Perturbed Vitamin A Status Induced by Iron Deficiency Is Corrected by Iron Repletion in Rats with Pre-Existing Iron Deficiency

ABSTRACT

Background

Although iron deficiency is known to interrupt vitamin A (VA) metabolism, the ability of iron repletion to restore VA metabolism and kinetics in iron-deficient rats is not well understood.

Objectives

In the present study, we examined the effects of dietary iron repletion on VA status in rats with pre-existing iron deficiency.

Methods

Weanling Sprague-Dawley rats were fed a VA-marginal diet (0.35 mg retinol/kg diet) containing either a normal concentration of iron [35 ppm, control group (CN)] or reduced iron (3 ppm, iron-deficient group, ID−); after 5 wk, 4 rats/group were killed for baseline measurements. A ³H-labeled retinol emulsion was administered intravenously to the remaining rats (n = 6, CN; n = 10, ID−) as tracer to initiate the kinetic study. On day 21 after dosing, n = 5 ID− rats were switched to the CN diet, generating an iron-repletion group (ID+). Blood samples were collected at 34 time points ≤92 d after dose administration, when all rats were killed and iron and VA status were determined.

Results

At baseline, ID− rats had developed iron deficiency, with a reduced plasma VA concentration (0.67 compared with 1.20 μmol/L in ID− and CN rats, respectively; P < 0.01) and a tendency toward higher liver VA (265 compared with 187 nmol in ID− and CN rats, respectively; P = 0.10). On day 92, iron deficiency persisted in ID− rats, accompanied by 2-times higher liver VA (456 nmol compared with 190 nmol in ID− and CN rats, respectively; P < 0.001) but lower plasma VA (0.64 compared with 0.94 μmol/L in ID− and CN rats, respectively; P = 0.05). ID+ rats not only recovered from iron deficiency, but also exhibited less liver VA sequestration (276 nmol) and normal plasma VA (0.91 μmol/L, not different from CN rats).

Conclusions

Our results suggest that iron repletion can remove the inhibitory effect of iron deficiency on hepatic mobilization of VA and restore plasma retinol concentrations in iron-deficient rats, setting the stage for kinetic studies of VA turnover in this setting.

See corresponding article on page 1982.

Introduction

Deficiencies of both vitamin A (VA) and iron have been recognized as worldwide public health issues for decades, especially affecting people in lower-income countries in Africa and Southeast Asia. Often regions that have a high prevalence of VA deficiency overlap with those having a high occurrence of iron deficiency (1–3). Although overall malnutrition and the prevalence of infectious diseases generally account for deficiencies of these micronutrients, some studies indicate that potential interactions between VA and iron, such as correlations between suboptimal hematological iron indexes and low serum retinol (4–6), development of anemia or aggravation of iron deficiency with VA deficiency (7, 8), and improvement in iron status or amelioration of anemia resulting from VA supplementation alone (9–11), may underlie the coexistence of these deficiencies.

Moreover, it was reported that, depending on dietary treatments, iron deficiency caused a decrease in plasma retinol, an elevation in liver VA, and an increase in the molar ratio between esterified and nonesterified retinol in liver (12, 13). By applying model-based compartmental analysis to the results from a rat model, Jang et al. (14) further demonstrated that iron deficiency may inhibit the mobilization of VA from the liver, which in turn may lead to a lower concentration of plasma retinol and increased liver VA storage.

Although previous studies have revealed the effects of iron deficiency on VA metabolism, the possible outcomes of iron repletion with dietary iron on VA kinetics in iron-deficient animals are not well explored. Therefore, to gain a better understanding of the interrelation between VA and iron metabolism, we investigated the impact of iron repletion on VA dynamics in iron-deficient rats. We hypothesized that iron repletion, using an iron-replete diet, could remove the unfavorable changes in VA metabolism caused by iron deficiency and restore body VA status in rats.

Methods

Animals and diets

Three lactating Sprague-Dawley rats with 10 pups each (pups were 14 d of age when the shipment arrived) were purchased from Charles River Laboratory. Upon arrival, the dams were fed a modified AIN-93G diet, which contained a marginal concentration of VA (0.35 mg retinol equivalents/kg diet) but a normal concentration of iron (35 mg Fe equivalents/kg diet), referred to as the VA-marginal diet (Research Diets, Inc.) (15). At 21 d of age, weanling pups were randomly assigned into 2 groups with similar mean body weight in each group, either assigned to the VA-marginal diet [control group (CN)] or to a VA-marginal plus iron-deficient diet [iron-deficient group (ID−)] containing a low amount of iron (3 mg Fe equivalents/kg diet). In both customized diets, the forms of iron and VA used were the same as in the AIN-93G diet. To control the body weight and VA intake, whereas ID− rats had free access to food and water, CN rats were food restricted to match the amount of diet consumed by ID− rats as described in a previous study (12), and daily food intake and body weight were recorded. Animals were housed individually in solid plastic cages in an environmentally controlled (12-h light/dark cycle) animal facility. All procedures were approved by the Institutional Animal Care and Use Committee of The Pennsylvania State University.

Study design

The study consisted of 2 parts conducted simultaneously: an assessment of iron and VA status (reported here) and a study of ³H-retinol kinetics (to be reported separately); because these parts were integrated, the preparation of the 11,12-[³H]retinol-labeled emulsion and its administration are described briefly here (Figure 1).

FIGURE 1.

Study design of the vitamin A kinetic study. The star denotes the time of isotope injection. CN, control group; ID−, iron-deficient group; ID+, iron-repletion group.

Dose preparation

The procedure was modified based on the method described previously (16). By mixing a known amount of 11,12-[³H]retinol (Perkin-Elmer) with Tween 20 (Sigma-Aldrich) and PBS (Mediatech, Inc.), we prepared the injection emulsion, in which the isotope concentration was ∼16.7 μCi/mL. The dose emulsion was prepared fresh on the day of injection and mixed on a vortex sufficiently before injection.

Kinetic studies

After 5 wk dietary treatment, n = 4 rats from each of the ID− and CN groups were killed to determine the iron and VA indexes at baseline, including plasma, liver, and spleen iron; hemoglobin and hematocrit; and plasma and liver VA. The remaining rats (n = 16) were anesthetized with isoflurane (Phoenix Pharmaceutical). For the kinetic analysis, described in the accompanying article, a defined volume of 11,12-[³H]retinol-labeled emulsion (0.1 mL/100 g body weight) was injected into the tail vein. After recovery from anesthesia, all rats were fed their same assigned diet. Serial blood samples were collected from the tail vein of each rat at 34 preselected time points, from 12 min after dosing ≤92 d. An aliquot of plasma separated from each blood sample was used for radioactivity measurement by liquid scintillation spectrometry. On day 21 after injection, half of the rats from the ID− group (n = 5/subgroup) were switched to the same VA-marginal diet as the CN group, which contained a normal concentration of iron (35 mg Fe equivalents/kg diet); these rats are referred to as the iron-repletion group (ID+). Whereas, the remaining half of the ID− group rats were continued on the VA-marginal plus iron-deficient diet. The mean body weights in the ID+ and ID− groups were balanced during the subgrouping process. On day 92, all rats were killed with carbon dioxide inhalation, and blood, liver, and spleen were collected. Aliquots of whole blood were used to measure hemoglobin and hematocrit. Plasma was separated and aliquots were stored at −20°C, whereas liver and spleen were frozen at −80°C until analysis.

Determination of iron indexes

Hemoglobin and hematocrit before the beginning and at the end of the kinetic study were determined in blood samples collected from tail veins. Hemoglobin concentrations were measured using the colorimetric cyanmethemoglobin method, with the analytical kit from Sigma-Aldrich. Hematocrit was determined by centrifugation (12000 × gat room temperature for 5 min) of blood drawn into heparinized microcapillary tubes (Drummond Scientific Company). Plasma iron was measured by the colorimetric method described by Walmsley et al. (17–19). Total iron-binding capacity (TIBC) was determined as outlined in Scott and Murray-Kolb (20). Transferrin saturation was calculated as the ratio of plasma iron concentration to TIBC, presented as a percentage (20). Liver and spleen nonheme iron concentrations were quantified as previously described (21).

Determination of VA mass quantification

Aliquots (100 μL) of plasma samples were used for VA mass quantification as previously described (22). A known proportion of hexanes were transferred into a 7-mL vial and an exact amount of internal standard, trimethylmethoxyphenyl-retinol (TMMP), was added to each vial. This hexane extract was dried under a stream of nitrogen and reconstituted in 100 μL methanol for ultra-high performance liquid chromatography analysis. Using a C-18 column, with the mobile phase as methanol:water (92.5:7.5 vol:vol, 0.6 mL/min), retinol and TMMP peaks were detected by UV absorbance at the wavelength of 325 nm. Retinol mass was calculated from the ratio of the integrated areas of the TMMP and retinol peaks, and the known amount of added TMMP standard. For tissue VA content determination, weighed portions of frozen liver and spleen samples (∼0.05 g liver, ∼0.1 g spleen) were first homogenized with ethanol. The following analysis procedure was the same as aforementioned for plasma.

Statistical analysis

Data are reported as means ± SEMs. Differences between 2 groups (CN, ID−) were determined by unpaired t test at baseline and between 3 groups (CN, ID−, and ID+) by 1-factor ANOVA, including a Geisser–Greenhouse correction for unequal variance, with Tukey's multiple comparisons test at the end of the study. Linear regression analysis was used to relate some of the iron and VA indexes. Changes over time within the same treatment group (CN and ID− groups only) were determined by unpaired t test. All statistical tests were performed using GraphPad Prism 6.0. A P value < 0.05 was considered statistically significant.

Results

Body weights in ID− and pair-fed CN and ID+ rats

By restricting the food intake of CN and ID+ rats, the body weights of the different groups were successfully controlled and no significant difference was found between the groups: after 5 wk dietary treatment (at the beginning of the kinetic study, defined as baseline, t₀ in Figure 1), body weights were nearly identical, with means of 224 g in each of the 3 groups. At the study's conclusion (92 d later, t_t), body weights had increased in all rats but were not different between the 3 groups. There was a significant enlargement of the spleen in ID− rats (P < 0.03), whereas this symptom was corrected in ID+ rats. In addition, ID− rats had a slightly but significantly (P < 0.0001) higher daily intake of VA than the other 2 groups because of their greater diet consumption (Table 1).

TABLE 1.

Body weights, diet and VA intakes, and tissue weights for CN, ID–, and ID+ rats¹

	CN	ID−	ID+
t 0 Body weight, g	224 ± 8.2	224 ± 18.3	224 ± 7.9
t t Body weight, g	425 ± 52.9	419 ± 29.2	450 ± 62.9
Diet intake, g/d	17.1 ± 0.2c	19.3 ± 0.2a	18.2 ± 0.2b
VA intake, μg/d	6.0 ± 0.1c	6.7 ± 0.1a	6.4 ± 0.1b
t t Liver weight, g	14.3 ± 1.6	14.5 ± 0.9	14.7 ± 1.9
t t Relative liver weight, %	3.4 ± 0.1	3.5 ± 0.0	3.3 ± 0.0
t t Spleen weight, g	0.7 ± 0.1b	1.1 ± 0.1a	0.8 ± 0.1ab
t t Relative spleen weight, %	0.2 ± 0.0b	0.3 ± 0.0a	0.2 ± 0.0b

¹Values are means ± SEMs, n = 5–6 rats/group. Labeled means in a row without a common letter differ, P < 0.05. CN, control group; ID−, iron-deficient group; ID+, iron-repletion group; t₀, at the beginning of the kinetic study; t_t, at the end of the kinetic study; VA, vitamin A.

Plasma and tissue iron and VA at baseline

After 5 wk dietary treatment, ID− rats developed iron deficiency as revealed by hemoglobin, hematocrit, and other iron indexes (Figure 2). In ID− rats, hemoglobin concentration was 2 times (Figure 2A) and hematocrit 1.3 times lower (Figure 2B) than their corresponding values in CN rats. Plasma iron concentration and transferrin saturation level were both significantly reduced in the ID− group compared with the CN group (Figure 2C, D) (P < 0.0005).

FIGURE 2.

Plasma hemoglobin (A), hematocrit (B), iron (C), and transferrin saturation (D) in CN, ID−, and ID+ rats at the beginning (t₀) and end (t_t) of the kinetic study. Data are means ± SEMs, n = 4–6 rats/group. *At t₀, different from control, P < 0.05. At t_t, mean values without a common letter differ, P < 0.05. CN, control group; ID−, iron-deficient group; ID+, iron-repletion group.

In the major iron-storage tissues, liver and spleen, iron concentrations were also significantly decreased. The liver of ID− rats had less than half the iron concentration present in the CN group (345 μg/g in ID− compared with 747 μg/g in CN) (Figure 3A), and likewise for spleen (92 μg/g in ID− compared with 179 μg/g in CN) (Figure 3B). Iron concentrations in these 2 organs were both significantly correlated with plasma iron concentration (Figure 4A, B).

FIGURE 3.

Liver (A) and spleen (B) iron concentrations in CN, ID−, and ID+ rats at the beginning (t₀) and end (t_t) of the kinetic study. Data are means ± SEMs, n = 4–6 rats/group. Tissue iron concentrations were measured in wet tissue. *At t₀, different from control, P < 0.05. At t_t, mean values without a common letter differ, P < 0.05. CN, control group; ID−, iron-deficient group; ID+, iron-repletion group.

FIGURE 4.

Correlations between plasma and liver iron (A), and between plasma and spleen iron (B), at the beginning of the kinetic study (t₀); and between plasma and liver iron (C), and between plasma and spleen iron (D), at the end of the kinetic study (t_t). Tissue iron concentrations were measured in wet tissue. CN, control group; ID−, iron-deficient group; ID+, iron-repletion group.

VA status at both the plasma and tissue levels was significantly affected by body iron condition (Figure 5). At baseline, ID− rats had significantly lower plasma retinol concentration (0.67 μmol/L) than the CN group (1.20 μmol/L) (Figure 5A), in spite of the fact that the daily VA intake from the diet was actually higher in the ID− group (Table 1). Although the ID− rats exhibited hyporetinolemia, VA mass detected in the liver of the ID− group tended to be greater than in the CN group (265 compared with 187 nmol, P = 0.10) (Figure 5B), and no such phenomenon was observed in the spleen (Figure 5C). These results thus suggested that iron deficiency resulted in malfunctioning of hepatic VA mobilization.

FIGURE 5.

Plasma (A), liver (B), and spleen (C) VA mass in CN, ID−, and ID+ rats at the beginning (t₀) and end (t_t) of the kinetic study. Data are means ± SEMs, n = 4–6 rats/group. *At t₀, different from control, P < 0.05. At t_t, mean values without a common letter differ, P < 0.05. CN, control group; ID−, iron-deficient group; ID+, iron-repletion group; VA, vitamin A.

Changes in indicators of iron status after consumption of an iron-sufficient diet

At the conclusion of the study, after 92 d iron-deficient diet in the ID− group, and 21 d iron-deficient diet followed by 71 d iron-sufficient diet in the ID+ group (t_t in Figure 1), indicators of iron status continued to be significantly lower in the ID− group than in CN rats, for hemoglobin (Figure 2A), hematocrit (Figure 2B), plasma iron concentration (Figure 2C), and plasma transferrin saturation (Figure 2D). These results confirmed that iron deficiency continued in ID− rats. In contrast, rats in the ID+ group showed significantly improved iron status compared with the ID− rats for each measure of iron status, and in fact reached a value for each plasma parameter that was indistinguishable from that of the CN group: in ID+ rats, hemoglobin increased to 17.1 g/dL (Figure 2A), hematocrit increased to 49.6% (Figure 2B), plasma iron increased to 4.0 μg/mL (Figure 2C), and plasma transferrin saturation level increased to 53.8% (Figure 2D).

In addition, findings in tissue iron content were noteworthy. Whereas ID− rats had less than half the amount of liver iron compared with CN rats at the beginning of the kinetic study (after 5 wk iron-poor diet), this phenomenon was reversed at the conclusion of the study, wherein a significantly higher iron concentration was observed in the liver of ID− rats (430 μg/g) than in either CN (252 μg/g) or ID+ rats (173 μg/g) (Figure 3A). At the same time, the relative reduction in spleen iron concentration in ID− rats, present at baseline, was further magnified at the end of the study, when mean concentrations equaled 138 μg/g in ID− compared with 1065 μg/g in CN rats (P < 0.01), whereas spleen iron in ID+ rats (969 μg/g) had returned to the concentration seen in the CN group (Figure 3B).

Plasma and liver VA are altered after consumption of an iron-sufficient diet

Because iron deficiency persisted in ID− rats, their plasma retinol remained lower than in CN rats at the end of the study (0.64 compared with 0.94 μmol/L, P = 0.05) (Figure 5A), whereas VA accumulated to a concentration more than double in the liver (P < 0.0001) (Figure 5B) of ID− rats. Although no significant correlation was detected between the plasma and liver VA concentrations at baseline (Figure 6A), the increased liver VA concentration was significantly associated with the decreased plasma VA concentration at the conclusion of the study (Figure 6B). Along with the recovery of iron status, plasma retinol in ID+ rats returned to a concentration indistinguishable from that of CN rats (0.91 compared with 0.94 μmol/L in ID+ and CN rats, respectively) (Figure 5A), and at the same time the amount of hepatic VA storage reduced by nearly half in the ID+ group as compared with the ID− group (276 nmol in ID+ compared with 456 nmol in ID− groups, P < 0.01) (Figure 5B). This altered VA status in ID+ rats, accompanied by improved iron status, indicated that iron repletion effectively stopped the sequestration of VA in the liver of rats with pre-existing iron deficiency.

FIGURE 6.

Correlations between plasma VA and liver VA at the beginning of the kinetic study (t₀) (A) and at the end of the kinetic study (t_t) (B). CN, control group; ID−, iron-deficient group; ID+, iron-repletion group; VA, vitamin A.

Discussion

Several interesting observations on the effect of iron repletion on iron status and on iron and VA status emerged from this study. It is important to emphasize that the iron-repletion strategy we selected for our study was based on restoring a normal concentration of dietary iron to rats that were moderately iron deficient owing to having previously consumed a diet low in iron (23, 24). Therefore, our study design provided a stringent test of whether adequate dietary iron is sufficient to correct previous iron deficiency, to concentrations similar to those in animals that have never been iron deficient, based on 1) hematologic measures of iron status; 2) measurements of iron itself in plasma and tissues; and 3) indexes of VA status and the distribution of VA in plasma and liver, the body's main organ of VA storage. Our study was conducted under a condition of mild VA deficiency induced by feeding the rats a VA-marginal diet, because it is often the case that inadequacies of VA and iron overlap in human populations (3). Our main findings under these experimental conditions were that iron repletion not only corrected the effects of moderate iron deficiency (the state of the ID− rats observed at baseline), but also had a significant impact on the distribution of VA, causing an increase in plasma VA to the concentration observed in the VA-marginal CN group, and concomitantly, a significant reduction in liver VA. These reciprocal changes suggest changes in the kinetics and distribution of retinol, which are further explored in the companion article (25).

Dietary strategy and weight control

The results of our study suggest that a low-iron diet and the resulting iron repletion result in moderate changes in food efficiency. Given the previous finding that iron-deficient rats exhibited poor growth and that a higher food intake was required for normal growth (26), in the current study, the rats consuming the control diet were food restricted to match the daily food intake of the ID− rats. Our pair-feeding strategy attained the desired goal of eliminating differences in body weight as a confounder. However, via pair feeding the mean daily diet intake and therefore VA intake were ∼13% higher in the ID− rats than in the CN rats, whereas the food intake of the ID+ rats lay between these 2 groups (Table 1); despite this, there was no difference in body weight between the 3 groups. Similar results were found in the study of Jang et al. (14), wherein their iron-deficient rats consumed 20% more food daily than did the food-restricted control rats, yet there was no difference in body weight. Together, these findings suggested that iron deficiency interferes with the efficiency of metabolism. Another physical observation worthy of discussion was that the spleen of ID− rats was significantly enlarged, both absolutely and relative to body weight, compared with CN rats. Splenic enlargement is usually reported in iron deficiency anemia, and similar results have also been reported by other researchers studying physiological changes due to iron deficiency (27, 28). In addition, the spleen weight in our ID+ rats fell between those of the CN and ID− groups, indicating that iron repletion successfully stopped the progression of spleen enlargement and partially reversed the effect of iron deficiency on the spleen size. However, because this effect was partial, it would be interesting to follow splenic changes over a longer period of iron repletion, to determine if complete reversal can be achieved.

Indexes of iron status during iron depletion and repletion

Based on hematologic indexes of iron status, we found that 5 wk consuming the low-iron diet (basal conditions at the start of the kinetic study) resulted in significant decreases in plasma hemoglobin and hematocrit, as well as in plasma iron concentration and plasma transferrin saturation level (Figure 2), and in tissue iron storage in liver and spleen (Figure 3). These data suggest that the rats consuming the iron-deficient diet successfully developed iron deficiency and could be used as a reliable model to study VA kinetics under iron-deficient conditions. At the conclusion of the study, whereas ID− rats continued to exhibit significantly lower hemoglobin and hematocrit than CN rats, ID+ rats showed improved hemoglobin and hematocrit (Figure 2). But interestingly, compared with their own baseline value, ID− rats exhibited a significant increase in hemoglobin (P < 0.002) and hematocrit (P < 0.0001), which possibly resulted from aging, because with age the animal's growth rate slows down, and therefore less iron is required to support tissue development, which in turn may spare more iron for the synthesis of hemoglobin; similar results were observed by other investigators (29, 30). In addition, the plasma iron concentration and transferrin saturation index in ID+ rats were restored to the levels in CN rats, both of which were significantly higher than in ID− rats (Figure 2). Moreover, dynamic changes in tissue iron were worth noting. In contrast to the baseline results, liver iron concentration in the ID− rats was 2.5 and 0.7 times higher than in CN and ID+ rats, respectively (Figure 3A), suggesting iron sequestration in liver. In addition, liver iron concentration in CN rats was significantly reduced at the end of the study as compared with this group's own baseline value (P < 0.001), suggesting there may have been an increased mobilization of liver iron to extrahepatic tissues. The opposite results were observed in the spleen as, at the end of the study, whereas CN rats had significantly increased spleen iron concentration compared with their own baseline (P < 0.05), ID+ rats showed a spleen iron concentration comparable with that of CN rats, and the spleen iron concentration remained lowest in ID− rats. These differences in tissue iron dynamics may reflect physiological changes under differences in iron status: in the iron-adequate state, the CN rats may increase the output of iron from liver to avoid iron overaccumulation and maintain the availability of iron in plasma to support the iron needs of peripheral tissues, whereas on the contrary, the ID− rats may downregulate their peripheral iron requirement to adapt to the limited dietary iron intake, thus more iron was stored in the liver. Although no thorough explanations could be provided for some of the aforementioned results, the similar patterns observed in the ID+ and the CN rats indicated that iron repletion did alter whole body iron metabolism and trafficking in the ID+ rats, making this group more similar to the CN rats. Overall, these data demonstrate the ability of improvement in dietary iron to nearly reverse changes in iron distribution in plasma and iron storage organs that were induced by a habitually low intake of dietary iron.

Impact of iron depletion and repletion on VA

Despite the fact that dietary iron concentration was the only independent variable in the current study, we found that VA status changed corresponding to the body's iron status. At baseline, iron-deficient rats showed significantly reduced plasma retinol concentration, even though the iron-deficient rats had a significantly higher dietary intake of VA (Table 1) and a relatively higher VA mass in the liver (Figure 5B). These findings were in agreement with previous studies (12–14), indicating that iron deficiency causes hyporetinolemia regardless of the liver VA store, which may lead to a false-negative prediction of body VA status based on plasma retinol. At the end of the study, plasma retinol remained lower in the ID− than in the CN rats, as would be expected without a change in the diet, but the ID+ rats’ plasma retinol concentration was restored, reaching a concentration comparable with that of the CN rats. Because all rats were fed a VA-marginal diet, these plasma retinol concentrations are, as expected, slightly low compared with previous studies in VA-adequate rats, in which plasma retinol concentrations of ∼1.54 μmol/L were reported (14). At the same time, VA continued to increase in the liver of the ID− rats, resulting in a significantly higher liver VA store in this group than both their own amount at baseline (P < 0.005) and that of CN rats. In contrast, liver VA storage in the ID+ rats declined to a concentration significantly lower than in ID− rats, and not statistically different from that of CN rats. This result suggests that VA was mobilized from the liver, but the process may have been still ongoing at the end of the study. This altered VA status in ID+ rats, accompanied by the improved iron status, suggests that iron repletion effectively stopped the sequestration of VA in the liver of rats with pre-existing iron deficiency. A similar pattern was observed in spleen: whereas the difference in VA mass between ID− and CN rats became greater at the study conclusion, VA residing in the spleen of ID+ rats was decreased, bringing its concentration closer to that present in the spleen of CN rats. Although our current findings did not elucidate the underlying mechanisms responsible for the observed changes in VA status, several possible explanations can be speculated. With the knowledge that iron is essential for a wide variety of physiological processes, including DNA, RNA, and protein synthesis (31), the restricted liver mobilization of VA in ID− rats may result from the decreased hepatic synthesis of retinol-binding protein (RBP), where RBP is required to form a binding complex with a retinol molecule before it can be secreted from the liver. Besides, enzyme activities associated with VA metabolism may also be altered by iron deficiency, driving the formation of retinyl ester, the storage form of VA, in the liver, while limiting the generation of free retinol.

Conclusions

Taking these results together, we speculate that iron deficiency may cause sequestration of VA in the liver, leading to a reduced plasma retinol concentration, and that with iron repletion based on an improved iron content of the diet, the ID+ rats were able to increase the mobilization of VA from the liver, which in turn increased their plasma retinol concentration and decreased the mass of VA retained in the liver. More information regarding the changes in the VA kinetics resulting from iron deficiency and repletion is presented in the accompanying article focusing on model-based compartmental analysis (25).

Notes

Supported by the Pennsylvania State University Graduate Program in Nutritional Sciences and by NIH grant HD-066982 (to ACR).

Author disclosures: The authors report no conflicts of interest.

Abbreviations used: CN, control group; ID−, iron-deficient group; ID+, iron-repletion group; RBP, retinol-binding protein; TIBC, total iron-binding capacity; TMMP, trimethylmethoxyphenyl-retinol; VA, vitamin A.