Introduction

Top Abstract Introduction Materials and methods Results Discussion References

Metallocenes are a class of compounds consisting of a transition metal ion contained complexed to a planar cyclopentadienyl ring structure. Dicyclopentadienyl iron (ferrocene) and its derivatives have been studied in biological systems for almost 30 years (Yeary, 1969). Even though the metallocenes can be thought of as a ferrous ion coordinated to two cyclopentadienyl (C₅H₅) radicals, it is clear that this simple ionic view is not sufficient to explain the stability of ferrocene (Lauber and Hoffmann, 1976). The stability of ferrocene is attributed to the 18- electrons filling the e_2g and a_1g nonbonding molecular orbitals. Ferrocene is aromatic and is more reactive toward electrophiles than benzene (Rausch et al., 1960). Addition of 3,5,5-trimethylhexanoyl (TMH)¹ to ferrocene results in the product TMH-ferrocene, which is less chemically reactive than ferrocene.

Studies using TMH-ferrocene have shown that this particular compound serves as an efficient iron delivery agent in rats in vivo (Düllmann et al., 1992; Braumann et al., 1992, 1994; Nielsen et al., 1993; Nielsen and Heinrich, 1993). Of all the available in vivo experimental methods of iron loading, TMH-ferrocene most closely recapitulates many of the features observed in the human disease hemochromatosis (Halliday and Searle, 1996). TMH-ferrocene can also serve to efficiently deliver iron to long-term primary rat hepatocytes (Cable et al., 1998). Work by Neilsen and Heinrich has shown that ⁵⁹Fe-TMH-ferrocene can be metabolized (Nielsen and Heinrich, 1993) and the released iron can be incorporated into ferritin both in the liver (Düllmann et al., 1992) and in the heart (Braumann et al., 1994). We postulated that in primary rat hepatocytes, iron would not be incorporated properly into ferritin and ferritin induction would not occur unless the iron was removed from the ferrocene nucleus (Cable et al., 1998). Even though the breakdown of other substituted-ferrocene derivatives has been observed in whole animals, as measured by the appearance of free ⁵⁹Fe-labeled iron in urine (Wenzel et al., 1977), the cellular mechanisms that are responsible for the breakdown of TMH-ferrocene have remained undefined.

We hypothesize that TMH-ferrocene is enzymatically broken down and that the enzyme system that is the most likely catalyst for this activity is cytochrome P-450. In this study we have used an instrument that measures iron by constant potential coulometry: the Ferrochem II analyzer. This instrument has a selective response to ionic forms of iron, while having a low response to iron bound covalently in molecules such as TMH-ferrocene or heme. We have utilized this selective response to measure whether ionic iron is released from TMH-ferrocene in the presence of biological samples and to determine how this process occurs.

	Materials and Methods

Top Abstract Introduction Materials and methods Results Discussion References

Materials. Bicinchoninic acid was purchased from Pierce (Rockford, IL). Benzyloxyresorufin, ethoxyresorufin, methoxyresorufin, pentoxyresorufin, and resorufin were purchased from Molecular Probes (Eugene, OR). 3,5,5-Trimethylhexanoylferrocene was a gift from Dr. Peter Nielsen (Universitätkrankenhaus Eppendorf, Hamburg, Germany). Ferrozine [3-(2-pyridyl)-5,6-bis(4-pheylsulfonic acid)-1,2,4-triazine], ethyl acetate, and Silica Gel 60 thin-layer chromatography (TLC) plates were obtained from Aldrich (Milwaukee, WI). Carbon monoxide (CO), scientific grade, was obtained from MG Industries (Malvern, PA). All other chemicals were purchased from Sigma (St. Louis, MO).

Rat Treatment Protocols. Young male Fisher F344 rats, 51 to 76 g, were treated to induce specific varieties of cytochrome P-450 by the following protocols: 20-methylcholanthrene (20-MC), an inducer of cytochrome P-450 1A (Bresnick et al., 1981), 20 mg/kg dissolved in corn oil, i.p. injection every day for 4 days; -naphtholflavone (-NF), an inducer of cytochrome P-450 1A and 2B (Sharma et al., 1979), 40 mg/kg dissolved in corn oil, i.p. injection every day for 4 days; phenobarbital (PB) (sodium salt), an inducer of cytochrome P-450 2B (Blouin et al., 1993), 0.1% (w/v) in water for 6 days; pregnenelone 16-carbonitrile (PCN), an inducer of cytochrome P-450 3A (Wortelboer et al., 1991), 300 mg/kg dissolved in corn oil, i.p. injection every day for 4 days; clofibrate, an inducer of cytochrome P-450 4A; 500 mg/kg dissolved in saline, i.p. injection 18 h; and isoniazid, an inducer of cytochrome P-450 2E, 0.1% (w/v) in water for 8 days. Untreated controls and vehicle controls consisting of rats injected with corn oil for 4 days also were included. The rats were sacrificed and their livers were removed and homogenized using a polytron homogenizer in Tris-buffered sucrose (250 mM sucrose, 50 mM Tris, pH 7.4). Rat liver postmitochondrial supernatants (S9) were prepared by differential centrifugation, 1000g for 10 min, to remove large cellular debris, followed by 10,000g for 10 min to pellet mitochondria. All steps were carried out at 4°C. The S9 supernatants were stored at 80°C. The additional microsomal preparation on the phenobarbital S9 supernatants was carried out by centrifugation at 100,000g for 1 h. The cytosolic fraction was removed and the microsomes were resuspended in phosphate-buffered saline and the mixture was made homogenous by brief sonication.

Measurement of Iron Using the Ferrochem II Analyzer. The Ferrochem II analyzer (ESA Inc., Chelmsford, MA) measures iron content by the use of constant potential coulometry. This instrument utilizes two test electrodes, T1 and T2, a reference silver/silver chloride electrode, and an uncontrolled counter electrode. The potential of T1 is set at +460 mV and that of T2 is set at +325 mV (Ferrochem II Instruction Manual). The oxidation of ferrous ions to ferric ions occurs at T1, and the reduction of ferric iron to ferrous iron occurs at T2. This instrument was designed to measure both total iron and iron-binding capacity in serum. Because the iron is assumed to be in an ionic state or can be placed readily in an ionic state due to the highly acidic reagent utilized in the instrument, the measurement of total iron from biological samples is relatively straightforward. Twenty-five microliters of sample is injected, and the instrument measures the total potential of the injected sample. The instrument is calibrated using a 17.9 µM iron standard supplied with the instrument and a blank containing copper, the only metal ion that could interfere with the iron measurements.

TMH-Ferrocenase (TMHFase) Activity. The activity of TMHFase was calculated from the difference of the iron concentrations from samples incubated in the presence of NADPH minus the iron concentration of the samples incubated in the absence of NADPH (Fig. 1). Unless otherwise noted, the TMH-ferrocene concentration was 200 µM and the incubation time was 5 min.

View larger version (27K): [in this window] [in a new window]   Fig. 1.   Analysis of different iron-containing compounds using the Ferrochem II analyzer. The indicated solutions of iron were tested for their response in the Ferrochem II iron analyzer (ESA). The solutions tested were: copper, the copper standard provided with the instrument to blank a response to copper; Fe std, the 17.9 µM iron standard provided with the instrument; water blank, sterile glass-distilled water produced in the Department of Microbiology and Immunology; water/DMSO blank, water containing 1% DMSO; TMH-ferrocene, a solution of 100 µM TMH-ferrocene diluted into water from a stock solution of 8 mg/ml in DMSO (yielding a final DMSO concentration of 0.04%); FeNTA, a solution of 100 µM ferric chloride complexed to 200 µM nitrilotriacetic acid; heme, a solution of 100 µM heme diluted into water from a 6.5-mg/ml stock solution in DMSO (yielding a final DMSO concentration of 1%). Visual inspection of dilutions of TMH-ferrocene or heme demonstrated that these compounds were stable in aqueous solution throughout the duration of the analysis. The offset white bars shown behind gray bars show the expected response based on the iron concentration in each of the preparations. If no white bar is present, the expected response was zero. The data represent means ± S.E.M., n = 3.

Other Enzyme Assays. Alkoxyresorufin dealkylase assays were performed as described elsewhere (Burke et al., 1985; Cable et al., 1994). Protein was measured using the bicinchoninic acid assay with bovine serum albumin as a standard (Smith et al., 1985).

Analysis of Ionic Iron Using Ferrozine. Iron released in the TMHFase activity assay was analyzed using the ferrous iron chelator, ferrozine (Cowart et al., 1993). Microsomes prepared from the livers of phenobarbital-treated rats were used as a source of TMHFase activity. TMH-ferrocene (200 µM) was incubated for 5 min at 37°C in either the presence or absence of NADPH. The samples were split into two aliquots, a 25-µl and a 50-µl aliquot. The 25-µl aliquot was analyzed on the Ferrochem II analyzer, and the 50-µl aliquot was diluted to 1 ml in the presence of 500 µM ferrozine (Cowart et al., 1993). The mixture was allowed to incubate for 30 min at room temperature in the dark to permit the ferrozine reaction to come to completion. Spectra were taken of both assay mixtures, in the absence and presence of NADPH, versus water, from 500 to 600 nm. The activity of TMHFase was calculated from the difference in the absorbance readings at 562 nm from samples incubated in the presence of NADPH minus the absorbance of samples incubated in the absence of NADPH using an extinction coefficient of 27.9 mM¹ cm¹ (Cowart et al., 1993).

Analysis of TMH-Ferrocene in the TMHFase Reaction. TMH-ferrocene (200 nmol) was incubated in the presence of microsomes isolated from phenobarbital-treated rats for 20 min. Reaction conditions were chosen so that there was excess TMHFase activity, 2-fold the amount needed to digest the amount of TMH-ferrocene in the reaction. TMH-ferrocene was extracted from the reaction using 1 ml of ethyl acetate. The organic phase was evaporated to dryness, and the resulting products were dissolved in 6 µl of ethyl acetate and spotted on a Silica Gel 60 TLC plate. The products were separated using a mobile phase of toluene/ethyl acetate (90:10) (Nielsen and Heinrich, 1993). The TMH-ferrocene standard shown in Fig. 3B was TMH-ferrocene extracted from the reaction mixture without any incubation. This standard had the same R_F as chemically purified TMH-ferrocene (Nielsen and Heinrich, 1993).

Oxygen Consumption and Anaerobiosis. A Clark-type oxygen electrode was used to measure oxygen concentrations. The reaction volume was scaled to fit the 1-ml volume of the oxygen electrode. Baseline oxygen measurements were taken both in the presence of TMHF and NADPH individually and were not different. Aliquots of the reaction mixture were removed for subsequent analysis on the Ferrochem II analyzer. Anaerobiosis was achieved using the TMHFase activity to deplete the reaction mixture of oxygen. An amount of TMHFase, in the form of microsomes isolated from phenobarbital, was chosen that depleted the amount of oxygen in the reaction mixture within 2 min. Aliquots were removed at zero time; at 2 min, when the oxygen was depleted; and then after an additional 3 min. The amount of iron released during each of the phases was calculated. There was sufficient substrate available to remain in excess for the duration of the reaction.

CO Incubation. Microsomes prepared from phenobarbital-treated rat liver were exposed to CO at room temperature in the presence of NADPH and the NADPH-generating system to reduce the heme iron to allow CO to bind (Kamataki and Kitagawa, 1974). The positive TMHFase control was incubated for the same time at room temperature without exposure to CO. The incubation was started by the addition of TMH-ferrocene and incubation at 37°C. The baseline sample was placed immediately on ice and yielded iron concentrations similar to those incubated in the absence of NADPH. The values obtained for the non-CO-treated positive controls were similar to those obtained under the normal reaction conditions, indicating that the incubation at room temperature did not affect TMHFase activity.

Statistical Analysis. Statistical analysis was carried out on a Macintosh computer using JMP 3.02 software (Cary, NC). Analysis of variance and Tukey-Kramer tests were done as appropriate. The null hypothesis was rejected for p  .05. Unless otherwise stated, data are presented as means ± S.E.M., n = 3.

	Results

Top Abstract Introduction Materials and methods Results Discussion References

Iron Measurements Using the Ferrochem II Analyzer. The Ferrochem II analyzes ionic forms of iron in samples. Because TMH-ferrocene contains ferrous iron coordinated to the -bonding molecular orbital electrons from the cytopentadiene rings, where it is not available for oxidation-reduction reactions, we wanted to test whether 100 µM TMH-ferrocene would produce a response in the Ferrochem II. Appropriate solvent control blanks were included. Ferric nitrilotriacetate (FeNTA, 100 µM) was used to provide iron in an ionic form as a positive control in addition to the 17.9 µM iron standard supplied by the manufacturer, ESA (Bedford, MA). Heme (100 µM) was included in the assay so that another molecule containing iron in a covalent bond could be tested for its ability to respond. Only the iron standard provided by ESA and FeNTA produced a response that was of the proper magnitude (Fig. 1). The responses of both 100 µM TMH-ferrocene and 100 µM heme were just above background. This indicates that ionic forms of iron produce a much greater response using the Ferrochem II analyzer than do molecules that contain covalently bound iron. It was not determined whether the small TMH-ferrocene response is a result of an actual response from TMH-ferrocene or possibly from a small amount of contaminating ionic iron contained in the TMH-ferrocene preparation. The differential response between TMH-ferrocene and ionic forms of iron using the Ferrochem II analyzer was utilized to measure the amount of iron released when TMH-ferrocene was incubated with various biological samples.

Iron Is Released from TMH-Ferrocene by Rat Liver S9 Supernatants. Rat liver S9 supernatants were tested for their ability to release iron from TMH-ferrocene. The cofactor mixture contained NADPH and an isocitrate/isocitrate dehydrogenase NADPH generating system, a system used for measuring cytochrome P-450-dependent activities (Burke et al., 1985). The release of iron from TMH-ferrocene occurred only in the presence of NADPH and the amount of iron in the assay mixture increased with time (Fig. 2A). The assay mixture contained detectable levels of iron (approximately 5 µM), about 50% of which was from the rat liver sample; the remaining 50% was from the TMH-ferrocene substrate (data not shown). The amount of iron in the NADPH cofactor mixture and the dimethyl sulfoxide (DMSO) used to dissolve the TMH-ferrocene was not above background (data not shown). The activity catalyzing the release of iron from TMH-ferrocene was termed TMHFase and was calculated from the difference between the amount of iron measured in the absence of NADPH and the amount of iron measured in the presence of NADPH after a 5-min incubation at 37°C (Fig. 2B).

View larger version (13K): [in this window] [in a new window]   Fig. 2.   Temporal characteristics of TMHFase activity in S9 supernatants from untreated male rats. TMHFase activity was measured using the Ferrochem II analyzer as described in Materials and Methods and in the legend of Fig. 1. Protein (100 ng) from control S9 supernatants was used as the source of enzyme, and the TMH-ferrocene concentration in the assay was 200 µM. S9 supernatants, TMH-ferrocene, and specific cofactors were mixed together on ice, and the enzyme activity was initiated by incubating at 37°C in a shaking water bath. The reaction was stopped by placing the samples on ice, and the amount of iron in the reaction mixture was measured immediately using the Ferrochem II analyzer. A, the actual iron concentration in the samples with respect to time. The samples were incubated in the absence [cofactor ()] or in the presence of NADPH and the isocitrate/isocitrate dehydrogenase NADPH-generating system [+cofactor ()]. When NADPH and the NADPH-generating system were present, an increase in the concentration of ionic iron with respect to time was observed. The samples without NADPH and the NADPH-generating system showed no change in ionic iron concentration in the assay sample within 30 min. The initial iron concentration in the assay can be attributed to two components: the background iron measured from the 200 µM TMH-ferrocene and the iron contained within the S9 supernatant. B, change in iron concentration. The cofactor readings were subtracted from the iron concentration observed in the samples incubated in the presence of NADPH. The assay was linear through 5 min.

Induction of TMHFase Activity by Phenobarbital. Several different prototypical cytochrome P-450-dependent compounds were tested for their ability to induce TMHFase activity in male rat livers. The ability of each of these compounds to catalyze the specific dealkylation of a series of alkoxyresorufin substrates first was measured to determine whether the predicted cytochrome P-450 activities were induced (Table 1). As expected, methoxyresorufin-O-demethylase (MROD) and ethoxyresorufin-O-de-ethylase (EROD) were induced by 20-MC and -NF (Burke et al., 1985; Lubet et al., 1985; Blouin et al., 1993). Isoniazid also induced MROD and EROD activities. Pentoxyresorufin-O-depentylase (PROD) and benzyloxyresorufin-O-debenzylase (BROD) activities were maximally induced by PB and to a slightly lesser extent by 20-MC and -NF. PCN, clofibrate, and isoniazid induced BROD activities about 2-fold. TMHFase activity was induced by PB to a greater extent than by any other treatment. PCN and clofibrate also significantly induced TMHFase activity, but not to the level that was observed in the presence of PB. None of the other treatments had any effect on TMHFase activity.

TMHFase Activity in the Microsomal Fraction. Microsomes were prepared from phenobarbital-treated rat liver S9 supernatants to test whether TMHFase activity was cytosolic or microsomal (as other cytochrome P-450-dependent activities). The TMHFase activity increased from 12.6 ± 0.5 nmol Fe release per mg protein/min in the S9 fraction to 43.4 ± 1.1 in the microsomal fraction, a purification of 3.5-fold. No residual TMHFase activity was observed in the cytosol (100,000g supernatant).

Analysis of Product and Substrate in the TMHFase Assay. To further verify the generation of free iron in the TMHFase assay, spectra were taken of both the blank (NADPH) and the incubation (+NADPH) mixture in the presence of 500 µM ferrozine (Fig. 3A). The amount of iron released calculated from the ferrozine data was 19.8 ± 2.2 nmol/ml/min (n = 3) and was comparable to the value of 19.3 ± 1.9 nmol/ml/min (n = 3) obtained using the Ferrochem II analyzer.

View larger version (12K): [in this window] [in a new window]   Fig. 3.   Analysis of product and substrate in the TMHFase assay. Microsomes from phenobarbital-treated rats were prepared as described in Methods. A, spectra of ferrozine-chelatable iron after reaction of TMH-ferrocene with microsomes in the presence of NADPH and the NADPH-generating system (+NADPH) or the absence of NADPH and the NADPH-generating system (NADPH). The spectra shown are versus water. B, analysis of TMH-ferrocene product by TLC. TMH-ferrocene is no longer visible in the presence of NADPH but remains in the absence of NADPH.

Kinetic Parameters of TMHFase Activity in Purified Microsomes. The kinetic parameters of TMHFase activity in purified microsomes was calculated from activity measurements produced at different substrate concentrations. The increasing rate of iron release that occurred with increasing substrate concentrations demonstrates that TMHFase activity follows Michaelis-Menten kinetics (Fig. 4). When an Eadie-Hofstee transformation was carried out on the data, TMHFase had a calculated K_m of 58.5 ± 7.7 µM and a V_max of 57.5 ± 4.1 nmol Fe released/mg protein/min (Fig. 4, inset). The best-fit line for the transformed data was significant at p = .0016. 

View larger version (13K): [in this window] [in a new window]   Fig. 4.   Kinetic parameters of TMHFase activity in purified phenobarbital-induced rat liver microsomes. Microsomes were prepared as described in Materials and Methods. The amount of iron released was measured using microsomes and the indicated TMH-ferrocene concentrations after a 5-min incubation at 37°C in the presence of NADPH as described in the legend to Fig. 1. The TMHFase activity was calculated by subtracting the background readings obtained in the absence of NADPH. Inset, Eadie-Hofstee transformation of the kinetic data, which was used to calculate the Km and Vmax. The data represent means ± S.E.M., n = 3.

Oxygen Consumption in the TMHFase Assay. TMHFase was tested for its ability to consume oxygen during the reaction conditions using a Clark-type electrode. The TMHFase reaction consumed oxygen at a ratio of 1.2 mol of O₂ consumed per mole iron released.

Inhibition of TMHFase Activity by Anaerobiosis. The effect of anaerobiosis on the TMHFase activity also was measured. TMHFase activity in the absence of oxygen was decreased by approximately 77% when compared with the activity measured in the presence of oxygen (Table 2).

Effects of Various Treatments on TMHFase Activity. We tested various treatments and chemicals for their ability to inhibit TMHFase activity in either S9 supernatants or in purified microsomes (Table 2). Boiling the S9 supernatant before assembly of the assay mixture destroyed TMHFase activity. No activity was detected when the complete assay mixture was held on ice instead of being incubated at 37°C. NADH could not substitute for NADPH in either the cytosolic or microsomal preparation (Table 2).

	Discussion

Top Abstract Introduction Materials and methods Results Discussion References

The detection of TMHFase activity was possible because of the development of an assay that utilized constant potential coulometry to measure forms of ionic iron within biological samples. We showed that the Ferrochem II analyzer has a greatly diminished response to iron that was covalently bound in molecules such as TMH-ferrocene and heme. The ability of this instrument to detect noncovalently bound forms of iron was utilized to measure the release of iron from TMH-ferrocene in the presence of rat liver postmitochondrial (S9) supernatants. The ability of the Ferrochem II analyzer to detect released iron was also shown to correlate with the amount of iron detected using ferrozine and with the loss of TMH-ferrocene substrate detected using TLC. The release of iron from TMH-ferrocene occurred in a time-dependent fashion and only in the presence of NADPH. Because there was a residual concentration (5 µM) of iron within the assay to start, contributed by both the rat liver sample and by the background reading from TMH-ferrocene, the amount of iron measured in the absence of NADPH was used as a blank and subtracted from the reading obtained in the presence of NADPH to obtain the activity of TMHFase within the sample. The requirement for a physiologic temperature and the heat-labile nature of the activity are consistent with TMHFase activity being associated with an enzyme.

To test initially whether cytochrome P-450 catalyzed the TMHFase activity, we utilized S9 supernatants from Fisher F344 rats that had been treated with a variety of prototypical cytochrome P-450 inducers. The greatest induction of TMHFase activity was observed in rats treated with phenobarbital. PCN and clofibrate also induced TMHFase activity but to a statistically lesser amount than phenobarbital. There was no increase in TMHFase activity in the presence of cytochrome P-4501A inducers 20-MC or -NF or the cytochrome P-450 2E inducer isoniazid. Because the largest induction of TMHFase activity was observed in the presence of phenobarbital and there was some induction observed in the presence of PCN and clofibrate, we conclude that a cytochrome P-450 2B isozyme is the most likely catalyst of TMHFase activity. However, the lack of effect of -NF on TMHFase activity indicates that TMHFase activity is not catalyzed by cytochrome P-450 2B1/2. The concept that a cytochrome P-450 catalyzes TMHFase activity is supported further by the partitioning of TMHFase to microsomes and inhibition of TMHFase activity by piperonyl butoxide and CO.

The kinetic parameters of TMHFase activity in phenobarbital-treated rat liver microsomes were determined. The enzyme has a high affinity for TMH-ferrocene (K_m = 58.5 µM), which is in the range of other high-affinity cytochrome P-450 substrates. The consumption of oxygen during the TMHFase reaction at a molar ratio of 1.2 mol O₂ per mole iron released is consistent with TMHFase activity being catalyzed by a cytochrome P-450 monooxygenase. The linearity of the Eadie-Hofstee-transformed data is consistent with TMHFase activity being catalyzed by a single enzyme. The combination of the low K_m of the TMHFase activity and the high lipophilicity of TMH-ferrocene suggests that TMH-ferrocene can be broken down readily in vivo.

The data showing that a cytochrome P-450-dependent TMHFase activity exists within rat liver microsomes provide a mechanism that is consistent with previous observations that iron can be released from TMH-ferrocene in various biological systems (Nielsen and Heinrich, 1993; Cable et al., 1998). The studies presented here add to the literature to date in which the metabolic activation of substituted and unsubstituted ferrocenes was studied. Unsubstituted ferrocene has been shown previously to be the substrate for a phenobarbital-inducible cytochrome P-450 hydroxylation (Hanzlik and Soine, 1978). In addition, ferrocene was able to undergo glucuronidation and sulfation, but neither hydroxylation, glucuronidation, nor sulfation removed iron from the ferrocene nucleus (Hanzlik and Soine, 1978). Studies using carboxylic acid derivatives of ferrocene and ruthocene indicated that iron was released from ferrocene but that ruthenium was not released from ruthocene (Wenzel et al., 1977). When the pharmacologic properties of ferrocene, 1,1'-bis(TMH)-ferrocene (Nielsen and Heinrich, 1993) and TMH-ferrocene were compared, it was apparent that most of the unsubstituted ferrocene was cleared in the urine (Nielsen and Heinrich, 1993). The bis(TMH)-ferrocene compound was also poorly metabolized. TMH-ferrocene was metabolized in vivo, probably in the liver, and iron was released from the ferrocene nucleus of TMH-ferrocene. We conclude from previous studies and the data presented in this study that enzyme-catalyzed iron release is more efficient when ferrocene derivatives are used compared to ferrocene.

The use of TMH-ferrocene as an iron donor in biological systems has several advantages over the use of other ionic forms of iron. First, the iron in TMH-ferrocene is not redox active until removed from the cyclopentadiene rings and should cause less oxidative damage to cells when compared with various iron salts such as ferric ammonium citrate or ferrous sulfate. Second, because TMH-ferrocene uptake does not depend on the normal physiologic uptake mechanisms utilized for other forms of iron, TMH-ferrocene can be used to increase the amount of iron added to a system, even when a high amount of iron is already present. Third, previous observations that iron from TMH-ferrocene is incorporated into ferritin (Wenzel et al., 1977; Nielsen and Heinrich, 1993; Cable et al., 1998) demonstrate that iron released from the ferrocene nucleus undergoes normal, physiologic processing and metabolism. These data further support the use of TMH-ferrocene to deliver iron to biological systems.