The Journal of Nutrition Nutrition and Disease

Disease Manifestations of Canine Distemper Virus Infection in Ferrets Are Modulated by Vitamin A Status1,2
Carey Rodeheffer,3 Veronika von Messling,4 Sylvain Milot,4 Franc xois Lepine,4 Amee R. Manges,5 3 and Brian J. Ward *
3 McGill University Health Centre Research Institute, Faculty of Medicine, Division of Infectious Diseases, Montreal General Hospital, Montreal, QC, Canada; 4Institut National de la Recherche Scientique-Institut Armand Frappier, Universite du Que bec, Laval, QC, Canada; 5Department of Epidemiology, Biostatistics and Occupational Health, McGill University, Montreal, QC, Canada

Downloaded from jn.nutrition.org at Universidade Federal de Viosa on April 3, 2009

Abstract
The measles virus (MV) causes half a million childhood deaths annually. Vitamin A supplements signicantly reduce measlesassociated mortality and morbidity. The mechanisms whereby vitamin A acts against MV are not understood and currently there is no satisfactory small animal model for MV infection. We report on the development of a ferret model to study antiviral activity of vitamin A against canine distemper virus (CDV). CDV is closely related to MV at the molecular level and distemper in ferrets mimics measles in humans. We infected vitamin A-replete (control) and vitamin A-depleted ferrets with CDV and assessed the ability of high-dose vitamin A supplements to inuence CDV disease. In control ferrets, CDV infection caused fever, rash, conjunctivitis, cough, coryza, and diarrhea. In contrast, control ferrets that were given 30 mg of vitamin A did not develop typical distemper after infection and exhibited only a mild rash. The supplement did not negatively affect ferret health and resulted in a 100% increase in serum and liver vitamin A concentrations. We also found that profound vitamin A deciency is inducible in ferrets and can be rapidly reversed upon high-dose vitamin A supplementation. Vitamin A deciency caused anorexia, diarrhea, cataracts, behavioral abnormalities, and ultimately death, with or without CDV infection. All ferrets that received vitamin A supplements, however, recovered uneventfully from CDV infection. These results replicate many aspects of the observations of vitamin A therapy in humans with measles and suggest that CDV infection in ferrets is an appropriate model for the study of the antiviral mechanism of vitamin A. J. Nutr. 137: 19161922, 2007.

Introduction
Measles virus (MV)6 is no longer endemic in North America, but remains the fth leading cause of death globally (1) resulting in an estimated 530,000 deaths annually (2). Although death due to MV itself is rare, infection is typically accompanied by immunosuppression and acquisition of secondary infections that can be fatal in 515% of cases in many regions of the developing world (3,4). The interaction between vitamin A status and MV in infected subjects is complex. In well-nourished, vitamin A-replete populations, acute measles results in a rapid decrease in circulating retinol levels that return to normal after the infection resolves (57). Isolated vitamin A deciency is a major
1 Supported by grants from the McGill University Health Centre Research Institute (to C.R.), Health Canada (to C. R.) and Canadian Institutes of Health Research (to B.J.W. and V.vM.). 2 Author disclosure: C. Rodeheffer, V. von Messling, S. Milot, F. Lepine, A. R. Manges, and B. J. Ward, no conicts of interest. 6 Abbreviations used: CDV, canine distemper virus; C, ferrets fed a control diet; C1VA, ferrets fed a control diet and supplemented with high-dose vitamin A; D, ferrets fed a vitamin Adecient diet; D1VA, ferrets fed a vitamin Adecient diet then supplemented with vitamin A and switched to the control diet; eGFP, enhanced green uorescent protein; IU, international units; MV, measles virus; PBMC, peripheral blood mononuclear cells. * To whom correspondence should be addressed. E-mail: brian.ward@mcgill.ca.

risk factor for severe measles disease and death (1). Finally, vitamin A supplements signicantly reduce measles-associated mortality and morbidity in populations both with and without endemic vitamin A deciency (5,8,9). Although the impact of vitamin A supplementation and therapy on measles can be great (10), the mechanisms that underlie these effects are not understood. To date, experiments to explore vitamin Ameasles interactions have been severely limited because clinical MV infection occurs only in humans and primates (1113). The term vitamin A is often used loosely to refer to a family of metabolically related retinoids including dietary retinol, precursors such as b-carotene, and retinoic acids involved in intracellular signaling. Dietary vitamin A is stored primarily in the liver until it is required by the tissues and is released as retinol or retinyl ester into the blood. In target tissues, retinol may be metabolized into a variety of compounds that mediate vision, fetal development, maintenance of the mucosal barrier, and immunity. Retinoic acids bind to nuclear retinoic acid receptors, which regulate gene transcription (14). To explain the action of vitamin A against measles or other viruses, it is generally proposed that vitamin A acts to potentiate adaptive immune responses or to maintain innate immune responses such as the mucosal barrier (1517). We showed that retinoids can also

1916

0022-3166/07 $8.00 2007 American Society for Nutrition. Manuscript received 24 March 2007. Initial review completed 7 May 2007. Revision accepted 24 May 2007.

directly interfere with MV replication in vitro through interaction with type-I interferon-signaling pathways (B. Ward, unpublished results). Canine distemper virus (CDV) and MV are both members of the Paramyxoviridae morbillivirus subfamily and share a high degree of genetic and functional similarity (18). CDV, like MV, is transmitted via the respiratory route and initially infects lymphocytes and monocytes, leading to systemic infection (19). In dogs and ferrets, CDV causes a disease that is highly reminiscent of measles in humans. Both are clinically dened by a maculopapular rash and fever, appearing 712 d after infection, and by at least one of the following symptoms: cough, coryza, conjunctivitis, or diarrhea (3,20,21). Similar to MV infection, CDV results in immunosuppression characterized by lymphopaenia, suppression of lymphoproliferation, loss of the delayedtype hypersensitivity response, and the acquisition of secondary infections (1,20,22). Survivors generate high-titer neutralizing antibodies that offer lifelong protection from reinfection (23,24). CDV infection of ferrets has recently been described as a tractable laboratory model to investigate Morbillivirus pathology (20,25). Ferrets may also be a useful model to investigate the interaction of vitamin A status and health. As in humans, ferret serum and liver vitamin A concentrations are sensitive to changing dietary levels of vitamin A (2628). Ferrets possess functional retinoic acid receptors and vitamin A metabolic pathways that resemble their human counterparts, although serum and liver vitamin A concentrations in ferrets are ;4- and 20-fold higher than corresponding levels in humans, respectively (2830). Ferrets have also been used to dene the molecular mechanisms of the unexpected increase in cancer risk following b-carotene supplementation in cancer prevention trials in humans (31,32). In this study, our principal goal was to examine the effects of vitamin A supplements on CDV morbidity in vitamin Areplete ferrets. We also sought to determine whether ferrets could be made vitamin A decient and to generate preliminary data on the effects of CVD infection with and without supplementation in these ferrets.

Downloaded from jn.nutrition.org at Universidade Federal de Viosa on April 3, 2009

FIGURE 1 Generation and CDV infection of vitamin A replete, deficient and supplemented ferrets. Twenty-seven ferrets were born of vitamin A-replete mothers at Harlan and Marshall Farms (1). Upon weaning at 6 wk of age, ferrets consumed either a vitamin A-replete (control, C) diet or a vitamin Adeficient diet (D) for an additional 12 wk (2). At 18 wk of age ferrets were infected with 1.5 3 104 tissue culture infectious doses of 5804P-eGFP/uN canine distemper virus; simultaneously, half the ferrets in each group received 30 mg retinyl palmitate (C1VA and D1VA). The supplemented ferrets were also switched to the control diet (3). CDV pathology was observed for 5 additional wk, or until terminal endpoints occurred, and ferrets were killed (4). Retinol concentrations were determined in serum (weekly after infection) and liver (upon termination). Three additional D ferrets were not infected, but observed for the same period (D, uninfected). (10 mg Telazol/kg; Fort Dodge Animal Health) on d 0, 7, 14, 21, and 28 from the jugular vein and divided it into uncoated, EDTA-coated, and heparin-coated tubes. Three D ferrets were not infected, but were bled at the same intervals to control for the effects of the diet and blood collection on ferret survival. All ferrets were killed on d 35 (6 3 d) by pentobarbital overdose unless killed earlier according to the dened terminal endpoints of the study. These endpoints included failure to eat for .48 h, 1520% weight loss, severe dehydration, severe pneumonia and/or diarrhea, or extreme lethargy and/or depression. Blood was collected by cardiac puncture and liver biopsies were taken for tissue vitamin A measurements. Ferret experiments were approved by the Institutional Animal Care and Use Committee of the Institut-Armand Frappier (Laval, QC). Clinical symptoms of CDV infection (rash, fever $38.5C, conjunctivitis, dehydration, diarrhea, cough, and coryza) and possible vitamin A deciency or toxicity were recorded daily (20,35,36). No data on hypervitaminosis A in ferrets are available but, based on observations in humans and domestic animals, redness, ushed skin, desquamation, vomiting, and evidence of increased intracranial pressure might all have been anticipated. The severity of CDV disease was assessed by assigning each symptom a score of 1.0, taking the sum of points for each ferret and calculating the mean. Viruses. After 12 wk of diet (d 0), anesthetized ferrets were inoculated intranasally with 1.5 3 104 tissue culture infectious doses of CDV strain 5804P-GFP/uN, which carries the gene for the enhanced version of the green uorescence protein (eGFP) in an additional transcription unit upstream of the N gene (25). In healthy, vitamin Areplete ferrets, this virus is not lethal but typically causes a disease that closely mimics measles in humans (rash, fever, conjunctivitis, cough, coryza, diarrhea, immunosuppression, and induction of neutralizing antibodies). Vitamin A in CDV-infected ferrets 1917

Materials and Methods

Ferrets and diets. Six-wkold male (n 6) and female ferrets (n 4) were obtained from Harlan, and an additional 22 male ferrets were obtained from Marshall Farms. Data were pooled from both males and females for vitamin A analyses, but females were excluded from analyses of CDV-induced morbidity because there was no evidence of successful infection (i.e., viremia or CDV disease) in these ferrets. Upon weaning, ferrets were divided into 2 equal groups and consumed, ad libitum, either a vitamin Adecient diet (TD.04590) or a vitamin Acontrol diet containing 18,000 international units (IU)/kg (5400 mg/kg) retinol palmitate (TD.04591; Harlan Teklad). These diets were similar to those previously published, except that the cottonseed oil was reduced to 75.0 g/kg from 105.0 g/kg and replaced with 30.0 g/kg soybean oil (33,34). Ferrets consumed their respective diets for ;12 wk before infection and were monitored daily for signs of vitamin A deciency (behavioral and/or physiological abnormalities). On the day of infection (d 0) and on d 1 after infection, half of the vitamin Areplete (control, C) ferrets received 50,000 IU (15 mg) vitamin A as retinyl palmitate (Aquasol A; Mayne Pharma) intramuscularly (total dose of 100,000 IU 30 mg). Similarly, half of the vitamin Adecient (D) ferrets received 30 mg Aquasol A. These ferrets were also switched to the control diet (Fig. 1). These 2 supplemented groups are hereafter described as the control 1 vitamin A (C1VA) and decient 1 vitamin A (D1VA) experimental groups. After infection, the ferrets consumed their respective diets for 5 additional weeks. We collected 3 mL blood from anesthetized ferrets

Vitamin A measurements. Serum retinol was measured in all ferrets on d 0, 7, and 35 after infection. Serum retinol was measured in only a subset of ferrets on d 14, 21, or 28 due to restrictions placed on blood collection frequency and volume by the Animal Care Committee. For statistical analyses, the results of these 3 timepoints were pooled. Serum was prepared from untreated whole blood (protected from light) and stored in opaque tubes at 220C. Upon thawing, serum vitamin A was extracted as described (26), except that the saponication time was 4 h, and hexane contained 0.45 mmol/L butylated hydroxyl toluene. We chose a 4-h saponication period because preliminary work suggested that variability in retinyl palmitate recovery in spiked triplicate serum samples was lower at this time (103 6 13%) than at 30 min (105% 6 53%) or 2 h (107% 6 34%). Frozen liver sections (0.20.45 g) were thawed, minced, and liver vitamin A was saponied and extracted as described (26). Hexane extracts stored at 4C overnight were evaporated under nitrogen and resuspended in 500 mL 50% acetonitrile:50% methanol for analysis. HPLC analysis of retinol. Retinol was detected at 325 nm in 20 mL of each sample by HPLC (HP 1100; Agilent) equipped with a 5 mm particle size C8 150 3 3mm Luna column (Phenomenex). Retinol was eluted with water (A) and acetonitrile (B) both containing 2 mmol/L of ammonium acetate (pH 7.4). The initial composition was 20% A, 80% B for 0.5 min, increasing to 100% B in 5 min and maintained for 10 min (400 mL/min). The calibration curve with retinol (retention time 8.5 min) was linear between 0.2 and 50 mmol/L. Virus quantitation. After lysis of erythrocytes, peripheral blood mononuclear cells (PBMC) were serially diluted and cultured on VerodogSLAMtag cells to detect cell-associated viremia as described (20). Viral titer (TCID50) per 1 3 106 PBMC was calculated by the Karber method (34). Epithelial virus expression was qualied by macroscopic imaging of ferrets with the MacroIllumination Imaging System (Lightools) as described (25). Virus expression was categorized as none (score 0), 1 or 2 spots of uorescence in a single location on the body (score 1), 1 or 2 spots of uorescence in multiple locations on the body (score 2), and uorescence covering the whole body (score 3). Statistical analyses. Data were analyzed with signicance determined at a 0.05 using SAS, version 9.2, and values in the text are expressed as means 6 SEM. Repeated measures 2-way ANOVA was performed to examine the relationships between experimental groups (C vs. C1VA and D vs. D1VA) and 3 outcome variables: 1) serum vitamin A, 2) weight, and 3) temperature following challenge with CDV. An interaction term was added to the models to estimate the combined effect of experimental group and time. Polynomial contrasts were included in the model to account for linear or quadratic effects using the command PROC MIXED command in SAS. Mann-Whitney (rank sum) unpaired, 2-tailed t test was performed to compare 2 groups with unequal variances. Kaplan-Meier survival curves were generated for each ferret group and compared using the Log rank test. Decient, uninfected ferrets were analyzed separately from decient, infected ferrets for survival analyses but not with respect to vitamin A concentrations.

Downloaded from jn.nutrition.org at Universidade Federal de Viosa on April 3, 2009

FIGURE 2 Vitamin A treatment (30 mg) on d 0 of infection increases serum vitamin A (A) and reduces morbidity due to CDV in control ferrets (B) with concomitant reduction in CDV-associated fever (C). Values are means 6 SEM, n 6. (A) There was a quadratic trend over time that was greater in C1VA than in C (P 0.02, repeated measures 2-way ANOVA). (B) **Different from C (P , 0.01, MannWhitney rank sum test). (C) Time (P , 0.001) and treatment (P 0.02) affected temperature (repeated measures 2-way ANOVA, with a significant interaction, P , 0.001). Grey line marks 38.5C (pyrexia).

Results
Supplements enhance vitamin A concentrations in control ferrets. The repeated measures ANOVA demonstrated a quadratic trend in time that was greater in the C1VA group than in the C group [F (1,30) 6.6, P 0.02] (Fig. 2A). Liver retinol concentrations in C1VA ferrets (2.07 6 0.35 mmol/g; n 6) were higher than those of C ferrets (0.69 6 0.07 mmol/g; n 6) at the end of the 35-d period (P 0.002). Vitamin A supplementation did not adversely affect the health of C1VA ferrets. Supplemented ferrets behaved and appeared similar to C ferrets, overt symptoms of vitamin A toxicity were not observed, and growth rate of supplemented ferrets did not differ from that of C ferrets (not shown). Thus, vitamin A supplementation con1918 Rodeheffer et al.

comitant with CDV infection resulted in a sustained increase in tissue storage of vitamin A and a transient increase in circulating serum vitamin A over a 35-d period. Infection of female ferrets. Four female ferrets were initially included in this study, one in each experimental group. Neither the C nor C1VA female ferret exhibited signs or symptoms of CDV disease or detectable CDV in their PBMC at any time during the experiment. As a result, the females were likely not infected at the dose used and were excluded from analysis of CDV morbidity data. Supplements reduce CDV morbidity in control ferrets. We hypothesized that vitamin A supplementation would reduce

CDV-associated morbidity and we predicted that C1VA ferrets would have less severe CDV disease than nonsupplemented C ferrets. MV infection in humans and CDV infection in ferrets cause an epidermal rash on the face (mouth, chin, ears, and nose), torso, and extremities, pyrexia (fever), conjunctivitis, coryza, cough, diarrhea, and dehydration that result from systemic virus infection as well as secondary infections (3,20,21). As expected, all C ferrets exhibited a whole-body rash and temperature $38.5C by d 12 after infection; 2 ferrets also displayed conjunctivitis or soft stool, and 4 ferrets displayed diarrhea, cough, edema, or dehydration. All C1VA ferrets also exhibited a whole-body rash, but none exhibited an elevated temperature or any other symptoms of CDV infection. The mean CDV morbidity score of C ferrets was higher than that of C1VA ferrets (P 0.003) (Fig. 2B). Signicant differences in temperature were observed over time [F (11, 100) 11.0, P , 0.001] and between experimental groups [F (1, 100) 7.1, P 0.02] in the C vs. C1VA groups following infection (Fig. 2C). An interaction was also observed for experimental group and time [F (11, 100) 11.0, P , 0.001]. The illness displayed by C1VA ferrets was so mild that it did not meet the clinical denition CDV disease (or MV disease in humans) and could not be unambiguously diagnosed as such in the absence of serological conrmation (3,20,21). Consistent with our hypothesis, vitamin A supplementation signicantly improved morbidity due to CDV infection in vitamin Areplete ferrets. Viremia is independent of vitamin A status. The mechanism of vitamin A against measles infection in humans is unknown, but retinol signicantly inhibits MV and CDV replication in vitro (B. Ward, unpublished data). We predicted that reduced CDV morbidity in C1VA ferrets would be concomitant with reduced viral load in this group. However, we did not detect differences in the peak or pattern of either viral titer per 106 PBMC or in the percentage of CDV-infected PBMC (not shown) from C and C1VA ferrets over the course of infection. Similarly, expression of virally encoded green uorescent protein was identical in C and C1VA groups (not shown). These data suggest that increasing concentrations of vitamin A in blood and tissue does not have major impact on CDV infection of PBMC or the skin epithelia. Vitamin A deciency in ferrets. We next hypothesized that vitamin A deciency in ferrets could be induced and subsequently rescued with vitamin A supplementation. After 12 wk of control or decient treatment, serum retinol concentrations in ferrets consuming the vitamin Adecient diet (0.47 6 0.09 mmol/L, n 15) were much lower than in the C ferrets (12.50 6 1.08 mmol/L, n 12; P 0.003). Analysis of liver retinol in a single D (0.055 mmol/g) and C (0.779 mmol/g) ferret at this timepoint suggested tissue vitamin A stores were also depleted. These observations demonstrated that 12 wk of the decient diet is sufcient to cause a profound state of vitamin A deciency, which was accompanied by negative health and behavioral effects. Prior to CDV infection, 8 of 14 D ferrets developed cataracts or corneal clouding, and 6 of 14 displayed either disorientation or head-bobbing behavior. Additionally, 100% of D ferrets had soft stool or diarrhea and anorexia. Although one C ferret developed corneal clouding during the same period, behavioral and gastrointestinal disorders were otherwise not observed in the C group. In addition, C ferrets exhibited linear growth during the entire experiment, whereas the growth of D ferrets ceased after

9 wk (Fig. 3A). Differences in body weight occurred over time [F (16, 151) 152.7; P , 0.001] following infection, and resulting in signicantly lower weights in the D ferrets than in the C and D1VA ferrets [F (1, 151) 12.7, P 0.004]. An interaction occurred for experimental group and time [F (13, 151) 12.8, P , 0.001]. In contrast, similar trends in body weight were observed over time [F (16, 129) 153; P , 0.001] and for both C and D1VA ferrets [F (1, 129) 8.0; P 0.02]; although D1VA ferrets remained smaller than C ferrets. An interaction

Downloaded from jn.nutrition.org at Universidade Federal de Viosa on April 3, 2009

FIGURE 3 Vitamin A treatment (30 mg) on d 0 of infection normalizes the reduced growth rate (A), reduced serum vitamin A concentration (B), and elevated death rate due to CDV (C) in vitamin A-deficient ferrets. Values are means 6 SEM, n 6 (C and C1VA), n 12 (D), and n 5 (D1VA). (A) Body weights after infection differed between D and C or D1VA. Time (P , 0.001) and treatment (P 0.004) affected body weight (repeated measures 2-way ANOVA with a significant interaction, P , 0.001) (B) There was a linear trend over time that was greater in D1VA than in D (P , 0.05, repeated measures 2-way ANOVA). yAll D ferrets died or were killed before d 35. This time point was excluded from statistical analyses, but the vitamin A concentrations for D1VA ferrets are shown. (C) ***Different from C (P , 0.001, log rank test). Vitamin A in CDV-infected ferrets 1919

occurred between C and D1VA ferrets over time [F (16, 129) 7.3; P , 0.001]. D1VA ferrets stopped displaying disorientation and/or head-bobbing behavior within 24 h of supplementation, but we did not observe a similar improvement of gastrointestinal or ocular symptoms in these ferrets, possibly due to concomitant CDV infection of epithelial tissues. Vitamin A supplementation completely reversed the deciency in the D1VA ferrets (Fig. 3B). The repeated measures ANOVA of the D ferrets compared with the D1VA ferrets showed a linear trend in time that was greater among the D1VA ferrets [F(1, 16) 16.6; P , 0.05]. Due to the mortality in D ferrets, the nal timepoint (35 d) had to be excluded in these analyses. Consistent with this observation, mean liver retinol concentrations measured at d 35 or upon ferret deaths were signicantly higher in D1VA ferrets (2.12 6 0.29 mmol/g, n 7) than in D ferrets (0.015 6 0.004 mmol/g, n 5; P 0.003). These results demonstrate that profound vitamin A deciency has an obvious negative impact on ferret health but is rapidly reversible upon high-dose vitamin A supplementation and replenishment of vitamin A in the diet. Profound vitamin A deciency is lethal in ferrets. All vitamin Adecient ferrets were killed prior to d 35 after infection whether or not they were concomitantly infected with CDV because they lost 1520% of their maximum weight, failed to eat for .48 h, or were severely dehydrated. The D ferrets had a median survival rate of 13 d after infection, whereas both D1VA and C1VA ferrets and all but one C ferret recovered from CDV infection and survived until d 35 (Fig. 3C; P , 0.0001). The survival of the D1VA ferrets is particularly striking given their profound state of deciency at the time of infection (Fig. 3B). Infected and uninfected D ferrets had equivalent mean survival times (not shown). Although the CDV-infected D ferrets had signs of severe CDV morbidity (not shown), the relative roles of deciency itself compared with CDV infection were difcult to determine because of the overlap of signs and symptoms (e.g., weight loss, anorexia, and neurologic changes).

Discussion
Despite the efcacy of the live attenuated vaccine, measles infection causes ;500,000 deaths per year (2). Although no 100% curative therapy exists, high-dose vitamin A supplements dramatically reduce mortality and morbidity rates, especially in young children at the greatest risk of infection (8,9,37). Because infection depletes vitamin A, even in populations where deciency is rare, and deciency itself is a risk factor for severe disease and mortality, high concentrations of circulating and tissue vitamin A may be benecial (47). We describe here, to our knowledge for the rst time, a tractable animal model to study the impact of vitamin A on Morbillivirus infection. High-dose vitamin A supplements transiently increased circulating vitamin A concentrations and caused a sustained increase in liver concentrations in the vitamin A-replete (C) ferrets (Fig. 2) without observable toxicity. The mean serum vitamin A concentration observed in C ferrets (12.50 6 3.75 mmol/L) is consistent with previously published values (2830) and 400 2000% greater than the normal range in humans (1.053.05 mmol/L) (38). Because vitamin A supplementation decreases MV-associated morbidity and mortality in humans, we predicted the same result in CDV-infected ferrets. Consistent with our hypothesis, un1920 Rodeheffer et al.

supplemented ferrets exhibited a febrile illness with conjunctivitis, cough, diarrhea, and/or dehydration that closely parallels the clinical manifestations of measles in humans (3,20,21). In contrast, supplemented (C1VA) ferrets developed only a transient illness with a mild rash and no fever. Given the genetic similarity between MV and CDV, as well as the remarkable consistency in the inuence of vitamin A on measles in humans and on CDV in ferrets in this study, it follows that this model will be useful for determining the mechanism of vitamin A against Morbillivirus infections in vitamin A-replete individuals. We suspected that vitamin A reduced morbidity by directly inhibiting CDV replication in vivo, as occurred in vitro for HIV (4144), herpes simplex virus (15), and MV (B. Ward, unpublished results). However, we did not detect any differences in viremia, the percentage of infected PBMC, or viral load (eGFP expression) in the skin of C and C1VA ferrets (not shown). As these assays do not directly measure kinetics of viral replication, vitamin A supplements could have had other effects such as delaying secretion from infected cells or blocking replication or clearance from secondary target tissues. Consistent with this last hypothesis, C1VA ferrets did not display respiratory, ocular, or gastrointestinal symptoms, whereas C ferrets displayed frank conjunctivitis, respiratory distress, and/or diarrhea. Despite repeated attempts, we were unable to collect adequate conjunctival impression samples to assess histopathologic changes in epithelial and goblet cell distribution described in humans with vitamin A deciency (45). Rodent models demonstrate that vitamin A deciency leads to a TH1-biased immune response, whereas vitamin A supplements result in stronger TH2 responses (4652). In humans, vitamin A deciency during infection may promote TH1 responses in some conditions (53,54). Although we have not yet measured the TH1/TH2 balance in the response to CDV in ferrets, our preliminary data suggest that vitamin A supplements did not signicantly alter the kinetics of cell-mediated immune responses such as the rash. The strain 5804P-eGFP/uN CDV used herein is a recombinant virus expressing eGFP as a separate transcriptional unit upstream of the N gene (25). In ferrets this virus typically causes rash, fever, and a range of other symptoms as well as immunosuppression. However, ferrets rapidly make neutralizing antibodies and recover fully from infection with this strain, unlike the parent CDV strain, which is rapidly lethal (severe symptoms typically appear within 12 d). Although the mechanism of attenuation of 5804P-eGFP/uN is not yet dened, we elected to use this nonlethal strain to detect possible effects of altered vitamin A status in either sense: i.e., decreased morbidity in C1VA vs. C ferrets or increased morbidity in D vs. D1VA ferrets. In light of our ndings, future experiments will include a trial of vitamin A supplementation in wild-type CDV infection. In humans, hyporetinemia (,0.7 mmol/L serum retinol) occurs in 2092% of all acute measles cases even in populations where vitamin A deciency is rare, although serum concentrations return to normal as the infection resolves (5,7,39). In contrast, serum retinol concentrations in C ferrets during the CDV infection remained relatively constant throughout infection. It is unknown whether retinol binding protein synthesis is suppressed during the acute phase of CDV infection in ferrets, as it is in humans with acute measles (40). Circulating vitamin A in ferrets consists of a higher proportion of retinyl esters than in humans (26), which could inuence turnover rates. Here, we measured total vitamin A concentration and not relative concentrations of retinol and retinyl esters. It may be necessary to measure retinol and retinyl esters separately to fully assess

Downloaded from jn.nutrition.org at Universidade Federal de Viosa on April 3, 2009

changes in vitamin A circulation during infection and supplementation. Our nal hypotheses predicted that profound vitamin A deciency could be induced in ferrets and could be rescued with high-dose vitamin A supplementation. Twelve weeks of the vitamin A decient diet was sufcient to severely deplete serum retinol and 17 wk was sufcient to deplete liver retinol stores (Fig. 3). Both serum and liver retinol were restored completely with 30 mg vitamin A and return to a vitamin A-replete diet. These data are consistent with the observation of a small but signicant decrease in ferret serum retinol after 5 wk of consuming a vitamin Adecient diet (26). We also observed that profound depletion of serum and tissue vitamin A is ultimately lethal in ferrets. Survival time of D ferrets after CDV infection varied with the supplier (;7 d for Marshall Farms and ;26 d for Harlan), raising the possibility that baseline vitamin A concentrations in the 2 groups differed. Vitamin A deciency led to anorexia and arrested growth, disorientation or head-bobbing behavior, cataract development, soft stool, diarrhea, and lethargy within 9 wk of starting the decient diet. Similar signs of deciency have been observed in cows, rats, and mice fed suboptimal quantities of vitamin A (16,49,55,56). Vitamin A supplementation, however, rapidly restored even the most severely affected ferrets to apparent normal health. Most ferrets were killed before the ultimate cause of death could be determined. In gnotobiotic rats, vitamin A deciency is associated with increased bacterial translocation from intestine to kidney, which is not observed in decient rats with a normal complement of intestinal bacteria (16). We predict that vitamin A deciency rendered ferrets susceptible to a physiological challenge that would not normally be lethal (i.e., underlying infection), as observed in decient humans who are more susceptible to bacterial, viral, and parasitic infections (17,38). Currently, we are measuring the kinetics of development of vitamin A deciency in ferrets to identify when ferrets lack overt symptoms of deciency but are at increased risk of severe morbidity and mortality from CDV. To our knowledge, this is the rst report of an animal model that mimics specic aspects of the interplay between vitamin A status and measles in humans. Vitamin A supplements can enhance vitamin A status in vitamin Areplete ferrets and reduce the severity of Morbillivirus infection, relative to nonsupplemented ferrets. This result closely replicates the observation that high-dose vitamin A supplements can reduce measles morbidity and mortality even in well-nourished populations (5,8,9). We also show, for the rst time, that profound vitamin A deciency can be induced in ferrets and is rapidly and completely reversed with vitamin A supplementation. We think that this model will allow detailed investigation of the cellular and molecular mechanisms of vitamin A against Morbillivirus infection in vivo.

5. 6.

7. 8. 9.

10. 11.

12.

13.

14. 15.

16.

17. 18.

19.

20.

21.

22.

23. 24. 25.

26.

Literature Cited
1. Permar SR, Grifn DE, Letvin NL. Immune containment and consequences of measles virus infection in healthy and immunocompromised individuals. Clin Vaccine Immunol. 2006;13:43743. Progress in reducing measles mortality worldwide, 19992003. MMWR Morb Mortal Wkly Rep. 2005 Mar 4;54:2003. Perry RT, Halsey NA. The clinical signicance of measles: a review. J Infect Dis. 2004;189: Suppl 1:S416. Nandy R, Handzel T, Zaneidou M, Biey J, Coddy RZ, Perry R, Strebel P, Cairns L. Case-fatality rate during a measles outbreak in eastern Niger in 2003. Clin Infect Dis. 2006;42:3228.

27. 28. 29.

2. 3. 4.

30. 31.

Hussey GD, Klein M. A randomized, controlled trial of vitamin A in children with severe measles. N Engl J Med. 1990;323:1604. Arrieta AC, Zaleska M, Stutman HR, Marks MI. Vitamin A levels in children with measles in Long Beach, California. J Pediatr. 1992;121: 758. Caballero B, Rice A. Low serum retinol is associated with increased severity of measles in New York City children. Nutr Rev. 1992;50:2912. Semba RD. On Joseph Bramhall Ellisons discovery that vitamin A reduces measles mortality. Nutrition. 2003;19:3904. Huiming Y, Chaomin W, Meng M, Yang H. Vitamin A for treating measles in children. Cochrane Database Syst Rev. 2005;Oct 19;(4): CD001479. Ellison JB. Intensive vitamin therapy in measles. BMJ. 1932;2:70811. Horvat B, Rivailler P, Varior-Krishnan G, Cardoso A, Gerlier D, Rabourdin-Combe C. Transgenic mice expressing human measles virus (MV) receptor CD46 provide cells exhibiting different permissivities to MV infections. J Virol. 1996;70:667381. Mrkic B, Pavlovic J, Rulicke T, Volpe P, Buchholz CJ, Hourcade D, Atkinson JP, Aguzzi A, Cattaneo R. Measles virus spread and pathogenesis in genetically modied mice. J Virol. 1998;72:74207. Niewiesk S. Cotton rats (Sigmodon hispidus): an animal model to study the pathogenesis of measles virus infection. Immunol Lett. 1999; 65:4750. Privalsky ML. The role of corepressors in transcriptional regulation by nuclear hormone receptors. Annu Rev Physiol. 2004;66:31560. Isaacs CE, Kascsak R, Pullarkat RK, Xu W, Schneidman K. Inhibition of herpes simplex virus replication by retinoic acid. Antiviral Res. 1997; 33:11727. Kozakova H, Hanson LA, Stepankova R, Kahu H, Dahlgren UI, Wiedermann U. Vitamin A deciency leads to severe functional disturbance of the intestinal epithelium enzymes associated with diarrhoea and increased bacterial translocation in gnotobiotic rats. Microbes Infect. 2003;5:40511. Ward BJ, Semba RD. Vitamin A and HIV infection. In: Watson RR, editor. Nutrition and AIDS. Boca Raton: CRC Press; 1994. p. 14166. von Messling V, Zimmer G, Herrler G, Haas L, Cattaneo R. The hemagglutinin of canine distemper virus determines tropism and cytopathogenicity. J Virol. 2001;75:641827. Appel MJ. Canine distemper virus. In: Horzinek M, editor. Virus infections of carnivores 1-virus infections of vertebrates. Amsterdam: Elsevier Science Publishers; 1987. p. 13359. von Messling V, Springfeld C, Devaux P, Cattaneo R. A ferret model of canine distemper virus virulence and immunosuppression. J Virol. 2003; 77:1257991. Hutchins SS, Papania MJ, Amler R, Maes EF, Grabowsky M, Bromberg K, Glasglow V, Speed T, Bellini WJ, Orenstein WA. Evaluation of the measles clinical case denition. J Infect Dis. 2004;189: Suppl 1:S1539. Ward BJ, Johnson RT, Vaisberg A, Jauregui E, Grifn DE. Cytokine production in vitro and the lymphoproliferative defect of natural measles virus infection. Clin Immunol Immunopathol. 1991;61:23648. Grifn DE, Bellini W. Measles virus. In: Fields BN, et al., editors. Fields virology. Philadelphia (PA): Lippincott-Raven; 1996. p. 1267312. Harder TC, Osterhaus AD. Canine distemper virusa morbillivirus in search of new hosts? Trends Microbiol. 1997;5:1204. von Messling V, Milosevic D, Cattaneo R. Tropism illuminated: lymphocyte-based pathways blazed by lethal morbillivirus through the host immune system. Proc Natl Acad Sci USA. 2004;101:1421621. Lederman JD, Overton KM, Hofmann NE, Moore BJ, Thornton J, Erdman JW. Ferrets (Mustela putoius furo) inefciently convert betacarotene to vitamin A. J Nutr. 1998;128:2719. Raila J, Gomez C, Schweigert FJ. The ferret as a model for vitamin A metabolism in carnivores. J Nutr. 2002;132:1787S9S. Sauberlich HE, Skala JH. Vitamins. Laboratory tests for the assessment of nutritional status. Cleveland: CRC Press, Inc.; 1974. p. 413. Wang XD, Krinsky NI. Identication and quantication of retinoic acid and other metabolites from beta-carotene excentric cleavage in human intestine in vitro and ferret intestine in vivo. Methods Enzymol. 1997;282: 11730. Fox JG. Biology and diseases of the ferret. 2 ed. Baltimore (MD): Lippincott Williams & Wilkins; 1998. Wang XD, Liu C, Bronson RT, Smith DE, Krinsky NI, Russell M. Retinoid signaling and activator protein-1 expression in ferrets given

Downloaded from jn.nutrition.org at Universidade Federal de Viosa on April 3, 2009

Vitamin A in CDV-infected ferrets

1921

32. 33.

34. 35.

36. 37. 38. 39.

40.

41.

42.

43.

44.

beta-carotene supplements and exposed to tobacco smoke. J Natl Cancer Inst. 1999;91:606. Wolf G. The effect of low and high doses of beta-carotene and exposure to cigarette smoke on the lungs of ferrets. Nutr Rev. 2002;60:8890. White WS, Peck KM, Ulman EA, Erdman JW, Jr. The ferret as a model for evaluation of the bioavailabilities of all-trans-beta-carotene and its isomers. J Nutr. 1993;123:112939. Payment P, Trudel M. Methods and techniques in virology. New York: Marcel Dekker; 1993. Bresee JS, Fischer M, Dowell SF, Johnston BD, Biggs VM, Levine RS, Lingappa JR, Keyserling HL, Petersen KM, et al. Vitamin A therapy for children with respiratory syncytial virus infection: a multicenter trial in the United States. Pediatr Infect Dis J. 1996;15:77782. Neuzil KM, Gruber WC, Graham BS. Vitamin A therapy for respiratory syncytial virus infection. Pediatr Infect Dis J. 1997;16:845. DSouza RM, DSouza R. Vitamin A for the treatment of children with measlesa systematic review. J Trop Pediatr. 2002;48:3237. Ward BJ, Humphrey JH, Clement L, Chaisson RE. Vitamin A status in HIV infection. Nutr Res. 1993;13:15766. Frieden TR, Sowell AL, Henning KJ, Huff DL, Gunn RA. Vitamin A levels and severity of measles. New York City. Am J Dis Child. 1992; 146:1826. Varavithya W, Stoecker B, Chaiyaratana W, Kittikool J. Vitamin A status of Thai children with measles. Trop Geogr Med. 1986;38: 35961. Maciaszek JW, Coniglio SJ, Talmage DA, Viglianti GA. Retinoidinduced repression of human immunodeciency virus type 1 core promoter activity inhibits virus replication. J Virol. 1998;72:58629. Hanley TM, Kiefer HL, Schnitzler AC, Marcello JE, Viglianti GA. Retinoid-dependent restriction of human immunodeciency virus type 1 replication in monocytes/macrophages. J Virol. 2004;78:281930. Semmel M, Macho A, Coulaud D, Alileche A, Plaisance S, Aguilar J, Jasmin C. Effect of retinoic acid on HL-60 cells infected with human immunodeciency virus type 1. Blood. 1994;84:24808. Kiefer HL, Hanley TM, Marcello JE, Karthik AG, Viglianti GA. Retinoic acid inhibition of chromatin remodeling at the human immunodeciency virus type 1 promoter. Uncoupling of histone acetylation and chromatin remodeling. J Biol Chem. 2004;279:4360413.

45. Wittpenn JR, West KP, Keenum D, Farazdaghi M, Humphrey J, Howard GR, Sommer A. Training manual: assessment of vitamin A status by impression cytology. Baltimore (MD): The Johns Hopkins University. 1988. 46. Stephensen CB, Rasooly R, Jiang X, Ceddia MA, Weaver CT, Chandraratna RA, Bucy RP. Vitamin A enhances in vitro Th2 development via retinoid X receptor pathway. J Immunol. 2002;168:4495 503. 47. Carman JA, Pond L, Nashold F, Wassom DL, Hayes CE. Immunity to Trichinella spiralis infection in vitamin A-decient mice. J Exp Med. 1992;175:11120. 48. Stephensen CB, Jiang X, Freytag T. Vitamin A deciency increases the in vivo development of IL-10-positive Th2 cells and decreases development of Th1 cells in mice. J Nutr. 2004;134:26606. 49. Austenaa LM, Carlsen H, Ertesvag A, Alexander G, Blomhoff HK, Blomhoff R. Vitamin A status signicantly alters nuclear factor-kappaB activity assessed by in vivo imaging. FASEB J. 2004;18:12557. 50. Albers R, Bol M, Bleumink R, Willems AA, Pieters RH. Effects of supplementation with vitamins A, C, and E, selenium, and zinc on immune function in a murine sensitization model. Nutrition. 2003;19:9406. 51. Ma Y, Ross AC. The anti-tetanus immune response of neonatal mice is augmented by retinoic acid combined with polyriboinosinic:polyribocytidylic acid. Proc Natl Acad Sci USA. 2005;102:1355661. 52. Yu S, Xia M, Xu W, Chu Y, Wang Y, Xiong S. All-trans retinoic acid biases immune response induced by DNA vaccine in a Th2 direction. Vaccine. 2005;23:51607. 53. Semba RD. Vitamin A and immunity to viral, bacterial and protozoan infections. Proc Nutr Soc. 1999;58:71927. 54. Jason J, Archibald LK, Nwanyanwu OC, Sowell AL, Buchanan I, Larned J, Bell M, Kazembe PN, Dobbie H, Jarvis WR. Vitamin A levels and immunity in humans. Clin Diagn Lab Immunol. 2002;9:61621. 55. Carta M, Stancampiano R, Tronci E, Collu M, Usiello A, Morelli M, Fadda F. Vitamin A deciency induces motor impairments and striatal cholinergic dysfunction in rats. Neuroscience. 2006; 193(4):116372. 56. Swanson KS, Merchen NR, Erdman JW Jr, Drackley JK, Orias F, Morin DE, Haddad MF. Inuence of dietary vitamin A content on serum and liver vitamin A concentrations and health in preruminant Holstein calves fed milk replacer. J Dairy Sci. 2000;83:202736.

Downloaded from jn.nutrition.org at Universidade Federal de Viosa on April 3, 2009

1922

Rodeheffer et al.