INTRODUCTION
Clearance of Heligmosomoides bakeri following a challenge infection in well-nourished mice is understood to involve both parasite-specific immunological responses (Urban et al. Reference Urban, Katona and Finkelman1991a, Reference Urban, Katona, Paul and Finkelmanb) and inflammatory responses in the gut (Shea-Donohue et al. Reference Shea-Donohue, Sullivan, Finkelman, Madden, Morris, Goldhill, Pineiro-Carrero and Urban2001), orchestrated primarily by the Th2 cytokines IL-4 and IL-13 (Urban et al. Reference Urban, Katona, Paul and Finkelman1991b; Gause et al. Reference Gause, Urban and Stadecker2003). The early stage of infection (L4) is relatively protected while it is embedded in the serosal musculature although its development and the timing of emergence into the gastrointestinal (GI) lumen are delayed during a challenge infection (Ey, Reference Ey1988). It is only once the adult parasites are in the lumen, however, that effector responses contribute to worm expulsion. Enhanced intestinal epithelial cell ion flux (Shea-Donohue et al. Reference Shea-Donohue, Sullivan, Finkelman, Madden, Morris, Goldhill, Pineiro-Carrero and Urban2001), increased mucosal permeability (Madden et al. Reference Madden, Whitman, Sullivan, Gause, Urban, Katona, Finkelman and Shea-Donohue2002) and increased intestinal smooth muscle contractility (Zhao et al. Reference Zhao, McDermott, Urban, Gause, Madden, Yeung, Morris, Finkelman and Shea-Donohue2003), together with Th2 immunological responses (Gause et al. Reference Gause, Urban and Stadecker2003) are believed to induce expulsion of H. bakeri from the GI tract. Parasite clearance presumably depends, therefore, on initial priming of the immune response during the first exposure to the parasite and rapid development of both immunological and inflammatory effector responses during challenge so that both are fully functional when the adults emerge into the GI lumen.
Protein energy malnutrition (PEM) often coexists with GI nematode infections (Koski and Scott, Reference Koski and Scott2001) because infections cause malnutrition and PEM increases susceptibility to infection (Scrimshaw and San Giovanni, Reference Scrimshaw and San Giovanni1997; Koski and Scott, Reference Koski and Scott2001). Immunity is considered to be a crucial causal link as both human and laboratory studies have demonstrated that PEM has wide-spread effects on both systemic and gut-associated lymphoid tissues, including reduced homing of mucosal lymphocytes to the gut (McDermott et al. Reference McDermott, Mark, Befus, Baliga and Suskind1982), reduced activity of natural killer cells (Ingram et al. Reference Ingram, Crouch, Douez, Croy and Woodward1995), reduced release of secretory IgA (Ha and Woodward, Reference Ha and Woodward1997), reduced cytokine production by both Th1 (Doherty et al. Reference Doherty, Golden, Remick and Griffin1994) and Th2 cells (Ing et al. Reference Ing, Su, Scott and Koski2000) and reduced numbers of memory cells (Najera et al. Reference Najera, Gonzalez, Toledo, Lopez, Cortes, Betancourt and Ortiz2001). At the level of the gut, protein deficiency alone induces villus and mesenteric lymph node (MLN) atrophy and crypt hypoplasia (Bell et al. Reference Bell, Hazell and Price1976; Syme, Reference Syme1982; Woodward and Miller, Reference Woodward and Miller1991; Deitch et al. Reference Deitch, Xu, Qi, Specian and Berg1992), decreases numbers of B lymphocytes (Bell et al. Reference Bell, Hazell and Price1976) and CD4+ T cells (Woodward et al. Reference Woodward, Hillyer and Hunt1999), reduces mast cell and goblet cell responses (Wells, Reference Wells1962, Reference Wells1963; Madi et al. Reference Madi, Jervis, Anderson and Zimmerman1970), increases permeability of the gut mucosal barrier (Cummins et al. Reference Cummins, Kenny, Duncombe, Bolin and Davis1987a, Reference Cummins, Duncombe, Bolin, Davis and Yongb; Deitch et al. Reference Deitch, Xu, Qi, Specian and Berg1992), and induces oxidative stress by increasing catalase activity and the concentration of thiobarbituric acid-reacting substances (Darmon, Reference Darmon1993) leading to intestinal dysfunction. Importantly, protein deficiency also prolongs survival of Heligmosomoides bakeri (=Heligmosomoides polygyrus, Nematospiroides dubius) (Slater and Keymer, Reference Slater and Keymer1988; Boulay et al. Reference Boulay, Scott, Conly, Stevenson and Koski1998; Ing et al. Reference Ing, Su, Scott and Koski2000), a murine nematode commonly used as a model for human and livestock GI nematode infections.
We have previously demonstrated (Ing et al. Reference Ing, Su, Scott and Koski2000) that protein deficiency blocks the protective response to H. bakeri, but how quickly protective responses are restored following protein re-feeding was not described. Protein re-feeding of uninfected rats for only 3 days has been shown to replete body weight, serum albumin, insulin-like growth factor 1, leucine concentrations and colon structure (Qu et al. Reference Qu, Ling, Tahan, Sierra, Onderdonk and Bistrian1996). Re-feeding also elevates crypt cell mitotic index in the mouse duodenum (Acra et al. Reference Acra, Bulus, Bogatcheva, Coffey and Barnard1998), rapidly restores non-specific inflammatory effectors in mice (Ikeda et al. Reference Ikeda, Saito, Kazuhiko, Inoue, Han, Furukawa, Matsuda and Hidemura2001) and activity of intestinal enzymes in rhesus monkeys (Rana et al. Reference Rana, Gupta, Malik, Katyal and Mehta1995, Reference Rana, Gupta, Katyal and Singh2003). However, the rapidity of restoration of immune responses to fungal infection has been shown to depend on the relative timing of infection and re-feeding (Oarada et al. Reference Oarada, Nikawa and Kurita2002).
The objective of the present study was to determine whether the timing of protein re-feeding influenced restoration of nutritional status and functional protection against the GI nematode, H. bakeri, in protein-deficient mice. We hypothesized that the presence of infection would prolong restoration of nutritional status. We further hypothesized that functional protection during the challenge infection might be impaired if protein deficiency extended throughout the period of primary infection and into the challenge infection, but that functional protection would be restored if re-feeding occurred prior to the challenge infection.
MATERIALS AND METHODS
Experimental design
To establish if protein re-feeding restored functional protective immunity against H. bakeri, nutritional, parasitological and immunological indicators measured in control mice, fed either a protein sufficient (PS) diet or a protein-deficient (PD) diet throughout the study, were compared with those of 4 groups of mice fed a PD diet for varying periods of time before they were transferred to a PS diet. Re-feeding (RF) commenced on the day of primary infection (RF0Pr), 10 days after the primary infection (RF10Pr), on the day of challenge infection (RF0Ch), and 7 days after challenge infection (RF7Ch).
Diets and mice
Two egg-white based semi-purified powdered diets were formulated (Table 1). In the 24% PS diet, energy from protein made up 24% of total calorie intake, an amount sufficient for maintenance of normal physiology, metabolism and growth of mice (NRC, 1995). In contrast, the PD diet contained 3% protein; previous experiments in our lab have shown that mice fed this diet have reduced body weight gain, prolonged H. bakeri survival and suppressed Th2 immune responses (Boulay et al. Reference Boulay, Scott, Conly, Stevenson and Koski1998; Ing et al. Reference Ing, Su, Scott and Koski2000). Both diets were identical for all nutrients except the concentrations of protein and carbohydrate, the latter of which was adjusted to ensure that both diets were isocaloric. The diet formulation was based on National Research Council mouse requirements (NRC, 1995) and ensured that a 30% reduction in food intake due to infection or protein deficiency would not generate other nutrient deficiencies.
Table 1. Composition of experimental diets (g/kg diet)
1 Ovalbumin (Dried Egg White Solids Type P-110), ICN Biochemicals Canada Ltd, Montreal, PQ.
2 ICN Biochemicals.
3 Vitamin mix (mg/kg diet): vitamin A acetate 8·0; cholecalciferol 0·025; α-tocopheryl acetate 60; menadione 1·15; biotin 3·6; folic acid 2·0; nicotinic acid 50·0; d-calcium pantothenate 32·0; riboflavin 24·0; thiamin hydrochloride 16·0; pyridoxine hydrochloride 4·0; cyanocobalamin 0·05; butylated hydroxytoluene 100; choline chloride 4000; alphacel 7699·175.
4 Mineral mix (mg/kg diet): CaHPO4 27·16; KHCO3 10·25; NaCl 2·54; MgSO4 4·95; CrK(SO4)2·12H2O 0·0384; CuCO3⋅Cu(OH) 0·0521; KIO3 0·001; FeSO4·7H2O 0·87115; MnCO3 0·10462; ZnCO3 0·09588; Na2SeO3 0·0004; Na2MoO4·2H2O 0·000378; KF⋅2H2O 0·0099; citric acid 0·42617.
Sixty, 3-week-old female BALB/c mice (Charles River Breeding Laboratories, Quebec, Canada) were housed individually in Nalgene cages (Fisher Scientific, Montreal, Canada) with stainless steel covers in a temperature controlled mouse room (22–25°C) with a 14 h:10 h light:dark cycle. Food and tap water were provided to the mice ad libitum. All mice were fed mouse chow (Mouse Chow 5015, Agribands Canada, Ontario, Canada) for 3 days and then acclimatized for 4 days to the PS diet provided in mouse Powder Feeders (Lab Products, Quebec, Canada) especially designed to reduce spillage of powdered diets. Then the mice were randomly assigned to one of the 6 diet groups (10 mice per group). PS and PD mice were fed their respective diets throughout the experiment, beginning 7 days before infection. All mice in the 4 re-feeding groups were fed the PD diet beginning 7 days before infection until the day of re-feeding, at which time they were fed the PS diet until the end of the study.
Infection protocol
H. bakeri maintained in outbred CD-1 mice (Charles River) were used as the source of infective third-stage larvae (L3). Mice were infected with 100±4 L3 by oral gavage, and the accuracy of the infection dose was determined by counting 10 sham doses. A standard anthelmintic-abbreviated immunizing protocol was used (Behnke and Robinson, Reference Behnke and Robinson1985) whereby the primary infection was drug-terminated with pyrantel pamoate (172 mg/kg body weight; Pfizer Canada Inc., Montreal, Canada) on days 9 and 14 after the primary infection; mice were then challenged with 100±4 L3 one week later (21 days after the primary infection). In well-nourished mice, parasites are normally expelled within 1 month post-challenge infection (p.c.i.) (Behnke and Robinson, Reference Behnke and Robinson1985). On days 16 and 26 p.c.i., 24 h egg production was measured from individual mice, using a modified McMaster technique (Scott, Reference Scott1988). Mice were killed on day 28 p.c.i. using Ketamine/Xylazine (50/5 mg/kg body weight) anaesthesia. Blood was collected from the tail vein (day 10 p.c.i.) or cardiac puncture (day 28 p.c.i.) and serum samples were stored at −20°C for later detection of total serum IgE, leptin, and alkaline phosphatases (ALP). All procedures were approved by the McGill's Macdonald Campus Animal Care Committee according to the Canadian Council on Animal Care Guidelines and Policies.
Nutritional status indicators
Body weight, food intake and feed efficiency (weight gain per day/food intake per day) were recorded every 2–3 days. For each diet group, weight gain and cumulative food intake were calculated for the 4 days prior to, and after, both primary and challenge infection, and the 4 days prior to, and after, commencement of re-feeding.
Serum leptin was assayed at day 28 p.c.i., using a mouse leptin ELISA kit (Linco Research, Missouri, USA, Cat. EZML-82K), according to the manufacturer's instructions. Briefly, 10 μl of standards and samples were added to mouse leptin microtitre plates which had been pre-coated with pre-titred capture antibodies. Then 50 μl of pre-titred anti-rodent leptin was added to each well and the 96-well plates were then incubated for 2 h. Then 100 μl of pre-titred biotinylated anti-mouse leptin antibody was added to each well, wells were incubated for 1 h, 100 μl of pre-titred streptavidin-horseradish peroxidase conjugate was added and wells were incubated for a further 30 min. The substrate 3,3′,5,5′-tetramethylbenzidine (TMB) was then added and incubated for 30 min and the reaction was stopped by adding 0·3 m hydrochloric acid. The plates were read with an EL 309 Microplate Autoreader (Bio-Tek Instruments, Inc. Winooski, Vermont) at 450 nm and 590 nm. The OD value was calculated by subtracting the reading at 590 nm from the reading at 450 nm and averaged across the duplicates for each sample.
Parasitological indicators
Worm fecundity was estimated by dividing 24 h egg release per mouse on day 26 p.c.i. by the total number of female worms on day 28 p.c.i. At necropsy, the small intestines were excised and the numbers of adult worms in the intestine were counted and sexed.
Indicators of immunity and tissue damage
Total serum IgE was measured at both days 10 and 28 p.c.i., by a two-site ELISA following the manufacturer's directions (BD PharMingen, Mississauga, Ontario). Immulon II plates (Becton Dickinson and Co., Oxnard, Ontario) were coated with rat anti-mouse IgE mAb (R35–72, BD PharMingen) and developed with biotinylated rat anti-IgE mAb (R35–92, BD PharMingen) and avidin-peroxidase (BD PharMingen). TMB was used as a substrate. After stopping the reaction with phosphoric acid, the plates were read with a Microplate Autoreader (EL309, Mandel Scientific Company Ltd, Guelph Ontario) at 450 nm. IgE concentration was computed according to an IgE standard curve with known concentrations of mouse IgE standard (C38-2, BD PharMingen) and the average of duplicate values is reported as μg/ml of serum.
Alkaline phosphatase concentrations (ALP) in serum were measured (Mulivor et al. Reference Mulivor, Boccelli and Harris1985) as an index of tissue damage. Total serum ALP was assayed at days 10 and 28 p.c.i., and serum concentrations of the intestinal and liver-kidney-bone (LBK) isoenzymes were assayed on day 28 p.c.i. All assays were done in triplicate, using the Promega AttoPhos® AP Fluorescent Substrate System (Promega; Madison, Wisconsin), according to the manufacturer's protocol. Absorbance readings were taken using the Wallac fluorimeter (Turku, Finland), set at 430–440 nm for the excitation filter and 550–560 nm for the emission filter.
Statistics
All results were expressed as mean±s.e.m. Body weight and food intake data were initially analysed by repeated measures ANOVA, and the change in weight and food intake during the 4 days before and after infection or re-feeding was subsequently analysed by one-way ANOVA. Data on parasite reproduction and worm burdens were analysed using Kruskal Wallis non-parameteric ANOVA. IgE, ALP and leptin were analysed using one-way or two-way ANOVA depending on whether the effect of time was included. Bonferroni or Scheffe's test was performed when one-way ANOVA revealed a significant main effect. Finally, regression analysis was used to determine whether there was a linear relationship between ALP or leptin and worm numbers. All data were analysed by SAS 9.01 (SAS Institute, Cary, NC) and P<0·05 was considered as significant.
RESULTS
Infection
Egg production (day 16 p.c.i.) was elevated in the PD mice, but was very low in PS mice, and also in mice from all of the re-feeding groups (Table 2). A similar result was observed for egg production at day 26 p.c.i., and for worm burdens and per capita fecundity at day 26 p.c.i. (Table 2). Together these results demonstrate that re-feeding even as late as 7 days after the challenge infection restored the functional protection that had been impaired by protein deficiency.
Table 2. Impact of re-feeding with protein sufficient diet on nutritional, parasitological and immunological outcomes (mean±s.e.)Footnote 1,2
1 PS mice fed protein sufficient diet throughout the experiment; RF0Pr mice fed protein deficient diet, then fed PS diet beginning on day 0 of primary infection; RF10Pr mice fed protein deficient diet, then fed protein sufficient diet beginning on day 10 of primary infection; RF0Ch mice fed protein deficient diet, then fed protein sufficient diet beginning on day 0 of challenge infection; RF7Ch mice fed protein deficient diet, then fed the protein sufficient diet beginning on day 7 of challenge infection; PD mice fed the protein deficient diet throughout the study.
2 Different lower case letters across a row indicate significant differences among experimental groups as revealed by post-hoc tests (P<0·05).
3 Average food intake between day of primary infection and day of re-feeding significantly higher than average food intake after protein re-feeding (P<0·001).
4 Post-challenge infection.
5 Serum IgE is significantly higher at day 10 p.c.i. than at day 28 p.c.i. (P=0·03).
Serum samples collected at days 10 and 28 p.c.i. were assayed for total IgE as a systemic index of immune response against H. bakeri. Although IgE did not differ between PS and PD mice at either time-point, suggesting that total IgE was not perturbed by PD at these times, serum IgE was higher at day 10 p.c.i., compared with day 28 p.c.i. (Table 2); also the highest IgE concentration was obtained in PD mice and the lowest was found in the RF10Pr group.
Body weight
Body weights during the pre-feeding week were similar among all diet groups. However, by day 6 post-infection (p.i.), the negative impact of protein deficiency on weight gain was evident (Fig. 1). By the end of the experiment, PD mice had lost body weight; however, mice in all re-feeding groups had regained their lost weight and had weights similar to PS mice (Fig. 1; Table 2). No catch-up growth occurred in the RF0Pr group (Fig. 2A), as body weights did not differ from those of PS mice at the time they were re-fed (Fig. 1). However, re-feeding induced rapid weight gain in RF10Pr and RF7Ch mice (Fig. 2A) and body weight was similar to that of the PS group within 4 days of re-feeding (Table 2). In contrast, in the case of the group re-fed on the same day as the challenge infection (RF0Ch), no significant improvement in weight (Fig. 2a) occurred over the 4 days following re-feeding, and full restoration of body weight did not occur until 13 days after re-feeding (Fig. 1; Table 2), indicating that the stress of larval challenge on the day of re-feeding interfered with the host's ability to benefit from the PS diet.
Fig. 1. Body weight over time in mice infected with Heligmosomoides bakeri on day 0, and again on day 21 following drug treatment on days 9 and 14. The time of primary infection, the first drug treatment and the challenge infection are indicated by the vertical lines. PS mice (- -◊- -) fed protein sufficient diet throughout the experiment; RF0Pr mice (––□––) fed protein deficient diet, then fed protein sufficient diet beginning on day 0 of primary infection; RF10Pr mice (––▲––) protein deficient diet, then fed protein sufficient diet beginning on day 10 of primary infection; RF0Ch mice (––■––) fed protein deficient diet, then fed protein sufficient diet beginning on day 0 of challenge infection; RF7Ch mice (––△––) mice fed the protein deficient diet, then fed the protein sufficient diet beginning on day 7 of challenge infection; PD mice (- -◆- -) fed the protein deficient diet throughout the study. The time of re-feeding is indicated by the square boxes.
Fig. 2. Influence of infection and re-feeding on weight gain and food intake. (A) Change in body weight over the 4 days prior to re-feeding (solid bars) and the 4 days following re-feeding (cross-hatched bars) in the 4 re-feeding groups (groups defined in legend to Fig. 1). Different lower-case letters indicate differences among all groups (P<0·05). Comparison of feed efficiency 4 days prior and after re-feeding showed the same pattern as body weight change. (B) Change in body weight over the 4 days prior to (solid bars) or after (cross-hatched bars) primary or challenge infection in mice fed either the protein sufficient (PS) and protein deficient (PD) diet. Diet effects (P=0·0009) shown by upper-case letters; differences in weight gain before and after challenge infection shown by lower-case letters (P=0·001). (C) Difference in cumulative food intake in the 4 days after primary infection compared with the 4 days prior to primary infection. Different lower-case letters indicate differences among groups (P<0·05). (D) Difference in cumulative food intake in the 4 days after challenge infection compared with the 4 days prior to challenge infection.
Food intake and feed efficiency
In order to determine whether the delayed weight gain in the RF0Ch mice may have resulted from the stress associated with the larval challenge, we compared both body weight change and feed efficiency during the 4 days prior to infection with data from the 4 days immediately after infection in both PS and PD groups. Food intake (Fig. 2D) and feed efficiency (data not shown) were significantly reduced immediately following challenge infection in all groups of mice, indicating that larval challenge induced anorexia. The challenge infection also caused PS mice to stop gaining weight, and caused significant weight loss in PD mice (Fig. 3B). Interestingly, food intake immediately after the primary infection was also reduced in PS mice, whereas food intake increased significantly in 3 of the 4 groups continuing to consume the PD diet (RF10Pr, RF0Ch, RF7Ch) (Fig. 2C). Also, the concentration of serum leptin on day 28 increased linearly with cumulative food intake during the last week of the experiment (P=0·049) although no overall significant difference was detected among the 6 groups (Table 2).
Fig. 3. Serum concentrations of total alkaline phosphatase (ALP) at day 10 (A), or day 28 p.c.i. (B) with Heligmosomoides bakeri and of intestinal ALP (C) and liver-bone-kidney (LBK) ALP (D) on day 28 p.c.i. (groups defined in legend to Fig. 1). Different lower-case letters indicate significant differences among groups (P<0·05).
Tissue damage
As general indicators of tissue damage resulting both from the infection and from the protein deficiency, sera were assayed for total ALP as well as intestinal and LBK isoenzymes. At day 10 p.c.i., total ALP of the PS mice was similar to that of all other diet groups (Fig. 3A), but was significantly higher in RF10Pr and RF0Ch than in RF7Ch and PD mice (Fig. 3A), suggesting that longer re-feeding led to higher total serum ALP. By day 28 p.c.i., total ALP was significantly higher in PD than PS mice, and total ALP was also significantly elevated in the RF0Ch mice (Fig. 3B), the group of mice where restoration of body weight had been prolonged (Fig. 1; Table 2). Significantly increased concentrations of LBK ALP and slightly higher intestinal ALP on day 28 p.c.i. underscored the elevations in total ALP seen in the PD and RF0Ch mice compared with the other re-feeding groups (Fig. 3C). We also found positive linear relationships between day 28 p.c.i. measurements of total (P=0·0045), intestinal (P=0·0001) and LBK (P=0·02) ALP and total days of consumption of the PD diet, suggesting lingering metabolic stress to specific organs.
DISCUSSION
Our experiment was designed to determine whether the benefits of re-feeding on restoration of functional protection (measured by expulsion of adults from the GI tract) required adequate protein at a critical time during either the primary or challenge infection. We chose 4 re-feeding times, 2 during the primary infection and 2 during the challenge infection. As protein status is restored within a few days of re-feeding (Qu et al. Reference Qu, Ling, Tahan, Sierra, Onderdonk and Bistrian1996), our first hypothesis that mice re-fed either on the day of primary infection (RF0Pr) or on the day immediately after drug treatment (RF10Pr) (when parasite antigen would likely still be present), would have mounted a fully competent priming response to H. bakeri was supported by our data on egg production, worm numbers and serum IgE levels. Our earliest indicators (serum IgE on day 10 p.c.i. and egg production day 16 p.c.i.) as well as all subsequent indicators (serum IgE on day 28 p.c.i., net and per capita egg production and worm numbers at day 28 p.c.i.) were all comparable to those of PS mice. Thus, protein deficiency throughout the primary infection did not reduce the ability of re-fed mice to reject a challenge infection. Although protein deficiency causes atrophy of lymphoid tissues including MLN (Deitch et al. Reference Deitch, Xu, Qi, Specian and Berg1992), interferes with germinal centre development (Bell et al. Reference Bell, Hazell and Price1976) and homing of lymphocytes to mucosal tissues (McDermott et al. Reference McDermott, Mark, Befus, Baliga and Suskind1982), sufficient antigen presentation apparently occurred during the primary infection to permit an adaptive response to challenge. In fact, we have shown that parasite-specific immune effectors are present during a primary infection in PD mice (Ing et al. Reference Ing, Su, Scott and Koski2000), albeit at significantly lower levels that those seen in the PS controls.
We had further hypothesized that the nutritional status of mice re-fed either on the day of challenge (RF0Ch) and more particularly, those re-fed 7 days later (RF7Ch), would not permit a fully operational immune response to the challenge infection, and thus that worm survival would be prolonged as it was in PD mice. This was not supported by our data. Mice in both challenge re-feeding groups were able to clear the challenge infection with the same efficiency as mice fed the PS diet throughout the study. Based on these observations, we suggest that immunological and inflammatory responses rapidly returned to normal but at an energetic cost to the nutritionally compromised animal. In well-nourished hosts, T cells release cytokines very rapidly after challenge (Garcia et al. Reference Garcia, Disanto and Stockinger1999; Rogers et al. Reference Rogers, Dubey and Swain2000) in a primed host. Protein re-feeding is also known to rapidly restore gut structure (Qu et al. Reference Qu, Ling, Tahan, Sierra, Onderdonk and Bistrian1996) and non-specific inflammatory responses (Ikeda et al. Reference Ikeda, Saito, Kazuhiko, Inoue, Han, Furukawa, Matsuda and Hidemura2001). In addition, the high proportion of quiescent CD4+ T lymphocytes found in uninfected PD mice (Woodward et al. Reference Woodward, Hillyer and Hunt1999) and malnourished children (Najera et al. Reference Najera, Gonzalez, Toledo, Lopez, Cortes, Betancourt and Ortiz2001) are presumably immediately responsive to re-feeding. Our results differ from those of Oarada et al. (Reference Oarada, Nikawa and Kurita2002) who showed that re-feeding on the day of infection effectively restored protection against fungal infection in mice, but re-feeding even one day later was less effective. It is important to note that clearance of H. bakeri cannot begin until the larval parasites have emerged into the intestinal lumen, an event that occurs after 7 days of challenge. This may explain why we detected no differences in worm numbers between mice re-fed on the day of challenge infection or 7 days after challenge infection.
The literature on the effect of protein deficiency on food intake is inconsistent. Some studies have reported increased food intake in mice fed a moderately low protein diet (Swick and Gribskov, Reference Swick and Gribskov1983; White et al. Reference White, He, Dean and Martin1994) whereas others have reported no change (Boulay et al. Reference Boulay, Scott, Conly, Stevenson and Koski1998; Ing et al. Reference Ing, Su, Scott and Koski2000) and still others found that food intake decreased (Mercer et al. Reference Mercer, Kelley, Humphries and Dunn1994). Discrepancies among studies may be related to the differences in degree of deficiency (Ing et al. Reference Ing, Su, Scott and Koski2000), other dietary components (Menaker and Navia, Reference Menaker and Navia1973; Boulay et al. Reference Boulay, Scott, Conly, Stevenson and Koski1998), or genetic background or age of animals (White et al. Reference White, Porter and Martin2000). None of these explanations, however, would account for why average daily intake of the PD diet prior to re-feeding exceeded average intake of the PS diet after re-feeding by 10–23%. Interestingly, however, we demonstrated that serum leptin was positively associated with food intake in these infected animals. Most studies have shown reduced serum leptin in PEM (Haluzik et al. Reference Haluzik, Kabrt, Nedvidkova, Svobodova, Kotrlikova and Papezova1999; Soliman et al. Reference Soliman, Elzalabany, Salama and Ansari2000; Palacio et al. Reference Palacio, Perez-Bravo, Santos, Schlesinger and Monckeberg2002) with the exception of Stapleton et al. (Reference Stapleton, Barden, McCarter, Mackrell, Freeman, Naama and Daly2003). Our data in infected mice, if anything, suggest higher leptin in PD mice than in PS mice. Elevated leptin has been reported during bacterial infection (Arnalich et al. Reference Arnalich, Lopez, Codoceo, Jim Nez, Madero and Montiel1999), during inflammation (Barbier et al. Reference Barbier, Cherbut, Aube, Blottiere and Galmiche1998; Matarese et al. Reference Matarese, Carrieri, La Cava, Perna, Sanna, De Rosa, Aufiero, Fontana and Zappacosta2005) and during the early stage of Nippostrongylus brasiliensis infection (Roberts et al. Reference Roberts, Hardie, Chappell and Mercer1999). Therefore, it is possible that adaptation to the PD diet together with prolonged infection-induced inflammation in PD mice led to increased leptin concentrations and higher food intakes in infected PD mice.
Weight loss is not normally reported in studies on H. bakeri (Slater and Keymer, Reference Slater and Keymer1986; Shi et al. Reference Shi, Scott, Stevenson and Koski1994; Kristan and Hammond, Reference Kristan and Hammond2000), perhaps because researchers focus on weight at the end of the study, rather than immediately after larval exposure. Our data clearly show a temporary cessation of weight gain at the time of larval challenge in PS mice, and actual weight loss in PD mice. We suggest that this could result from 3 factors. First, food intake was reduced in all diet groups during the 4 days after larval challenge, compared with the 4 days preceding challenge infection. Second, protein deficiency alone increases intestinal paracellular permeability (Rodriguez, Reference Rodriguez1996), and within 24 h of infection, the penetration of L3 larvae into the mucosa causes haemorrhaging and net water loss (Sukhdeo and Mettrick, Reference Sukhdeo and Mettrick1984; Kamal et al. Reference Kamal, Dehlawi, Rosa Brunet and Wakelin2002). Thus, the water loss resulting both from active penetration of L3 larvae and leakage due to protein deficiency may explain the weight loss immediately following infection. Finally, reduced weight gain may also be a consequence of the allocation of resources to the rapid immunological response triggered by larval challenge. IL-4 production is extremely high within 8 h of challenge, IL-13 first peaks at day 3 p.c.i. and concentrations of both are substantially higher than during a primary infection (Finkelman et al. Reference Finkelman, Morris, Orekhova, Mori, Donaldson, Reiner, Reilly, Schopf and Urban2000). Mucosal mast cell activation and degranulation are also induced within 3 days of larval challenge (Ing et al. Reference Ing, Su, Scott and Koski2000), even in PD mice. Therefore, we posit that the reduced weight gain observed over the 4 days following challenge infection reflects water loss and anorexia, together with preferential allocation of resources to host responses rather than to growth.
Nutritional status was restored rapidly in the RF7Ch group but much more slowly in the RF0Ch group where indicators of tissue damage (both total and intestinal ALP) were higher even on day 28 p.c.i. than those of PS mice. Our data thus indicate that protein re-feeding that begins as late as 7 days after challenge infection is able to restore functional protection measured by egg production approximately 1 week later, and by worm numbers present on day 28 p.c.i. Catch-up growth following re-feeding of protein-deficient rats normally occurs within 3 days (Qu et al. Reference Qu, Ling, Tahan, Sierra, Onderdonk and Bistrian1996), an observation consistent with our data for the RF10Ch and RF7Ch groups of mice. However, larval challenge dramatically delayed the catch-up growth of mice re-fed on the same day as larval challenge (RF0Ch mice). Compared with a primary infection, challenge infection induces more prominent changes both in the parasite-specific immune responses and in the architecture and physiology of the gut. At the local level, IL-4 expression by MLN cells remains elevated for 2 weeks (Ing et al. Reference Ing, Su, Scott and Koski2000), mucosal mastocytosis and eosinophilia is elevated as early as day 3 p.c.i. and persists at least for 28 days (Ing et al. Reference Ing, Su, Scott and Koski2000; Shea-Donohue et al. Reference Shea-Donohue, Sullivan, Finkelman, Madden, Morris, Goldhill, Pineiro-Carrero and Urban2001), and mucosal mast cell protease-1 (mMCP-1), a peripheral marker of mast cell degranulation, increases to a peak on day 14 p.c.i. (Ing et al. Reference Ing, Su, Scott and Koski2000). Intestinal glands are enlarged and an intense cellular reaction with neutropolymorphonuclear cells, plasma cells, and eosinophils is evident in the mucosal epithelium, along with an increase in number and size of goblet cells and in the amount of mucus present in the lumen of the intestine (Cywinska et al. Reference Cywinska, Czuminska and Schollenberger2004). In addition, elevated mucosal permeability of jejunal epithelial cells along with reduced Na+-linked glucose absorption and increased Cl-secretory response to prostaglandin E2 occur on day 10 p.c.i. (Shea-Donohue et al. Reference Shea-Donohue, Sullivan, Finkelman, Madden, Morris, Goldhill, Pineiro-Carrero and Urban2001). Thus, it is likely that the demands on host protein imposed by this cascade of responses delays the ability of the host to divert the additional dietary protein into body growth, and that this explains the very much longer catch-up growth period observed in the mice re-fed on the same day as larval challenge.
Concentrations of ALP have been reported in a variety of studies examining protein deficiency and re-feeding as well as helminth infection in rodents, rabbits, monkeys and humans, but no consistent pattern has emerged. Some studies report that ALP is reduced by protein deficiency (Pond et al. Reference Pond, Ellis and Schoknecht1992; Rana et al. Reference Rana, Gupta, Malik, Katyal and Mehta1995) whereas others indicate the opposite (Rana et al. Reference Rana, Gupta, Katyal and Singh2003; Sidhu et al. Reference Sidhu, Garg, Morgenstern, Vogt, Butz and Dhawan2005). Similarly, ALP is reduced in Trichostrongylus colubriformis (Stankiewicz et al. Reference Stankiewicz, Cabaj, Pernthaner, Jonas and Rabel1996) and Clonorchis sinensis (Hong et al. Reference Hong, Park, Seo, Choi, Chai and Lee1994) infection but elevated in Schistosoma mansoni (Mansour et al. Reference Mansour, Farid, Bassily, Salah and Watten1982) and H. bakeri (Boulay et al. Reference Boulay, Scott, Conly, Stevenson and Koski1998) infection. These apparently conflicting results may stem from differences in the species, the infection and/or the severity of deficiency. In the classic model, protein deficiency in humans induces a reduction in serum ALP that progresses with duration and severity of the deficiency. In mild protein deficiency in children, ALP rapidly returns to control levels, but in more severe deficiency, ALP remains depressed for a period after re-feeding, then rebounds and may increase to levels above that of normal children (Schwartz, Reference Schwartz1956). This sequence of events in humans is consistent with our mouse data. Those mice re-fed early experienced only limited dietary protein deprivation, and early re-feeding would have resulted in a rapid return to control levels by day 10 p.c.i. In contrast, both our PD mice and those re-fed later were protein deficient for a longer period of time and would have experienced a more severe deficiency that might explain the low day 10 ALP levels in our PD and RF7Ch mice. The relatively higher values in RF10Pr and RF0Ch would be consistent with a rebound ALP that is somewhat elevated following more prolonged protein deficiency. By day 28 p.c.i., ALP levels were restored to normal in all groups except PD mice and the RF0Ch group. It is easy to attribute the elevated ALP in the PD group of mice to continuing damage induced by PD (Li et al. Reference Li, Specian, Berg and Deitch1989) and infection (Boulay et al. Reference Boulay, Scott, Conly, Stevenson and Koski1998). What is more surprising is the elevated ALP on day 28 p.c.i. in the RF0Ch group, the only group of mice where re-feeding began on the same day as larval challenge, and where restoration of body weight took 3 times as long as in all other re-feeding groups. Apparently, larval challenge concurrent with re-feeding caused a substantially higher physiological stress in these animals. This is supported by the elevated LBK ALP known to be associated with hepatic involvement in resolution of tissue damage due to infection.
Based on our results, several key points emerge. We have confirmed that protein deficiency impairs the ability of mice to reject a challenge infection with H. bakeri, and have shown for the first time that re-feeding a protein sufficient diet during either a primary or challenge infection restores host clearance of this GI nematode. We have also shown that catch-up growth following re-feeding allows mice to attain normal body weights within 4 days, unless they are simultaneously exposed to H. polgyrus L3. Based on these observations together with our previous results, we suggest that even a 3% diet provides adequate protein for priming of a parasite-specific immune response. So long as a protein-sufficient diet is available for at least a week during the challenge infection, mice are able to mount the appropriate immunological and/or inflammatory responses required to reject the parasite. We suggest that the protein deficiency–protein re-feeding paradigm may be extremely useful in defining the thresholds of various effectors needed for protection.
This study was supported by an Equipe grant from the Fonds FCAR pour l'aide et le soutien à la recherche and the Natural Sciences and Engineering Council of Canada (Grants 3585 and 3623). Research at the Institute of Parasitology is supported by a Regroupement stratégique from FQRNT, a provincial funding agency.
