Potential conflict of interest: Nothing to report.
Effect of an oral iron chelator or iron-deficient diets on uroporphyria in a murine model of porphyria cutanea tarda†
Article first published online: 13 SEP 2007
Copyright © 2007 American Association for the Study of Liver Diseases
Volume 46, Issue 6, pages 1927–1834, December 2007
How to Cite
Gorman, N., Zaharia, A., Trask, H. S., Szakacs, J. G., Jacobs, N. J., Jacobs, J. M., Balestra, D., Sinclair, J. F. and Sinclair, P. R. (2007), Effect of an oral iron chelator or iron-deficient diets on uroporphyria in a murine model of porphyria cutanea tarda. Hepatology, 46: 1927–1834. doi: 10.1002/hep.21903
- Issue published online: 28 NOV 2007
- Article first published online: 13 SEP 2007
- Manuscript Accepted: 29 JUN 2007
- Manuscript Received: 9 MAR 2007
- Department of Veterans Affairs
- National Institutes of Health. Grant Numbers: ES 06263, AA 12898
Porphyria cutanea tarda is a liver disease characterized by elevated hepatic iron and excessive production of uroporphyrin (URO). Phlebotomy is an effective treatment that probably acts by reducing hepatic iron. Here we used Hfe(−/−) mice to compare the effects on hepatic URO accumulation of two different methods of hepatic iron depletion: iron chelation using deferiprone (L1) versus iron-deficient diets. Hfe(−/−) mice in a 129S6/SvEvTac background were fed 5-aminolevulinic acid (ALA), which results in hepatic URO accumulation, and increasing doses of L1 in the drinking water. Hepatic URO accumulation was completely prevented at low L1 doses, which partially depleted hepatic nonheme iron. By histological assessment, the decrease in hepatic URO accumulation was associated with greater depletion of iron from hepatocytes than from Kupffer cells. The L1 treatment had no effect on levels of hepatic cytochrome P4501A2 (CYP1A2). L1 also effectively decreased hepatic URO accumulation in C57BL/6 Hfe(−/−) mice treated with ALA and a CYP1A2 inducer. ALA-treated mice maintained on defined iron-deficient diets, rather than chow diets, did not develop uroporphyria, even when the animals were iron-supplemented either directly in the diet or by iron dextran injection. Conclusion: The results suggest that dietary factors other than iron are involved in the development of uroporphyria and that a modest depletion of hepatocyte iron by L1 is sufficient to prevent URO accumulation. (HEPATOLOGY 2007.)
Porphyria cutanea tarda (PCT) is the most common human porphyria and manifests as a skin disease characterized by massive hepatic production and an accumulation of uroporphyrin (URO). Most cases are sporadic rather than familial.1, 2 There are several associated risk factors, including the consumption of alcoholic beverages and increased body iron stores.1–3 PCT patients usually display hepatic siderosis, and the disease is responsive either to phlebotomy, which presumably acts through iron depletion, or a treatment with low-dose chloroquine or hydroxychloroquine.1, 2 In addition, some PCT patients have quite high levels of hepatic iron associated with the common mutations in the HFE gene (C282Y and H63D),4, 5 which are also found in hereditary hemochromatosis (type I).
The hepatic accumulation of URO in PCT is attributed to the oxidation of the heme pathway intermediate, uroporphyrinogen, to URO (Fig. 1). In rodents, this oxidation is catalyzed mainly by cytochrome P4501A2 (CYP1A2)6–8 and is thought to lead to the inactivation of uroporphyrinogen decarboxylase through the formation of an inhibitor of this enzyme9 (Fig. 1). Several factors that influence this sequence of events have been identified through investigations with experimental animal models and liver cell cultures. These factors include the hepatic content of CYP1A2, the amount of hepatic iron, the activity of 5-aminolevulinic acid (ALA) synthase (the first and rate-limiting enzyme of the heme pathway), the hepatic concentration of ascorbic acid,10 and other factors influenced by the genetic background of the mice.1, 11 The absolute requirement for CYP1A2 is shown by the finding that Cyp1a2(−/−) mice are totally resistant to uroporphyria produced in several protocols, including the administration of excess iron.7, 8
In early animal studies, iron deficiency produced in mice by defined iron-deficient diets was shown to prevent uroporphyria induced by 2,3,7,8-tetrachlorodibenzo-p-dioxin or polychlorinated biphenyls.12–14 In addition, the treatment of mice or rats with desferrioxamine partially decreased hepatic URO accumulation caused by hexachlorobenzene.15, 16 The purpose of this study was to determine the effectiveness of reducing hepatic iron in preventing the development of uroporphyria. Hfe(−/−) mice, which accumulate URO when administered ALA,17 were used to model the elevated hepatic iron levels found in PCT patients.17, 18 In initial experiments, we found that iron-deficient defined diets completely prevented uroporphyria, as previously reported by others.12–14 However, mice raised on these defined diets, even when supplemented with iron, still did not become uroporphyric. Therefore, we evaluated the effectiveness of iron deficiency caused by the removal of hepatic iron with the oral iron chelator deferiprone (L1), a clinically used iron chelator.19 The complete prevention of uroporphyria was achieved by an intermediate dose of L1, which only partially depleted hepatic iron.
Materials and Methods
129S6/SvEvTac (129) Hfe(−/−) mice were generated in the laboratory of Dr. N. Andrews (Children's Hospital, Boston, MA)18 and were bred at the Veterans Affairs animal facility. Wild-type 129 mice were purchased from Taconic (Germantown, NY). Hfe(−/−) mice in a C57BL/6 (B6) background were generated in the laboratory of Dr. W. Sly (Department of Biochemistry, St. Louis University)20 and were bred at the Veterans Affairs animal facility. Wild-type B6 mice were purchased from the National Cancer Institute. Only male mice (8-10 weeks old, 20-25 g at the start of the experiment) were used. Animals were housed in standard plastic cages, were maintained on 12-hour light/dark cycles, and, unless indicated otherwise, had free access to rodent chow (Teklad LM-485, containing 250-300 mg Fe/kg of diet) from Harlan Teklad (Madison, WI). L1 was purchased from Fisher Scientific (Pittsburgh, PA) and dissolved in drinking water at the concentrations indicated in the figure legends. ALA hydrochloride (Sigma, St. Louis, MO) was dissolved in drinking water at 2 mg/mL. 3,3′,4,4′,5-Pentachlorobiphenyl (PCB126; Ultrascientific, North Kingston, RI) was injected intraperitoneally (75 μg/kg body weight), as indicated, after dissolving in dimethyl sulfoxide and diluting in corn oil. Iron dextran solution (100 mg Fe/mL; Phoenix Pharmaceutical, St. Joseph, MO) was injected, as indicated, at a dose of 500 mg Fe/kg body weight intraperitoneally, 3 days prior to the injection of PCB126 and the start of the ALA treatment.
The defined diets used with the 129 mice were from Harlan Teklad: diet #80396, an iron-deficient diet containing 10-20 mg of Fe/kg of diet, or diet #01583, an iron-sufficient diet that was the same diet as #80396, but supplemented with FeSO4 (200 mg Fe/kg of diet). In some experiments with B6 mice, the defined diets were from ICN (Aurora, OH): either iron-deficient diet #902199 containing 7-25 mg Fe/kg of diet or iron-sufficient diet #902199 supplemented with FeSO4 (200 mg Fe/kg of diet).
All animal protocols were approved by the local Veterans Affairs animal use committee.
Preparation of the Liver Fractions.
Liver homogenates and microsomes were prepared as described.21
The porphyrin composition in the liver homogenates was determined spectrophotofluorometrically22 and, for some samples, was confirmed by high-performance liquid chromatography.23 The protein concentrations were determined by the method of Lowry et al.,24 with bovine serum albumin used as a standard. Microsomal methoxyresorufin demethylase (MROD) and uroporphyrinogen oxidation (UROX) activities were measured as described.25 Nonheme Fe in liver homogenates and mouse diets was measured by a modification of the method of Torrance and Bothwell as described.17
The results are presented as means ± the standard deviation (SD). The significance was determined by a one-way analysis of variance, except when the SDs were significantly different; then, a nonparametric test was used. P < 0.05 was considered significant.
Effect of L1 on Hepatic URO Accumulation and Nonheme Iron in ALA-Treated 129 Hfe(−/−) Mice.
Previously, it has been shown that Hfe(−/−) mice of this strain develop uroporphyria following continual treatment with the porphyrin precursor, ALA.17 To deplete hepatic iron, Hfe(−/−) mice were additionally treated with increasing concentrations of L1 in the drinking water starting 2 weeks before ALA (Fig. 2). In the absence of the L1 treatment, all ALA-treated mice accumulated hepatic URO, as expected17 (Fig. 2A). Without the ALA treatment, the hepatic URO accumulation was less than 1 nmol/g of liver. The lowest dose of L1 (0.5 mg/mL) produced an 80% decrease in URO accumulation but only a 20% decrease in nonheme iron (Fig. 2B). L1 at 1 mg/mL totally prevented URO accumulation, even though nonheme iron levels were still significantly higher than those of wild-type mice not treated with L1 (P < 0.05; Fig. 2B). L1 at 2 mg/mL decreased hepatic nonheme iron to about the same level as that found in wild-type 129 mice (Fig. 2B).
Histological Assessment of Hepatic Iron Depletion.
In Hfe(−/−) mice of the 129 substrain used here, stainable iron is present in both hepatocytes and Kupffer cells.17 Therefore, we determined histologically which cell types were iron-depleted by L1. Although wild-type 129 mice have stainable iron in Kupffer cells but not in hepatocytes, they do not accumulate hepatic URO when treated with ALA.17 Figure 3A-D shows representative Perls-stained sections of livers from 129 Hfe(−/−) mice following treatment with increasing doses of L1. Figure 3E summarizes the quantitative histological assessment of the iron staining in hepatocytes and Kupffer cells. At the lowest L1 dose used (0.5 mg/mL), there was about a 45% decrease in iron staining in hepatocytes, but there was no decrease in Kupffer cells, and there was an 80% decrease in URO (Fig. 2A). At this L1 concentration, the overall hepatocyte iron staining compared with that of untreated mice was reduced in 5 of the 6 mice. The concentration of L1, which completely suppressed URO accumulation (1 mg/mL; Fig. 2A), did not completely eliminate stainable hepatocyte iron (Fig. 3C). L1 did not detectably decrease Kupffer cell staining until the concentration reached 3 mg/mL (Fig. 3).
Effect of L1 on Hepatic CYP1A2.
Hepatic URO accumulation is highly dependent on the level of CYP1A2, especially in the range of its constitutive expression.7, 26 Therefore, hepatic CYP1A2 was assessed in L1-treated mice to ensure that the chelation of iron did not decrease CYP1A2. Two specific microsomal enzyme activities, MROD and UROX,26 were used to assess CYP1A2. Even at the highest L1 concentration used in this study (3 mg/mL), there was no effect of the L1 treatment on CYP1A2-catalyzed enzyme activities (Fig. 4).
Effect of L1 on URO Accumulation in B6 Hfe(−/−) Mice.
To ensure that the effects of L1 on hepatic URO accumulation were not restricted to Hfe(−/−) mice of a single genetic background, an experiment similar to that shown in Fig. 2 was performed with B6 Hfe(−/−) mice, which have only 40% of the hepatic Fe of 129 Hfe(−/−) mice.27 The mice were also treated with a moderate dose of a CYP1A2 inducer, PCB126, because untreated B6 Hfe(−/−) mice, though expressing constitutive levels of CYP1A2, do not accumulate URO when treated with ALA alone (data not shown). At the PCB126 dose used, cytochrome P4501A1 (CYP1A1) is only slightly increased.10 Figure 5 shows that L1 at 2 mg/mL significantly decreased the levels of both hepatic URO and nonheme iron in B6 Hfe(−/−) mice. URO was not detected at L1 concentrations greater than 2 mg/mL, but these doses of L1 caused a massive accumulation of another porphyrin, protoporphyrin, exceeding 1000 nmol/g of liver, which was accompanied by liver damage, as indicated by lymphocyte infiltration, focal hepatic necrosis, and cholestatic hepatitis.
Effects of Iron-Deficient Defined Diets.
Dietary iron deficiency was tested as an alternative to oral iron chelation. 129 Hfe(−/−) mice were treated with ALA alone for 8 weeks and were maintained on normal laboratory chow or on a defined diet that was either iron-deficient or supplemented with ferrous sulfate. Table 1 shows that the mice maintained on the chow diet became highly uroporphyric, as expected, whereas the mice maintained on the defined iron-deficient diet accumulated no URO. Mice maintained on the iron-supplemented defined diet did not develop uroporphyria, although the hepatic nonheme iron content was twice that of mice maintained on the chow diet (Table 1). Most of the extra iron was detected in hepatocytes, as determined histologically. As expected, the hepatic nonheme iron content of the mice on the iron-deficient diet was much lower than that of mice on the chow diet (Table 1). Other experiments, similar to those shown in Table 1, showed that a defined diet from a different source (see the Materials and Methods section), even when supplemented with iron, was also unable to support the development of uroporphyria. In these other experiments, wild-type B6 mice were treated with PCB126 to induce CYP1A2 under conditions similar to those of the experiment in Fig. 5. In all these experiments with defined diets, there was no effect on either constitutive or PCB126-induced levels of CYP1A2 in comparison with chow-fed mice (data not shown).
|Diet||Dietary Fe Content (mg/kg Diet)||Hepatic URO (nmol/g Liver)||Hepatic Nonheme Fe (μg of Fe/g Liver)|
|Regular chow||245 ± 20a||310 ± 30a,b||560 ± 85a,b|
|+Fe||220 ± 20b||1 ± 0.8a||1220 ± 110a|
|−Fe||20 ± 1a,b||<1b||90 ± 35b|
To ensure that the problems with defined diets were not caused by impaired intestinal absorption of iron, wild-type B6 mice treated with PCB126 and ALA were also administered iron by an injection with iron dextran (500 mg Fe/kg).26 In iron dextran–treated mice maintained on the defined diets, there was no URO accumulation, but in iron dextran–treated mice maintained on the chow diet, there was massive URO accumulation (Table 2). Even though no hepatic URO accumulated in mice on the defined diet administered iron dextran, there was a large accumulation of iron in the livers of these animals. It is concluded that the use of defined diets prevents the uroporphyria that develops in mice on chow diets and that this is not due to hepatic iron insufficiency or decreased iron absorption from the defined diets.
|Diet||Iron Dextran||Dietary Fe Content (mg Fe/kg Diet)||Hepatic URO (nmol/g Liver)||Hepatic Nonheme Fe (μg Fe/g Liver)|
|Regular chow||No||245 ± 20a||75 ± 30a,b||74 ± 13a,b,c|
|Yes||245 ± 20||320 ± 40a||2640 ± 275a|
|+Fe||No||240 ± 30b||<1b||73 ± 13d|
|Yes||220 ± 20||<1||2115 ± 215c|
|−Fe||No||18 ± 2a,b||<1||28 ± 9b,d|
The normal therapy for PCT, phlebotomy, acts by depleting hepatic iron. One purpose of the present study was to determine the effectiveness of decreasing hepatic iron in preventing the development of uroporphyria in an animal model of PCT, Hfe(−/−) mice treated with ALA. Two different methods to achieve iron deficiency were compared: the chelation of iron by the administration of an oral iron chelator, L1, and the production of iron deficiency by feeding iron-deficient defined diets. The results here demonstrated for the first time in an animal model of PCT that relatively low doses of an oral iron chelator completely prevented hepatic uroporphyria without large decreases in hepatic iron and without evidence of hepatic toxicity.
The L1 dose response studies furnished new insights into the quantitative relationships between total hepatic iron, its distribution in hepatocytes and Kupffer cells, and the minimal iron depletion required for the prevention of uroporphyria. In 129 Hfe(−/−) mice, a low dose of L1 (1.0 mg/mL drinking water) completely prevented URO accumulation but caused only 40% depletion of hepatic iron (Fig. 2). In particular, hepatic nonheme iron remained elevated with respect to the levels in wild-type mice that do not accumulate URO (Fig. 2). Thus, hepatic iron loading may still be present without an associated increase in URO accumulation. This conclusion is supported by previous studies with wild-type B6 mice treated with increasing doses of iron dextran where considerable hepatic iron loading was necessary before URO accumulation occurred.10, 26
It needs to be emphasized that the hepatic iron detected either histologically or as total nonheme iron (Figs. 2B and 3) was probably mainly present as inert iron stored as ferritin or hemosiderin.17 We observed that although low concentrations of L1 (1 mg/mL) completely prevented URO accumulation, they caused only partial iron depletion from hepatocytes and no depletion of Kupffer cell iron (Fig. 3). This suggests that L1 preferentially depletes iron from a different pool of hepatic iron, that is, the postulated free iron pool that is involved in the oxidative process that is thought to result in inactivation of uroporphyrinogen decarboxylase and hence the subsequent development of uroporphyria.1, 2
The question arises as to whether the action of L1 on URO accumulation is mainly due to the depletion of the hepatic active iron pool through excretion of the iron-L1 chelate. L1 may also prevent URO accumulation by the formation of an Fe-L1 chelate that is not redox-active.19, 28 Other iron chelates such as Fe–ethylene diamine tetraacetic acid29 and Fe-nitrilotriacetate30 stimulate CYP1A2-catalyzed microsomal oxidation of uroporphyrinogen to URO (Fig. 1), a reaction that is considered to play a critical role in the development of uroporphyria in vivo.7, 8 In preliminary experiments (N. Gorman, P. Sinclair, unpublished observations), the 50%-60% stimulation of microsomal UROX caused by the nitrilotriacetate chelate of iron was completely prevented by the addition of 0.15 mM L1 (in a 3:1 molar ratio of L1 to the Fe). In the absence of added Fe-nitriloacetate, this L1 concentration had no effect on uroporphyrinogen oxidation. These results suggest that L1 not only may deplete hepatic iron but may also prevent uroporphyria in vivo by the formation of the non–redox-active L1-Fe chelate.
Although the data presented here suggest that the mechanism by which L1 prevents URO accumulation is due to the chelation of iron, other possibilities exist. Nonspecific toxicity was unlikely because no pathology was evident histologically in the experiments with 129 mice, even at L1 doses as high as 3 mg/mL. Another possible effect of L1 on URO accumulation could be due to zinc chelation, which can occur when iron is completely chelated.31 However, it appears that with low concentrations of L1, in the presence of excess chelatable iron, only a small amount of free chelator is available to chelate zinc.32 Therefore, because URO accumulation was reduced maximally when the liver still contained detectable iron (Figs. 2 and 3), it is unlikely that the effect of L1 on hepatic URO accumulation is due to zinc chelation.
In previous studies designed to assess the effects of Fe deficiency on uroporphyria in rodents, iron-deficient diets were used to produce iron deficiency, and uroporphyria was induced by the administration of polyhalogenated aromatic hydrocarbons.12–14 Uroporphyria developed in mice on the normal chow diet but not in mice on the iron-deficient defined diets. It was concluded that the prevention of uroporphyria in animals on the defined diets was due to a lack of iron in the defined diet. Our results suggest that the decrease in hepatic URO accumulation in animals on the defined diets is due not to a deficiency of iron but rather to some other effect caused by the use of defined diets because the addition of iron to the defined diet did not restore the susceptibility to uroporphyria (Table 1). Jones et al.13 also found that the development of uroporphyria caused by a treatment with 2,3,7,8-tetrachlorodibenzo-p-dioxin was delayed in mice fed defined diets, even when those diets were iron-supplemented. Similar conclusions concerning the inadequacies of defined diets for supporting the development of uroporphyria have been reached by others (A. G. Smith, MRC Toxicology Unit, Leicester UK, personal communication, 2006). These findings suggest that unidentified dietary factors are involved in the development of uroporphyria; they may be either essential components lacking in the defined diets or agents present in the defined diets that block the development of uroporphyria. If further studies identify such dietary components, this could be important for understanding the complex etiology of experimental uroporphyria and possibly be relevant to the etiology and therapy of human uroporphyria.
Another important finding was that iron limitation by L1 or dietary deficiency did not limit the synthesis of sufficient heme to maintain CYP1A2 levels. Although previous studies concluded that there was no deficiency in CYP1A2 in mice made iron-deficient with defined diets,12–14 the assays used in those studies were less specific for CYP1A2 than the MROD and UROX activities used here.7, 26
The results of this study have additional implications for understanding the clinical disease PCT. Our results may be relevant to the mechanism of the accepted practice of phlebotomy until the serum ferritin indicates borderline iron deficiency. The question raised is whether our findings might indicate that a more moderate reduction of hepatocyte iron by phlebotomy or L1 therapy might suffice. At this time, this is not known and would require further study. Whether L1 or some other oral iron chelator might have some clinical application in PCT also requires further study, especially in cases of PCT for which phlebotomy cannot be used.
In summary, two methods of producing Fe deficiency have been compared for their effects on the development of experimental uroporphyria. Defined diets, whether iron-deficient or iron-supplemented, protected against uroporphyria. Oral iron chelation with L1 prevented uroporphyria produced in chow-fed mice with only partial depletion of hepatocyte iron. Although supporting the central role of iron, these results implicate dietary iron factors in experimental and clinical uroporphyria and indicate that oral iron chelation may be an effective and relatively simple therapy for PCT.
The authors thank Terry Mattoon and Linda Wilmot for the histological preparation and staining of the samples and Dawn Carbonneau and Debbie Gilfeather for contributions to the animal work. They also thank Dr. George H. Elder for helpful discussions and suggestions.
- 1Porphyria cutanea tarda and related disorders. In: KadishKM, SmithKM, GuilardR, eds. The Porphyrin Handbook II. Vol. 14. Boston, MA: Academic Press; 2003.: 62–97..
- 11Porphyria caused by chlorinated AH receptor ligands and associated mechanisms of liver injury and cancer. In: KadishKM, SmithKM, GuilardR, eds. The Porphyrin Handbook II. Vol. 14. Boston, MA: Academic Press; 2003.: 169–210..
- 23Measurement of heme concentration. In: MainesMD, CostaLG, ReedDJ, SassaS, SipesIG, eds. Current Protocols in Toxicology. Vol. 1. New York, NY: John Wiley & Sons; 2000.: 8.3.1–8.3.7., , .