An infant formula toxicity and toxicokinetic feeding study on carrageenan in preweaning piglets with special attention to the immune system and gastrointestinal tract

Food and Chemical Toxicology 77 (2015) 120–131

Contents lists available at ScienceDirect

Food and Chemical Toxicology j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e / f o o d c h e m t o x

An infant formula toxicity and toxicokinetic feeding study on carrageenan in preweaning piglets with special attention to the immune system and gastrointestinal tract ☆ M.L. Weiner a,*, H.E. Ferguson b, B.A. Thorsrud c, K.G. Nelson c, W.R. Blakemore d, B. Zeigler c, M.J. Cameron c,1, A. Brant c,1, L. Cochrane c, M. Pellerin c, B. Mahadevan b a

TOXpertise, LLC, Princeton, NJ 08540, USA Abbott Nutrition, Columbus, OH 43219, USA c MPI Research, Mattawan, MI 49071, USA d Celtic Colloids Inc., Topsham, ME 04086, USA b

A R T I C L E

I N F O

Article history: Received 14 October 2014 Accepted 27 December 2014 Available online 12 January 2015 Keywords: Carrageenan Infant formula Piglet Gastrointestinal tract Immune system

A B S T R A C T

A toxicity/toxicokinetic swine-adapted infant formula feeding study was conducted in Domestic Yorkshire Crossbred Swine from lactation day 3 for 28 consecutive days during the preweaning period at carrageenan concentrations of 0, 300, 1000 and 2250 ppm under GLP guidelines. This study extends the observations in newborn baboons (McGill et al., 1977) to piglets and evaluates additional parameters: organ weights, clinical chemistry, special gastrointestinal tract stains (toluidine blue, Periodic Acid– Schiff), plasma levels of carrageenan; and evaluation of potential immune system effects. Using validated methods, immunophenotyping of blood cell types (lymphocytes, monocytes, B cells, helper T cells, cytotoxic T cells, mature T cells), sandwich immunoassays for blood cytokine evaluations (IL-6, IL-8, IL1β, TNF-α), and immunohistochemical staining of the gut for IL-8 and TNF-α were conducted. No treatmentrelated adverse effects at any carrageenan concentration were found on any parameter. Glucosuria in a few animals was not considered treatment-related. The high dose in this study, equivalent to ~430 mg/ kg/day, provides an adequate margin of exposure for human infants, as affirmed by JECFA and supports the safe use of carrageenan for infants ages 0–12 weeks and older and infants with special medical needs. © 2014 Elsevier Ltd. All rights reserved.

Abbreviations: ADI, acceptable daily intake; ANOVA, one-way analysis of variance; BW, body weight; CGN, carrageenan; cPs, centipoise; EMEA, European Medicines Agency; FAO, Food and Agriculture Organization of the United Nations; FDA, Food and Drug Administration; GIT, gastrointestinal tract; GLP, Good Laboratory Practice; H&E, hematoxylin and eosin; ICH, International Conference on Harmonisation; IL, interleukin cytokines; IM, intramuscular; IPCS, International Programme on Chemical Safety; JECFA, Joint FAO/WHO Expert Committee on Food Additives; kDa, kiloDaltons; K3EDTA, potassium ethylene diamine tetraacetic acid; LD, lactation day; LMT, low molecular weight tail; MSE, mean square error; Mw, weight-average molecular weight; PAS, Periodic Acid–Schiff; PGN, poligeenan; NBF, neutral buffered formalin; SD, standard deviation; TNF-α, tumor necrosis factor alpha; USA, United States of America; WHO, World Health Organization. ☆ The work included in this paper was partially presented at the 2014 Society of Toxicology Meeting in Phoenix, AZ in the following abstracts: Weiner M.L., et al. (2014) Safety of Carrageenan in Infant Formula: A 4-Week Toxicity Study in Preweaning Piglets. Toxicol. Sci. 138 (1) 341 (Abstract); Zeigler B., et al. (2014) Safety of Carrageenan in Infant Formula: A 4-Week Study of the Potential Immune System Effects. Toxicol. Sci. 138 (1), 342 (Abstract); Blakemore, W.R., et al. (2014) Safety of Carrageenan in Infant Formula: A 4-Week Toxicokinetic Evaluation in Preweaning Piglets. Toxicol. Sci. 138 (1) 343 (Abstract). * Corresponding author. President, Toxpertise LLC, 100 Jackson Avenue, Princeton, NJ 08540, USA. Tel./fax: +1 908 938 2733. E-mail address: [email protected] (M.L. Weiner). 1 No longer at MPI. http://dx.doi.org/10.1016/j.fct.2014.12.022 0278-6915/© 2014 Elsevier Ltd. All rights reserved.

M.L. Weiner et al./Food and Chemical Toxicology 77 (2015) 120–131

1. Introduction Carrageenan (CGN) is a high molecular weight polymer derived from certain Rhodophyceae (red seaweeds). Food-grade CGN is primarily used to bind water, promote gel formation, improve palatability, thicken and stabilize structure for food products by binding with protein. Chemically, CGN has a molecular backbone of repeating galactose units that may have sulfate groups attached with a weight average molecular weight (Mw) of 200– 800 kDa. Regulatory agencies specify a viscosity of ≥5 cPs (centipoise), for food grade CGN.1 This viscosity is equivalent to ~100–170 kDa Mw. CGN has been widely used as a food additive for decades in numerous foods, including infant formula, and its safety is based on a large database of studies (Weiner, 2014). In Canada, CGN is approved at a maximum level of 0.05% in infant formula as a suspension agent for calcium salts in lactose-free infant formula, based on milk protein, and at a level of 0.1% in formula based on isolated amino acids and/or protein hydrolysates (Canada Gazette, 2004). This latter formula is used for infants with special medical needs, such as allergies to cow’s milk, prematurity and health concerns. CGN was reviewed by the United Nations’ World Health Organization (WHO) Food and Agriculture Organization’s (FAO) Joint Expert Committee on Food Additives (JECFA) in 2008. JECFA (2008) reaffirmed the Acceptable Daily Intake (ADI) for CGN as “Not Specified”. However, JECFA raised2 concerns regarding the safety of CGN for use in infant formula for infants under the age of 12 weeks. This concern was based on lack of data on potential effects of CGN ingestion on the neonatal immune system and gastrointestinal tract (GIT). Injection of CGN (intravenous or intraperitoneal) is known to cause inflammatory immune responses in animal models (see Weiner, 2014). This has raised questions whether orally administered carrageenan affects the immune system. Earlier work by McGill et al. (1977) had shown CGN to be safe to infant baboons fed infant formula with up to 1220 ppm CGN (~432 mg/kg/day) from birth for 112 days. There were no effects on health, growth, blood, fecal or urine parameters or any histopathological effects on the GIT tissues (McGill et al., 1977). McGill et al. (1977) did not specifically look at immune system effects or absorption of CGN. JECFA (2008) considered McGill et al. (1977) to lack a full evaluation of the GIT because no special stains, such as toluidine blue, were used to visualize mucosal mast cells, indicators of immune system responses, and because potential CGN absorption was not evaluated. The present study in preweaning piglets sought to evaluate (1) the potential absorption of CGN in the GIT, (2) the presence of CGN in serum following ingestion of swine-adapted infant formula containing CGN via toxicokinetic analysis and (3) to assess the impact of CGN on the developing immune system. No regulatory study guidelines for the evaluation of constituents in infant formula exist (International Programme on Chemical Safety (IPCS), 2009). The

1 Commission Regulation. (2012), 2031/2012 of 9 March 2012. Official Journal of the European Union L83 55 L 83/140 – L 83/141; Food and Agriculture Organization of the United Nations. (2007) Combined Compendium of Food Additive Specifications; CGN – 68th session of JECFA, 2007; Food Chemicals Codex. (2013). FCC Monographs, Edition 8, Supplement 2. 219–228; Japan Food Additives Association (2009) Japan’s Specifications for Food Additives, Eighth Edition, the Ministry of Health and Welfare, 344. 2 “The JECFA Secretariat clarified that it was not the existence of data raising any specific concerns, rather the lack of data on the potential impact of carrageenan on the immature gastrointestinal and immune systems which led to the conservative conclusion. As a general principle, JECFA considered that the ADI was not applicable to infants under the age of 12 weeks, in the absence of specific data to demonstrate the safety of these substances for this age group.” (Codex Committee on Food Additives, 2008).

121

Table 1 Study design for the 28-day study. Group numbera

Dose volume: mL/kg bw/day

Main study 1 500 2 500 3 500 4 500 Toxicokinetic study 5 500 6 500 7 500 8 500

CGN dose concentration: ppm

Number of males per group

Number of females per group

0 300 1000 2250

6 6 6 6

6 6 6 6

0 300 1000 2250

3 3 3 3

3 3 3 3

a Corresponding main study and toxicokinetic groups were combined to facilitate data interpretation. All animals were dosed from Lactation Day 2 (Study Day 1) for 28 consecutive days.

preweaning piglet chosen for this study is considered a good model to evaluate nutritional status, growth and development and basic physiology, including the GIT and immune system, which have been shown to resemble humans (Barrow, 2012; Guilloteau et al., 2010; Helm et al., 2007; Odel et al., 2014; Penninks et al., 2012). The dosing period (0–28 days) was chosen because it corresponds to the preweaning period and is equivalent to about the first 23 months of life for a human (Barrow, 2012; Buelke-Sam, 2002). Toluidine blue was used to visualize mucosal mast cells and Periodic Acid–Schiff (PAS) reagent was used to stain the goblet cells of the jejunum for evaluation of possible toxicological effects. A detailed immune system evaluation, including immunophenotyping, cytokine evaluation and immunohistochemistry of the GIT, was conducted. This study is the first time that CGN has been evaluated for safety in infant formula using new technologies for immune system parameters; detailed GIT staining and measurement of CGN plasma absorption in an established animal model for the preweaning period. This study provides a thorough evaluation of this major food additive for its use in infant formula under Good Laboratory Practice (GLP) guidelines and also demonstrates the value of the preweaning piglet model for safety evaluation of food additives used in infant formula. 2. Materials and methods The study was conducted by MPI Research, Mattawan, MI, USA in accordance with GLP Regulations3 and was based on current guidelines and guidance documents for preclinical juvenile studies for drugs (Buelke-Sam, 2002; European Medicines Agency (EMEA), 2008; International Conference on Harmonisation (ICH) Harmonised Tripartite Guidelines, 2010; United Stated Food and Drug Administration (FDA), 2000; United States Food and Drug Administration (FDA) Center for Drug Evaluation and Research (CDER), 2006). Table 1 outlines the study design for this study. Table 2 summarizes the frequency of animal observations (mortality, morbidity, clinical observations, body weight, and feed consumption) and types and frequency of all parameters evaluated, the times of collection and animal groups studied. A separate set of animals was designated for the toxicokinetic phase to avoid excessive handling of any particular animal, since piglets are prone to stress when handled too much. 2.1. Test material characterization The test material sample of κ/λ-CGN (FMC Lot. 90303011) was fully characterized for molecular weight (Mw), percentage of the Mw below 50 kDa, known as the Low Molecular Weight Tail (LMT) as per European requirements, and met interna-

3 United States Food and Drug Administration (FDA) Good Laboratory Practice (GLP). Regulations, 21 CFR Part 58; Organisation for Economic Co-operation and Development (OECD) 2002. Consensus Document “The Application of the Organisation for Economic Co-operation and Development Principles of GLP to the Organisation and Management of Multi-Site Studies” ENV/JM/MONO (2002); Japanese Ministry of Health, Labour and Welfare (MHLW) Good Laboratory Practice Standards Ordinance No. 21, March 26, 1997.

122

M.L. Weiner et al./Food and Chemical Toxicology 77 (2015) 120–131

Table 2 Study parameters evaluated. Parameter

Groups evaluated

Times evaluated

Test article analyses for homogeneity Test article analyses for concentration and stability Mortality Observations Body weight (BW) Feed consumption Feed efficacy Hematology: leukocyte count (total and absolute differential), erythrocyte count, hemoglobin, hematocrit, mean corpuscular hemoglobin, mean corpuscular volume, mean corpuscular hemoglobin concentration (calculated), absolute reticulocytes, platelet count, blood cell morphology. Coagulation: prothrombin time, activated partial thromboplastin time. Clinical chemistry: alkaline phosphatase, total bilirubin, aspartate aminotransferase, alanine aminotransferase, gamma glutamyl transferase, urea nitrogen, creatinine, total protein, albumin, globulin and albumin/globulin ratio (calculated), glucose, total cholesterol, triglycerides, electrolytes (sodium, potassium, chloride), calcium, phosphorus. Urinalysis: osmolality, color and appearance, specific gravity, pH, protein, glucose, bilirubin, ketones, blood, urobilinogen, microscopy of centrifuged sediment. Toxicokinetic evaluations: serum analyzed for the presence and amount of low molecular weight CGN Plasma cytokine evaluation: IL-1β, IL-6, IL-8, TNF-α Peripheral blood immunophenotyping for lymphocytes (CD45+), monocytes (CD45+, CD14+), B cells (CD45+, CD21+), helper T cells (CD45+, CD3+,CD4+), cytotoxic T cells (CD45+, CD3+,CD8+), mature T cells (CD45+, CD3+) Organ weights: adrenal gland, brain, spleen, epididymides, esophagus, heart, kidneys, large intestine (cecum, colon, rectum), liver, small intestine (duodenum, jejunum, ileum)b Histopathology: the adrenals, aorta, brain, bone, epididymides, esophagus, eyes, gallbladder, GALT, heart, kidneys, large intestine (cecum, colon, rectum), larynx, liver, lungs, lymph nodes, mammary gland, nerve, ovaries, oviducts, pancreas, pituitary gland, prostate gland, salivary glands, seminal vesicles, skeletal muscle, skin, small intestine (duodenum, jejunum, ileum), spinal cord, spleen, stomach (nonglandular, fundus, pylorus), testes, thymus, thyroid gland, tongue, trachea, ureter, urinary bladder, uterus with cervix, vagina, gross lesions, masses. Special stains for GIT tissues (toluidine blue and PAS stains) Immunohistochemistry of the GIT (stomach, duodenum, jejunum, ileum, proximal colon, distal colon) for evaluation of tissue cytokines (IL-8, TNF-α)

1–8 1–8 1–8 1–8 1–8 1–8 1–4 1–4

300 and 2250 ppm for first preparationa All concentrations for first and last preparationsa Daily Daily Daily Daily For study period days 1–28 Day 14 and 29

1–4

Day 14 and 29

1–4

Terminal necropsy

5–8 1–4 1–4

Day 3 (1 hour after initiation of the third daily dose) and Day 28 (1 hour after initiation of the first daily dose) Day 14 and 29 Day 14 and 29

1–4

Terminal necropsy

1 and 4

Terminal necropsy

1–4

Terminal necropsy

a b

Formulation samples from the first and last preparations on study were analyzed for carrageenan. Organ weights were determined on key standard organs and additional organ weights for the sections of the GIT were included.

tionally recognized CGN food additive specifications (Blakemore et al., 2014a). The Mw of the test material sample was 664–732 kDa with a LMT of 0.3–3.9% and a viscosity of 80 cPs (Blakemore et al., 2014a).

2.2. Animals Domestic Yorkshire Crossbred Swine (farm pigs) were received on Lactation Day (LD) 2 from Bailey Terra Nova Farms, Schoolcraft, MI, USA. The day piglets were born was designated as Lactation Day (LD) 0. This day varied slightly among litters since not all sows delivered on the same day. The piglets nursed from the sow at the animal supplier for at least 48 hours to allow them to consume maternal colostrum for its beneficial effects (Farmer and Questnel, 2009; Guilloteau et al., 2010). After the first 48 hours, the piglets were injected with an iron supplement and a broad spectrum antibiotic (Excede®). An additional iron supplement injection was given about one week after the first injection. Additional antibiotics (5 mg/kg) were given via intramuscular (IM) injection weekly during the study. Both the iron supplement and antibiotics were administered prophylactically to insure good health of the piglets during the study. Piglets are sensitive to gastrointestinal changes during development; thus, antibiotics were recommended by the laboratory veterinarian to maintain good health during the study as a standard practice. The antibiotic treatment is crucial to the proper health and growth of the piglets at this age, which follows proper veterinary practice. In addition, since the controls and carrageenan groups are treated the same, this allows for proper interpretation of the data. The antibiotic is not given orally, but via IM injection, which would avoid direct contact with the GIT and the microflora throughout the GIT. Following the injections at 48 hours, the piglets were transported to nearby MPI Research Laboratories in heated vans and identified with animal numbers. There was no acclimation period in this study; experimental diets were introduced on the day of the animals’ arrival at the testing facility, which was designated as Study Day 1. Animals were randomized by sex into treatment groups using a standard by weight measured value randomization procedure. Piglets were housed individually in mobile stainless steel cages with plastic coated flooring and rubber mats in an environmentally controlled room. Supplemental heat was provided, as needed. At randomization, piglets were assigned to the control and treatment groups as outlined in Table 1.

2.3. Dose selection Dose levels selected for the study were 0, 300, 1000 and 2250 ppm CGN in formula. The low dose (300 ppm) represents the concentration of CGN in commercial infant formula. The middle dose (1000 ppm) represents the concentration of CGN in commercial hydrolyzed protein and amino acid formula for special needs infants. The dose levels were also based on a feasibility study in which 2-day old Göttingen minipigs were fed 0 or 3000 ppm CGN for 3 days (Beck, M.J., 2011, unpublished study). A 10-day range-finding study at doses of 0, 300, and 3000 ppm CGN was also conducted in 2-day old Göttingen minipigs (Beck, M.J., 2012, unpublished study). Reduced food consumption and BW gain were noted at the high dose (3000 ppm) in the 10day study. Due to the high viscosity of the infant formula containing 3000 ppm CGN and the possible effect of viscosity on formula palatability, 2250 ppm CGN was selected as the high dose in the present study. The work of Davidson and Swithers (2005) supports the use of a lower viscosity, rather than higher viscosity, to avoid reduced BW gain and caloric intake due to high viscosity. Thus, dose levels were selected based on the 3-day and 10-day range-finding studies (unpublished studies) and the potential confounding effects of high viscosity formulations on growth (Davidson and Swithers, 2005).

2.4. Control and test material formulations All formulations were prepared by Abbott Nutrition, Columbus, OH, USA in singleuse Tetra Paks (250 mL each) containing the test material at concentrations of 0, 300, 1000 or 2250 ppm CGN. Formulations were stored refrigerated (2 to 8 °C) when not used. Final swine-adapted infant formula composition was consistent with the nutritional composition of mature sow’s milk (Klobasa et al., 1987). Ingredients used are standard to the human infant formula industry (e.g. bovine milk proteins, plantderived oils, and vitamin, mineral, and trace element mixtures). The formulations also contained lecithin and monoglycerides, commonly utilized to provide emulsification and stabilization to infant formula. The basal swine-adapted infant formula was considered to be adequate to support normal growth and health of the piglets (Xu, 2003); see Blakemore et al. (2014a) for the specific formula composition used in this study.

M.L. Weiner et al./Food and Chemical Toxicology 77 (2015) 120–131

Dosing formulations were evaluated for homogeneity and stability and verification of CGN concentrations in swine-adapted infant formula using validated analytical methods (Blakemore et al., 2014b). 2.5. Test material administration Piglets received control or treated swine-adapted infant formula six times per day, every 3 hours from individual feeding containers at a dose volume of 500 mL/ kg body weight/day for 28 consecutive days. The dose volume was based on published literature (Dilger and Johnson, 2010) and experience during the 10-day rangefinding study (unpublished study). The control and test material formulations were allowed to warm to room temperature for at least 30 minutes prior to use. The dose concentrations were 0.5, 3.0, or 10.0 g/L/day for the three treatment groups (~83.33 mL/ kg body weight/dose). Fresh control and test material formulations were put into use once daily and feeding devices were changed daily for cleaning. Individual doses were based on the most recent body weight measurements. Feed consumption and body weight were determined daily on each animal. Feed efficacy was calculated for days 1–28. 2.6. Clinical pathology parameters Blood samples (~4.0 mL) were collected from the anterior vena cava through the thoracic outlet prior to dosing (non-fasted) for standard hematology, clinical chemistry and coagulation parameters. The samples were collected into tubes containing either potassium ethylene diamine tetraacetic acid (K3EDTA) for evaluation of hematology parameters or sodium citrate for evaluation of coagulation parameters. A serum separator was used for the clinical chemistry samples. Urinalysis on Day 29 was done by cystocentesis.

123

one-way analysis of variance (ANOVA) and utilized Dunnett’s comparison (Dunnett, 1955) of each treatment group with the control group. When Levine’s test was significant (p ≤ 0.01), comparisons with the control group were made using Welch’s t-test (Welch, 1937) with a Bonferroni correction (Snedecor and Cochran, 1982). Historical data indicate that leukocyte counts (total and differential) are not normally distributed; therefore, these data were log-transformed prior to being analyzed, as described above. Historical data for urinalysis endpoints indicate that these endpoints have unpredictable distribution characteristics; therefore, a non-parametric test was used. Feed efficiency was also statistically analyzed using this approach. Specifically, the data were rank transformed and Dunnett’s test was used on the transformed data to compare each treatment group having a non-zero sample size with the control group. Results of all pair-wise comparisons were reported at the 0.05 and 0.01 significance levels, and all endpoints were analyzed using two-tail tests.

3. Results 3.1. Homogeneity and stability of test material sample Analytical evaluations of the first preparation at the low and high dose concentrations from the top, middle and bottom strata, confirmed that good homogeneity at the established acceptance criteria (within 100 ± 20%) were met: recovery of 85%–89% with the percent relative standard deviation of 1.93% to 2.60% (≤5%). CGN was stable during the course of the experiment based on recoveries of 84.8%, 88.4%, and 89.6% for low, mid and high doses, respectively, at the start, and 82.2%, 93.8% and 84.8%, respectively, at the end of the study.

2.7. Immune system parameters The methods for these parameters were developed and validated in a separate, independent study using known immunotoxic compounds (Thorsrud et al., 2014). Using these methods, the potential effects of CGN were evaluated on immunophenotyping, cytokine levels (IL-1β, IL-6, IL-8, TNF-α) and immunohistochemistry (IL-8 and TNF-α). Blood samples were collected on Day 14 and prior to the terminal necropsy (Day 29). Animals were not fasted prior to blood collection. Samples were placed in tubes containing sodium heparin. Whole blood specimens were analyzed for the leukocyte phenotypes by flow cytometry. For each sample, absolute cell counts and cellular percentage values were determined for each particular cell phenotype. A dual platform method was used to determine absolute counts and data were recorded as cells/μL. The number of cells bearing each phenotype was determined independently. Tissues from the GIT were collected for evaluation of tissue cytokines for immunohistochemistry using the method of Thorsrud et al. (2014). 2.8. Toxicokinetic evaluation The methods developed for this study to evaluate swine plasma for the presence of the LMT of CGN, used poligeenan (PGN) as a surrogate for the LMT of CGN (Blakemore et al., 2014c). Based on CGN’s chemistry, reactivity and strong binding to proteins, PGN is an ideal surrogate for the low Mw components of CGN (LMT), which are more likely to be absorbed, compared to the high molecular weight components (Blakemore et al., 2014a). The method used swine plasma to which graded levels of PGN were added in vitro as a surrogate for CGN (LMT) in developing the bioanalytical method. The signal responses for PGN in the plasma samples of control and CGN-treated piglets were measured using a standard curve developed by spiking various concentrations of PGN into swine plasma (Blakemore et al., 2014c). No animals were fed or exposed to PGN.

3.2. Mortality and clinical observations There was no treatment-related mortality during the study, though there were a few incidental deaths early in the study, necessitating the use of replacement animals. These deaths were across dose levels, including controls, and were not related to test material administration. Soft and/or watery feces were the only clinical observations noted in the study, and were noted in all dose groups of both sexes, including controls. Although the incidence of soft/ watery stools was slightly increased in CGN-treated animals, these animals showed good growth during the study. Thus, the incidence in soft/watery stools in treated animals was not considered adverse. 3.3. Body weight and feed consumption Figs. 1 and 2 plot the results of mean body weight values for males and females, respectively, for each dose group for study days 1–28. There were no statistically significant changes in mean body weight for either male or female piglets at any CGN concentration, compared to the vehicle control. Male and female piglets showed good feed consumption during the 4-week dosing period without any adverse treatment-related changes.

2.9. Organ weights and macroscopic and histopathology evaluations At terminal necropsy, organ weights were recorded. Macroscopic examinations and histopathology evaluations were conducted at study termination. Histopathology evaluations were conducted on a standard selection of tissues fixed with 10% neutral buffered formalin (NBF) and stained with hematoxylin and eosin (H&E) stain to encompass the full range of potential affected organs (Table 2). Microscopic evaluations included multiple sections of the GIT fixed with 10% NBF for H&E stain; one section of the jejunum was fixed in 10% NBF for Periodic Acid– Schiff (PAS) staining to evaluate the goblet cells in the villi and crypts of the jejunum; and additional GIT samples were fixed in Carnoy’s solution for toluidine blue staining for mucosal mast cell evaluation.

3.4. Compound consumption Table 3 summarizes the compound consumption based on the daily feed consumption, daily BWs and administered concentration of CGN. The calculated compound consumption showed a good correlation to the targeted concentrations with similar ratios between low, middle and high doses for compound consumption and targeted doses.

2.10. Statistical analysis

3.5. Feed efficacy

Levine’s test (Milliken and Johnson, 1992) was used to assess the homogeneity of group variances for each of the following endpoints and for all collection intervals: body weights, food consumption, hematology (except leukocyte counts), coagulation, clinical chemistry, and organ weights (absolute weights, relative to body weights, relative to brain weights). When Levine’s test was not significant (p ≥ 0.01), a pooled estimate of the variance Mean Square Error (MSE) was computed from a

Feed efficacy was comparable among control and treated groups and did not show an adverse treatment-related effect. Mean feed efficacy values as percentage with Standard Deviations (SD) for days 1–28 were 15.19 ± 1.03, 16.87 ± 1.46, 14.69 ± 1.47 and 15.36 ± 2.57 for control, low, mid- and high dose males, respectively; and

124

M.L. Weiner et al./Food and Chemical Toxicology 77 (2015) 120–131

Fig. 1. Group mean male body weights for piglets from day 1 to day 28.

15.25 ± 0.98, 15.88 ± 1.67, 14.71 ± 1.07 and 14.89 ± 1.43 for control, low, mid- and high dose females, respectively.

3.6. Hematology, coagulation and clinical chemistry parameters Supplementary Table S1 summarizes hematology and coagulation results for males and females. There were no treatmentrelated effects among hematology and coagulation parameters in any treatment group on Day 14 or Day 29. Supplementary Table S2 summarizes the clinical chemistry results for males and females. There were no treatment-related effects among clinical chemistry parameters in any treatment group on Day 14 or Day 29.

3.7. Urinalysis Supplementary Table S3 summarizes the urinalysis results on Volume, Specific Gravity, pH and Osmolality for males and females. There were no effects on these parameters in CGN-treated males or females. In addition, glucosuria was noted in 1/6 males and 3/6 females receiving 2250 ppm CGN. No hyperglycemia was noted in these animals. No other treatment-related effects were observed among urinalysis parameters in any group.

3.8. Organ weights, macroscopic and histopathologic evaluation Supplementary Table S4 summarizes the mean organ weights for males and females. Mean absolute rectal organ weights of males

of the high dose group were decreased significantly (28%, p < 0.05), compared to controls. However, in the absence of associated microscopic findings; in the absence of similar changes in the remainder of the gastrointestinal tract, and in the absence of similar findings in high dose females, this finding was considered to be due to biological variation and not definitively test article related. All other organ weight findings in these swine were typical of those seen at the testing facility in swine of the same strain, sex, and age, or were considered incidental. There were no test material-related microscopic findings in any tissue. All microscopic findings were typical of those seen in swine of this strain, sex, and age, or were considered incidental. Fig. 3 shows the H&E stained colon of a control (Fig. 3A) and a high dose animal (Fig. 3B) at study termination under high power magnification (100×). There were no differences found between control and high dose animals. Toluidine blue staining of the gastrointestinal tissues for mast cell assessment showed no differences between control and high dose groups in either sex. Fig. 4 shows the toluidine blue-stained colon of a control animal (Fig. 4A) and a high dose animal (Fig. 4B), indicating no differences in histopathology. PAS staining of the jejunum for goblet cell assessment showed no differences between control and high dose groups in either sex. Fig. 5 shows the PAS-stained jejunum of a control animal (Fig. 5A) and a high dose group animal (Fig. 5B) at study termination. The Supplementary materials have additional photomicrographs from control and high-dose animals at the low magnification (40×) (see Figs. S1–S3 for tissues stained by H&E, Toluidine blue, PAS, respectively).

M.L. Weiner et al./Food and Chemical Toxicology 77 (2015) 120–131

125

Fig. 2. Group mean female body weights for piglets from day 1 to day 28.

3.9. Immunophenotyping, cytokine and immunohistochemical staining Table 4 summarizes the results of the immunophenotyping assessment. None of the cells types analyzed showed a biologically significant difference in relative cell percentage or absolute cell counts after CGN administration at any dose level. No biologically significant effects across gender or time interval (Day 14 or Day 29) were observed. Administration of CGN to neonatal piglets did not induce any changes to the peripheral leukocyte patterns, as evaluated by immunophenotyping, compared to controls. Values of all cellular subset data were within range of naïve Yorkshire-Crossbred piglet data (Thorsrud et al., 2014). Table 5 summarizes the results of the cytokine analyses for IL1β, IL-6, IL-8 and TNF-α. Concentrations of porcine IL-1β were below the Lower Limit of Quantitation (LLOQ) for all samples tested, except for one sample. A single high dose male animal had a detectible level of IL-1β. Concentrations of IL-6 were
sex or dose group. The concentration of IL-8 was measurable above the LLOQ in all samples tested. While there was an overall decline in IL-8 levels from Day 14 to Day 29, there was no discernible treatment-related effect on IL-8 levels. The concentration of TNF-α was measurable above the LLOQ in all samples tested with the exception of two samples. There was no treatment-related effect on levels of TNF-α. There were no effects of CGN on cytokine levels for all four cytokines measured, compared to control animals. Figs. 6 and 7 show sections of the proximal colon of control and high dose animals stained for IL-8 and TNF-α, respectively, at 100× magnification. Low magnification (40×) photomicrographs for these tissues can be found in the Supplementary materials as Figs. S4 and S5 for IL-8 and TNF-α, respectively. There were no effects of CGN on the immunohistochemical staining of any GIT tissues for IL-8 and TNF-α at any dose level. Positive controls for IL-8 and TNF-α immunohistochemical staining were derived from Thorsrud et al. (2014) and used in this evaluation, and are shown in the Supplementary material (Figs. S6 and S7) (courtesy of Thorsrud et al. 2014). 3.10. Toxicokinetic analysis

Table 3 CGN consumption in the 28-day study. Dose group: ppm

300 1000 2250 a

8.

CGN consumption: mg/kg bw/day ± SDa Males

Females

51.71 ± 4.06 192.86 ± 18.38 430.27 ± 67.33

55.57 ± 6.88 202.53 ± 12.72 448.25 ± 59.98

Number of measures = 9 for all groups, except the 1000 ppm males which was

Table 6 summarizes the individual animal results of the plasma analyses of the signal for PGN, used as a surrogate for the LMT of CGN (and expressed in the table as PGN/LMT) using the method developed and validated at MPI Research (Blakemore et al., 2014c). The toxicokinetic analyses were conducted both at the beginning of the study on day 3 and again at the end of the study on day 28 in order to assess if there was any potential accumulation of the LMT of carrageenan in plasma.

126

M.L. Weiner et al./Food and Chemical Toxicology 77 (2015) 120–131

(A)

(B)

Fig. 3. Hematoxylin and eosin stain (100×). (A) Left, Colon from piglet fed for 28 days with formula without CGN added. (B) Right, Colon from piglet fed for 28 days with formula with 2250 ppm CGN.

Although PGN has a lower Mw than CGN, CGN has a small fraction or LMT with Mw below 50 kDa (Blakemore et al., 2014a, 2014c). A spectral response was detected with the same transition and retention time as PGN in plasma samples of all male animals on days 3 and 28 and one high dose female animal on day 3. The remaining values for the females were below level of LLOQ of ≤10 μg/mL PGN signal on days 3 and 28. Values of the detected signal were relatively low (ranging from <10 to 47.3 μg/mL), comparable among the dose groups, and did not follow a relationship to treatment. This in vivo response cannot be PGN as no animal was exposed to PGN, and this in vivo response cannot be low molecular weight CGN (LMT) as the responses of the control animals with no CGN exposure were of similar magnitude. Thus, these signal concentrations must be some endogenous moiety in the plasma of preweaning piglets with a higher incidence in males.

(A)

4. Discussion The results confirm that CGN exposure during the preweaning period at maximally feasible concentrations does not have adverse effects on the growth and development of neonates. CGN had no effects on BW, BW gain, feed consumption, feed efficacy, hematology or clinical chemistry parameters or complete histopathology when fed in infant formula at 2250 ppm from LD2 for 28 consecutive days, a period (~4 weeks of age) corresponding to the time of nursing to weaning for piglets (Buelke-Sam, 2002). If CGN was irritating to the gastrointestinal tract of these piglets, then it would be expected that histopathological findings consistent with an irritant, such as inflammation, erosion, villous degeneration, edema, and even ulceration would be present (Parfitt and Driman, 2006), but none of these findings were seen in the CGN-dosed animals. Similarly, none of the special stains or

(B)

Fig. 4. Toluidine blue stain (100×). (A) Left, Colon from piglet fed for 28 days with formula without CGN. (B) Right, Colon from piglet fed for 28 days with formula with 2250 ppm CGN.

M.L. Weiner et al./Food and Chemical Toxicology 77 (2015) 120–131

(A)

127

(B)

Fig. 5. PAS stain (100×). (A) Left, Jejunum from piglet fed for 28 days with formula without CGN. (B) Right, Jejunum from piglet fed for 28 days with formula with 2250 ppm CGN.

immunohistochemical stains found any evidence of immune activation in the gastrointestinal tract. The high dose is considered the highest feasible dose concentration, based on viscosity. The present study extends the parameters studied previously in neonatal baboons (McGill et al., 1977) to include immune system organ histopathology, GIT immunohistochemical staining for IL-8 and TNF-α and GIT special stains (toluidine blue and PAS stains); immunophenotyping of blood leukocytes and blood cytokine evaluations for IL-1β, IL-6, IL-8 and TNF-α. The levels of IL-1β and IL-6 were below the limit of quantification regardless of dose or sex. The levels of IL-8 and TNF-α were determined for all the animals in all of the treatment groups. No significant change in the levels of IL-8 or TNF-α were observed regardless of dose or sex. Based on these observations, dosing piglets with CGN had no effect on circulating levels of the cytokines. Similarly, the immunohistochemical staining of the intestinal tract showed no significant local change in the levels of IL-8 or TNF-α positive staining, regardless of dose or sex. Based on the results on these immune system endpoints, CGN does not affect the immune system of the neonatal pig when fed at 2250 ppm in infant formula. The presence of glucosuria observed in the high dose animals at study termination was not correlated with an increase in blood glucose or renal pathology. Hall (2013) indicates that glucosuria may occur in animals if an agent induces a state of hyperglycemia that exceeds the renal capacity for glucose reabsorption. Since CGN-treated animals did not have hyperglycemia, this explanation is not the cause of the glucosuria. The literature indicates that other possible causes of glucosuria include a pharmacologic effect, which may include the modulation of glucose reabsorption in the proximal tubule or a direct renal tubular injury or dysfunction (Hall, 2013). Microscopic evaluation of the kidneys from the CGN-treated animals did not reveal any adverse pathological findings. Since there was a lack of adverse effects, the glucosuria did not have an adverse impact on the treated animals and is not considered toxicologically meaningful. Previous chronic and subchronic dietary studies in adult animals also show no clinical pathology findings, including hyperglycemia, and no organ histopathology at doses of 500 mg/kg/day CGN in Rhesus monkeys for 7.5 years (Unpublished Report from Albany Medical College, 1983); at 5% CGN in the diet of rats for 40 weeks (Abraham et al., 1985), and at 5% in diet of rats for 90 days, equivalent to 3394–3867 mg/kg/day CGN (Weiner et al., 2007). Glucose in urine was not evaluated in the infant baboon study by McGill et al. (1977).

Bhattacharyya et al. (2012) reported that C57BL/6J male mice exposed to CGN in drinking water at 10 mg/L for 18 days had an impaired ability to clear glucose from the blood based on a glucose tolerance test, compared to control male mice. Bhattacharyya et al. (2012) report a statistically significant lag in the rate of glucose uptake in blood, relative to control animals. It was also reported that CGNtreated animals had a statistically significant resistance to insulin based on an insulin tolerance test after 33 days of exposure. However, the plasma glucose levels in the CGN-treated mice under the test conditions of Bhattacharyya et al. (2012) were within the range of normal glucose values reported by Andrikopoulos et al. (2008) for this mouse strain. Thus, although it appears that CGN in drinking water produced a lag in glucose uptake and insulin resistance in adult mice, the absolute levels of glucose remained within a normal range. Moreover, Bhattacharyya et al. (2012) did not demonstrate a dose–response to CGN and did not characterize or analyze the CGN test sample. The test sample was a commercial sample from Sigma Chemical Co. McKim (2014) has shown that a Sigma Chemical Co. CGN sample may have as much as 36.5% unidentified sugar diluent. Since CGN is a high Mw polymer that binds tightly to food proteins, absorption of CGN across the GIT in adult animals is insignificant (Uno et al., 2001; Weiner, 2014). Thus, it is difficult to explain the findings of Bhattacharyya et al. (2012) on blood glucose and insulin tolerance which both presume absorption of CGN. The relevance of the Bhattacharyya et al. (2012) study to human diabetes is unknown and the accuracy and completeness of these data are not adequate for risk assessment, as discussed in the reviews of McKim (2014) and Weiner (2014), or relevant to the use of CGN in infant formula. Pigs have been shown to resemble humans with respect to the anatomy and physiology functions of the GIT and immune system (Barrow, 2012; Odel et al., 2014; Rothko˝tter et al., 2005). Gut permeability to macromolecules decreases before or during the early neonatal period. The term “gut closure” has been used to describe the decrease in gut permeability to proteins and other macromolecules at or following birth. For example, in humans gut closure occurs in utero (National Research Council, 2004); whereas piglets undergo gut closure within the first 48 hours after birth, resulting in very limited or no absorption of macromolecules (Barrow, 2012; Rothko˝tter et al., 2005). In this study the piglets were allowed to nurse for the first 48-hours after birth to allow for the beneficial effects of colostrum and for completion of gut closure. It was

128

Table 4 Summary of peripheral blood leukocyte phenotyping values for males and females on days 14 and 29.a Endpoint

Lymphocytes (% of leukocytes) Monocytes (cells/μL)** Monocytes (% of leukocytes) Mature T cells (cells/μL)* Mature T cells (% of lymphocytes) Helper T cells (cells/μL)* Helper T cells (% of lymphocytes) Cytotoxic T cells (cells/μL)* Cytotoxic T cells (% of lymphocytes) B cells (cells/μL)* B cells (% of lymphocytes)

Day 14 Day 29 Day 14 Day 29 Day 14 Day 29 Day 14 Day 29 Day 14 Day 29 Day 14 Day 29 Day 14 Day 29 Day 14 Day 29 Day 14 Day 29 Day 14 Day 29 Day 14 Day 29 Day 14 Day 29

Control: 0 ppm

CGN: 300 ppm

CGN: 1000 ppm

CGN: 2250 ppm

Male

Female

Male

Female

Male

Female

Male

Female

6120 ± 977 5945 ± 1152 49.16 ± 8.83 56.05 ± 11.77 1481 ± 709 978 ± 422 11.26 ± 2.68 8.88 ± 3.30 3787 ± 625 3916 ± 907 62.18 ± 5.39 65.88 ± 7.36 1915 ± 28 2013 ± 676 31.52 ± 3.83 33.48 ± 5.80 635 ± 173 745 ± 138 10.24 ± 1.96 12.87 ± 2.74 987 ± 291 864 ± 268 16.34 ± 4.94 14.45 ± 3.43

5797 ± 656 5844 ± 1605 49.83 ± 6.25 57.50 ± 9.54 1300 ± 249 833 ± 321 11.17 ± 2.24 10.65 ± 5.81 3584 ± 580 3808 ± 815 61.78 ± 5.89 67.87 ± 14.07 1756 ± 32 1931 ± 635 30.30 ± 4.09 33.43 ± 13.02 630 ± 146 773 ± 276 10.83 ± 1.91 13.42 ± 2.63 1141 ± 241 728 ± 255 19.72 ± 3.62 12.30 ± 2.14

5338 ± 841 6222 ± 549 46.35 ± 7.48 53.03 ± 8.97 1235 ± 184 1245 ± 494 10.43 ± 3.85 10.43 ± 3.85 3267 ± 660 4220 ± 455 61.40 ± 8.35 67.95 ± 6.41 1632 ± 3 1977 ± 244 30.65 ± 4.41 31.85 ± 3.61 559 ± 169 745 ± 88 10.47 ± 2.37 12.03 ± 1.67 976 ± 299 1061 ± 269 18.13 ± 3.67 17.12 ± 4.66

5635 ± 930 6183 ± 1383 49.82 ± 8.08 58.85 ± 14.42 1202 ± 171 1169 ± 316 10.57 ± 0.94 10.98 ± 2.27 3378 ± 298 3959 ± 1120 61.17 ± 10.31 64.52 ± 13.61 1781 ± 19 1970 ± 736 32.45 ± 7.08 31.85 ± 8.46 588 ± 73 773 ± 221 10.52 ± 0.97 12.53 ± 2.30 736 ± 225 786 ± 317 13.08 ± 3.48 12.63 ± 3.45

5456 ± 434 6053 ± 680 62.40 ± 7.06 62.40 ± 7.06 1202 ± 422 975 ± 167 12.04 ± 2.86 10.12 ± 2.23 3473 ± 451 4077 ± 604 64.26 ± 11.55 67.28 ± 6.38 1745 ± 34 1960 ± 312 32.28 ± 7.45 32.35 ± 3.37 725 ± 60 867 ± 137 13.40 ± 2.00 14.35 ± 1.95 982 ± 570 1025 ± 317 17.58 ± 9.21 17.20 ± 5.63

4413 ± 641 6247 ± 920 54.30 ± 7.27 71.25 ± 8.94 1135 ± 381 753 ± 169 13.57 ± 1.95 8.57 ± 2.44 2743 ± 312 4510 ± 645 62.78 ± 7.75 72.30 ± 4.10 1378 ± 1 2441 ± 697 31.45 ± 3.38 38.65 ± 6.31 505 ± 121 807 ± 201 11.43 ± 2.27 12.87 ± 2.75 820 ± 213 867 ± 164 18.48 ± 3.66 14.35 ± 4.47

5855 ± 908 7546 ± 1852 50.77 ± 7.20 59.14 ± 6.83 1381 ± 338 1083 ± 317 11.47 ± 2.33 9.90 ± 3.66 3517 ± 761 5080 ± 871 60.52 ± 11.57 67.68 ± 7.60 1777 ± 42 2444 ± 690 30.57 ± 6.73 32.92 ± 5.84 633 ± 133 1003 ± 331 11.02 ± 2.83 12.82 ± 2.87 821 ± 197 976 ± 427 13.93 ± 1.76 13.43 ± 3.61

5081 ± 543 4833 ± 1315 50.90 ± 6.49 49.97 ± 18.23 1253 ± 361 1008 ± 363 12.31 ± 2.83 9.71 ± 3.11 3084 ± 450 3244 ± 1068 60.76 ± 7.10 66.43 ± 10.01 1561 ± 29 1743 ± 889 30.70 ± 4.82 34.88 ± 11.55 532 ± 100 594 ± 275 10.47 ± 1.53 11.95 ± 2.23 982 ± 287 831 ± 466 19.36 ± 5.55 16.82 ± 7.24

a Mean ± SD for Groups 1–4. * Absolute cell count calculated based off % of lymphocytes for each cell type. ** Absolute cell count calculated based off % of leukocytes for each cell type.

M.L. Weiner et al./Food and Chemical Toxicology 77 (2015) 120–131

Lymphocytes (cells/μL)**

Interval

M.L. Weiner et al./Food and Chemical Toxicology 77 (2015) 120–131

Table 5 Cytokine levels of male and female animals on days 14 and 29.a Times analyzed

IL-1β: pg/mL Day 14: Male Day 14: Female Day 29: Male Day 29: Female IL-6: pg/mL Day 14: Male Day 14: Female Day 29: Male Day 29: Female IL-8: pg/mL Day 14: Male Day 14: Female Day 29: Male Day 29: Female TNF-α: pg/mL Day 14: Male Day 14: Female Day 29: Male Day 29: Female

Study group: CGN: ppm 0

300

1000

2250









852 (382) 998 (444) 316 (93) 397 (227)

687 (341) 1380 (939) 311 (136) 540 (481)

1160 (757) 944 (273) 408 (182) 223 (55)

823 (511) 1020 (570) 282 (99) 438 (225)

43 (7.1) 42 (6) 51.4 (15) 51.5 (32)

29.8 (12) 44.3 (10) 66.7 (27) 48 (6)

38.7 (13) 41 (9) 55.6 (29) 62 (22)

44.3 (9.6) 49 (13) 50.4 (19) 56.5 (27)

a
important to measure plasma CGN concentration in this study to determine if CGN is absorbed and systemically bioavailable in the neonate. Plasma was collected shortly after the initiation of dosing on study day 3 and again at the end of the study on day 28 in the animals designated for serum analyses (Groups 5–8). In order to measure CGN concentrations in piglet plasma, a method was developed by Blakemore et al. (2014c), and validated at MPI Research using PGN as a surrogate for the LMT of CGN. CGN binds strongly to plasma proteins making it extremely difficult for the quantitative separation of CGN from plasma for analysis. Since PGN, a low Mw polygalactan (Mw ~20 kDa), prepared from CGN by severe acid hydrolysis at elevated temperatures, and with a nearly identical molecular structure as CGN, and similar Mw as the LMT, does not bind as strongly

(A)

129

to plasma proteins and can be quantitatively extracted from plasma, it is an appropriate and ideal surrogate for the detection of the LMT of CGN for plasma toxicokinetic analysis (Blakemore et al., 2014c). The method of Blakemore et al. (2014c) has a bioanalytical lower limit of quantitation (LLOQ) of ≤10 μg/mL. The results of the plasma analyses found the presence of a response signal at the same transition and retention time as PGN in all male samples, regardless of dose group, including control males, and one high dose female animal. The levels of the response among males were highest in a control and a low dose male, 40 and 47 μg/mL, respectively. These results indicate that the response signal in the plasma samples is not the LMT of CGN, as none of the control animals were exposed to CGN and all the treated animals were exposed to increasing amounts of CGN, based on the respective dose levels, but the plasma levels of the detected signal showed no dose–response. In conclusion, the response observed would be more consistent with a signal produced from an unknown endogenous moiety in the plasma of preweaning piglets with a higher incidence in males. It is concluded that CGN is not absorbed from the neonatal gut in preweaning piglets. In addition, the results of the present study found no systemic effects of CGN to indicate that absorption of significant amounts of CGN occurred. The present study verifies the piglet model as a sensitive, robust model for studying the toxicological properties of ingredients in infant formula. The piglet model has been used extensively in research to study the effects of modulation of the constituents of formula and the effects of novel additives based on the similar physiology, development and ontogeny of the pig and humans. The evaluation of CGN in the piglet model demonstrates the safety of CGN to pre-weaning piglets. The estimated human exposure to 0.03% CGN (300 ppm) in formula has been estimated at 6.7–42.5 mg/kg/ day and at 21.7–142.5 mg/kg/day when used at 0.1% CGN (1000 ppm) in special needs formula (JECFA, 2008). The present study confirms and extends the work of McGill et al. (1977) in infant baboons; both studies independently derived a maximal feasible dose that is remarkably similar (~430 mg/kg/day). The exposure in both the piglet and infant baboon with no toxicological effects of ~430 mg/ kg/day is well above the dose consumed by human infants. JECFA (2014) calculated the margins of exposure (MOE) between the NOAEL of 430 mg/kg/day CGN and human infant exposures to be from 2 to 12 for infants on a mg/kg/day basis and from 2 to 8 on a concentration basis (2250 ppm) at 2–4 weeks of age. JECFA (2014)

(B)

Fig. 6. IL-8 immunohistochemical stain (100×). (A) Left, Colon from piglet fed for 28 days with formula without CGN. (B) Right, Colon from piglet fed for 28 days with formula with 2250 ppm CGN.

130

M.L. Weiner et al./Food and Chemical Toxicology 77 (2015) 120–131

(A)

(B)

Fig. 7. TNF-α immunohistochemical stain (100×). (A) Left, Colon from piglet fed for 28 days with formula without CGN. (B) Right, Colon from piglet fed for 28 days with formula with 2250 ppm CGN added.

concluded that CGN is safe in infant formula at the use rate of 300 ppm and at 1000 ppm, for special medical purposes, based on this study, the unlikelihood of absorption and the toxicology database on CGN. This study extends the safety data on CGN to include additional data on the GIT and immune system, and supports the safe use of CGN in infant formula for humans.

The authors thank FMC Corporation and the International Formula Council for funding this research.

consulting firm providing advice on toxicological and risk assessment issues to private firms. Dr. Weiner acted as a consultant to FMC Corporation under a cost reimbursable contract on this project. Coauthor, Mr. William Blakemore, also acted as a consultant to FMC Corporation under a cost reimbursable contract on this project. FMC Corporation is a manufacturer of CGN and products containing CGN. This paper is the professional work product of the co-authors. The FMC Corporation was given the opportunity to review the paper and offer comments on the paper. Those comments did not alter the professional opinions of the co-author. The conclusions drawn are not necessarily those of the FMC Corporation.

Conflict of interest

Transparency document

The first author of this paper is identified on the cover page. Myra Weiner is the owner and president of TOXpertise, LLC, a

The Transparency document associated with this article can be found in the online version.

Funding

Table 6 Individual animal results for the presence of PGN/LMT in plasma after 3 days and 28 days of administration.a Treatment group: ppm CGN

0

300

1000

2250

Animal number: female

549 550 551 555 556 557 561 562 563 567 568 569

Animal number: male

552 553 554 558 559 560 564 565 566 570 571 572

Signal concentration for PGN/LMT: (μg/mL) Females

Males

Day 3

Day 28

Day 3

Day 28

<10 <10 <10 <10 <10 <10 <10 <10 <10 <10 <10 16.8

<10 <10 <10 <10 <10 <10 <10 <10 <10 <10 <10 <10

40.6 32.4 15.4 47.3 13.7 32.7 23.4 32.5 14.5 28.9 27.7 36.0

29.9 15.0 28.8 25.8 <10 15.3 11.7 <10 14.5 13.4 15.4 16.0

a The signal for PGN, which was used as a surrogate for the LMT of CGN, is expressed in the table as PGN/LMT. The lower level of quantitation (LLOQ) was 10 μg PGN per mL with a quantification range from 10 (LLOQ) to an Upper Limit of quantitation (ULOQ) of 100 μg PGN per mL. The “Signal Concentration for PGN/LMT” is measured using the standard curve developed by spiking various concentrations of PGN into swine plasma (Blakemore et al., 2014c). As stated earlier, this in vivo response cannot be PGN as no animal was exposed to PGN, and this in vivo response cannot be low molecular weight CGN (LMT) as the response of the control animals with no CGN exposure are of similar magnitude to those of dosed animals. Based on the above, these signal responses must be from an unknown endogenous moiety in the plasma of preweaning piglets with a higher incidence in males. Note: PGN is used as a surrogate for the low Mw tail of CGN (LMT of CGN). However, note that no animals were exposed to PGN.

M.L. Weiner et al./Food and Chemical Toxicology 77 (2015) 120–131

Acknowledgements The authors thank Eunice M. Cuirle (FMC), Christopher J. Sewall (FMC), Theresa Hedrick (IFC) and James M. McKim (IONTOX, LLC) for technical discussions and to Brenda Frantz of MPI Research Laboratories for her diligent project management and coordination of this study. Thanks to FMC Corporation for donation of the test sample and to Abbott Nutrition for the preparation, packaging and donation of the swine-adapted infant formula formulations. Appendix: Supplementary material Supplementary data to this article can be found online at doi:10.1016/j.fct.2014.12.022. References Abraham, R., et al., 1985. Chronic and subchronic effects of various forms of carrageenan in rats. Ecotoxicol. Environ. Saf. 10, 173–183. Andrikopoulos, S., et al., 2008. Evaluating the glucose tolerance test in mice. Am. J. Physiol. Endocrinol. Metab. 295 (6), E1323–E1332. Barrow, P.C., 2012. Use of the swine pediatric model. In: Hoberman, A.M., Lewis, E.M. (Eds.), Pediatric Non-Clinical Drug Testing: Principles, Requirements, and Practice. John Wiley & Sons, Inc., Hoboken, NJ, pp. 213–230. Bhattacharyya, S., et al., 2012. Exposure to the common food additive carrageenan leads to glucose intolerance, insulin resistance and inhibition of insulin signaling in HepG2 cells and C57BL/6J mice. Diabetologia 55 (1), 194–203. Blakemore, W.R., et al., 2014a. Carrageenan analysis. Part 1: characterisation of the carrageenan test material and stability in swine-adapted infant formula. Food Addit. Contam. Part A Chem. Anal. Control Expo. Risk Assess. 31 (10), 1661–1669. doi:10.1080/19440049.2014.955536. Blakemore, W.R., et al., 2014b. Carrageenan analysis. Part 2: quantification in swine-adapted infant formula. Food Addit. Contam. Part A Chem. Anal. Control Expo. Risk Assess. 31 (10), 1670–1672. doi:10.1080/19440049.2014.955537. Blakemore, W.R., et al., 2014c. Carrageenan analysis. Part 3: quantification in swine plasma. Food Addit. Contam. Part A Chem. Anal. Control Expo. Risk Assess. 31 (10), 1673–1677. doi:10.1080/19440049.2014.955538. Buelke-Sam, J., 2002. Postnatal Evaluation in Developmental and Juvenile Toxicity Studies, presented at the Henry Stewart Conference; Meeting FDA, EPA, OECD and ICH Regulatory Requirements for Reproductive Toxicology Data Submissions. Georgetown University Conference Center, Washington, DC. Canada Gazette, 2004. 138, No. 20: May 15, Item A.5. Codex Committee on Food Additives, 2008. Report of the 40th session (2008) of the CCFA to the 31st session (2008) of the Codex Alimentarius Commission. ALINORM 08/31/12, May. Davidson, T.L., Swithers, S.E., 2005. Food viscosity influences caloric intake compensation and body weight in rats. Obes. Res. 13, 537–544. Dilger, R.N., Johnson, R.W., 2010. Behavioral assessment of cognitive function using a translational neonatal piglet model. Brain Behav. Immun. 24, 1156–1165. Dunnett, C.W., 1955. A multiple comparison procedure for comparing several treatments with a control. J. Am. Stat. Assoc. 50, 1096–1121. European Medicines Agency (EMEA), 2008. Guideline on the need for non-clinical testing in juvenile animals of pharmaceuticals for paediatric indications. EMEA Guideline, 24 Jan 2008. Farmer, C., Questnel, H., 2009. Nutritional, hormonal and environmental effects of colostrums in sows. J. Anim. Sci. 87 (Suppl. 1), 56–65. Guilloteau, P., et al., 2010. Nutritional programming of gastrointestinal tract development. Is the pig a good model for man? Nutr. Res. Rev. 23, 4–22.

131

Hall, R.L., 2013. Principles of clinical pathology. In: Sahota, P.S., Popp, J.A., Hardisty, J.F., Gopinath, C. (Eds.), Principles of Clinical Pathology in Toxicologic Pathology: Nonclinical Safety Assessment. CRC Press, Boca Raton, FL, pp. 133–174. Helm, R.M., et al., 2007. Diet regulates the development of gut-associated lymphoid tissue in neonatal piglets. Neonatology 91, 248–255. International Conference on Harmonisation (ICH) Harmonised Tripartite Guidelines, 2010. Guidance for industry on non-clinical safety studies for the conduct of human clinical trials and marketing authorization for pharmaceuticals. ICH M3(R2), 2010 Jan. International Programme on Chemical Safety (IPCS), 2009. Environmental Health Criteria 240, Principals and methods for the risk assessment of chemicals in food. . JECFA: Joint FAO/WHO Expert Committee on Food Additives, 2008. 68th Report of the Joint FAO/WHO Expert Committee on Food Additives; Evaluation of Certain Food Additives and Contaminants. WHO Technical Report Series 947, 32–37. JECFA: Joint FAO/WHO Expert Committee on Food Additives, 2014. 79th Meeting, Geneva, 17–26 June 2014, Summary and Conclusions, JECFA79/SC. . Klobasa, F., et al., 1987. Composition of sow’s milk during lactation. J. Anim. Sci. 64, 1458–1466. McGill, H.C., et al., 1977. Carrageenan in formula and infant baboon development. Gastroenterology 73, 512–517. McKim, J.M., 2014. Food additive carrageenan: part I: a critical review of carrageenan in vitro studies, the potential pitfalls, and implications for human health and safety. Crit. Rev. Toxicol. 44 (3), 211–243. Milliken, G.A., Johnson, D.E., 1992. Analysis of Messy Data. Chapman and Hall, London. National Research Council, 2004. Infant Formula: Evaluating the Safety of New Ingredients. National Academies Press, Washington, DC. Odel, J., et al., 2014. The suckling piglet as an agrimedical model for the study of pediatric nutrition and metabolism. Annu. Rev. Anim. Biosci. 2, 419–444. Parfitt, J.R., Driman, D.K., 2006. Pathological effects of drugs on the gastrointestinal tract: a review. Hum. Path. 38, 527–536. Penninks, A.H., et al., 2012. Juvenile immunodevelopment in minipigs. In: Hoberman, A.M., Lewis, E.M. (Eds.), Pediatric Non-Clinical Drug Testing: Principles, Requirements, and Practice. John Wiley & Sons, Inc., Hoboken, NJ, pp. 231–254. Rothko˝tter, H.J., et al., 2005. The pig as a model for developmental immunotoxicology. In: Holladay, S.D. (Ed.), Developmental Immunotoxicology. CRC Press, Boca Raton, FL, pp. 103–114. Snedecor, G.W., Cochran, W.C., 1982. Statistical Methods, seventh ed. Iowa State University Press, Ames, IA. Thorsrud, B., et al., 2014. The development and validation of methods for evaluating the immune system in pre-weaning piglets. Toxicol. Sci. 138 (1), 845, Abstract. United Stated Food and Drug Administration (FDA), 2000. Guidance for Industry: E11 Clinical Investigation of Medicinal Products in the Pediatric Population. . United States Food and Drug Administration (FDA) Center for Drug Evaluation and Research (CDER), 2006. Guidance on Nonclinical Safety Evaluation of Pediatric Drug Products Nonclinical Safety Evaluation of Pediatric Drug Products. U.S. Department of Health and Human Service, Rockville, MD. Uno, Y., et al., 2001. Molecular weight and fecal excreted quantity of carrageenan administered to rats in blended feed. Jpn. J. Food Chem. 8, 83–93. Weiner, M.L., 2014. Food additive CGN: part II: a critical review of carrageenan in vivo safety studies. Crit. Rev. Toxicol. 44 (3), 244–269. Weiner, M.L., et al., 2007. A 90-day dietary study of kappa carrageenan with emphasis on the gastrointestinal tract. Food Chem. Toxicol. 45, 98–106. Welch, B.L., 1937. The significance of difference between two means when the population variances are unequal. Biometrika 29, 350–362. Xu, R.J., 2003. Xu, R.J., Cranwell, P. (Eds.), Composition of Porcine Milk in the Neonatal Pig: Gastrointestinal Physiology and Nutrition. Nottingham University Press, Nottingham, UK, pp. 213–246.