Effect of Iron Overload and Iron Deficiency on Atherosclerosis in the Hypercholesterolemic Rabbit

Materials

Sodium-heparin Vacutainers (for plasma preparation), Vacutainers without addition (for serum preparation), needles, and plastic syringes were purchased from Becton Dickinson. Acrodisc LC13 syringe filters were from Gelman Sciences, HPLC columns from Supelco, Chelex-100 resin from Bio-Rad Laboratories, and 15(S)-hydroperoxy-(5Z,8Z,11Z,13E)-eicosatetraenoic acid from Cayman Chemical Co. Organic solvents (HPLC grade) were purchased from Fisher Scientific. All other chemicals were obtained from Sigma Chemical Co. CuSO₄ · 5H₂O was prepared in Chelex-treated water adjusted to pH 3.0 and diluted from a 10 mmol/L stock solution just before use in experiments.

Animals and Diets

Twenty-eight male NZW rabbits (≈3 kg body weight) were divided into 4 groups of 7 animals each and given the following treatments throughout the 9-week study: (1) Rabbits in the FeD group were iron depleted by twice-weekly phlebotomy from a marginal ear vein (15 to 30 mL of blood per week). This method has been shown to produce mild anemia without excessive iron deficiency.²³ (2) Rabbits in the FeO were given intramuscular injections of iron dextran (INFeD; Schein Pharmaceutical, Inc) in doses of 25 mg iron (0.5 mL of a 50 mg/mL solution of iron dextran) every 3 days for the first 3 weeks of the study (total dose of 150 mg iron, or ≈50 mg/kg body weight) and every 5 days for an additional 6 weeks (total dose by the end of the study was 350 mg iron, or ≈120 mg/kg body weight). A relatively moderate degree of FeO was chosen so as not to confound the detrimental effects due to excessive FeO alone with the effects of increased body iron stores on atherosclerosis. (3) The control group of rabbits was given intramuscular injections (0.5 mL) of sterile saline. (4) An additional group of 7 rabbits was left untreated (the NC group).

All rabbits were fed standard Agway rabbit chow (Ralston Purina Co) for the first 3 weeks of the study. The control, FeO, and FeD groups were fed the standard chow diet containing 1% (wt/wt) cholesterol for an additional 6 weeks. Cholesterol-supplemented chow was prepared weekly by dissolving pure cholesterol in ether and spraying the appropriate amount over the standard chow. The NC group of rabbits continued to be fed the standard chow. All rabbits were allowed to consume the chow and water ad libitum. Blood was obtained from a marginal ear vein at 3, 6, and 9 weeks of the study. Plasma was prepared by centrifugation (1000g) for 15 minutes at 4°C and either used immediately or stored at −70°C (protected from light) for subsequent assays.

At the end of the study, fasted animals were killed with pentobarbital (120 mg/kg body weight) injected into a marginal ear vein. The rabbits were incised midventrally; blood was collected from the left ventricle into both heparinized and nonheparinized Vacutainers; and the liver, aortic arch, and thoracic aorta were excised. The aorta was subdivided so that same-sized sections from the same location of the aorta of each animal were used for the respective assays. Tissue and plasma or serum samples were either used immediately or frozen at −70°C until the time of analysis (within 6 weeks). All animal studies were approved by the Boston University Medical Centre Institutional Animal Care and Use Committee.

Iron Status

Serum iron, TIBC, and transferrin iron saturation were measured spectrophotometrically in nonhemolyzed serum samples as described.²⁴ Colorimetric determination of hemoglobin concentration in whole blood was performed with Sigma kit procedure No. 525 based on the method of Drabkin and Austin.²⁵ Total liver and thoracic aorta iron were measured by atomic absorption spectroscopy as previously described by us⁸ and the results expressed in micrograms of iron per gram of wet tissue.

Markers of Lipid Peroxidation

Fresh plasma, liver, and thoracic aorta samples were analyzed for acylated F₂-isoprostanes, a marker for in vivo lipid peroxidation.^{8 26} Lipids were extracted by a modified Folch procedure and base hydrolyzed.²⁷ After purification and derivatization, the resulting free F₂-isoprostanes were measured by gas chromatography/negative-ion chemical ionization mass spectrometry as described.²⁷ Neutral lipid hydroperoxides in n-hexane extracts of fresh plasma samples were measured by HPLC with isoluminol chemiluminescence detection as described.^{28 29}

Antioxidant Status

Ascorbic acid was measured in fresh plasma extracts by HPLC with electrochemical detection.³⁰ α-Tocopherol in plasma was extracted into n-hexane and stored at −20°C until assayed by HPLC with electrochemical detection.³¹ Because levels of circulating α-tocopherol are influenced by increases in serum lipid levels,³² plasma α-tocopherol levels were expressed as the ratio of α-tocopherol (in micromoles per liter) to total cholesterol plus triglycerides (in millimoles per liter) as described.³³

For analysis of α-tocopherol levels in the liver and thoracic aorta, the samples were cleaned of extraneous tissue and homogenized in 1.0 mL of ice-cold 10 mmol/L PBS, pH 7.4, containing 5.0 mmol/L of the metal chelator DTPA to prevent ex vivo lipid peroxidation. A sample of tissue homogenate was subjected to base hydrolysis by addition of 1.0 mL of 10N NaOH and 1% SDS for analysis of tissue protein contents. The remaining tissue homogenate was then extracted with 1.0 mL of methanol and 5.0 mL n-hexane, and the lipid-soluble antioxidants in the hexane extract were measured as described.³¹

Plasma Lipids

Plasma total cholesterol and triglycerides were quantified enzymatically as previously described.^{34 35} HDL cholesterol was measured after phosphotungstic acid–MgCl₂ precipitation of apoB-containing lipoprotein fractions.³⁶ The difference between total cholesterol and HDL cholesterol was calculated as the combined cholesterol in the VLDL and LDL fractions.

Fatty Streak Quantification

The extent of atherosclerotic lesion development in the aortic arch and thoracic aorta was determined by methods previously described.³⁷ Aortic segments were pinned at in situ length, fixed for en face preparation with 4% phosphate-buffered paraformaldehyde (pH 7.0) at room temperature for 1 hour, and then stored at 4°C overnight. For each en face preparation, the segments were cleaned of adventitia and rinsed in 60% isopropanol, and the inner aortic surface was stained for 25 minutes with a saturated solution of oil red O in 60% isopropanol. The aorta was counterstained with Gill’s type hematoxylin (1:20 dilution) for 1 minute. The area of oil red O staining (lesion area) was quantified with a Bioscan Optimas 4.02 image analysis (Aldus Corp), and the degree of lesion formation was expressed as a percentage of the total area examined (20 mm²).

LDL Oxidation

At the time when the rabbits were killed, LDL was prepared from fresh plasma by the method of Chung and colleagues³⁸ by using single vertical-spin discontinuous density gradient ultracentrifugation as previously described by us.³⁹ Standard incubation of LDL for susceptibility to oxidation was performed at a concentration of 0.1 mg LDL protein per milliliter in the presence of 1.25 μmol/L CuCl₂ at 37°C. Lipid peroxidation was determined by assay of sample absorbance at 234 nm (diene conjugation) at 10-minute intervals with a Hitachi U-2000 spectrophotometer equipped with a thermostatic six-cell holder. LDL susceptibility to lipid peroxidation was quantified by the duration of the lag phase before the propagation phase of diene conjugation as described by Esterbauer and colleagues⁴⁰ as well as by the rate of lipid peroxidation during the initiation and propagation phases.

Statistics

Unless otherwise specified, all results are expressed as mean±SD. For comparisons between the different groups, unpaired Student’s t test was used. Pearson’s correlations were used to determine relationships between covariates. Statistical significance was accepted if the null hypothesis was rejected at the P<.05 level.

Iron, Cholesterol, and Antioxidant Levels in NC Rabbits

Groups of 7 rabbits were either given intramuscular injections of iron dextran (FeO group) or saline (control group) or were phlebotomized (FeD group) for 3 weeks. In the FeD group, significantly decreased serum iron levels (P<.01), transferrin iron saturation (P<.01), and hemoglobin levels (P<.01) were observed compared with the control group (Fig 1⇓). Iron dextran injections (a total of 150 mg of iron per animal) in the FeO group did not significantly affect serum iron or hemoglobin levels but significantly (P<.05) decreased TIBC and increased (P<.05) transferrin saturation (Fig 1⇓) values compared with those in control animals. FeO animals also exhibited a 23% decrease in plasma cholesterol levels compared with control animals (P<.05), whereas FeD did not affect plasma cholesterol levels (Table 1⇓). As also shown in Table 1⇓, neither FeO nor FeD had an effect on plasma levels of either the water-soluble antioxidants (ascorbic acid and uric acid) or the lipid-soluble antioxidant α-tocopherol. In addition, no plasma lipid hydroperoxides were detected (<5 nmol/L) in any of the groups (data not shown).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1.

Effects of FeO and FeD on circulating iron levels. Three groups of rabbits were iron depleted, iron loaded, or used as control animals and first fed a standard chow and then a cholesterol-containing chow as described in the legend to Table 2⇓. Blood samples were collected at weeks 3 (standard chow), 6, and 9 (cholesterol-containing chow). Serum was assayed for iron levels (A) and TIBC (B), and transferrin saturation was calculated (C). Blood was assayed for hemoglobin levels (D). Values for a fourth group of 7 rabbits fed standard chow for 9 weeks (NC rabbits) were as follows (mean±SD): serum iron (μmol/L)=27.3±4.7; TIBC (μmol/L)=51.9±5.1; transferrin saturation (%)=52.4±6.6; and hemoglobin (g/dL)=14.5±1.35. Values shown in the figure represent mean±SE (n=7). (▵), control; (▪), FeO; (□), FeD. *P<.05, ‡P<.01 vs control.

Lipoproteins, Iron, and Antioxidant Status in Hypercholesterolemic Rabbits

Cholesterol feeding (1% wt/wt) for 6 weeks led to a significant increase in plasma total and VLDL plus LDL cholesterol levels in all treatment groups (control, FeO, and FeD) compared with NC animals (Table 2⇓). This effect was less marked in the FeO group, and the difference between the FeO and the control group was of marginal statistical significance (P=.055). There was no difference in the plasma levels of HDL cholesterol between the groups (Table 2⇓). Triglyceride levels were higher in the cholesterol-fed rabbits than in the NC animals but not different between the treatment groups (control, FeO, and FeD; Table 2⇓).

The iron status of the treatment groups was assessed after 3 weeks of cholesterol feeding (week 6 of the study) and at the end of the study (week 9; Fig 1⇑). At week 6, levels of serum iron and blood hemoglobin were significantly lower in the FeD group than in control rabbits (P<.05 and P<.01, respectively). Furthermore, significantly decreased TIBC (P<.05) and increased transferrin saturation (P<.01) values were observed in the FeO rabbits compared with control animals (Fig 1⇑). At week 9, the differences in transferrin saturation between the groups were similar to those observed at week 6 of the study (Fig 1C⇑). Interestingly, there was a decrease in serum iron levels as a result of cholesterol feeding in the control group (P=.077, week 3 versus week 9). As a result of this decrease, serum iron levels at week 9 were no longer different between FeD and control animals but were now significantly (P<.01) different between the FeO and control animals (Fig 1A⇑). Furthermore, a significant decrease in the TIBC level (P<.05) was observed in control rabbits from the start of cholesterol feeding to the time when the rabbits were killed. Because high concentrations of plasma lipids interfere with the hemoglobin assay, hemoglobin levels could not be determined at week 9.

As shown in Table 3⇓, cholesterol feeding also led to a significant (P<.05) reduction in lipid-standardized plasma α-tocopherol levels but had no effect on plasma levels of ascorbic acid or F₂-isoprostanes, a marker of lipid peroxidation. Neither FeO nor FeD significantly affected plasma levels of F₂-isoprostanes compared with the cholesterol-fed control group (Table 3⇓).

As expected, liver iron levels significantly (P<.01) increased in the iron dextran–treated group and decreased in the phlebotomy group compared with the control group. Similar but nonsignificant trends were observed for aortic iron levels (Fig 2A⇓). As shown in Fig 2B⇓, aortic α-tocopherol levels did not differ between any of the groups. However, liver α-tocopherol levels were higher in the control group than in the NC group (P<.05), an observation that probably reflects increased hepatic cholesterol levels as a result of cholesterol feeding.^{20 41} Within the treatment groups, hepatic levels of α-tocopherol were significantly (P<.01) higher in the FeO than the control rabbits (Fig 2B⇓). F₂-isoprostane levels in the thoracic aorta did not differ significantly among any of the groups. However, hepatic F₂-isoprostane levels in the control group were significantly (P<.05) lower than in the NC group (Fig 2C⇓).

• Download figure
• Open in new tab
• Download powerpoint

Figure 2.

Effects of FeO and FeD on liver and thoracic aorta levels of iron, α-tocopherol, and F₂-isoprostanes. Rabbits were treated as described in the legend to Table 2⇑. At the end of the study, liver (shaded bars) and thoracic aorta (open bars) levels of iron (A), α-tocopherol (B), and F₂-isoprostanes (C) were measured. Values represent mean±SE (n=7). *P<.01 vs control; ‡P<.05 vs NC rabbits.

Aortic Atherosclerotic Lesion Formation

Lesions were not observed in the NC rabbits but were present in both the aortic arch, and, to a lesser degree, in the thoracic aorta of the cholesterol-fed animals (Fig 3⇓). There was a significant (P<.05) 56% decrease in aortic arch lesion coverage in the FeO group compared with the control group. In contrast, FeD led to a 32% increase in aortic arch lesions, but this difference was not statistically significant (P=.19). Thoracic aorta lesion coverage was lower in the FeO group than in the control rabbits (P=.09). Aortic arch but not thoracic aorta lesions were significantly (P=.01) and positively correlated with plasma cholesterol levels (r=.54).

• Download figure
• Open in new tab
• Download powerpoint

Figure 3.

Effects of FeO and FeD on aortic arch and thoracic aortic lesions in hypercholesterolemic rabbits. Rabbits were treated as described in the legend to Table 2⇑. At the end of the study, the aortic arch and thoracic aorta were excised, fixed, and stained with oil red O. The area of oil red O staining (lesion area) was quantified and the degree of lesion formation expressed as a percentage of the total area examined (20 mm²). Values represent mean±SE (n=7). *P<.05 vs control.

Ex Vivo LDL Oxidation

The susceptibility of LDL to oxidation was determined ex vivo by using LDL concentrations standardized to protein (0.1 mg LDL protein per milliliter). No significant differences in either the lag phase or the R_i of lipid peroxidation were observed between any of the groups. However, cholesterol feeding led to a dramatic increase in the R_p of lipid peroxidation (P<.001, control versus NC group; Table 4⇓). Because the animals were fed cholesterol and the cholesterol contents of LDL may vary between animals, we measured the cholesterol content in isolated LDL. As expected, the LDL from control animals contained significantly (P<.001) higher levels of cholesterol per milligram of LDL protein than the LDL from NC rabbits (Table 4⇓). We also found that the LDL cholesterol content was positively correlated with total plasma cholesterol (r=.52, P=.028) and R_p (r=.62, P=.006) and negatively correlated with the lag phase (r=−.54; P=.021). These data are in agreement with previous observations of a significant negative correlation between the lag phase and LDL cholesterol content in humans.⁴² To adjust for cholesterol content, the lag phase, R_i, and R_p were expressed per micromole of LDL cholesterol (Table 4⇓). This analysis showed that the LDL from control rabbits was significantly more susceptible to ex vivo oxidation than the LDL from NC animals, as indicated by a decreased lag phase (P<.05) and an increased R_p (P<.001). However, FeO or FeD did not affect the susceptibility of LDL to ex vivo oxidation compared with controls (Table 4⇓).