The Journal of Nutrition Vol. 128 No. 7 July 1998, pp. 1077-1083

Iron Deficiency Reduces the Hydrolysis of Cell Membrane Phosphatidyl Inositol-4,5-Bisphosphate during Splenic Lymphocyte Activation in C57BL/6 Mice^1,2,3

Solo R. Kuvibidila⁴, B. Surendra Baliga^*, Raj P. Warrier, and Robert M. Suskind

Louisiana State University, School of Medicine, Department of Pediatrics, Division of Hematology/Oncology, New Orleans, LA 70112 and ^* University of South Alabama, College of Medicine, Department of Pediatrics, Mobile, AL 36617

	ABSTRACT

Abstract Introduction Methods Results Discussion References

Iron deficiency impairs lymphocyte proliferation in humans and laboratory animals by unknown mechanisms. In this study, we investigated whether this alteration can be attributed in part to impaired hydrolysis of cell membrane phosphatidyl inositol-4,5-bisphosphate (PIP2), a required early event of T-lymphocyte activation. The study involved 46 iron-deficient (ID), 26 control (C) and 23 pair-fed (PF) mice, and ID mice that were repleted for 3 (n = 16), 7 (n = 17) or 14 d (n = 18). Mice were killed after 40-63 d (mean, 48 d) of consuming the test diet (0.09 mmol/kg iron) or the control diet (0.9 mmol/kg). The mean (±SEM) hemoglobin concentrations were 57 ± 16.7, 176 ± 2.6 and 181 ± 9.7 g/L for ID, C and PF groups, respectively. After splenic lymphocytes were labeled in vitro with ³H-myoinositol for 3 h, PIP2 hydrolysis was estimated by measuring the radioactivity recovered as a mixture of inositol mono-, di- and triphosphate (IP) from concanavalin A (0, 1, 2.5, 5 and 10 mg/L) activated cells. Although cells from ID mice and those from mice repleted for 3 d incorporated slightly more radioactivity in cellular phospholipids than did cells from C or PF mice, less (P < 0.005) was recovered as IP than in controls, suggesting impaired conversion of the precursor to PIP2. At almost all incubation periods (10-120 min) and mitogen concentrations, the rate of PIP2 hydrolysis expressed as the ratio of radioactivity obtained in Con A-treated to untreated cells was significantly (P < 0.05) reduced in cells from ID mice compared with those obtained from C and PF mice. For cells that were activated for 60 min or less, iron repletion for 14 d significantly (P < 0.05) improved the rate of PIP2 hydrolysis. PIP2 hydrolysis positively and significantly (P < 0.05) correlated (r = 0.27-0.56) with indicators of iron status. Mitogenic response was also significantly (P < 0.05) reduced in ID but not PF mice, and it was corrected by iron repletion for 3, 7 or 14 d. Lymphocyte proliferation positively (r = 0.27-0.37, P < 0.01) correlated with indices of iron status and IP ratios. The data suggest that reduced PIP2 hydrolysis contributes to impaired blastogenesis in iron deficiency.

	INTRODUCTION

Abstract Introduction Methods Results Discussion References

Cell membrane phospholipids play a major role in cell signaling as precursors of second messengers (Michel et al. 1992, Pantaleo et al. 1987). During lymphocyte activation, within seconds after the binding of a mitogen or antigen to the T cell receptor, three major categories of events occur: the hydrolysis of cell membrane phospholipids, specifically phosphatidyl inositol-4,5-bisphosphate (PIP2),⁵ increases in free cytoplasmic calcium and phosphorylation of membrane and cytoplasmic proteins by protein kinase C (PKC) (Abbas et al. 1991). PIP2 hydrolysis, which is catalyzed by phospholipase C, leads to two end products, diacylglycerol and inositol-1,4,5-triphosphate (IP3). Both of these end products are believed to activate PKC (May et al. 1985, Shinomura et al. 1991, Wolf et al. 1985).

Together with other investigators, we have previously observed that iron deficiency, the most common single nutrient deficiency in the world, impairs lymphocyte proliferation in humans and laboratory animals (reviewed by Dallman 1987 and Kuvibidila et al. 1989). The mechanisms are not fully understood. Although reduced activity of ribonucleotide reductase, the iron-dependent enzyme required for the biosynthesis of deoxynucleotides, may be the reason for impaired lymphocyte proliferation and associated functions, no one has ever studied the effects of in vivo iron deficiency on some of the early events of lymphocyte activation. In a study conducted in laboratory mice, we observed that PKC activation was impaired in spleen cells by iron deficiency but not undernutrition per se (Kuvibidila et al 1991). We concluded that one of the mechanisms by which iron deficiency reduced lymphocyte proliferation involved reduced translocation or activation of PKC. However, it is unknown whether reduced PIP2 hydrolysis could explain at least in part poor PKC activation and impaired blastogenesis in lymphocytes from iron-deficient mice. This study was therefore designed to investigate the effects of iron deficiency on PIP2 hydrolysis.

	MATERIALS AND METHODS

Abstract Introduction Methods Results Discussion References

Materials.  All reagents were of high quality (95% pure). They were received from the following sources: Sigma (St Louis, MO), inositol-free RPMI-1640, ammonium formate, lithium chloride, concanavalin A (Con A), sodium bicarbonate, -mercaptoethanol, hemoglobin standard, Drabkin's reagents and iron assay kits; GIBCO (Grand Island, NY), RPMI-1640 with 25 mmol/L HEPES, fetal calf serum, nonessential amino acids, sodium pyruvate and penicillin/streptomycin; Bio-Whittaker (Walkersville, MD), L-glutamine; Mallinckrodt (Paris, KY), chloroform; Curtin Matheson Scientific (Houston TX), methanol; Harlan Teklad (Madison, WI), the iron deficient test diet (TD 80396) and the deficient diet supplemented with iron; NEN Dupont (Boston, MA), ³H-myoinositol (specific activity 23.5 Ci or 867.7 Gbq/mmol) and ³H-thymidine [6.7 Ci (247.9 GBq/mmol)]; BioRad (Hercules, CA), AG1-X8 resin, formate form; ICN (Costa Mesa, CA), cytoscint.

Experimental design and feeding of mice.  The overall protocol involved 146 C57BL/6 female mice, 21-23 d of age. They were purchased from Charles River breeding laboratories (Wilmington, MA). Upon receipt, they received the control diet that contained 0.9 mmol of iron (Fe) per kg diet and sterile deionized water for 4-5 d. After the adjustment period, mice were anesthetized by diethyl-ether inhalation (in a desiccator) and weighed. Approximately 100 µL (two full hematocrit tubes) of blood were drawn from the retro-orbital plexus. Both hemoglobin and hematocrit were determined by the cyanmethemoglobin method and centrifugation, respectively.

After determining the hematologic indices and body weight, mice were divided into three groups and assigned to the following dietary treatment groups: control (C, n = 26), iron deficient (ID, n = 97), and pair-fed (PF, n = 23). The initial body weight (g, mean ± SEM) of the groups was 15.2 ± 0.38, 15.0 ± 0.16 and 15.5 ± 0.31 in C, ID and PF mice, respectively. Hemoglobin (Hb, g/L) concentration was 155.2 ± 3.78, 146.8 ± 1.57, and 147.5 ± 3.15 in C, ID and PF mice, respectively; and hematocrit was 0.48 ± 0.003, 0.47 ± 0.004 and 0.47 ± 0.005, respectively. Means for these variables were not significantly different among the three groups. Iron deficiency was induced by feeding mice a diet that contained only 0.09 mmol iron/kg diet. The C and PF groups received the same diet except that it was supplemented with 0.9 mmol iron/kg diet in the form of ferrous sulfate. The compositions (protein, energy, vitamin and mineral concentrations) of both the deficient and the control diets were identical except for the iron content (Table 1).

Although C and ID mice had free access to their diets 24 h/d, PF mice received the control diet in amounts equal to the mean that ID mice had consumed during the preceding 24 h. To detect the development of anemia, ~100 µL of blood was drawn by retro-orbital plexus once every 2 wk from all mice. When the Hb concentration of ID mice decreased to <70 g/L, 49 mice were given the control diet (repletion protocol) for 3 d (R3, n = 16), 7 d (R7, n = 17) or 14 d (R14, n = 18). The average feeding period was 48 d (range 40-63 d), depending on hemoglobin concentrations of the mice assigned to the low iron diet. During the entire feeding period, all mice were housed in sterile microisolator system cages (product 109EI, Laboratory Products, Maywood, NJ) and they received only sterile deionized water. The study was approved by the Institutional Committee for Animal Care and Use of Louisiana State University School of Medicine.

Evaluation of iron status at the end of the feeding period.  At the time of killing, mice were anesthetized by ether inhalation for 30-60 s. After as much blood as possible was drawn from the retro-orbital plexus, mice were killed by cervical dislocation. In each experiment, between one and four mice were killed in each dietary treatment group. Hemoglobin and hematocrit were measured by the cyanmethemoglobin method and centrifugation, respectively. The livers were removed, weighed and immediately frozen at 40°C until used for iron assay. Liver iron was prepared from liver homogenates by a modified method normally used for food digestion in nitric acid in a microwave (Model MDS-2000; CEM Corporation 1991). Each liver was homogenized in 2 mL of deionized water at 4°C. One milliliter of the homogenate was mixed with 4 mL of 3 mol/L nitric acid and the samples were heated at 90°C for 3 h. After being cooled to room temperature, the samples were centrifuged at 29,000 × g for 10 min at 4°C (Sorvall, Newton, CT). The supernatant was transferred to a clean plastic culture tube. The pellet was resuspended in 2 mL of nitric acid, heated again for an additional 3 h and centrifuged. The second supernatant was transferred to another culture tube. Preliminary experiments, which involved six livers obtained from iron-sufficient mice treated with nitric acid four times, showed that up to 97% of iron was extracted in the first two treatments (mean ± SEM iron concentration in µmol/g liver, 1.112 ± 0.22). The fraction from the fourth treatment contained only between 0.004 and 0.025 µmol Fe/g liver). As a result of these preliminary data, liver homogenates for the mice reported in this study were treated with nitric acid only twice. Iron concentration in each supernatant was assayed by colorimetry using a kit purchased from Sigma (kit # 565).

To determine the iron concentration of control and the iron-deficient test diets, six samples were subjected to nitric acid digestion at 90°C four times, using the same procedure as that used for liver homogenates. The iron concentration in the extracts was measured by colorimetry (Sigma kit # 565). The iron concentration of the iron-deficient test diet was 0.087 mmol/kg diet and was recovered from the first two fractions. The concentration of the control diet was 0.84 mmol/kg diet (not significantly different from the expected levels of 0.90 mmol/kg). Samples of both diets were also treated with nitric acid for 10 min in a microwave (Model MDS-2000) as described in the manual on Microwave Sample Preparation (CEM Corporation 1991). Iron was measured by atomic absorption spectrophotometry (Perkin-Elmer Model 3030B, Norwalk, CT). The mean concentration of five samples was 0.10 and 0.79 mmol/kg for the iron-deficient diet and control diet, respectively.

Preparation of spleen cell suspension and labeling of cell membrane phosphatidyl inositol-4,5-bisphosphate with ³H-myoinositol.  Spleens were removed under sterile conditions. Single-cell suspensions were prepared by standard techniques in serum-free RPMI-1640 medium that was supplemented with 10 g/L bovine serum albumin, 50 mg/L streptomycin and 50,000 units of penicillin/L (subsequently referred to as wash medium) (Kuvibidila et al. 1991). After the cells were washed at 670 × g, 4°C for 10 min, the pellets were resuspended in 1 mL ice-cold sterile deionized water for 20 s to lyse red blood cells. Cells were further washed twice under the same conditions. The pellets were then resuspended in 2 mL wash medium. Total and viable cell counts were performed under a light microscope and hemocytometer after dilution of the cells in a solution of trypan blue (4 g/L). Viable cell counts were (%, mean ± SEM): 96.8 ± 0.5, 97.5 ± 0.3, 91.7 ± 1.2, 88.2 ± 2.6, 96.2 ± 0.7 and 95.0 ± 0.8, in C, PF, ID, R3, R7 and R14 groups, respectively. For PIP2 hydrolysis, spleen cells were labeled with ³H-inositol (370 kBq/10⁷ viable cells) in myoinositol-free and serum-free RPMI-1640 for 3 h. After being washed three times to remove free ³H-myoinositol, the pellets were resuspended in myoinositol-free medium. Because IP3 is rapidly converted to inositol diphosphate (IP2) and inositol monophosphate (IP1), which is later metabolized to myoinositol, 10 mmol/L lithium chloride was added to the inositol-free medium. This compound inhibits myoinositol-1 phosphatase (Berridge et al. 1982). The end product that we measured in the hydrolysis of PIP is therefore a mixture of IP1, IP2 and IP3, and is referred to as IP throughout the text, table and figures.

Measurement of PIP2 hydrolysis.  PIP2 hydrolysis was studied by the methods described by van Tits et al. (1991). Ten million cells were incubated with or without 5 mg/L Con A for various periods. In several experiments, spleen cells from two mice from the same group were pooled to obtain sufficient cells to study as many time points as possible. In a pilot study, the effects of Con A concentrations (1, 2.5, 5 and 10 mg/L) on PIP2 hydrolysis in spleen cells that were incubated for either 30 min or 120 min were also assessed. The reaction was stopped by adding 1.5 mL ice-cold methanol/chloroform (2:1) to each tube, vortexing immediately and incubating on ice for 60 min. After the addition of 500 µL of ice-cold chloroform and vortexing, 500 µL of ice-cold deionized water was again added. The tubes were vortexed and centrifuged at 670 × g (IEC-6000R, Needham Heights MA) at 4°C for 10 min. The upper layer was directly transferred to a column containing 1 mL AG1-X8 Dowex-1 resin, 200-400 mesh, formate form. To remove free ³H-myoinositol and ³H-glycerolphosphorylinositides, the columns were washed with 5 mL ice-cold distilled water followed by 5 mL 25 mmol/L ammonium formate dissolved in 0.1 mol/L formic acid, respectively. The inositol phosphates (a mixture of IP3, IP2 and IP1) were eluted by washing the columns with 2 mol/L ammonium formate dissolved in 0.1 mol/L of formic acid. Three 1-mL fractions were collected and mixed with scintillation fluid (cytoscint). The radioactivity was measured on a liquid scintillation counter (LKB model 1219, Wallac, Turku, Finland).

Lymphocyte proliferation.  For lymphocyte proliferation, 1 × 10⁶ viable cells were mixed with either 2.5 or 5 mg/L Con A. For background DNA synthesis, culture medium was added to cells instead of the mitogen. The culture medium was composed of RPMI-1640, 100 mL/L fetal calf serum, 50 mg/L streptomycin, 50,000 units/L penicillin, 2 mmol/L L-glutamine, 1 mmol/L sodium pyruvate, 0.1 mmol/L nonessential amino acids and 50 µmol/L -mercaptoethanol. The iron concentration measured by colorimetry (Sigma, kit # 565) of the culture medium, plain RPMI-1640 (not supplemented with serum, bovine serum albumin, apotransferrin or antibiotics) and inositol-free medium was 4.48, 1.23 and 0.79 µmol/L, respectively. Mitogen-treated and untreated cells were incubated at 37°C, 5% CO₂ for 48 h before being pulsed with 37 kBq ³H-thymidine per 200,000 cells. After 24 h, the cultures were harvested (PHD Mash Harvester, Cambridge Technology, Watertown MA). Each filter disk was transferred to a scintillation vial containing 2 mL of cytoscint. The radioactivity incorporated into DNA was measured on an LKB  scintillation counter.

Calculations and statistical analysis.  Descriptive analysis (mean ± SEM), ANOVA and Pearson's correlation coefficient were calculated by the use of the Microstatistical program (Ecosoft, Indianapolis, IN) and as described in the literature (Munro and Page 1993). For the comparison of indices of iron status, body weight, cell viability, ³H-myoinositol cellular uptake and radioactivity recovered as IP, one-way ANOVA was used. Because data on ³H-myoinositol uptake were skewed, they were first transformed by decimal logarithm (log₁₀) before ANOVA was performed. Results on PIP2 hydrolysis were analyzed in two steps, first by comparing only C, PF and ID groups and second by comparing all six study groups. To determine if there was an interaction between the effect of dietary treatment and that of incubation period on PIP2 hydrolysis of C, ID and PF groups, two-way ANOVA was performed. When ANOVA revealed significant differences among groups, Scheffé's test was performed to determine which pairs of groups were significantly different (Munro et al. 1986). The level of significance was set at P < 0.05.

	RESULTS

Abstract Introduction Methods Results Discussion References

At the time of killing, hematocrits, hemoglobin concentrations and liver iron concentrations of the ID mice were <50% those of control mice (Table 2). The differences among the C, ID and PF groups, or those among the six study groups, were significant (P < 0.0001). Although the indices of iron status of PF mice were not different from those of C mice, they were significantly higher than those of ID mice (P < 0.01). At least 7 d were required to restore the hematologic measurements and liver iron concentrations to normal levels. Mice in the ID and R7 groups weighed significantly (P < 0.05) less than those in the C group. There was no significant difference in the food intake (mean ± SEM, g/48 d) among the C (125 ± 3), PF (117 ± 2) and ID (117 ± 2) groups.

Table 3 summarizes the results on the radioactivity incorporated into spleen cells after 3 h of incubation with ³H-myoinositol. Spleen cells from ID and R3 mice incorporated 35 and 76% more radioactivity in cellular phospholipids than those from C mice, respectively. Although the small difference persisted even after taking into account cell viability, it was not significant when all six groups were compared by ANOVA. In contrast to ³H-myoinositol uptake, the percentage of radioactivity recovered as IP from the aqueous phase (of cell extracts loaded on Dowex columns) and incubated with Con A for 120 min, was significantly different (P < 0.05) among the six study groups. The percentage was significantly higher in cells obtained from C than in those from ID and all repleted mice (Table 3). However, the percentages were not significantly different among PF, ID and repleted mice or between C and PF mice.

There were significant (P < 0.05) differences among C, PF and ID groups in baseline rate of PIP2 hydrolysis expressed in becquerels (Fig. 1A). In nonactivated cells, the levels of IP produced were higher in ID than in C and PF groups. This difference persisted even after taking into consideration cell viability (Fig. 1A). In contrast, no significant difference was observed between C and PF groups. Upon activation of cells with Con A, PIP2 hydrolysis increased in each group with time of incubation. Increases were significant in all groups at 60 and 120 min incubation time compared with baseline (P < 0.01). No significant difference was observed in total radioactivity among the three groups in Con A-treated cells. When results were expressed as the difference between Con A-treated and untreated cells, values were significantly (P < 0.05) lower in cells from ID mice than in those from PF mice at 30 and 60 min (Fig. 1B). No significant differences were observed in cells from ID and C mice, or those from C and PF mice. When results were expressed as ratios of IP obtained in activated cells to IP in nonactivated cells, values were significantly lower in cells from ID mice than in those from C or PF mice at 30, 60 and 120 min (Fig. 1C). No significant differences were observed when cells were activated for 10 (Fig. 1C), or 2 or 5 min (data not shown).

View larger version (19K): [in this window] [in a new window]   View larger version (18K): [in this window] [in a new window]   View larger version (18K): [in this window] [in a new window]   Fig 1. Hydrolysis of cell membrane phosphatidyl inositol-4,5-bisphosphate (PIP2) by spleen cells expressed in becquerels (Figs. 1A and B) and as ratios of inositol phosphates (IP) obtained from Con A-treated (5 mg/L) to untreated cells from control (C), iron-deficient (ID) and pair-fed (PF) mice (Fig. 1C). Figures 1A, B, solid lines: radioactivity released per 107 viable cells, dashed line: radioactivity based on 107 viable and nonviable (total) spleen cells. Figure 1B: difference between radioactivity obtained from Con A-treated and untreated cells. Values are means ± SEM; n = 10-28 C, 10-38 ID and 10-22 PF. Figures 1A, B, ANOVA and Scheffé's test were performed on log10 becquerels. *Indicates significantly different from baseline (P < 0.01). Means at a particular time with unlike letters are significantly different (P < 0.05) from one another. No significant dietary treatment × incubation time interaction.

Refeeding ID mice the iron-supplemented diet for 3, 7 or 14 d did not significantly lower basal IP production expressed in becquerels or increase IP production in Con A-treated cells in respect to ID group (Fig. 2A). PIP2 hydrolysis was greater than baseline in the R7 group at 60 and 120 min of incubation (P < 0.05). No significant differences were observed among ID and repleted mice when results were expressed as the difference in radioactivity (becquerels) between Con A-treated and untreated cells (data not shown). When results were expressed as ratios of IP produced in Con A-treated to untreated cells, iron repletion for 7 and 14 d generally improved PIP2 hydrolysis at 30, 60 and 120 min incubation time in respect to ID mice (Fig. 2B). However, the improvement was significant (P < 0.05) only for cells from the R14 group that were activated for 30 and 120 min. IP ratios of cells obtained from C (and PF) mice and that were activated for 60 min or less were not significantly different from those obtained from R7 and R14 mice. However, for cells that were activated for 120 min, the ratios of cells from C mice were significantly (P < 0.05) higher than those of cells obtained from R7 and R14 mice, suggesting that PIP2 hydrolysis is not fully restored to normal levels even after 2 wk of iron repletion. Iron repletion for 3 d did not increase IP ratios in respect to cells obtained from iron-deficient mice.

View larger version (18K): [in this window] [in a new window]   View larger version (21K): [in this window] [in a new window]   Fig 2. Cell membrane phosphatidyl inositol-4,5-bisphosphate (PIP2) in 107 viable spleen cells obtained from control (C), iron-deficient (ID) and pair-fed (PF) mice that were repleted for 3 d (R3), 7 d (R7) and 14 d (R14). Values are means ± SEM; n = 10-23 C, 8-33 ID, 6-12 R3, 5-10 R7 and 7-14 R14. Figure 2A: results expressed as absolute radioactivity in becquerels. Figure 2B: ratios of inositol phosphates (IP) produced in Con A-treated to untreated cells. At any incubation time, means with unlike letters are significantly different (P < 0.05) from one another. * Significantly higher than baseline (P < 0.05). No significant dietary treatment × incubation time interaction.

In spleen cells that were incubated with 1, 2.5 or 5 mg/L Con A for 30 min, the rate of PIP2 hydrolysis in cells from ID mice was lower than that in cells from C and PF mice (P < 0.01, Fig. 3). When cells were incubated with 2.5 and 5 mg/L Con A for 120 min, the difference among the three groups of mice persisted (P < 0.01; data not shown).

View larger version (14K): [in this window] [in a new window]   Fig 3. Hydrolysis of cell membrane phosphatidyl inositol-4,5-bisphosphate (PPI2) in 107 viable spleen cells from control (C), iron-deficient (ID) and pair-fed (PF) mice that were treated with different concentrations of concanavalin A (Con A). Values are means ± SEM; n = 3-8 C, 8-12 ID, and 4-8 PF. At a particular Con A concentration, values followed by unlike letters are significantly different (P < 0.05).

For cells that were activated by Con A for 10 min, there were no correlations between IP ratios and either Hb or liver iron concentrations (Table 4). In contrast, for cells that were activated for 30, 60 and 120 min, IP ratios positively and significantly (P < 0.05) correlated with indicators of iron status.

Both iron deficiency and underfeeding in pair-fed mice were associated with higher baseline lymphocyte proliferation studied by ³H-thymidine uptake than control mice and mice that were repleted for 14 d (Table 5, P < 0.05). Background ³H-thymidine uptake by splenic lymphocytes was not significantly different among iron-deficient mice and mice that were repleted for 3 or 7 d. Significant differences were observed in lymphocyte proliferation in response to both concentrations of Con A among the six groups (Table 5). Iron deficiency, but not underfeeding in pair-fed mice was associated with reduced blastogenic response (P < 0.05). After 3, 7 or 14 d of iron repletion, the response did not differ from that of controls. Lymphocyte proliferation positively and significantly correlated with hemoglobin (r = 0.266, P < 0.01), hematocrit (r = 0.283, P < 0.01), liver iron concentration (r = 0.267, P < 0.01), IP ratios at 30 min of incubation (r = 0.378, P < 0.001) and 60 min of incubation (r = 0.265, P < 0.01).

	DISCUSSION

Abstract Introduction Methods Results Discussion References

This study confirms our previous observation as well as that of other investigators that iron deficiency impairs lymphocyte proliferation (Chandra and Saraya 1975, Kemahli et al. 1988, Kuvibidila et al. 1983a and 1991, Latunde-Dada and Young 1992). However, we are unaware of any study showing reduced PIP2 hydrolysis in lymphocytes from iron-deficient laboratory animals or humans. Although the basal radioactivity associated with IP that was recovered in spleen cells from ID mice was not lower than that in cells from C and PF mice, the magnitude of increase (expressed as IP ratios) after cells were activated with Con A was significantly affected. When the baseline IP was subtracted from IP produced in activated cells, significant differences were observed between cells from ID mice and PF mice, further supporting the point that PIP2 hydrolysis is reduced in iron deficiency. There may be a sightly enhanced spontaneous PIP2 hydrolysis in spleen cells from ID mice, but these cells are less responsive to the stimulus (Con A) than those from C and PF mice. This observation is very important because what is crucial in cell-mediated immune responses is how well the host responds to stimuli and increases the functions of various cells of the immune system, not just the activity of resting cells. The lack of decrease in PIP2 hydrolysis and lymphocyte proliferation from PF mice suggests that it is iron deficiency per se, not underfeeding, which is more likely responsible of the alteration observed in cells from ID mice.

PIP2 hydrolysis plays a crucial role in lymphocyte activation and subsequent proliferation in cells from humans and laboratory animals (Desai et al. 1990, Michel et al. 1992, Pantaleo et al. 1987). The end products of PIP2 hydrolysis play a major role as second messengers in many biochemical reactions in lymphoid as well as nonlymphoid organs such as the brain, liver, pancreas and neutrophils (Batty and Nahorski 1992, Berridge 1987, Thomas et al. 1984). Conditions that stimulate cell membrane PIP2 hydrolysis are known to activate the final immune response such as interleukin-2 secretion (Desai et al. 1990) or non-immune response (e.g., muscle contraction, Berridge 1987), and those that inhibit PIP2 hydrolysis would certainly be expected to abrogate these same responses. Assuming that the same process takes place in vivo, reduced hydrolysis of cell membrane phospholipids in iron-deficient subjects (humans or laboratory animals) would affect immune responses and other biological processes. As mentioned in the introduction, one of the end products of PIP2 hydrolysis (diacylglycerol) is directly involved in protein kinase C activation. Low concentrations of this second messenger may lead to inhibition of protein kinase C translocation, and hence lymphocyte proliferation or any other biochemical reactions mediated by protein kinase C. Our data therefore suggest that part of the mechanism by which iron deficiency alters lymphocyte proliferation is reduced PIP2 hydrolysis. Results obtained from the same mice used in this study suggest that in parallel to reduced PIP2 hydrolysis, protein kinase C activation is also impaired, and iron repletion corrects it (Kuvibidila and Baliga 1997).

There are several other ways by which iron may alter lymphocyte proliferation and associated functions in addition to the possible reduction of PIP2 hydrolysis. Two of the most likely mechanisms are differences in the proportion or percentage of immunocompetent T cells in blood or lymphoid organs and reduced activity of ribonucleotide reductase, the iron-dependent enzyme required for the biosynthesis of deoxynucleotides. In humans (Chandra and Saraya 1975, Galan et al. 1992, Krantman et al. 1982) and mice (Kuvibididla et al.. 1983a and 1990), iron deficiency was associated with significantly lower percentages of immunocompetent T cells in blood or spleen. In mice, the percentage of B cells identified by fluorescein-conjugated anti-mouse immunoglobulin G was also reduced in iron-deficient but not pair-fed mice, suggesting an increased percentage of non-T and non-B cells (Kuvibidila et al.. 1990). However, enrichment of the T-cell fractions by nylon wool columns or the B-cell fractions by depleting T cells with anti-theta antibodies and guinea pig complement did not restore blastogenesis in cells from ID mice (Kuvibidila et al. 1983a) or cytotoxicity of splenic lymphocytes against mastocytoma cells (Kuvibidila et al. 1983b). In fact, the difference between C and ID mice became greater, which implies that a simple difference in the number of immunocompetent cells cannot fully explain the altered immune responses including lymphocyte proliferation associated with iron deficiency. Decreased activity of ribonucleotide reductase has been reported in white blood cells in iron-deficient humans (Hoffbrand et al. 1976). Studies are underway in our laboratory to determine whether the activity of this enzyme is also reduced in spleen cells, T-cell subsets and B cells from iron-deficient mice.

The mechanisms by which iron deficiency alters PIP2 hydrolysis remain to be elucidated. Phospholipase C, the enzyme that catalyzes PIP2 hydrolysis, contains zinc rather than iron (Little and Otnäss 1975). The effects of iron on PIP2 hydrolysis are therefore indirect and could be related to changes in the cell membrane composition. Given the higher ³H-myoinositol incorporation into cellular phospholipids of ID than C mice, we speculate that there may be a defect in the biosynthesis of PIP2 in iron-deficient cells, a difference in intracellular inositol concentration or both. An increased degradation of inositol triphosphate in cells from iron-deficient mice is very unlikely because of the presence of lithium chloride in the culture medium. This compound inhibits myoinositol phosphatase, the enzyme responsible for the conversion of inositol monophosphate to myoinositol (Allan and Exton 1993, Berridge et al. 1982). The data presented in this paper do not allow us to determine what type of defects or changes occur in cell membrane composition as well as biosynthesis. However, we know that red blood cells from iron-deficient human subjects have a different cell membrane phospholipid profile, i.e., more total cell membrane phospholipids but less cholesterol and phosphatidylserine than those from iron-sufficient subjects (Kirilenko and Paramonava 1992, Kirilenko et al. 1993). Experiments are underway to analyze specific cell membrane phospholipid concentration of lymphocytes from iron-deficient and control mice.

In summary, our data suggest that reduced hydrolysis of cell membrane phospholipids contributes to impaired lymphocyte proliferation associated with iron deficiency. However, other mechanisms, including decreased activity of iron-dependent enzymes and iron containing enzymes, are not ruled out.

	FOOTNOTES

Manuscript received 10 June 1997. Initial reviews completed 9 July 1997. Revision accepted 11 March 1998.

	ACKNOWLEDGMENTS

We wish to thank Carole Lachney and Jessye Hilliard for their technical assistance during the preparation of this manuscript.

	LITERATURE CITED

Abstract Introduction Methods Results Discussion References

• Abbas, K., Lichtman A. H. & Jordan J. S. (1991) Molecular basis of T cell antigen recognition and activation. In: Cellular and Molecular Immunology (Abbas, A. B., Lichtman, A. H. & Pober, J. S., eds.), pp. 139-167. W. B. Saunders, Philadelphia, PA.
• Allan C. J., Exton J. H. Quantification of inositol phospholipid breakdown in isolated rat hepatocytes. Biochem. J. 1993; 290:865-872
• Batty I. H., Nahorski S. R. Analysis of ³H-inositol phosphate formation and metabolism in cerebral-cortical slices. Biochem J. 1992; 288:807-815
• Berridge M. J. Inositol triphosphate and diacylglycerol: two interacting second messengers. Annu. Rev. Biochem. 1987; 56:159-193[Medline]
• Berridge M. J., Downes C. P., Hanley M. R. Lithium amplifies agonist-dependent phosphatidylinositol responses in brain and salivary glands. Biochem. J. 1982; 206:587-595[Medline]
• CEM Innovators in Microwave Technology (1991) Microwave Sample Preparation: Applications and Manual. CEM Corporation, Matthews, NC.
• Chandra R. K., Saraya A. K. Impaired immunocompetence associated with iron deficiency. J. Pediatr. 1975; 86:899-902[Medline]
• Dallman P. R. Iron deficiency and the immune response. Am. J. Clin. Nutr. 1987; 46:329-334[Abstract/Free Full Text]
• Desai D. M., Newton M. E., Kadlecek T. K., Weiss A. Stimulation of the phosphatidylinositol pathway can induce T cell-activation. Nature (Lond.) 1990; 348:66-69[Medline]
• Galan P., Thibault H., Preziosi P., Hercberg S. Interleukin-2 production in iron deficient children. Biol. Trace Elem. Res. 1992; 32:421-426[Medline]
• Hoffbrand A. V., Ganeshaguru K., Hooton J.W.L., Tattersall M.H.N. Effects of iron deficiency and desferroxamine on DNA synthesis. Br. J. Haematol. 1976; 33:517-526[Medline]
• Kemahli A. S., Babacan E., Cavdar A. O. Cell mediated immune responses in children with iron deficiency and combined iron and zinc deficiency. Nutr. Res. 1988; 8:129-136
• Kirilenko N. P., Paramonova I. V. Erythrocyte membrane lipid composition in patients with iron deficiency during therapy. Gematol. Transfuziol. 1992; 37:20-23[Medline]
• Kirilenko N. P., Sliusar N. N., Matsul V. M., Kargapolov A. V. Phospholipid and glycerol levels in women's erythrocytes depending on body iron reserves. Gematol. Transfuziol. 1993; 38:31-33[Medline]
• Krantman H. J., Young S. R., Ank B. J., O'Donnel M. C., Rachelefsky G. S., Stiehman R. E. Immune function in pure iron deficiency. Am. J. Dis. Child. 1982; 136:840-844[Abstract/Free Full Text]
• Kuvibidila, S. & Baliga, B. S. (1997) Effects of iron deficiency and iron repletion on protein kinase C activity. FASEB J. 11: A651 (abs.).
• Kuvibidila S., Baliga B. S., Murthy K. K. Impaired protein kinase C activation as one of the possible mechanisms of reduced lymphocyte proliferation in iron deficiency. Am. J. Clin. Nutr. 1991; 54:944-950[Abstract/Free Full Text]
• Kuvibidila S., Baliga B. S., Nauss K. M., Suskind R. M. Impairment of blastogenic response of splenic lymphocytes from iron deficient mice: in vivo repletion. Am. J. Clin. Nutr. 1983a; 37:15-25[Abstract/Free Full Text]
• Kuvibidila S., Baliga B. S., Suskind R. M. The effects of iron deficiency anemia on cytolytic activity of mice spleen and peritoneal cells against allogenic tumor cells. Am. J. Clin. Nutr. 1983b; 38:238-244[Abstract/Free Full Text]
• Kuvibidila, S., Baliga, B. S. & Suskind, R. M. (1989) Consequences of iron deficiency on infection and immunity. In: Textbook of Gastroenterology and Nutrition in Infancy, 2nd ed. ( Lebenthal, E., ed.), pp. 423-431. Raven Press, New York, NY.
• Kuvibidila S., Dardenne M., Savino W., Lepol F. Influence of iron deficiency anemia on selected thymus functions in mice: thymulin biological activity, T-cell subsets, and thymocyte proliferation. Am. J. Clin. Nutr. 1990; 51:228-232[Abstract/Free Full Text]
• Latunde-Dada, G. O. & Young, S. P. (1992) Iron deficiency and immune responses. Scand. J. Immunol. 36 (suppl. 11): 207-209.
• Little C., Otnäss A. B. The metal ion dependence of phospholipase C from Bacillus cereus. Biochim. Biophys. Acta 1975; 391:326-333[Medline]
• May W. S. Jr., Sahyoun N., Wolf M., Cuatrecasas P. Role of intracellular calcium mobilization in the regulation of protein kinase C-mediated membrane processes. Nature (Lond.) 1985; 317:549-551[Medline]
• Michel M. C., van Tits L.J.H., Trenn G., Sykora J., Brodde O. E. Dissociation between phytohaemagglutinin-stimulated generation of inositol phosphates and Ca²⁺ increase in human mononuclear leucocytes. Biochem. J. 1992; 285:137-141
• Munro, B. H. & Page, E. B. (1993) Statistical Methods for Health Care Research, 2nd ed., pp. 99-125. J. B. Lippincott Company, Philadelphia, PA.
• Munro, B. H., Visintainer, M. A. & Page, E. B. (1986) Differences among group means: one-way analysis of variance. In: Statistical Methods for Health Care Research, 1st ed., pp. 174-199. J. B. Lippincott Company, Philadelphia, PA.
• Pantaleo G., Olive D., Poggi A., Kozumbo W. J., Moretta L., Moretta A. Transmembrane signalling via T11 dependent pathway of human T cell activation: evidence for the involvement of 1,2 diacylglycerol and inositol phosphates. Eur. J. Immunol. 1987; 17:55-60[Medline]
• Shinomura T., Asaoka Y., Oka M., Yoshida K., Nishizuka Y. Synergistic action of diacylglycerol and unsaturated fatty acid for protein kinase C activation: its possible implications. Proc. Natl. Acad. Sci. U.S.A. 1991; 88:5149-5133[Abstract/Free Full Text]
• Thomas A. P., Alexander J., Williamson J. R. Relationship between inositol polyphosphate production and the increase of cytosolic free calcium induced by vasopressin in isolated hepatocytes. J. Biol. Chem. 1984; 259:5574-5584[Abstract/Free Full Text]
• van Tits L.J.H., Michel M. C., Motulsky H. J., Maisel A. S., Brodde O. E. Cyclic AMP counteracts mitogen-induced inositol phosphate generation and increases in intracellular Ca²⁺ concentrations in human lymphocytes. Br. J. Pharmacol. 1991; 103:1288-1294[Medline]
• Wolf M., Levine H. III, May W. S. Jr., Cuatrecasas P., Sahyoun N. A model for intracellular translocation of protein kinase C involving synergism between Ca²⁺ and phorbol esters. Nature (Lond.) 1985; 317:546-551[Medline]

0022-3166/98 $3.00 ©1998 American Society for Nutritional Sciences