Retinoid- and carotenoid-enriched diets influence the ontogenesis of the immune system in mice

Abstract

Vitamin A (VA) has been identified as an important factor for the development of the immune system, especially during ontogenesis. It has been shown that antibody secretion and proliferation of lymphocyte populations depend on retinoids. In the present study we investigated the influence of a base VA diet and diets enriched with VA, β-carotene and lycopene, on the ontogenesis of the immune system in mice. We examined the absolute and relative concentrations of splenic B lymphocytes (CD45R/B220), T lymphocytes (CD3⁺) and their subpopulations (CD4⁺ and CD8⁺), and measured serum immunoglobulin G (IgG) concentrations in the offspring of supplemented dams at different ages (1, 3, 5, 7, 14, 21 and 65 days). The experimental diets resulted in higher numbers of T and B lymphocytes after VA and carotenoid enrichment, when compared, at various time-points, with the base diet. Higher values of total serum IgG were found in the β-carotene-enriched diet group on day 7. On days 7 and 14, the enriched diets induced significant alterations in the percentages and total numbers of splenic lymphocytes in comparison to the base diet. Our results confirm that supplementation with VA and carotenoids affect the immune-cell function during ontogenesis and suggest a possible role of these nutritional factors on the development of the immune system.

Introduction

The sequence of events occurring during T-cell development in humans and mice is comparable, thus murine studies are applicable to our understanding of the immune response in neonatal development. However, it is important to consider the differences in the kinetics of T-cell development between mice and humans.¹

Development of the immune system starts during the embryonic stage. The influx of lymphoid cells occurs, in a cyclical manner, from progenitor cells at programmed times of fetal development, leading to self-perpetuating and differentiating populations of T lymphocytes.^2,3 The incoming prethymic murine cells derive first from the earliest haematopoietic centres (yolk sac in mice), later on from fetal liver and finally, during postnatal development, from the bone marrow.^4–6 In terms of the lymphocyte maturation rate in the spleen, the T-cell frequency increases about twofold during the first 5 days of life; by day 10, splenic T cells represent only 5% of the total percentage in adults; the adult rates are reached by day 16. Mice neonates have a complete, but naïve, T-cell repertoire and are competent to recognize a full array of antigens; however, their number of antigen-presenting cells (APCs) and distribution in the peripheral lymphoid organs is not fully developed.¹ Splenic B lymphocytes show a twofold increase in percentage during the first 2 weeks after birth; the numbers peak on day 16 and this level is maintained until adulthood.⁷ In terms of antibody production, the earliest B-cell precursors express only mRNA encoding immunoglobulin M (IgM), the other isotypes begin to appear 3 days after birth.⁸

Neonatal mice are exposed to a high level of maternal immunoglobulin G (IgG); this exposure continues for 2–3 weeks after birth by absorption of maternal milk immunoglobulin in the small intestine.⁹ Natural antibodies of maternal origin may play a role in the B-lineage development of the offspring, because the infusion of normal IgG to adult mice markedly modulates B-lineage cells and antibody repertoire.^10–12 IgG concentrations in serum from newborn mice show a threefold increase from day 3 to day 7, which indicates a marked effect of milk transmission. IgG concentrations peak on day 14 and decrease from days 21–35 after birth.¹³

Retinoids play an important role in cell growth, differentiation and gene regulation, and are important nutritionally relevant modulators of the immune system.¹⁴ The active vitamin A (VA) derivative, retinoic acid (RA), increases thymocyte differentiation^15,16 and enhances the lymphocyte response to mitogens.^17–19 Retinoids have also been shown to stimulate antibody production in vivo and in vitro.^20–22 On the other hand, VA-deficient mice show a reduced IgG1 and IgG3 response, which is restored by RA.^23,24 RA has been demonstrated to inhibit B-cell proliferation and immunoglobulin E (IgE) synthesis in cell-culture studies via RARs receptors.²⁵ Maternal VA supplementation (intraperitoneal) increased serum IgM and IgG1 levels in the progeny.²⁶ Respectively, the transfer of VA from mother to offspring by milk, and the VA status of dams and suckling neonates, is influenced by maternal VA intake during lactation.^27,28

Nutritionally relevant carotenoids, such as lycopene and β-carotene, have been proposed to enhance cell-mediated immune responses,^29–32 but the exact mechanism of action is still unclear. One mechanism associated with the carotenoid-modulating action may be enhancement of the cell-surface expression of the major histocompatibility complex (MHC) class II molecules,³³ owing to the ability of carotenoids to quench singlet oxygen.³⁰ Another mechanism involved in the prevention of oxidative damage has been suggested to be the modulation of prostaglandin E₂ production via cyclooxygenase and lipooxygenase pathways.³⁴

The aim of our study was to investigate whether nutritionally relevant retinoids and carotenoids affect the development of the cellular and humoral immune system via dietary intake (maternal milk and continuing throughout weaning). We studied the effects of VA and carotenoid supplementation on the immune system in a mouse model at different postnatal and adult stages.

Materials and methods

Experimental animals and diets

All animal experiments were performed in the facilities of the Max Rubner Laboratory of the German Institute of Human Nutrition (DIFE, Bergholz-Rehbrücke, Germany). The experiment was approved by the respective ethical authorities from the Land Brandenburg.

NMRI adult (8–10 week old) mice (Mus musculus) were obtained from Tierzucht Schönwalde (Schönwalde, Germany). The animals were housed under controlled conditions of room temperature (21 ± 1°), relative humidity (55 ± 5%) and 12-hr light/dark cycles (with light between 10·00 hr and 22·00 hr). The animals were mated between 18·00 hr and 09·00 hr and, after detection of a vaginal plug, the pregnant mice were separated and housed in individual cages. On day 21 after birth, the pups were separated from their mother and housed in different cages.

The dams and pups were fed according to the scheme in Table 1. Food and water was administered ad libitum. The composition of the base diet was 120 g/kg casein, 100 g/kg sucrose, 640 g/kg wheat starch, 50 g/kg sunflower oil, 20 g/kg cellulose, 20 g/kg mineral mix (Mineral-Spurenelemente-Vormischung C1000 mix; Altromin, Lage, Germany) and 20 g/kg vitamin mix (Vitamin-Vormischung C1000 mix; Altromin, Lage, Germany).³⁵ The vitamin A content is shown in Table 1 and no detectable amounts of carotenoids were reported to be present in the vitamin mix.

Table 1.

Feeding scheme (vitamin A/carotenoid content of the diets)

Base diet	4500 RE/kg of vitamin A from the vitamin mix35
Vitamin A-supplemented diet	4500 RE/kg of vitamin A from the vitamin mix and 324 000 RE as retinyl palmitate(2160 mg/kg diet) supplemented44, 45
β-Carotene-supplemented diet	4500 RE/kg vitamin A from the vitamin mix and 50 000 RE as β-carotene (300 mg/kg diet supplemented)46
Lycopene-supplemented diet	4500 RE/kg vitamin A from the vitamin mix and lycopene (300 mg/kg diet supplemented)46, 47

RE, Retinol equivalents. Vitamin A was administered as retinyl palmitate purchased from Sigma Chemical Co. (Taufkirchen, Germany); β-carotene (beadlets containing 10%β-carotene) was a gift from Hoffman-LaRoche, (Basel, Switzerland); and lycopene, Lyc-O-Mato (70·9% lycopene), was a gift from LycoRed (Beer-Sheva, Israel).

The litter size was standardized to eight pups to avoid fluctuations in the pups' dietary intake. The mothers were randomly assigned to a specific diet and began to receive it 1 day before giving birth; the dams received the experimental diets until day 21 after birth (the day of birth was considered as day 0), and the pups received the experimental diets after separation from their mothers (weaning).

Sample collection

On days 1, 3, 5, 7, 14, 21 and 65 after birth, the progeny were analysed. On days 1 and 3, seven or eight organs/blood were pooled to obtain a minimum amount of material for analysis; one sample (n = 1) was assigned for those days. On day 5, four organs were pooled (n = 2). On day 7, the pool consisted of two organs (n = 4). On days 14, 21 and 65, the organs and blood were collected from individual mice. On the assigned days, the animals were killed by decapitation. Blood was centrifuged at 1300 g for 3 min. Serum was obtained and stored at −20° until required for analysis. Spleen was collected for immunological examination, stored on ice and analysed on the same day.

Cell preparation and immunophenotyping

Mononuclear cells were harvested from the spleen by gently pressing the tissue through a fine nylon mesh sieve (100 µm) (Falcon^®; BD Pharmigen, San Diego CA) and the cell suspension was washed with phosphate-buffered saline (PBS) in the absence of Ca²⁺ and Mg²⁺. The spleen mononuclear cells were counted and 1–5 × 10⁶ cells were stained with fluorochrome-labelled monoclonal antibodies (mAbs). The cell suspensions were mixed with 100 µl of labelled mAb in buffer [PBS containing 10% bovine serum albumin (BSA)] (1 : 200 dilution). Two-colour staining using fluorescein isothiocyanate (FITC) and phycoerythrin (PE) was performed using the following mAb combinations: FITC-conjugated rat IgG2a; κ monoclonal immunoglobulin isotype control with R-PE-conjugated rat IgG2a; κ monoclonal isotype standard; FITC-conjugated rat anti-mouse CD8a(Ly-2) with R-PE-conjugated rat anti-mouse CD4(L3T4); and FITC-conjugated rat anti-mouse CD3 molecular complex with R-PE-conjugated rat anti-mouse CD45R/B220. All mAbs were purchased from BD PharMingen. After incubation (20 min at 4° in the dark), the cells were washed. The supernatant was removed and the pellet was resuspended in 500 µl of PBS containing 2% paraformaldehyde (Sigma, Munich, Germany). The fixed cells were stored at 4° until required for cytometric analyses.

The cells were acquired using a fluorescence-activated cell sorter (FACSscan; Becton-Dickinson, San Jose, CA) and analysed using cellquest 3.3 software (Becton-Dickinson). A gate was set to record 50 000 events, using side (SSC) and forward (FSC) light scatter parameters to permit identification of the morphological cell populations. The lymphocytes were gated again and analysed using FL1/FSC and FL2/FSC parameters. The FL1/FSC dot-plot expressed FITC and the FL2/FSC expressed PE. For each staining experiment the appropriate isotype controls were included to evaluate non-specific staining. The different lymphocyte subpopulations, extracted from the spleen, were calculated as absolute and percentage measurements, based on calculations of the cellquest 3.3 software.

Immunoglobulin analysis

The IgG assay was performed using an enzyme-linked immunosorbent assay (ELISA) protocol, according to the standard procedures in place in our laboratory. ELISA plates were coated overnight with goat anti-mouse IgG (heavy and light chain specific; Calbiochem, Darmstadt, Germany), diluted 1 : 1400 in a carbonate-coating buffer. The plates were washed with PBS and then blocked for 2 hr with PBS containing 1% BSA. After washing, the standard and samples were applied and incubated for 1 hr. As a standard, mouse IgG serum (Calbiochem) was used at a 1 : 200 000 dilution in PBS. The samples from days 1–7 consisted of pooled sera (see sample collection). Different sample dilutions were made, according to the age of the mice: day 1, 1 : 5000; day 3, 1 : 25 000; days 5 and 7, 1 : 50 000; days 14, 21 and 65, 1 : 100 000. The plates were washed and then incubated with a second antibody – goat anti-mouse IgG heavy and light chain biotin-conjugated (Calbiochem), at a 1 : 2000 dilution in PBS. Streptavidin (Dako, Denmark) was used to complete the antibody detection. o-Phenylenediamine dihydrochloride, one tablet/10 ml of citrate buffer (Sigma) was added as a substrate to allow development of the colour reaction. After 10 min, the optical density of the samples was read in a ELISA Microplate Reader™ (Bio-Rad, Munich, Germany) at 490 nm.

Statistics

Data for the lymphocyte percentages, cell numbers and ratios, and the IgG concentrations are expressed as means and standard errors. Statistical analysis was performed using the SPSS 11.0 (SPPS Inc., Chicago, IL) software for Windows; a P-value of 0·05 was used to determine statistical significance. Cell percentages and numbers, and IgG concentrations, were compared among the treatment groups using univariate analysis of variance (anova) models to estimate the effects of diet, age and diet–age interaction. Pairwise multiple comparisons among the treatment groups were conducted by using the Dunnett's t-test.

Results

In NMRI mice, the relative percentages, as well as the numbers of lymphocyte subpopulations, increased, as expected, with age, regardless of the dietary treatment given to dams and pups (Figs 1–4). However, the magnitude of increase in lymphocyte numbers and percentages, observed with increasing age and at specific time-points, was influenced by dietary treatment (% CD3⁺, P = 0·01; % CD4⁺, P = 0·05; % CD45R/B220⁺, P = 0·02). In all groups the serum IgG concentrations increased during the week after birth, reaching adult levels by day 14 and slightly decreasing after weaning (Fig. 5). In the control (base diet) group, there was a peak in the T-cell subset (CD4⁺) on day 7 followed by a rapid increase of T-cells (CD3⁺, CD4⁺) after day 14. By contrast, there was a lower proportion, as well as a lower total cell number, of cytotoxic cells (CD8⁺) than T-helper cells (CD4⁺); however, an increase in cell number/percentage from day 14 was observed, which was maintained until adulthood. Additionally an increase in CD4⁺ cell numbers was also indicated by a high CD4 : CD8 ratio of the base diet (control) group, observed on day 7 (11·3 ± 1·8), in comparison to lower ratios on day 14 (4.6 ± 2.4), day 21 (3.3 ± 0.2) and day 65 (4.4 ± 0.8) (Table 2, A). In all experimental groups, and during the whole study period, the percentage and numbers of B cells were higher than those of T cells; this was also observed directly after birth (CD45R/B220⁺: 1·26%/3·6 × 10⁴ total splenic mononuclear cells versus CD3⁺: 0·11%/3·0 × 10³ total splenic mononuclear cells). CD4 : CD8 ratios between the different groups on day 7 showed high variations, whereas on days 14, 21 and 65, all ratios were similar (Table 2, A). The CD3 : CD45R/B220 ratios after dietary supplementation showed small, non-significant variations on days 7, 14 and 21 and similar ratios on day 65 (Table 2, B).

Figure 1.

Percentage of total splenic mononuclear cells in mice after receiving the base and vitamin A (VA)-supplemented diet and during postnatal development. CD3⁺, CD4⁺, CD8⁺ and CD45R/B220 cells were detected in the spleen on days 1 (n = 1), 3 (n = 1), 5 (n = 2), 7 (n = 4), 14 (n = 4), 21 (n = 4) and 65 (n = 4) after birth. The percentage of positive cells is shown. *P < 0·05.

Figure 4.

Total splenic mononuclear cells in mice after receiving the base and supplemented β-carotene and lycopene diets and during postnatal development. CD3⁺, CD4⁺, CD8⁺ and CD45R/B220 cells were detected in the spleen on days 1 (n = 1), 3 (n = 1), 5 (n = 2), 7 (n = 4), 14 (n = 4), 21 (n = 4) and 65 (n = 4) after birth. The percentage of positive cells is shown. *P < 0·05.

Figure 5.

Total serum immunoglobulin G (IgG) concentrations (mg/ml) in mice neonates during postnatal development. The IgG concentrations were determined in mice after receiving the base, vitamin A (VA)- and carotenoid-supplemented diets on days 1 (n = 2), 3 (n = 2), 5 (n = 5), 7 (n = 4), 14 (n = 4), 21 (n = 4) and 65 (n = 4) after birth. *P < 0·05.

Table 2.

Ratios from percentage splenic lymphocyte subpopulations

Diet	Age
	Day 7	Day 14	Day 21	Day 65
A. CD4 : CD8*
Base	11·3 + 1·8	4·6 + 2·4	3·3 + 0·2	4·4 + 0·8
Vitamin A supplemented	63·6 + 57·1†	6·9 + 2·6	4·9 + 0·2	6·6 + 1·1
β-Carotene supplemented	4·6 + 1·4	4·2 + 0·7	4·0 + 0·6	4·1 + 1·1
Lycopene supplemented	4·1 + 1·9	6·7 + 2·9	3·4 + 0·3	4·3 + 0·5
B. CD3 : CD45R/B220*
Base	0·04 + 0·0	0·02 + 0·0	0·48 + 0·4	0·53 + 0·1
Vitamin A supplemented	0·21 + 0·2	0·04 + 0·0	0·03 + 0·0	0·61 + 0·3
β-Carotene supplemented	0·18 + 0·1	0·18 + 0·0	0·20 + 0·0	0·53 + 0·1
Lycopene supplemented	0·16 + 0·1	0·02 + 0·0	0·22 + 0·1	0·52 + 0·1

^*Values are expressed as mean + standard error of the mean (SEM).

^†P < 0·05.

VA supplementation

Higher percentages of CD3⁺ on day 7 (P = 0·01) and of CD4⁺ on day 14 (P = 0·01) were observed when compared to the percentages of T cells in the base diet (Fig. 1). However, on other days no significant differences were observed for the percentages of the different T-cell subpopulations. In terms of total cell numbers, mice receiving VA-supplemented diets showed higher levels of all subsets on days 3, 5 and 14 (Fig. 2); CD4⁺ values on day 3 (P = 0·05) and CD8⁺ values on day 5 (P = 0·01) were statistically significantly higher. Increased CD4 : CD8 ratios were observed at all time-points, but statistical significance was only obtained on day 7 (P < 0·05). In addition, an increased CD3 : CD45R/B220 ratio on day 7 and a decreased ratio on day 21 were observed (Table 2). By contrast, the total serum IgG concentrations did not differ after VA supplementation compared with the base feed group (Fig. 5).

Figure 2.

Total splenic mononuclear cells in mice after receiving the base and vitamin A (VA)-supplemented diets and during postnatal development. CD3⁺, CD4⁺, CD8⁺ and CD45R/B220 cells were detected in the spleen on days 1 (n = 1), 3 (n = 1), 5 (n = 2), 7 (n = 4), 14 (n = 4), 21 (n = 4) and 65 (n = 4) after birth. The percentage of positive cells is shown. *P < 0·05.

β-carotene supplementation

β-carotene supplementation showed a significant effect on percentages of the CD4⁺ (P < 0·05), CD8⁺ (P = 0·03), and CD45R/B220 (P = 0·04) subpopulations in comparison to those of the base diet. Significantly increased percentages and total numbers of CD3⁺ and CD8⁺ cells on days 7 and day 14 were observed by comparison to controls (Figs 3 and 4). However, the CD4 : CD8 ratios remained essentially unchanged between day 7 and day 65 and were similar to those of the base diet group at most of the time-points studied, except for day 7 when a lower CD4 : CD8 ratio was observed. The CD3 : CD45R/B220 ratio was increased on days 7 and 14 in comparison to that of the base diet group, but it did not reach statistical significance (Table 2, B). After supplementation, the serum IgG concentration on day 7 was significantly higher (P = 0·03) than that of the base diet group.

Figure 3.

Percentage of total splenic mononuclear cells (SMC) in mice after receiving the base and supplemented β-carotene and lycopene diets and during postnatal development. CDCD3⁺, CD4⁺, CD8⁺ and CD45R/B220 cells were detected in the spleen on days 1 (n = 1), 3 (n = 1), 5 (n = 2), 7 (n = 4), 14 (n = 4), 21 (n = 4) and 65 (n = 4) after birth. The percentage of positive cells is shown. *P < 0·05.

Lycopene supplementation

The percentage and total number of CD3⁺ cells on day 7 (P < 0·05, P = 0·03) were significantly higher after lycopene supplementation in comparison to those of the base diet group (Fig. 3). The CD4 : CD8 ratios throughout the study period showed no statistical significance when compared with those of the base diet group (Table 2, A). An increase in the percentage and total cell numbers of B lymphocytes was observed on day 7 (marginal P = 0·06 for both). On day 14, B cells were elevated (P = 0·04) in relation to the base diet group (Fig. 4). CD3 : CD45R/B200 ratios were higher compared to those of the base diet group on days 7 and 21, but this was not statistically significant (Table 2). In general, when comparing the lycopene-supplemented diet with the base diet, higher numbers of T- and B-cell subsets were observed on days 21 and 65 (P = not significant).

Discussion

In the present study we investigated the effect of different diets given to dams and pups, on the splenic lymphocyte populations and total serum IgG levels in NMRI mice. We found comparable results for lymphocyte counts and total IgG production in the different dietary groups, as reported in previous studies.^13,36 While lymphocyte appearance, in terms of age, was in agreement with previous studies, the percentages and total cell numbers of lymphocyte subpopulations were lower in our study than those reported in CD-1 and C57Bl mice,^7,36 regardless of the dietary intake. The differences, reported here, in the percentages and cells numbers in the different dietary groups were also reflected by the CD4 : CD8 and CD3 : CD45R/B220 ratios, which differed from those reported in previous studies.^7,36 This may be caused by differences in mouse strain and dietary treatment, e.g. in the study of Fagoaga the dams and progeny were fed a high protein diet,³⁶ while the mice in our study were fed a normal protein diet.

The reduced percentages and total cell numbers of CD4⁺ and CD8⁺ observed during the neonatal period suggested that neonates have reduced T-helper and cytotoxic activity, which is in agreement with previous work.⁶ We and others⁷ observed that the CD4 : CD8 ratio in neonates is twofold higher than that of adult mice. A reduced antibody (IgG) production was also observed in neonates, which can be explained by the reduced B-cell numbers, but may also be a result of the reduced T-helper cell activity.

In general, we observed, throughout the study, a steady increase of cell numbers in mice receiving the base diet,⁷ which was comparable to the mice receiving the supplemented diets. Surprisingly, the retinyl palmitate-supplemented diet resulted in a decreased number of CD3⁺, CD8⁺ and CD45R/B220⁺ cells on day 7. This decrease may be the result of an excess of retinoids, which lead to increased apoptosis of lymphocytes.³⁷ The percentage distribution of T lymphocytes mainly increased after day 14, whereas B-lymphocyte numbers increased steadily after day 1, as indicated by a strong increase of the CD3 : CD45R/B220 ratio after day 21.

Our results show that VA and carotenoid supplementation affect the development of the mouse immune system by influencing the numbers and percentages of different lymphocyte subpopulations, as well as the concentration of plasma IgG during the postnatal period. Our attention is focused on VA and β-carotene supplementation on days 7 and 14, when significant changes in the percentages and total cell amounts of CD3⁺, CD4⁺ and CD8⁺ were observed. An effect on VA transmission via the maternal milk^27,28,38 may be linked to the alteration of immune parameters. In another study (A. L. Garcia et al., in preparation) we analysed the serum VA levels in pups from VA-supplemented dams and found increased VA levels during the lactation period at days 3, 5, 7 and 14. These observations suggest a physiological effect of VA supplementation in the progeny via the mothers' milk. The effects of VA supplementation reported here may be related to the stimulation of the RXR pathway, which has been shown to enhance T helper 2 (Th2) cell development.³⁹ However, it should also be considered that lymphocyte expansion might be related to the stimulatory/inhibitory effects on lymphocyte physiology attributed to the retro-retinoid metabolites (14-hydroxy-retro-retinol and anhydroretinol).^40,41 Furthermore, the immunomodulatory effect of β-carotene treatment may also be attributed to pro-vitamin A properties, which concurs with previous studies carried out in humans (where an increased number of helper cells was observed) and is in agreement with our experiments showing increased numbers of CD3⁺, CD4⁺ and CD8⁺ cells.⁴²β-carotene has been proven to enhance immune functions, via an independent pathway, by enhancing the cell-surface expression of APC cells, such as the adhesion molecules intercellular adhesion molecule-1 and leucocyte-function-associated antigen-3.⁴³ Another possible pathway could be via the inhibitory action of β-carotene on the cyclooxygenase or lipooxygenase activities.³⁴

Lycopene supplementation also exerted immunomodulatory effects, but at later time-points: days 21 and 65. This carotenoid has been previously studied^32,43 and has been proven to act on the immune system (possibly by similar carotenoid-like actions) but is less active than β-carotene. We may interpret our results in terms of time effect, as increased numbers of cells in all subsets were observed, but only after a longer supplementation time. After lycopene supplementation, the higher numbers of B cells observed were in agreement with higher numbers of T-helper cells (higher CD4⁺ total cell numbers).

We conclude, from this study, that nutritionally relevant carotenoids and retinoids are involved in the development of the humoral and cellular immune system and may be relevant for human immunological disorders that are manifested during postnatal development; this is currently under investigation by our groups.