⁺ T cells from malaria-nonexposed donors expressed TGF-β mRNA when stimulated with P. falciparum schizont antigens (13). Furthermore, TGF-β and many other cytokines were secreted into the culture supernatants when human platelets, monocytes, and T lymphocytes were incubated with P. falciparum-parasitized erythrocytes (PRBCs) (37). However, no role of TGF-β in modulation of human malaria was indicated by these studies. A recent publication showed that TGF-β levels during murine malaria infection were inversely correlated with severity of disease (27). However, the downregulatory role of TGF-β is indicated in resistance to infection with a variety of microorganisms, including protozoal parasites (16, 18, 33, 40, 42). Since TGF-β is a multifunctional cytokine, the participation of this molecule in mechanisms of pathogenesis of malaria or immune protection merits further investigation.

In the present study, by using genetically resistant and susceptible strains of mice, we show the role of TGF-β in resistance to infection with P. chabaudi subsp. chabaudi AS. The results suggested that production of TGF-β above the physiological level was involved in failure of IFN-γ- and nitric oxide (NO)-dependent resistance of the mouse to acute infection with blood-stage malaria.

MATERIALS AND METHODS

Parasite and infection of mice.

P. chabaudi subsp. chabaudi AS was maintained by syringe passage every week in mice (20). PRBCs were collected by heart puncture, washed and suspended in phosphate-buffered saline (PBS). Female C57BL/10 Sn Slc (B10) and BALB/c Cr Slc (BALB/c) mice 5 to 7 weeks old were obtained from Japan SLC Co., Hamamatsu, Japan. Mice were infected intraperitoneally (i.p.) with 5 × 10⁴ PRBCs and monitored by counting the percentage of parasitemia on Diff-Quik-stained tail blood smears. In all of the experiments, mice were kept under specific-pathogen-free conditions and handled according to the guidelines for animal experimentation of the National Institute of Infectious Diseases.

rTGF-β and its administration.

Recombinant TGF-β (rTGF-β), prepared in transformed Chinese hamster ovary cells (9), was obtained from Genentech, Inc. (South San Francisco, Calif.). The bioactivity of the rTGF-β preparation was tested by the growth inhibition assay with CCL-64 mink lung epithelial cells (8). The purified preparation of rTGF-β showed a specific activity of 2.3 × 10⁷ U/mg. rTGF-β was diluted in PBS and injected i.p. (10 μg) 1 h before infection and every 2 days until day 8 postinfection.

Anti-TGF-β MAb and its administration.

A Wistar rat was immunized with a total of 400 μg of rTGF-β, and spleen cells were fused with SP2/0-Ag 14 myelomas by using polyethylene glycol 1000 (Sigma Chemical Co., Tokyo, Japan) and then cultured under standard conditions. Antibody activity in hybridoma supernatants was detected by an enzyme linked-immunosorbent assay (ELISA). A hybridoma which secreted neutralizing antibody of the IgG1 isotype was obtained (clone 15-1). Clone 15-1 MAb was partially purified from the culture supernatant by using a Procep-G column (Bioprocessing, Inc., Princeton, N.J.). To obtain 50% inhibition of bioactivity of 1 U of rTGF-β, 2.5 ng of the 15-1 MAb was required in the blocking assay of the TGF-β-dependent growth inhibition of CCL-64 cells. The MAb (1.0 mg) was injected into the mice i.p. −2, −1, 0, 2, 4, and 6 days postinfection. An isotype-matched rat MAb with unrelated specificity was used as a control.

RT-PCR of cytokine mRNAs.

Mouse spleens were obtained at various times after infection and immediately frozen in liquid nitrogen. Total RNA was extracted from the spleens by the guanidium thiocyanate-phenol-chloroform method (4) by using RNAzol B (Cinna/Biotecx, Houston, Tex.), followed by alcohol precipitation. RNA preparations were resuspended in diethylpyrocarbonate-treated distilled water (5 μg/20 μl). mRNA amplification was performed with reverse transcriptase PCR (RT-PCR) (Takara RNA-PCR; Takara Biomedicals, Ohtsu, Shiga, Japan) according to the manufacturer’s instructions.

The following oligonucleotides were used: TGF-β, 5′-TGGACCGCAACAACGCCATCTATGANTIGENAAAACC-3′ and 5′-TGGANTIGENCTGAANTIGENCAATANTIGENTTGGTATCCANTIGENGGCT-3′ (Clontech Laboratories, Inc., Palo Alto, Calif.); IFN-γ, 5′-TGGANTIGENGAACTGGCAAAANTIGENGATGGT-3′ and 5′-TTGGGACAATCTCTTCCCCAC-3′ (26); inducible NO synthase (iNOS), 5′-CATGGCTTGCCCCTGGAANTIGENTTTCTCTTCAAANTIGEN-3′ and 5′-GCANTIGENCATCCCCTCTGATGGTGCCATCG-3′ (12); and glyceraldehyde-3′-phosphate dehydrogenase (G3PDH), 5′-CTGGTGCTGANTIGENTATGTCGTG-3′ and 5′-CANTIGENTCTTCTGANTIGENTGGCANTIGENTG-3′ (17). The PCR was performed in the presence of 0.2 μM each primer, 1.25 U of Taq polymerase (Takara), and PCR buffer supplied in a total volume of 50 μl. Thirty cycles of amplification were run under conditions of denaturation at 94°C for 0.5 min, annealing at 60°C for 1.5 min, and elongation at 68°C for 1 min. The PCR products were electrophoresed on a 1.5% agarose gel and stained by ethidium bromide.

Measurement of cytokines.

Blood was obtained from the orbital venous plexus or by cardiac puncture at various times from mice infected with the parasite. Blood was allowed to clot for 30 min at room temperature and centrifuged at 1,200 × g for 10 min, and the sera obtained were stored at −20°C.

TGF-β and IFN-γ levels were determined with ELISA kits with antihuman TGF-β MAb (Morinantigena Bioscience Institute, Yokohama, Kanagawa, Japan) or antimouse IFN-γ MAb (Genzyme, Cambridge, Mass.) according to the manufacturers’ instructions Sera were acid activated for determination of TGF-β.

Production of NO was assessed by determination of NO₂⁻ and NO₃⁻ in the mouse sera (30). Briefly, infected mouse serum was mixed with distilled water (for measurement of NO₂⁻) or NADPH and nitrate reductase (both from Sigma, for measurement of NO₃⁻) and kept at room temperature for 20 to 30 min, followed by addition of the Griess reagent (15). Precipitates were removed by addition of 10% trichloracetic acid and centrifugation and the A₅₄₀ of the supernatant was read. Standards of nitrite and nitrate were prepared in pooled normal mouse serum.

Flow cytometry (FACS) analysis.

Mouse spleen cells were washed in RPMI 1640 and suspended in PBS containing 5% bovine serum albumin and 0.01% NaN₃. Fluorescein isothiocyanate (FITC)-labeled MAb against mouse CD4 (clone RM4-5) and I-A (clone AF6-120.1), phycoerythrin-labeled MAb to mouse Ly6A/E (clone D7) and CD11b (clone M1/70) were obtained from Pharmingen (San Diego, Calif.). The cells (2 × 10⁵/0.1 ml) were double stained with 1 μg of MAb clones RM-4-5 and D-7 or AF6-120.1 and M1/70 on ice for 20 min, washed, and analyzed on a FACScan fluorescence-activated cell sorter (FACS) by using Lysis II software (Becton Dickinson, San Jose, Calif.). CD4⁺ T cells or CD11b⁺ cells were gated, and the amounts of Ly6A/E and I-A antigens were compared in the histogram, respectively.

Mean fluorescein intensity (MFI) (linear conversion of log₁₀ fluorescence) was determined after correction for nonspecific fluorescence of controls by using Lysis II software.

Reproducibility and statistical analysis.

Similar experiments were repeated more than three times, and a representative result is shown. Data are expressed as means ± standard deviations. Differences were analyzed by one-way analysis of variance (ANOVA) or Student’s t test. P values of <0.05, <0.01, or <0.001 were considered to be significant.

RESULTS

Time course of infection and production of IFN-γ and NO.

B10 and BALB/c mice presented resistant and susceptible profiles in response to P. chabaudi subsp. chabaudi AS infection, respectively (Fig. 1). B10 mice experienced a transient peak parasitemia 9 days postinfection, but the infection was controlled spontaneously. In contrast, BALB/c mice developed a higher level of parasitemia, and all animals died within 12 days postinfection. In the resistant B10 mice, a significant amount of IFN-γ was revealed in the serum 7 days postinfection, and it was followed by the appearance of NO (approximately 220 nM) 3 days later (Fig. 2). In contrast, the amount of IFN-γ in the serum of the infected BALB/c mouse was only 3 ng/ml at the peak (7 days after infection), in contrast to 40 ng/ml in the infected B10 mouse serum. Furthermore, NO was undetectable in the serum of BALB/c mice throughout the infection period.

FIG. 1.

Parasitemias and survival rates of P. chabaudi subsp. chabaudi AS-infected mice. Groups of six BALB/c (triangles) and B10 (circles) mice were infected i.p. with 5 × 10⁴ AS PRBCs (solid symbols), and survival rates (open symbols; SR) were monitored. Results of parasitemia are expressed as means ± standard deviations.

FIG. 2.

Levels of IFN-γ and NO in the infected mouse serum. Groups of six BALB/c (triangles) and B10 (circles) mice were infected i.p. with 5 × 10⁴ P. chabaudi subsp. chabaudi AS PRBCs and bled from the orbital venous plexus, and sera were collected. The amounts of IFN-γ (open symbols) and NO (solid symbols) were determined by ELISA.

Production of TGF-β by AS-susceptible but not AS-resistant mice.

Figure 3 indicates the amounts of TGF-β in the sera of the infected mouse. When uninfected, both malaria-resistant B10 mice and malaria-susceptible BALB/c mice showed similar levels of serum TGF-β (9.1 ± 1.9 and 8.8 ± 1.6 ng/ml, respectively). In B10 mice, this physiological level of TGF-β did not change until day 6 postinfection. However, the TGF-β level of B10 mice decreased significantly after a marked level of IFN-γ production was revealed in the serum (7 days postinfection and later). In the susceptible BALB/c mice, the amount of TGF-β increased remarkably, with a peak at day 5 postinfection, and then returned to the physiological range by day 7 postinfection. The peak level of TGF-β in the infected BALB/c mice was approximately 2.4 times the control level.

FIG. 3.

Amounts of TGF-β in the acid-treated sera of infected mice. Groups of six BALB/c (open circles) and B10 (solid circles) mice were infected i.p. with 5 × 10⁴ AS PRBCs and bled from the orbital venous plexus, and sera were collected. The amount of TGF-β was determined by ELISA. Results are expressed as means ± standard deviations. Data were analyzed by ANOVA. ∗, P < 0.05; ∗∗, P < 0.01..

These data demonstrated a sequential regulation of cytokine production; in the infected B10 mice, because no TGF-β production was induced over the physiological level, sequential production of IFN-γ and NO was induced, while in BALB/c mice, increased production of TGF-β led to a relative lack of IFN-γ and NO production.

Splenic expression of mRNA of TGF-β, IFN-γ, and iNOS.

mRNA levels of TGF-β, IFN-γ, and iNOS in the mouse spleen were determined by RT-PCR at a variety of times after infection (Fig. 4). When uninfected, neither IFN-γ nor iNOS mRNA could be detected in the spleens. IFN-γ and iNOS mRNAs were detected in the spleens of B10 mice 6 to 8 days postinfection. The time of expression of IFN-γ and iNOS mRNAs preceded detection of IFN-γ and NO in the serum of the B10 mouse by 1 to 2 days. In the susceptible BALB/c mice, there was marginal expression of IFN-γ mRNA only on day 6 after infection, whereas iNOS mRNA was undetectable at all times.

FIG. 4.

Splenic expression of mRNA of IFN-γ, iNOS, and TGF-β in mice infected with 5 × 10⁴ P. chabaudi subsp. chabaudi AS PRBCs. Spleens of B10 and BALB/c mice were collected from uninfected (day 0) and 4, 6, and 8 days postinfected mice, and RNA was prepared for RT-PCR. mRNA of G3PDH was also investigated as a control. M, marker.

In contrast, expression of TGF-β mRNA was demonstrated in the spleens of both strains of uninfected mice. Interestingly, these physiological levels of TGF-β transcripts of resistant as well as susceptible strains of mice did not change throughout the infection period, irrespective of heightened (4 to 6 days postinfection in BALB/c mice) or lowered (7 days postinfection and later in B10 mice) levels of TGF-β production in the serum.

Effects of rTGF-β administration in the resistant mouse.

Administration of rTGF-β to the infected C57BL/10 mice resulted in a marked reduction of IFN-γ production (Fig. 5A). The amount of IFN-γ produced was only 7% of that produced by the PBS-injected control B10 mice 7 days after infection. IFN-γ could not be detected in the serum 5 and 10 days after infection (data not shown), indicating that administration of TGF-β did not alter the time of IFN-γ production. In the rTGF-β-administered B10 mice, the amount of serum NO was also significantly suppressed to 16% of that of the control mice 10 days after infection (Fig. 5B). Administration of rTGF-β rendered B10 mice susceptible to infection, as indicated by unrelenting parasitemia and 100% mortality by 12 days after infection (Fig. 5C).

FIG. 5.

In vivo effect of rTGF-β administration on serum cytokine production and infection profile of B10 mice. Groups of six B10 mice were infected with 5 × 10⁴ P. chabaudi subsp. chabaudi AS PRBCs on day 0. Mice were administered 10 μg of rTGF-β (TGF-β) or PBS 1 h before infection and every 2 days until day 8 postinfection. (A) Amount of serum IFN-γ 7 days after infection. (B) Amount of serum NO 10 days after infection. (C) Survival rates (circles; SR) and parasitemias (triangles; PRBCs) of the infected mice injected with rTGF-β (solid symbols) or PBS (open symbols). Results are expressed as means ± standard deviations. Data were analyzed by Student’s t test. ∗∗∗, P < 0.001.

Effect of anti-TGF-β MAb administration in the susceptible mice.

To neutralize the activity of endogenously produced TGF-β of infected BALB/c mice, an excess amount of anti-TGF-β-neutralizing MAb was injected. This treatment resulted in significant increases in IFN-γ (P < 0.05) and NO (P < 0.001) production in the serum of BALB/c mice 7 and 10 days after infection, respectively (Fig. 6A and B). Although a considerable amount of IFN-γ was also demonstrated in the serum of the control MAb-injected BALB/c mice (8.1 ± 2.1 ng/ml), the amount of IFN-γ appeared to be insufficient to induce detectable levels of NO. Neutralization of the endogenously produced TGF-β of the susceptible BALB/c mice resulted in a delay in the development of parasitemia, spontaneous decline of the parasitemia, and survival of the infection (Fig. 6C). The infection profile closely resembled that of the resistant mice.

FIG. 6.

In vivo effect of anti-TGF-β MAb administration on serum cytokine production and infection profile of BALB/c mice. Groups of six BALB/c mice were infected with 5 × 10⁴ P. chabaudi subsp. chabaudi AS PRBCs on day 0. Anti-TGF-β MAb (clone 15-1) was injected into the mice i.p. −2, −1, 0, 2, 4, and 6 days postinfection. An isotype-matched rat MAb (control [Cont] IgG1) with unrelated specificity was used as control. (A) Amount of serum IFN-γ 7 days after infection. (B) Amount of serum NO 10 days after infection. (C) Parasitemias (triangles; PRBCs) and survival rates (circles; SR) of the infected mice injected with clone 15-1 (solid symbols) or control MAb (open symbols). Results are expressed as means ± standard deviations. Data were analyzed by Student’s t test. Because the amount of NO in the control IgG1 group was less than the lower limit of the test, the concentration for the group was postulated to be 1 μM, the lowest concentration of the standards, and statistical analysis was performed. ∗∗, P < 0.01.

Activation of CD4⁺ T and CD11b⁺ cells in the infected mouse and their blockade by rTGF-β.

FACS analysis of the splenic CD4⁺ T cells from a B10 mouse 4 days postinfection showed a significantly higher level of expression of Ly6A/E antigen, which is known to increase on activated CD4⁺ T cells (5). A representative result is shown in Fig. 7A. MFI increased to 493 (498 ± 36 in three experiments), in contrast to 232 (236 ± 11) in uninfected mouse spleen cells, indicating that CD4⁺ T cells were strongly activated after infection. However, this CD4⁺ T-cell activation was completely prevented by injection of rTGF-β into the infected mice (MFI = 231 [233 ± 18]). The ANOVA showed a statistically significant difference (P < 0.001) between MFI values of the PBS-treated group and those of the uninfected or TGF-β-administered group. In contrast to the infected B10 mice, no CD4⁺ T-cell activation was revealed in BALB/c mice 4 days after infection (data not shown).

FIG. 7.

Effect of rTGF-β administration on in vivo activation of CD4⁺ T cells and CD11b⁺ cells in spleens of 4-day-postinfected B10 mice. B10 mice were infected with 5 × 10⁴ AS PRBCs on day 0 and administered 10 μg of rTGF-β or PBS 1 h before infection and every 2 days after infection. Spleen cells from uninfected (and rTGF-β–nonadministered) mice were also used. Cells were double stained with FITC-labeled anti-CD4 and PE-labeled anti-Ly6A/E MAbs (A) phycoerythrin-labeled anti-CD11b and FITC-labeled anti-I-A MAbs (B). CD4⁺ T cells (A; R1) or CD11b⁺ cells (B; R2) were gated as indicated in the dot plots, and then the intensities of the phycoerythrin-labeled anti-Ly6A/E MAb and FITC-labeled anti-I-A MAb, respectively, in the histogram were compared. The dot plots show data from the uninfected mouse to indicate areas of gating.

FACS analysis was also carried out to demonstrate expression of the major histocompatibility complex (MHC) class II I-A antigen on the surface of splenic CD11b⁺ cells obtained from 4-day-postinfected B10 mice. Mac-1⁺ cells express an increased amount of MHC class II antigen when activated (28). The result indicated a significant increase of I-A antigen on the surface of Mac-1⁺ cells after infection (MFI = 302 in the experiment indicated in Fig. 7B and 300 ± 25 in three experiments) compared with the level in uninfected control mice (MFI = 145 [148 ± 16 in three experiments]). However, this enhancement of I-A antigen expression on the splenic Mac-1⁺ cells was completely suppressed when the infected B10 mice were injected with rTGF-β (MFI = 148 [139 ± 30]). ANOVA showed a statistically significant difference (P < 0.001) between the MFI values of the PBS group and those of the uninfected or TGF-β-administered group. In contrast to the infected B10 mice, we failed to show increased expression of I-A antigen on the surface of Mac-1⁺ cells of BALB/c mice 4 days after infection (data not shown).

DISCUSSION

In clinical and epidemiological studies, it was evident that the level of TGF-β in serum was significantly depressed in the acute phase of P. falciparum infection (39) and that a higher level of TGF-β was present in the placentas of women living in an area in which malaria is holoendemic (11). However, no precise mechanism or role of TGF-β in regulation of manifestation of human malaria was indicated in these studies. In the present study, the role of TGF-β in induction of failure to protect against infection with blood-stage P. chabaudi subsp. chabaudi AS was demonstrated with resistant and susceptible mouse models.

Strain variations in the level of resistance to infection with AS have been investigated, and it was shown that, when survival time was used as criterion, C57BL/6J and other strains of mice were resistant to the parasite, whereas BALB/c and other strains of mice were susceptible (34, 41). We studied the level of resistance and susceptibility to infection with several strains of mouse malaria parasites in a wide variety of inbred, outbred, and congenic strains of mice (unpublished observations) and found that C57BL/10 Sn Slc (B10) and BALB/c Cr Slc (BALB/c) mice were representative strains of resistance and susceptibility to infection with AS, respectively.

P. chabaudi-susceptible mice produced low to background levels of IFN-γ (41), whereas resistant B10 mice produced a remarkably large amount of IFN-γ shortly after infection, followed by NO, which is known to have a potent antiparasitic effect (19). However, the precise role of IFN-γ in resistance to the infection is still controversial. Both susceptible and resistant mouse strains produced IFN-γ after infection, suggesting that susceptibility was not due to a defect in IFN-γ production (23). It was suggested that IFN-γ exerted the effect on the course of parasitemia and the outcome of infection by different mechanisms (7). Loci controlling peak parasitemia in susceptible and resistant mice were investigated, and several chromosomal regions were identified as suggestive linkages (10). Among them, a locus on chromosome 8 contained genes associated with host response to infection, interleukin-15, and the class A scavenger receptor-encoding Scvr locus. We have shown that the IFN-γ- and NO-dependent resistance to AS was regulated by CD4⁺ T cells alone in the resistant B10 mice. In contrast, despite their normal CD4⁺ T-cell activity, BALB/c mice failed to produce meaningful amounts of IFN-γ and NO, and, hence, continuous multiplication of the parasites was induced, resulting in a fatal outcome (36).

Although TGF-β and TGF-β mRNA showed similar levels in both strains of mice when uninfected, a marked difference was induced after infection. For levels of TGF-β, a transient increase was induced only in susceptible BALB/c mice, while in B10 mice, a significant decrease was induced after the appearance of IFN-γ in the serum. Unexpectedly, however, the level of TGF-β in serum of both strains of the infected mouse did not reflect quantitatively differential levels of cellular TGF-β mRNA. In fact, splenic mRNA of TGF-β showed similar levels in both susceptible and resistant strains of mice when uninfected, and these levels of expression of physiological TGF-β transcripts were not affected by infection. These results are in agreement with the concept that the expression of TGF-β transcripts is unlikely to be indicative of actual protein secretion. TGF-β is secreted as a biologically inactive form, and posttranscriptional events are pivotal in regulating the production of biologically active TGF-β (18). In the present paper, when sera were not acid activated, no TGF-β or a very low level, if any, of TGF-β was detected in sera from both strains of mice before and after infection (data not shown).

In contrast, close associations were demonstrated between expression of splenic mRNAs of IFN-γ and iNOS and levels of IFN-γ and NO in serum. In B10 mice, but not BALB/c mice, levels of mRNA of IFN-γ and iNOS in the spleens increased remarkably 1 to 2 days prior to enhancement of serum IFN-γ and NO levels, respectively.

Administration of rTGF-β rendered resistant B10 mice unable to control the infection. These mice showed a relative lack of IFN-γ and NO production, a higher level of parasitemia, and a fatal outcome. In contrast in the AS-infected BALB/c mice, neutralization of endogenous TGF-β activity by administration of anti-TGF-β MAb resulted in marked production of IFN-γ and NO and acquisition of resistance to the infection. The infection profile of the TGF-β-neutralized BALB/c mice was similar to that of genetically resistant B10 mice. These results provided direct evidence that TGF-β is the key molecule in induction of mouse susceptibility to blood-stage AS.

Resistance of C57BL/10 mice to the parasite was mediated solely by a CD4⁺ T-cell-dependent mechanism (36), and activated CD4⁺ T cells expressed an increased amount of Ly6A/E antigen (5). The number of spleen cells increased to about twofold or more 4 to 5 days postinfection, and one of the major populations of the increased spleen cells was CD11b⁺ cells (data not shown). CD11b⁺ cells played a role in protection against malaria infection (21) and expressed an increased amount of MHC class II antigen when activated (28). In this paper, it was demonstrated by FACS analyses that CD4⁺ T cells and CD11b⁺ cells of the infected B10 mice were activated shortly after infection, and this could be suppressed completely by administration of rTGF-β. In contrast, in BALB/c mice, no such activations were demonstrated. Thus, it was suggested that, in the infected BALB/c mice, endogenously produced TGF-β suppressed activation of CD4⁺ T cells and CD11b⁺ cells and production of meaningful amounts of IFN-γ and NO and hence led to the failure of resistance to AS.

In accordance with the observations presented in this paper, the downregulatory role of TGF-β was shown in mouse leishmaniasis, in which production of active TGF-β contributed to the blunting of CD4⁺ T cells, to the subsequent lack of Th1 cell development and promotion of the development of Th2 cells, and to activation of latent infection (1, 31, 32). Also, TGF-β has been indicated to have a downregulatory role in resistance to infection by a variety of microorganisms, including protozoal parasites (16, 18, 33, 40, 42). In Trypanosoma cruzi infection, many TGF-β-producing cells were revealed in tissues throughout the acute and chronic phases of infection in both the resistant and susceptible models (33, 42). In a recent publication, Omer and Riley (27) indicated that, in mice infected with variety of murine Plasmodium species, lethal infections were accompanied by low levels of TGF-β, while self-resolving infections were accompanied by high levels of TGF-β. BALB/c mice used by Omer and Riley were given aminobenzoic acid in drinking water and survived the infection for more than 3 weeks. However, the BALB/c mice we used showed a typical profile of susceptibility to infection with AS. These differences might result in discrepancies between the findings obtained in this paper and those obtained by Omer and Riley. TGF-β controls immune responses by a complex and often context-dependent manner (22). These findings together suggested an involvement of TGF in pathogenesis and resistance of the infected hosts by complicated mechanisms.