Malnutrition Alters the Innate Immune Response and Increases Early ...

4 downloads 7889 Views 164KB Size Report
Lymph node cells from the malnourished, infected mice produced increased levels ..... nous carbonate (0.1 g/kg), chromium potassium sulfate 12H2O (0.02 g/kg), ...... Harrison, L. H., T. G. Naidu, J. S. Drew, J. E. de Alencar, and R. D. Pearson.
INFECTION AND IMMUNITY, Aug. 2001, p. 4709–4718 0019-9567/01/$04.00⫹0 DOI: 10.1128/IAI.69.8.4709–4718.2001 Copyright © 2001, American Society for Microbiology. All Rights Reserved.

Vol. 69, No. 8

Malnutrition Alters the Innate Immune Response and Increases Early Visceralization following Leishmania donovani Infection GREGORY M. ANSTEAD,1,2 BYSANI CHANDRASEKAR,2 WEIGUO ZHAO,1,2 JUE YANG,1,2 LUIS E. PEREZ,1,2 AND PETER C. MELBY1,2,3* Medical Service, Department of Veterans Affairs Medical Center, South Texas Veterans Health Care System,1 and Departments of Medicine,2 and Microbiology,3 University of Texas Health Science Center at San Antonio, San Antonio, Texas 78229-3900 Received 4 January 2001/Returned for modification 19 February 2001/Accepted 29 April 2001

Malnutrition is a risk factor for the development of visceral leishmaniasis. However, the immunological basis for this susceptibility is unknown. We have developed a mouse model to study the effect of malnutrition on innate immunity and early visceralization following Leishmania donovani infection. Three deficient diets were studied, including 6, 3, or 1% protein; these diets were also deficient in iron, zinc, and calories. The control diet contained 17% protein, was zinc and iron sufficient, and was provided ab libitum. Three days after infection with L. donovani promastigotes, the total extradermal (lymph nodes, liver, and spleen) and skin parasite burdens were equivalent in the malnourished (3% protein) and control mice, but in the malnourished group, a greater percentage (39.8 and 4.0%, respectively; P ⴝ 0.009) of the extradermal parasite burden was contained in the spleen and liver. The comparable levels of parasites in the footpads in the two diet groups and the higher lymph node parasite burdens in the well-nourished mice indicated that the higher visceral parasite burdens in the malnourished mice were not due to a deficit in local parasite killing but to a failure of lymph node barrier function. Lymph node cells from the malnourished, infected mice produced increased levels of prostaglandin E2 (PGE2) and decreased levels of interleukin-10. Inducible nitric oxide synthase activity was significantly lower in the spleen and liver of the malnourished mice. Thus, malnutrition causes a failure of lymph node barrier function after L. donovani infection, which may be related to excessive production of PGE2 and decreased levels of IL-10 and nitric oxide. weight-for-age (WA), height-for-age, and weight-for-height (13). However, in previous mouse models of malnutrition, there have been no efforts to relate morphometric measures of nutritional status to either human anthropometric scales or immunocompetence. Weight-for-age determination is advantageous because it can be measured unambiguously, provides a synthesis of linear growth and body proportion (13), and correlates with probability of death in children in developing countries (36). In this study, murine WA was correlated with specific measures of host defense and risk for visceral L. donovani infection. A second component of the study was to examine possible defects in the mediator network of the innate immune system produced by malnutrition, because it is the early events that are likely to determine whether the inoculated parasites are controlled locally or disseminate to visceral organs. The innate immune system provides a first line of defense against pathogens and instructs the differentiation of Th0 cells into Th1 and Th2 cells (18). It has been estimated that the innate immune system provides protection against 98% of the pathogens that are encountered (34). Malnutrition has been associated with an increased risk of many infections (3); however, there are no animal models that specifically examine the effect of malnutrition on the innate immune response to infection. The third part of the study was to investigate the mechanism of visceralization (the process whereby parasites disseminate from the site of cutaneous inoculation and the draining lymph nodes to the liver, spleen, and bone marrow). To reach this goal, we needed to develop a more natural animal model of visceral leishmaniasis, using the vector stage of the parasite (promastigote) and the intradermal route of infection. Al-

Globally, protein-energy malnutrition is the most frequent cause of immunodeficiency (58). Epidemiologic and experimental studies have documented an increased risk for visceral leishmaniasis, caused by intracellular protozoan parasites of the Leishmania donovani complex, in the malnourished host (1, 2, 26). However, the immunologic basis for this association has not been established and standardized experimental models have not addressed this important issue. In this study, our goal was to investigate the mechanisms of the malnutrition-related susceptibility to visceral leishmaniasis. There were three components of the study. First, we needed to create a murine model of malnutrition that was relevant to human malnutrition in developing countries. Although the mouse has been extensively used in animal models of malnutrition, there is no standard murine model of protein-energy malnutrition (69). Human malnutrition is complex, typically involving deficiency of protein and energy with superimposed deficits of other nutrients. Zinc deficiency usually accompanies protein-energy malnutrition (19). Iron deficiency is highly prevalent in developing countries and may accompany zinc deficiency due to a common risk factor, cereal-based diets with little meat (61). Thus, in this model, in addition to protein and energy, zinc and iron were selected as deficient nutrients. Much of the vast body of data that has been collected on human malnutrition is based on anthropometric measures, i.e., * Corresponding author. Mailing address: Division of Infectious Diseases, Department of Medicine, University of Texas Health Science Center at San Antonio, 7703 Floyd Curl Dr., Mailcode 7881, San Antonio, TX 78229-3900. Phone: (210) 567-4614. Fax: (210) 567-4670. E-mail: [email protected]. 4709

4710

ANSTEAD ET AL.

though the immunopathogenesis of murine visceral leishmaniasis has been investigated, previous studies have used an unnatural (intravenous) route of infection and/or the mammalian host stage of the parasite (amastigote) as the infecting inoculum (35, 40, 43, 48, 65). The intravenous route of inoculation negates the immune processing that would occur in the skin and in the draining lymph node and therefore does not accurately reflect the immune response that would occur in a natural cutaneous infection. Furthermore, models of visceral leishmaniasis that utilize the intravenous route of inoculation do not facilitate an understanding of the mechanism of visceralization. Selection of either amastigote or promastigote as the infective inoculum may also have an important bearing on the course of infection, because these two stages of the parasite possess disparate virulence factors (9) and produce different immune responses (20). In this study, we establish a murine model of polynutrient deficiency that is similar to human malnutrition in developing regions of the world. We demonstrate that the malnourished mouse has an altered innate immune defense and is at increased risk of visceralization following cutaneous L. donovani infection. MATERIALS AND METHODS Mice. Weanling (3-week-old) female nu/⫹ BALB/c mice were obtained from the Veterinary Medical Unit breeding colony of the Department of Veterans Affairs Medical Center, South Texas Veterans Health Care System, San Antonio, Tex. Individual mice were identified by ear notching. Diets and feeding. Mice received a 3-day acclimation on standard mouse chow (Teklad LM-485; Harlan Teklad, Madison, Wis.) after weaning and prior to the change to experimental diets. The mice had free access to water. They were housed in groups of four, and a low-trace-element bedding was used (Alpha-Dri; Shepard Speciality Papers, Kalamazoo, Mich.). Experimental diets were formulated by Harlan-Teklad and a detailed description of their composition is shown in Table 1. There were four diets, A, B, C, and D, which contained 17, 6, 3, and 1% protein, respectively. Control diet A was sufficient in iron and zinc. Diets B, C, and D were iron and zinc deficient. Mice assigned to diet group A received 4.5 g of mouse chow/day. Mice in the deficientdiet groups B, C, and D received 3.0 g of food/day. Food intake was recorded on a twice-weekly basis. In the first experiment, mice were randomly selected and maintained on the four experimental diets A to D (10 mice per group) for 4 weeks. In subsequent experiments, mice were initially weight matched and maintained on diets A (17% protein) and C (3% protein) for 6 weeks prior to infection. Parasites and experimental infection. Metacyclic L. donovani (IS strain; MHOM/SD/00/S-2D) promastigotes were obtained from stationary-phase cultures by negative selection with peanut agglutinin (30, 60). In brief, promastigotes from cultures of spleen tissue from mice or hamsters previously infected with L. donovani were harvested after 5 days. The parasites were grown in complete M199 medium (cM199) (15% fetal calf serum (FCS), 1% [vol/vol] 10 mM adenine in 50 mM HEPES, 0.25% [wt/vol] hemin in 50% [vol/vol] triethanolamine, 1% penicillin-streptomycin [each at 10,000 IU/ml]) for 3 days and then passaged in fresh medium. The promastigotes were harvested on day 6, and the metacyclic forms were isolated. Then 5 ⫻ 106 metacyclic promastigotes in Dulbecco’s modified Eagle’s medium (DMEM) were inoculated intradermally into both hind footpads. Quantitation of parasite burdens. The tissue parasite burdens were determined by quantitative limiting-dilution culture in biphasic blood agar-cM199 (60). After 3 days, infected mice were euthanized by cervical dislocation and the footpads, popliteal lymph nodes, spleen, and livers were harvested and weighed. To harvest the footpads, the infected feet were cleansed in an iodine disinfectant and rinsed with 70% ethanol. The lymph nodes, footpads, and portions of the spleen and liver were homogenized between the frosted ends of two microscope slides in 1 ml of cM199. For limiting-dilution culture, the tissue was diluted as follows: liver and spleen, 2 mg/ml; footpad, 0.01 mg/ml; and lymph nodes, 0.1 mg/ml. To isolate bone marrow, both femurs were isolated and the ends were snipped. The marrow cavity was flushed with DMEM 2% FCS. The cell suspen-

INFECT. IMMUN. TABLE 1. Composition of experimental dietsa Nutrient (unit)

Egg white solids (g/kg) % of proteinb Dextrose (g/kg) CaHPO4 (g/kg) CaCO3 (g/kg) Potassium citrate (g/kg) NaCl (g/kg) Ferric citrate (g/kg) Zinc carbonate (g/kg) Iron (ppm)c Zinc (ppm)d

Amt in: Diet A (control)

Diet B

Diet C

Diet D

210 17 643 20.2 0.98 5.5 1.0 0.59 0.054 100 30

75 6 774 20.8 0.8 10.0 2.8 0 0 10 1

37.5 3 807 20.9 0.78 11.3 4.0 0 0 10 1

12.5 1 830 21.0 0.75 12.1 4.7 0 0 10 1

a All diets contained identical amounts of corn oil (90 g/kg), cellulose (15 g/kg), Teklad vitamin mix 40060 (10 g/kg), biotin (0.002 g/kg), TBHQ antioxidant (0.017 g/kg), potassium sulfate (1.9 g/kg), magnesium oxide (1.5 g/kg), manganous carbonate (0.1 g/kg), chromium potassium sulfate 䡠 12H2O (0.02 g/kg), cupric carbonate (0.016 g/kg), potassium iodate (0.001 g/kg), and sodium selenite 䡠 5H2O (0.0005 g/kg). Teklad vitamin mix 40060 consists of p-aminobenzoic acid (11.0 g/kg), ascorbic acid (101.7 g/kg), biotin (0.044 g/kg), vitamin B12 (3.0 g/kg), calcium panthothenate (6.6 g/kg), choline dihydrogen citrate (349.7 g/kg), folic acid (0.2 g/kg), inositol (11.0 g/kg), menadione (5.0 g/kg), niacin (9.9 g/kg), pyridoxine (2.2 g/kg), riboflavin (2.2 g/kg), thiamine (2.2 g/kg), vitamin A palmitate (4.0 g/kg), vitamin E acetate (24.2 g/kg), and cornstarch (467 g/kg). b The minimum daily protein content that supports weight gain in mice is 13.6%; optimal weanling weight gain is achieved at approximately 18% protein (37). c Normal growth and hematopoiesis in mice requires 25 to 100 ppm of iron (51). d Minimum zinc requirements in mice are 12 ppm; however, due to low bioavailability, 30 ppm is typically added (54).

sion was centrifuged, the supernatant was removed, and the pellet was resuspended in 1.2 ml of cM199. For limiting-dilution culture, the tissue homogenates (at the above concentrations) or the bone marrow suspension were plated in 96-well blood agar plates and cultured at 26°C for 10 days (three replicates for footpad and lymph nodes; six for the spleen, liver, and bone marrow). The reciprocal of the highest dilution positive for motile parasites was considered to be the concentration of parasites per milligram of tissue. For lymph nodes, liver, and spleen, the total organ parasite burden was calculated by multiplying this concentration by the whole-organ weight. Total parasite burdens were not calculated for footpad and bone marrow because the total amount of tissue could not be harvested. Buffy coat was cultured for L. donovani in one experiment. Samples of blood from the retro-orbital plexus were taken up in four heparincoated capillary tubes, and the pooled buffy coat layers from a single mouse were centrifuged. The pellet was treated with erythrocyte lysis buffer and washed with DMEM–2% FCS. For culture, one-third of the pooled buffy coat layers were placed in 200 ␮l of medium in the first well of each row of a 96-well plate (three replicates). Determination of nitric oxide production by resident peritoneal cells. Resident peritoneal cells were obtained from mice by peritoneal lavage with DMEM–1% FCS. One million cells from each mouse were resuspended into 1 ml of DMEM–10% FCS. The cells were stimulated with 20 U of mouse gamma interferon (IFN-␥) (PharMingen) per ml and 1 ␮g of lipopolysaccharide (LPS) (Escherichia coli O111:B4; Sigma, St. Louis, Mo.) per ml for 24 h at 37°C under 5% CO2. Supernatants were tested for total nitrite by the Greiss reaction, after conversion of nitrate to nitrite with nitrate reductase (nitrate/nitrite colorimetric assay kit; Cayman Chemical Co., Ann Arbor, Mich.). Inducible and constitutive nitric oxide synthase activity. Nitric oxide synthase (NOS) activity in the liver and spleen was determined as previously described (8), using the NOSdetect kit from Stratagene (La Jolla, Calif.). Inducible NOS (iNOS; NOS2) and constitutive NOS (cNOS; NOS1 and NOS3) enzymatic activities were determined by measuring the extent of conversion of L-[3H]arginine to L-[3H]citrulline in the presence (NOS2) or absence (cNOS) of Ca2⫹ chelators. Frozen liver and spleen tissue was homogenized in an ice-cold buffer containing 25 mM Tris-HCl, 1 mM EDTA, 1 mM EGTA, 1 mM dithiothreitol, 100 ␮g of phenylmethylsulfonyl fluoride per ml, 10 ␮g of leupeptin per ml, 10 ␮g of soybean trypsin inhibitor per ml, and 2 ␮g of aprotinin per ml (pH 7.4). The homogenate was centrifuged at 100,000 ⫻ g for 30 min, and the supernantant was used for the NOS assay. Bicinchoninic acid protein assay reagent (Pierce, Rockford, Ill.) was used to determine

VOL. 69, 2001

MALNUTRITION AND VISCERAL LEISHMANIASIS

4711

FIG. 1. (A) Growth characteristics of weanling mice fed the four experimental diets A to D for 4 weeks. Mice (n ⫽ 10 for each diet group) were changed to the experimental diets shortly after weaning and continued for 28 days. Diet A (control) contained 17% protein, was zinc and iron sufficient, and was provided ad libitum. Diets B, C, and D contained 6, 3, and 1%, protein, respectively, and were zinc, iron, and calorie deficient (see Table 1). Shown are growth curves over 28 days after the change to the experimental diets. The significance of the differences in mean values of percent weight gain on day 28 between the control (diet A) and the malnourished (diets B, C, and D) mice was determined by Student’s t test. Data are given as mean and SEM. (B) Effect of malnutrition on early visceralization of L. donovani to the liver and spleen. Mice (n ⫽ 6 for each diet group) were fed the four experimental diets A to D described for panel A and were infected in the hind footpad with 5 ⫻ 106 metacyclic promastigotes on day 28 after initiation of the experimental diet. The mice were euthanized 3 days after infection, and the parasite burden in the liver and spleen was determined by quantitative limiting-dilution culture. Parasite visceralization was considered to have occurred if parasites were detected in the culture of the lowest dilution of tissue. Percent visceralization is defined as the number of organs (spleen or liver) with detectable parasites divided by the total number of organs. The significance of the differences in the proportion of mice that showed visceralization between the normal control group (diet A) and the malnourished groups (diets B, C, and D) was determined using the chi-square test. the protein concentration, with bovine serum albumin as the standard. The reaction was initiated by incubating 10 ␮g of supernatant with 80 nM L-[3H]arginine (Amersham, Arlington Heights, Ill.) for 30 min at 25°C and pH 7.4 in a solution consisting of 50 mM Tris-HCl, 1 mM EGTA, 1 mM dithiothreitol, 1 mM NADPH, 100 ␮M 6-R-5,6,7,8-tetrahydrobiopterin, and 10 ␮M flavin adenine dinucleotide. After addition of 0.5 ml of buffer containing 20mM HEPES (pH 5.5) and 2 mM EGTA, the mixture was applied to a 1-ml column (DOWEX AG 50 W-X8, Na⫹-form; Bio-Rad, Hercules, Calif.) preequilibrated with a buffer containing 20mM HEPES (pH 5.5) and 2 mM EGTA. L-[3H]citrulline was eluted twice with 0.5 ml of distilled water. The radioactivity of this 1-ml eluate was determined by liquid scintillation counting. Ex vivo lymph node and spleen cell culture. Lymph nodes and spleen tissue were collected from the infected mice and homogenized as for the parasite burden quantitation. Lymph node cells and splenocytes (5 ⫻ 106/ml) were cultured without exogenous stimulation in DMEM–10% FCS for 24 h. Supernatants were collected at 12 and 24 h. For the measurement of transforming growth factor (TGF-␤1), splenocytes were cultured in DMEM–2% FCS. Cytokine assays. Cytokine levels were quantitated by enzyme-linked immunosorbent assay (ELISA) as specified by the manufacturers. For interleukin-10 (IL-10), IL-12, and IFN-␥, sandwich ELISA was done with reagents supplied by PharMingen (San Diego, Calif.). The following kits were used for other mediators: tumor necrosis factor alpha (TNF-␣) (Endogen, Woburn, Mass.), TGF-␤1 (Promega, Madison, Wis.), and prostaglandin E2 (PGE2) (R&D Systems, Minneapolis, Minn.). Nunc-Immuno plates (Nalge Nunc International, Denmark) were used in the ELISA analyses. Statistics. Results are expressed as the mean and standard error of the mean. Group means were analyzed using Student’s two-tailed t test. The chi-square test was used to compare proportions. Data were considered statistically significant at P ⱕ 0.05. Regression analyses were performed using the Stat View statistical software (version 5.0.1) (SAS Institute, Cary, N.C.).

RESULTS Malnutrition decreases weight-for-age in proportion to the level of dietary protein deficiency. At the end of 4 weeks of

feeding, mice in the deficient-diet groups B, C, and D showed decreased weight-for-age and an altered growth curve compared to the control mice (Fig. 1A). Mice on the control diet (group A) received 4.5 g of chow/day; each mouse ate 3.3 g per day. Mice in diet groups B (6% protein) and C (3% protein) received 3.0 g of food/day and consumed 2.9 g/mouse/day (12% calorie deficit with respect to group A). Mice in the 1% protein diet group (D) ate 2.7 g per day (18% calorie deficit with respect to group A). The intake of diet group D was less than the other diet groups from the onset, suggesting decreased palatability of this chow. The respective percent weight gain (mean ⫾ standard error of the mean) for the mice in groups A, B, C, and D was 37.5% ⫾ 5.93%, 22.1% ⫾ 3.36%, 5.59% ⫾ 3.53%, and ⫺26.8% ⫾ 2.36%, the respective P values for the comparison of the weights at day 28 for groups B, C, and D compared to group A were 0.04, ⬍0.001, and ⬍0.001. Diet D (1% protein) produced a weight loss similar to a previous murine model of wasting protein-energy malnutrition (70). Diet C (3% protein) was selected for further study because it produced a flat growth curve similar to that observed in malnourished human weanlings (33) and a moderate level of malnutrition. By analogy to the Go ´mez anthropometric classification of human malnutrition (23), a murine scale of malnutrition was derived based on WA, using the control diet group (A) in each experiment as the standard. WA was calculated as (sacrifice weight of the animal in diet groups B, C, or D/expected weight based on mice in the control diet group A) ⫻ 100%. Three factors complicated the calculation of an expected weight: in-

4712

ANSTEAD ET AL.

INFECT. IMMUN.

exact weight matching between the control and experimental diet groups, variation in the growth of mice in the control group, and the observation that smaller mice gain proportionally more weight than larger mice. To adjust for these factors, a regression plot of percent weight gain versus initial weight for the control diet group provided an equation that could be used to obtain an expected percent weight gain for mice in diet groups B, C, and D (equation 1): Expected percent weight gain ⫽ 224.1 ⫺ 11.5 ⫻

(1)

(initial weight in grams), R ⫽ 0.98 The expected weight for the mice in groups B, C, and D is shown in equation 2. Expected weight, diet groups B, C, and D ⫽

(2)

initial weight in grams ⫹ (initial weight in grams) (expected percent weight gain/100) In our study, mice on diet B were an average of 82.5% ⫾ 2.25% WA; on diet C, they were 67.5% ⫾ 3.29% WA; and on diet D, they were 46.0% ⫾ 0.53% WA. The Go ´mez classification uses the following categories: 75 to 90% WA, mild malnutrition; 60% to 75% WA moderate, malnutrition; and ⬍60% WA severe malnutrition (23). Thus, diet B produced mild malnutrition, diet C produced moderate malnutrition, and diet D produced severe malnutrition. Malnutrition produces increased visceralization after cutaneous L. donovani infection. In five experiments, malnutrition consistently resulted in an increased parasite burden in the liver and spleen, assessed by either the percentage of animals with a measurable spleen or liver parasite burden or the total calculated organ parasite burdens. The data for two experiments are shown: a comparison of the four diet groups, and a more in-depth analysis of parasite burdens in mice fed control diet A (17% protein) and diet C (3% protein). The degree of visceralization of L. donovani was influenced by the level of malnutrition. Mice from each of the four diet groups were infected intradermally with L. donovani promastigotes, and parasite burdens were determined after 3 days in the lymph nodes, spleen, and liver. The total parasite burdens in the lymph nodes were similar in diet groups A, B, and C (3,360 ⫾ 790, 2,440 ⫾ 950, and 2,270 ⫾ 630, respectively), whereas the total lymph node parasite burden in diet group D was lower than in the well-nourished counterparts (574 ⫾ 220; P ⫽ 0.018 versus group A). However, the malnourished groups B, C, and D showed greater dissemination of infection to the spleen and liver compared to the well-nourished group A (Fig. 1B). The increased rate of visceralization did not depend on a higher parasite burden in the local draining lymph node. In fact, dissemination was more apparent when the total lymph node parasite burden was lower (i.e., group D). This result is in accord with a loss of lymph node barrier function induced by malnutrition (see below). A regression plot of the average total lymph node parasite burden (PBLN) versus the average WA for groups A, B, C, and D gives the equation PBLN ⫽ ⫺1435.6 ⫹ 48.8 (WA), with R2 ⫽ 0.93 and P ⫽ 0.036 (for group A, a WA of 100 was used). Thus, the average WA value is inversely

correlated with and closely predicts the average total lymph node parasite burdens. To further assess the role of local parasite control on early visceralization in the malnourished host, footpad (skin), lymph node, and visceral parasite burdens were determined in diet groups A and C. The malnutrition-increased visceral parasite burden was not dependent on an increased footpad parasite burden. After 6 weeks of experimental feeding, parasite burdens in the footpads (per milligram) were not significantly different between the two diet groups at 3 days postinfection. The parasite burdens in the lymph nodes, on a per-milligram basis, were similar in the two groups; however, due to the larger lymph node weight in the well-nourished animals, the total lymph node parasite burden was significantly higher. In a cutaneous-inoculation model, it is assumed that the majority of parasites reach visceral sites hematogenously after egress from the local draining lymph node. To quantify the effect of the lymph node as a barrier to dissemination, we defined two indices, the percent parasite nodal escape (%PNE) and percent lymph node barrier function (BFLN) (equations 3 and 4): %PNE ⫽ (PBliver ⫹ PBspleen)/

(3)

(PBliver ⫹ PBspleen ⫹ PBLN) (100%) where PBliver, PBspleen, and PBLN are the total liver, spleen, and lymph node parasite burdens, respectively, and BFLN ⫽ 100 ⫺ %PNE

(4)

In conceptual terms, %PNE represents the percentage of measured extradermal parasites that have breached the lymph node barrier and BFLN is the percentage of measured extradermal parasites retained in the lymph node. For the lymph node barrier concept to be valid, the average total measured extradermal parasite burden (total parasite burden of lymph nodes, liver, and spleen combined) should not differ significantly between the two groups. In fact, this was the case. The total measured extradermal parasite burdens were statistically equivalent in the well-nourished and malnourished animals, but the liver and spleen parasite burdens were higher in the malnourished animals, when calculated as both the parasites per milligram of tissue and the total organ parasite burden (Table 2). The percent PNE was 4.0 ⫾ 1.3 for the well-nourished animals and 39.8 ⫾ 8.84 for the malnourished animals (P ⫽ 0.009). The respective %BFLN values were 96.0 ⫾ 1.34 and 60.2 ⫾ 8.84. In an independent experiment comparing mice from the control diet group (A) and the 3% protein group (C) (n ⫽ 12 animals per group), the parasite burdens per microliter in the bone marrow were higher in the malnourished mice (66.3 ⫾ 30.7 and 1.01 ⫾ 0.237, respectively; P ⫽ 0.057). In the malnourished group (diet C), parasites were detected in the bone marrow of 12 of 12 mice, whereas in the well-nourished group (diet A), parasites were cultured from only 8 of 12 animals. We were unable to detect parasites by culture of the buffy coat from mice in either of the diet groups. Nitric oxide (NO) production by resident peritoneal cells and tissue NOS2 activity is impaired by malnutrition. Because NO is an important host defense and immunoregulatory molecule, we measured NO production by resident peritoneal cells

VOL. 69, 2001

MALNUTRITION AND VISCERAL LEISHMANIASIS

TABLE 2. Parasite burdens of L. donovani-infected mice on diet A and diet C a Site

Footpad (per mg) Lymph node (per mg) Lymph node (total) Liver (per mg) Liver (total) Spleen (per mg) Spleen (total) tedPBb tvPBc

Parasite burden with: Diet A

Diet C

333 ⫾ 63 4870 ⫾ 860 43,600 ⫾ 7,000 2.1 ⫾ 1.1 2,300 ⫾ 1,100 1.0 ⫾ 0.33 143 ⫾ 50 46,000 ⫾ 8,590 2,400 ⫾ 1,280

857 ⫾ 303 5,720 ⫾ 1,350 23,900 ⫾ 5,600 19.5 ⫾ 5.3 15,100 ⫾ 4,300 6.6 ⫾ 2.0 406 ⫾ 105 39,400 ⫾ 6,090 15,500 ⫾ 4,270

P value

0.12 0.60 0.04 0.008 0.01 0.01 0.04 0.97 0.03

a Three days after infection (5 ⫻ 106 metacyclic promastigotes injected into each footpad; n ⫽ 12 mice per group). b Total extradermal parasite burden (total parasite burden in lymph nodes, liver, and spleen). c Total visceral parasite burden (total parasite burden in liver and spleen).

from mice on the four diets. After stimulation with IFN-␥ plus LPS, resident peritoneal cells from mice in diet groups A, B, C, and D produced progressively less NO as the percentage of dietary protein decreased (Fig. 2A). There was a correlation between resident peritoneal cell NO production and the average WA value of diet groups A, B, C, and D, as expressed by the equation: NO production (micromolar) ⫽ 0.030(WA) ⫺ 0.42, with R ⫽ 0.76 and P ⫽ 0.002. (For control group A, the average NO production and WA ⫽ 100 were used.) To determine if the in vitro findings of impaired generation of NO by resident peritoneal cells from malnourished mice also occurred in an in vivo infection, we measured hepatic and splenic cNOS and NOS2 activity in the well-nourished (diet A, 17% protein) and malnourished (diet C, 3% protein) mice 3 days after infection. As with the resident peritoneal cells, there was a significant reduction of the splenic and hepatic NOS2 enzyme activity in the malnourished mice compared to the well-nourished mice (Fig. 2B). In contrast, the cNOS activity was identical in the two groups in both organs. Considering the large differences in visceral parasite burdens in the two diet groups A and C and considering that NOS2 expression is related to the level of visceral L. donovani infection (43), a more meaningful comparison would be the level of NOS2 activity that has been normalized to the intensity of the stimulus (parasite burden). Parasite burden-adjusted NOS2 activity was calculated by dividing the NOS2 activity (in nanomoles per minute per milligram) by the parasite burden per milligram (Fig. 2C). In the spleen and liver, the parasite burden-adjusted NOS2 activities were 14- and 52-fold higher, respectively, in the well-nourished mice than in the malnourished mice. The spleen had a higher parasite burden-adjusted NOS2 activity than the liver, and this difference was amplified by malnutrition. Relationship of inflammatory mediators in the malnutrition-related increase in parasite visceralization. To understand the potential mechanisms for the increased visceralization in the malnourished mice, we determined the early (3 days postinfection) production of a number of inflammatory mediators by cells from the draining lymph nodes and spleens of infected mice in both the malnourished group C and the control group A. Lymph node cells from infected mice in groups A and C were cultured ex vivo without exogenous stimulation,

4713

and the culture supernatants were collected at 12 and 24 h (Table 3). Levels of IL-10 were significantly higher in mice from the well-nourished group at both time points. IL-12 levels were not significantly different. IFN-␥ levels were measured in several independent experiments and were variable, without a consistent difference between diet groups A and C. PGE2 synthesis was 2.8-fold higher in malnourished group C than in control group A (P ⫽ 0.005). To determine a relationship between cytokine production, NOS2 activity, and parasite burdens in the spleen, levels of IL-10, IL-12, IFN-␥, TNF-␣, and TGF-␤ were determined in the supernatants of splenocytes from infected mice that were cultured ex vivo (without exogenous stimulation) for 24 h. Splenocytes from mice from the two diet groups produced comparable levels of IL-10, IL-12, and TGF-␤ (Table 4). IFN-␥ was undetectable (lower level of detection ⫽ 0.16 ng/ ml). However, splenocytes from the infected malnourished animals (group C) produced significantly less TNF-␣ than did those from the well-nourished animals (group A). Regression analysis revealed a correlation between splenocyte TNF-␣ levels and NOS2 activity (R ⫽ 0.48, P ⫽ 0.034), suggesting that the impaired TNF-␣ production in the malnourished mice contributed to a reduction of NOS2 activity. Correlation of visceral parasite burden and percent parasite nodal escape with lymph node mediator levels. To relate visceral parasite burdens to lymph node mediator levels, several measures of visceralization, including the log-transformed total visceral parasite burden [log tvPB ⫽ log (PBliver ⫹ PBspleen)], total liver parasite burden, total spleen parasite burden, and %PNE, were plotted against lymph node levels of PGE2, IL-10, IL-12, and IFN-␥ (Table 5). PGE2 levels correlated (R ⫽ 0.65) and IL-10 levels negatively correlated (R ⫽ ⫺0.54 and ⫺0.47, for 12- and 24-h measurements) with log tvPB. Levels of IL-12 and IFN-␥ did not correlate with visceralization. From the coefficient of determination (R2), 42% of the variation in the total visceral parasite burden or total liver parasite burden is related to the level of lymph node PGE2 production. Similarly, PGE2 production correlated with %PNE (R ⫽ 0.69). However, the total spleen parasite burden did not correlate with lymph node PGE2 or IL-10 levels, probably because splenic parasite burden made up only a small fraction of the total visceral parasite burden (2.6% in the malnourished mice, 6.0% in the control mice) and, as discussed below, because dissemination to the spleen may occur after visceralization to the liver. DISCUSSION In this paper, we describe a murine model of malnutrition and demonstrate that malnutrition alters the innate immune response and leads to early visceralization following cutaneous L. donovani infection. This finding corroborates the epidemiologic observations that malnutrition is a risk factor for the development of visceral leishmaniasis (2, 26). Several features of this model are noteworthy. A polynutrient-deficient diet was used because human malnutrition typically involves multiple, simultaneous nutrient deficiencies. This study is the first to derive a scale of murine malnutrition based on body measurements (murinometrics) and relate this to a human scale of malnutrition. This model used the naturally infective stage of

4714

ANSTEAD ET AL.

INFECT. IMMUN.

the parasite (metacyclic promastigote) injected intradermally, as it would be delivered by the sand fly to initiate a natural infection. This allowed us to investigate the immunologic factors that may play a role in the dissemination of infection from the skin and draining lymph node to the visceral organs. We focused on early inflammatory events after infection, when antigen-specific immunity is not operative and the host defense depends on innate mechanisms. Early immunological events in the course of leishmaniasis are critical to the ultimate outcome of the infection (39). We found that the early visceralization observed in the malnourished host is not due to defective local parasite killing in the footpad or lymph node but results from failure of the draining lymph node to act as a barrier. Thus, we propose that the draining lymph node acts like a “firewall”, preventing or delaying pathogen dissemination, and that malnutrition causes a disruption of this barrier. The concept of a lymph node barrier has been presented in the oncology literature (17), but there have been no efforts to formalize this concept for infections. There are several examples of lymphoid tissue acting as a barrier to infection. In a murine model of L. donovani infection, Melby et al. (using a less sensitive culture method) were unable to culture parasites from liver and spleen after cutaneous inoculation, despite a significant lymph node parasite burden (43). In murine L. chagasi infection, after cutaneous inoculation of 107 promastigotes, parasites were detectable in spleen and liver at the same level as occurred following a low-dose (104) intravenous challenge (D. McMahon-Pratt, L. Soong, M. Colmenares, K. Pestana-Goldsmith, and L. Munstermann, Abstr. 48th Ann. Meet. Am. Soc. Trop. Med. Hyg. abstr. 552, 1999). Similarly, Dunn and North (16) observed a 1,000-fold lymph node barrier effect in listeriosis. In murine L. major infection, the integrity of the lymph node barrier depends on the mouse strain and inoculum size. In the BALB/c mouse, parasites disseminated rapidly from the site of subcutaneous footpad injection to the lymph nodes and visceral organs. However, in the C57BL/6, CBA/J, and C3H/HeJ mouse strains, the parasites remained localized in the footpad and local lymph node (39). Our study was performed with BALB/c mice; it is unclear if other mouse strains would show different results in this model of parasite visceralization. In the BALB/c mouse, cutaneous inoculation of small numbers of L. major parasites (103) did not lead to dissemination, but with inoculation of 106 organisms, the lymph node barrier was breached (67). Several studies have shown that protein malnutrition or zinc

FIG. 2. (A) NO production by resident peritoneal cells from mice in the four diet groups A to D. Resident peritoneal cells were cultured in medium alone or with IFN-␥ (20 U/ml) and LPS (1 ␮g/ml) for 24 h. Supernatants were tested for total nitrate and nitrite by the Greiss reaction. The results are expressed as the mean and standard error of the mean of the nitrate/nitrite concentration from the stimulated and unstimulated cultures (n ⫽ four mice per group). Differences in mean

values were determined by Student’s t test. (B) Splenic and hepatic NOS enzyme activity in control (diet A, 17% protein) and malnourished mice (diet C, 3% protein) 3 days after infection with L. donovani. NOS2 and cNOS activity in liver and spleen tissue homogenates was determined as described in Materials and Methods. Data are expressed as the mean ⫾ and standard error of the mean of enzyme activity per gram of tissue. Differences in mean values were determined by Student’s t test. (C) Parasite burden-adjusted NOS2 activity. The splenic and hepatic NOS2 activity was adjusted to the parasite burden per milligram of the same tissue. Data are expressed as the mean and standard error of the mean (n ⫽ 11 mice per group), and significant differences between the control and malnourished mice were determined by Student’s t test.

VOL. 69, 2001

MALNUTRITION AND VISCERAL LEISHMANIASIS

TABLE 3. Levels of mediators produced ex vivo by lymph node cells from L. donovani-infected mice on diet A and diet C Mediator (culture period)

IL-10 (12 h) IL-10 (24 h) IL-12 (12 h) IFN-␥ (24 h) PGE2 (24 h) a b

Level of mediator (pg/ml) with: Diet A

a

249 ⫾ 19.0 440 ⫾ 45 79.9 ⫾ 29.6 3,370 ⫾ 550 180 ⫾ 36

Diet C

b

99.0 ⫾ 19.1 147 ⫾ 35 28.8 ⫾ 9.6 2,350 ⫾ 610 498 ⫾ 81

P value

ⱕ0.001 ⱕ0.001 0.13 0.28 0.005

Mean ⫾ standard error of the mean of 9 to 11 mice. Mean ⫾ standard error of the mean of 8 to 10 mice.

deficiency leads to increased PGE2 production (56, 63, 73). In our study, levels of PGE2 produced by lymph node cells correlated with the total visceral parasite burden and the %PNE. The known biological activities of PGE2 suggest several mechanisms by which it may promote the dissemination of parasitized leukocytes from the lymph node: it down-regulates adhesion molecule expression (53), increases macrophage release of matrix metalloproteinases (68), stimulates leukocyte chemokinesis (71), and antagonizes the activity of macrophage migration inhibitory factor (38). These factors have been proposed as contributors to tumor cell metastasis (64), and PGE2 production was demonstrated to increase metastases in several tumor models (22). PGE2 has been previously implicated in the immunopathogenesis of leishmaniasis. Macrophages infected with L. donovani release increased quantities of PGE2 in a rapid and sustained manner (57). Splenocytes from susceptible BALB/c mice infected with L. major produced more PGE2 than did splenocytes from resistant mice (66). De Freitas et al. reported that blocking PGE2 synthesis with indomethacin during L. major infection decreased lesion parasite burden and increased NO and IFN-␥ production (12). In an L. donovani infection model, indomethacin given on day 1 decreased liver parasite burdens on days 14 and 28; initiation of indomethacin therapy 2 weeks after infection had little effect (47). A role for prostanoids in the development of metastatic lesions in L. major infection has also been suggested (62). Our study supports the general observation that increased PGE2 production promotes the progression of Leishmania infection and that the effect occurs early in the course of infection. The malnourished mice produced less IL-10 in response to infection, and the level of lymph node IL-10 production inversely correlated with %PNE and total visceral parasite burden, suggesting that this cytokine decreases early visceralization. The effects of malnutrition on IL-10 production have not TABLE 4. Levels of cytokines produced ex vivo by splenocytes from L. donovani-infected mice on diet A and diet C Cytokine

IL-10 IL-12 TNF-␣ TGF-␤1 a b

Level of cytokinea (pg/ml) with: Diet Ab

Diet Cb

102 ⫾ 51 97.9 ⫾ 32.1 229 ⫾ 10.1 141 ⫾ 45

34.9 ⫾ 18 96.3 ⫾ 28.8 175 ⫾ 10.1 106 ⫾ 41

Measured in supernatants after a 24-h culture. Mean ⫾ standard error of the mean of 11 mice per group.

P value

0.23 0.97 0.001 0.6

4715

TABLE 5. Correlation of log total visceral parasite burden and %PNE with mediator levels in the lymph node supernatants Parasite burden or %PNE

Log tvPBa Log tvPB Log tvPB Log tvPB Log tvPB Total liver parasite burden %PNE %PNE %PNE

Mediator

PGE2 IL-10 (12 IL-10 (24 IL-12 IFN-␥ PGE2 PGE2 IL-10 (12 IL-10 (24

h) h)

h) h)

R

R2

P value

0.65 ⫺0.54 ⫺0.47 ⫺0.3 ⫺0.05 0.65 0.69 ⫺0.47 ⫺0.47

0.42 0.35 0.22 0.090 0.003 0.42 0.47 0.22 0.22

0.007 0.007 0.04 0.23 0.81 0.007 0.002 0.04 0.03

a Log-transformed total visceral parasite burden (total parasite burden of liver and spleen).

been previously described. Classically, IL-10 is associated with deactivation of macrophages, NK cells, and Th1 cells (46). In L. major infection, early IL-10 production by draining lymph node cells is higher in the susceptible BALB/c mice than in healer mice; however, IL-10 did not play a role in Th-cell differentiation (10). Recent studies indicate that IL-10 may be an early proinflammatory mediator. In several macrophage systems, IL-10 enhanced NO production (32, 52). Furthermore, IL-10 augments monocyte phagocytosis, maturation, oxidative burst, and expression of ␤2-integrins and intercellular cell adhesion molecule 1 (5, 6, 27) and inhibits tumor metastasis (72). The high lymph node PGE2 levels, coupled with low IL-10 levels, observed in the malnourished mice are consistent with the known coregulation of these mediators (45). We observed decreased NO production by resident peritoneal cells and decreased hepatic and splenic NOS2 activity in the malnourished mice. The equivalent levels of cNOS in the liver and spleen of mice in groups A and C suggest that the deficit in NO production does not arise from a dietary arginine deficit, because this would affect both NOS isoforms. Other investigators have noted deficits in NO production during malnutrition (7, 15, 28), and the excessive production of PGE2 observed in malnutrition may be one mechanism of NOS2 down-regulation (25). NO is a principal microbicidal molecule in murine models of leishmaniasis, and a 50% decrease in macrophage NO production abolishes leishmanicidal activity (42). Nevertheless, several considerations suggest that the increased early visceral parasite burden (⬍3 days postinfection) observed in the malnourished mice is not primarily due to decreased NO-dependent parasite killing. First, in the tissue exposed to parasites for the longest period i.e., the footpad, parasite burdens were not significantly different in the two diet groups. The total parasite burdens in the lymph node were higher in the well-nourished mice, again suggesting that malnutrition did not impair parasite killing at this early time point. Furthermore, the total measured extradermal parasite burdens (lymph node, liver, and spleen) were equivalent in mice from groups A and C but the well-nourished mice retained a greater percentage of extradermal parasites within the lymph nodes. Intracellular killing of Leishmania is a slow process, requiring 24 h in a fully primed macrophage (49). Thus, the kinetics of macrophage priming and killing suggest that these are not likely to account for the large differences in visceral parasite burden observed at this early time point. The time course of

4716

ANSTEAD ET AL.

parasite replication also does not explain the much higher visceral parasite burdens observed in the malnourished mice. In nonadherent macrophages that are permissive for Leishmania growth, amastigote numbers increased only 10-fold over 8 days (50). We propose that NO may retard visceralization by a mechanism distinct from its microbicidal effect by modulating PGE2 production. The macrophage cyclooxygenase-2 (COX-2; the enzyme responsible for PGE2 production) and NOS2 systems exhibit a complex coregulation. Low concentrations of PGE2 enhance NO production, whereas high concentrations reduce it (44). In like manner, COX-2 activity is enhanced by low NO levels but inhibited by high concentrations of NO (4, 11, 24). The high PGE2 and low NO production observed in malnourished mice and the low PGE2 and high NO levels in the well-nourished mice fit this pattern of regulation. In the malnourished Leishmania-infected mouse, macrophage release of PGE2, which is rapid (57), probably precedes NOS2 induction, which is slow (21). Thus, the excess, early PGE2 release in the malnourished mice may account for the decreased NO production, and the lower NO concentration leads to an enhancement of PGE2 synthesis. Diefenbach et al. found that NOS2 knockout mice showed visceralization of L. major in the first 24 h after intradermal inoculation; in wild-type mice, parasites were found only at the site of inoculation and draining lymph nodes (14). The timescale of intracellular killing does not account for this rapid dissemination. Although PGE2 levels were not measured, a possible explanation for the increase in early visceralization is the absence of a negative regulatory effect of NO on COX-2 (4, 41). In contrast to the slow timescales of Leishmania intracellular replication and of macrophage activation and intracellular killing, the release of PGE2 and chemotactic factors and the up-regulation of adhesion molecules is rapid (31, 55, 57). Thus, the putative mechanisms that retain parasitized leukocytes in the local draining lymph node or allow their release are rapid compared to intracellular killing and parasite replication and may therefore determine early visceralization. We propose that the increased early visceral parasite burden observed in the malnourished mice is due to increased trafficking of parasitized leukocytes from the draining lymph node to the visceral organs, which is enhanced by lymph node PGE2 production and modulated by NO and IL-10. Our data suggest that visceralization of parasites from the draining lymph node may be a two-step process, because parasite dissemination to the spleen lags behind that to the liver and because the parasite burdens per milligram in the spleen are consistently lower than in the liver. Likewise, Hill observed the same sequence in disseminating cutaneous L. major infection (29). In tumor metastasis, the liver is commonly the first site of systemic dissemination after the local lymph node (64). Organ localization following parasite dissemination may depend on the specific adhesion molecules for parasitized leukocytes in the lymphoid organs (59). In conclusion, we describe a murine model of polynutrient malnutrition and derive a scale of murine malnutrition based on WA. WA measures in mice are biologically relevant indices of malnutrition because (i) there was greater L. donovani visceralization as WA decreased, (ii) there were progressively decreasing levels of NO production as WA decreased, and

INFECT. IMMUN.

(iii) the total lymph node parasite burden decreased as WA decreased (i.e., there was a higher degree of parasite nodal escape). Furthermore, the early visceralization observed in malnutrition is not likely to be due to defective microbicidal mechanisms in the footpad or lymph node but results from a failure of the draining lymph node to act as a barrier against dissemination. Parasite visceralization in the malnourished host is associated with increased PGE2 production and decreased IL-10 levels, conditions previously demonstrated to favor tumor cell metastasis and loss of leukocyte adhesion. The infected malnourished mice displayed decreased NOS2 activity, which probably served to enhance PGE2 synthesis. Further investigations of the role of these inflammatory mediators in the malnutrition-related failure of lymph node barrier function and the visceralization of L. donovani are in progress. ACKNOWLEDGMENTS This work was supported by funding from the U.S. Department of Veterans Affairs (P.C.M. and G.M.A.), the National Foundation for Infectious Diseases (G.M.A.), and the American Heart Association, Texas Affiliate (B.C). We thank Hector A. Flores and Clinton K. Murray for technical assistance and Gabriel Fernandes, Sunil K. Ahuja, Seema S. Ahuja, Robert A. Clark, and Marlon P. Quin ˜ones for helpful discussions. REFERENCES 1. Actor, P. 1960. Protein and vitamin intake and visceral leishmaniasis in the mouse. Exp. Parasitol. 10:1–20. 2. Badaro, R., T. C. Jones, R. Lorenco, B. J. Cerf, D. Sampaio, E. M. Carvalho, H. Rocha, R. Teixeira, and W. D. Johnson, Jr. 1986. A prospective study of visceral leishmaniasis in an endemic area of Brazil. J. Infect. Dis. 154:639– 649. 3. Berkowitz, F. E. 1992. Infections in children with severe protein-energy malnutrition. Pediatr. Infect. Dis. J. 11:750–759. 4. Brunn, G., C. Hey, I. Wessler, and K. Racke. 1997. Endogenous nitric oxide inhibits leukotriene B4 release from rat alveolar macrophages. Eur. J. Pharmacol. 326:53–60. 5. Buchwald, U. K., H. F. Geerdes-Fenge, J. Vockler, S. Ziege, and H. Lode. 1999. Interleukin-10: effects on phagocytosis and adhesion molecule expression of granulocytes and monocytes in a comparison with prednisolone. Eur. J. Med. Res. 4:85–94. 6. Calzada-Wack, J. C., M. Frankenberger, and H. W. Ziegler-Heitbrock. 1996. Interleukin-10 drives human monocytes to CD16 positive macrophages. J. Inflamm. 46:78–85. 7. Chan, J., Y. Tian, K. E. Tanaka, M. S. Tsang, K. Yu, P. Salgame, D. Carroll, Y. Kress, R. Teitelbaum, and B. R. Bloom. 1996. Effects of protein calorie malnutrition on tuberculosis in mice. Proc. Natl. Acad. Sci. USA 93:14857– 14861. 8. Chandrasekar, B., P. C. Melby, D. A. Troyer, and G. L. Freeman. 2000. Differential regulation of nitric oxide synthase isoforms in experimental acute chagasic cardiomyopathy. Clin. Exp. Immunol. 121:112–119. 9. Chang, K. P., G. Chaudhuri, and D. Fong. 1990. Molecular determinants of Leishmania virulence. Annu. Rev. Microbiol. 44:499–529. 10. Chatelain, R., S. Mauze, and R. L. Coffman. 1999. Experimental Leishmania major infection in mice: role of IL-10. Parasite Immunol. 21:211–218. 11. Clancy, R., B. Varenika, W. Huang, L. Ballou, M. Attur, A. R. Amin, and S. B. Abramson. 2000. Nitric oxide synthase/COX cross-talk: nitric oxide activates COX-1 but inhibits COX-2-derived prostaglandin production. J. Immunol. 165:1582–1587. 12. De Freitas, L. A., L. M. Mbow, M. Estay, J. A. Bleyenberg, and R. G. Titus. 1999. Indomethacin treatment slows disease progression and enhances a Th1 response in susceptible BALB/c mice infected with Leishmania major. Parasite Immunol. 21:273–277. 13. de Onis, M., C. Monteiro, J. Akre, and G. Glugston. 1993. The worldwide magnitude of protein-energy malnutrition: an overview from the WHO Global Database on Child Growth. Bull W. H. O. 71:703–712. 14. Diefenbach, A., H. Schindler, M. Rollinghoff, W. M. Yokoyama, and C. Bogdan. 1999. Requirement for type 2 NO synthase for IL-12 signaling in innate immunity. Science. 284:951–955. (Erratum, 284:1776). 15. Dong, W., M. K. Selgrade, I. M. Gilmour, R. W. Lange, P. Park, M. I. Luster, and F. W. Kari. 1998. Altered alveolar macrophage function in calorierestricted rats. Am. J. Respir. Cell Mol. Biol. 19:462–469. 16. Dunn, P. L., and R. J. North. 1991. Early gamma interferon production by natural killer cells is important in defense against murine listeriosis. Infect. Immun. 59:2892–2900.

VOL. 69, 2001 17. Engzell, U., C. Rubio, B. Tjernberg, and J. Zajicek. 1968. The lymph node barrier against Vx2 cancer cells before, during and after lymphography. A preliminary report of experiments on rabbits. Eur. J. Cancer 4:305–312. 18. Fearon, D. T., and R. M. Locksley. 1996. The instructive role of innate immunity in the acquired immune response. Science 272:50–53. 19. Filteau, S. M., and B. Woodward. 1982. The effect of severe protein deficiency on serum zinc concentration of mice fed a requirement level or a very high level of dietary zinc. J. Nutr. 112:1974–1977. 20. Fowell, D. J., A. E. Wakil, and R. M. Locksley. 1995. Interleukin-12 in murine leishmaniasis—match, flame or fuel? Res. Immunol. 146:566–575. 21. Fujihara, M., N. Ito, J. L. Pace, Y. Watanabe, S. W. Russell, and T. Suzuki. 1994. Role of endogenous interferon-beta in lipopolysaccharide-triggered activation of the inducible nitric-oxide synthase gene in a mouse macrophage cell line, J774. J. Biol. Chem. 269:12773–12778. 22. Fulton, A. M., S. Z. Zhang, and Y. C. Chong. 1991. Role of the prostaglandin E2 receptor in mammary tumor metastasis. Cancer Res. 51:2047–2050. 23. Gomez, F., R. R. Galvan, S. Frerenk, J. C. Munoz, R. Chavez, and J. Vasquez. 1956. Mortality in second and third degree malnutrition. J. Trop. Pediatr. 2:77–83. 24. Habib, A., C. Bernard, M. Lebret, C. Creminon, B. Esposito, A. Tedgui, and J. Maclouf. 1997. Regulation of the expression of cyclooxygenase-2 by nitric oxide in rat peritoneal macrophages. J. Immunol. 158:3845–3851. 25. Harbrecht, B. G., Y. M. Kim, E. A. Wirant, R. L. Simmons, and T. R. Billiar. 1997. Timing of prostaglandin exposure is critical for the inhibition of LPSor IFN-gamma-induced macrophage NO synthesis by PGE2. J. Leukoc. Biol. 61:712–720. 26. Harrison, L. H., T. G. Naidu, J. S. Drew, J. E. de Alencar, and R. D. Pearson. 1986. Reciprocal relationships between undernutrition and the parasitic disease visceral leishmaniasis. Rev. Infect. Dis. 8:447–453. 27. Hashimoto, S., M. Yamada, K. Motoyoshi, and K. S. Akagawa. 1997. Enhancement of macrophage colony-stimulating factor-induced growth and differentiation of human monocytes by interleukin-10. Blood 89:315–321. 28. Hill, A. D., H. Naama, J. Shou, S. E. Calvano, and J. M. Daly. 1995. Antimicrobial effects of granulocyte-macrophage colony-stimulating factor in protein-energy malnutrition. Arch. Surg. 130:1273–1277, 1278. 29. Hill, J. O. 1986. Pathophysiology of experimental leishmaniasis: pattern of development of metastatic disease in the susceptible host. Infect. Immun. 52: 364–369. 30. Howard, M. K., G. Sayers, and M. A. Miles. 1987. Leishmania donovani metacyclic promastigotes: transformation in vitro, lectin agglutination, complement resistance, and infectivity. Exp. Parasitol. 64:147–156. 31. Issekutz, A. C., and T. B. Issekutz. 1993. Quantitation and kinetics of blood monocyte migration to acute inflammatory reactions, and IL-1 alpha, tumor necrosis factor-alpha, and IFN-gamma. J Immunol. 151:2105–2115. 32. Jacobs, F., D. Chaussabel, C. Truyens, V. Leclerq, Y. Carlier, M. Goldman, and B. Vray. 1998. IL-10 up-regulates nitric oxide (NO) synthesis by lipopolysaccharide (LPS)-activated macrophages: improved control of Trypanosoma cruzi infection. Clin. Exp. Immunol. 113:59–64. 33. Jelliffe, D. B., and E. F. P. Jelliffe. 1978. Human milk in the modern world: psychosocial, nutritional, and economic significance, p. 241–299. Oxford University Press, Oxford, United Kingdom. 34. Jones, G. E. 2000. Cellular signaling in macrophage migration and chemotaxis. J. Leukoc. Biol. 68:593–602. 35. Karulin, A. Y., M. D. Hesse, M. Tary-Lehmann, and P. V. Lehmann. 2000. Single-cytokine-producing CD4 memory cells predominate in type 1 and type 2 immunity. J. Immunol. 164:1862–1872. 36. Kielmann, A. A., and C. McCord. 1978. Weight-for-age as an index of risk of death in children. Lancet i:1247–1250. 37. Knapka, J. J., K. P. Smith, and F. J. Judge. 1977. Effect of crude fat and crude protein on reproduction and weaning growth in four strains of inbred mice. J. Nutr. 107:61–71. 38. Koopman, W. J., M. H. Gillis, and J. R. David. 1973. Prevention of MIF activity by agents known to increase cellular cyclic AMP. J. Immunol. 110: 1609–1614. 39. Laskay, T., A. Diefenbach, M. Rollinghoff, and W. Solbach. 1995. Early parasite containment is decisive for resistance to Leishmania major infection. Eur. J. Immunol. 25:2220–2227. (Erratum, 25:2726.) 40. Lehmann, J., K. H. Enssle, I. Lehmann, A. Emmendorfer, and M. L. Lohmann-Matthes. 2000. The capacity to produce IFN-gamma rather than the presence of interleukin-4 determines the resistance and the degree of susceptibility to Leishmania donovani infection in mice. J. Interferon Cytokine Res. 20:63–77. 41. Levi, G., L. Minghetti, and F. Aloisi. 1998. Regulation of prostanoid synthesis in microglial cells and effects of prostaglandin E2 on microglial functions. Biochimie 80:899–904. 42. Liew, F. Y., S. Millott, C. Parkinson, R. M. Palmer, and S. Moncada. 1990. Macrophage killing of Leishmania parasite in vivo is mediated by nitric oxide from L-arginine. J. Immunol. 144:4794–4797. 43. Melby, P. C., Y. Z. Yang, J. Cheng, and W. Zhao. 1998. Regional differences in the cellular immune response to experimental cutaneous or visceral infection with Leishmania donovani. Infect. Immun. 66:18–27. 44. Milano, S., F. Arcoleo, M. Dieli, R. D’Agostino, P. D’Agostino, G. De

MALNUTRITION AND VISCERAL LEISHMANIASIS

45. 46. 47.

48.

49.

50. 51. 52.

53.

54. 55. 56. 57.

58. 59. 60.

61. 62. 63.

64. 65. 66. 67. 68. 69. 70.

4717

Nucci, and E. Cillari. 1995. Prostaglandin E2 regulates inducible nitric oxide synthase in the murine macrophage cell line J774. Prostaglandins 49:105–115. Minghetti, L., E. Polazzi, A. Nicolini, and G. Levi. 1998. Opposite regulation of prostaglandin E2 synthesis by transforming growth factor-beta 1 and interleukin 10 in activated microglial cultures. J. Neuroimmunol. 82:31–39. Moore, K. W., A. O’Garra, R. de Waal Malefyt, P. Vieira, and T. R. Mosmann. 1993. Interleukin-10. Annu. Rev. Immunol. 11:165–190. Murray, H. W., S. M. Carriero, and D. M. Donelly. 1986. Presence of a macrophage-mediated suppressor cell mechanism during cell-mediated immune response in experimental visceral leishmaniasis. Infect. Immun. 54: 487–493. Murray, H. W., J. J. Stern, K. Welte, B. Y. Rubin, S. M. Carriero, and C. F. Nathan. 1987. Experimental visceral leishmaniasis: production of interleukin 2 and interferon-gamma, tissue immune reaction, and response to treatment with interleukin 2 and interferon-gamma. J Immunol. 138:2290–2297. Nacy, C. A., and C. L. Diggs. 1981. Intracellular replication of Leishmania tropica in mouse peritoneal macrophages: comparison of amastigote replication in adherent and nonadherent macrophages. Infect. Immun. 34:310– 313. Nacy, C. A., M. S. Meltzer, E. J. Leonard, and D. J. Wyler. 1981. Intracellular replication and lymphokine-induced destruction of Leishmania tropica in C3H/HeN mouse macrophages. J. Immunol. 127:2381–2386. National Research Council Subcommittee on Laboratory Animal Nutrition. 1995. Nutrient requirements of laboratory animals, 4th ed. p. 46. National Academy Press, Washington, D.C. Nemoto, Y., T. Otsuka, H. Niiro, K. Izuhara, K. Yamaoka, H. Nakashima, and Y. Niho. 1999. Differential effects of interleukin-4 and interleukin-10 on nitric oxide production by murine macrophages. Inflamm. Res. 48: 643–650. Noguchi, K., K. Iwasaki, M. Shitashige, S. Murato, and I. Ishikawa. 1999. Prostaglandin E2 downregulates interferon-gamma-induced intercellular adhesion molecule-1 expression via EP2 receptors in human gingival fibroblasts. Inflammation 23:481–493. Panemangalore, M., and F. N. Bebe. 1996. Zinc diets: deficiency and excess, p. 190–213. In R. R. Watson (ed.). Trace elements in laboratory rodents. CRC Press, Inc., Boca Raton, Fla. Pober, J. S., and R. S. Cotran. 1990. Cytokines and endothelial cell biology. Physiol. Rev. 70:427–451. Redmond, H. P., J. Shou, C. J. Kelly, S. Schreiber, E. Miller, P. Leon, and J. M. Daly. 1991. Immunosuppressive mechanisms in protein-calorie malnutrition. Surgery 110:311–317. Reiner, N. E., and C. J. Malemud. 1985. Arachidonic acid metabolism by murine peritoneal macrophages infected with Leishmania donovani:in vitro evidence for parasite-induced alterations in cyclooxygenase and lipoxygenase pathways. J. Immunol. 134:556–563. Revillard, J. P., and G. Cozon. 1990. Experimental models and mechanisms of immune deficiencies of nutritional origin. Food Addit. Contam. 7:S82– S86. Ruoslahti, E., and D. Rajotte. 2000. An address system in the vasculature of normal tissues and tumors. Annu. Rev. Immunol. 18:813–827. Sacks, D., and P. Melby. 1998. Animal models for the analysis of immune responses to leishmaniasis, p. 19.2.1–19.2.20. In J. Coligan, A. Kruisbeek, D. Margulies, E. Shevach, and W. Strober (ed.), Current protocols in immunology. John Wiley & Sons, Inc., New York, N.Y. Sandstead, H. H. 1994. Understanding zinc: recent observations and interpretations. J. Lab. Clin. Med. 124:322–327. Scott, P. A., and J. P. Farrell. 1981. Experimental cutaneous leishmaniasis. I. Nonspecific immunodepression in BALB/c mice infected with Leishmania tropica. J. Immunol. 127:2395–2400. Skerrett, S. J., W. R. Henderson, and T. R. Martin. 1990. Alveolar macrophage function in rats with severe protein calorie malnutrition. Arachidonic acid metabolism, cytokine release, and antimicrobial activity. J. Immunol. 144:1052–1061. Sleeman, J. P. 2000. The lymph node as a bridgehead in the metastatic dissemination of tumors. Recent Results Cancer Res. 157:55–81. Smelt, S. C., C. R. Engwerda, M. McCrossen, and P. M. Kaye. 1997. Destruction of follicular dendritic cells during chronic visceral leishmaniasis. J. Immunol. 158:3813–3821. Soares, M. B., J. R. David, and R. G. Titus. 1997. An in vitro model for infection with Leishmania major that mimics the immune response in mice. Infect. Immun. 65:2837–2845. Solbach, W., and T. Laskay. 2000. The host response to Leishmania infection. Adv. Immunol. 74:275–317. Tsujii, M., S. Kawano, and R. N. DuBois. 1997. Cyclooxygenase-2 expression in human colon cancer cells increases metastatic potential. Proc. Natl. Acad. Sci. USA 94:3336–3340. Woodward, B. 1998. Protein, calories, and immune defenses. Nutr. Rev. 56: S84–S92. Woodward, B. D., and R. G. Miller. 1991. Depression of thymus-dependent immunity in wasting protein-energy malnutrition does not depend on an

4718

ANSTEAD ET AL.

altered ratio of helper (CD4⫹) to suppressor (CD8⫹) T cells or on a disproportionately large atrophy of the T-cell relative to the B-cell pool. Am. J. Clin. Nutr. 53:1329–1335. 71. Zeng, L., S. An, and E. J. Goetzl. 1996. Selective regulation of RNK-16 cell matrix metalloproteinases by the EP4 subtype of prostaglandin E2 receptor. Biochemistry 35:7159–7164. 72. Zheng, L. M., D. M. Ojcius, F. Garaud, C. Roth, E. Maxwell, Z. Li, H. Rong,

Editor: J. D. Clements

INFECT. IMMUN. J. Chen, X. Y. Wang, J. J. Catino, and I. King. 1996. Interleukin-10 inhibits tumor metastasis through an NK cell-dependent mechanism. J. Exp. Med. 184:579–584. 73. Zijlstra, R. T., B. A. McCracken, J. Odle, S. M. Donovan, H. B. Gelberg, B. W. Petschow, F. A. Zuckermann, and H. R. Gaskins. 1999. Malnutrition modifies pig small intestinal inflammatory responses to rotavirus. J. Nutr. 129:838–843.