Induction of Ferritin Synthesis in Human Lung Epithelial Cells Treated with Crocidolite Asbestos

ARCHIVES OF BIOCHEMISTRY AND BIOPHYSICS

Vol. 340, No. 2, April 15, pp. 369–375, 1997 Article No. BB979892

Induction of Ferritin Synthesis in Human Lung Epithelial Cells Treated with Crocidolite Asbestos Ruihua Fang and Ann E. Aust1 Department of Chemistry and Biochemistry, Utah State University, Logan, Utah 84322-0300

Received August 16, 1996, and in revised form December 24, 1996

Crocidolite asbestos is a known human carcinogen containing 27% iron by weight. It has previously been shown that iron was mobilized intracellularly from crocidolite after treatment of human lung epithelial cells (A549) and that the toxicity of the fibers was directly related to how much mobilized iron was in the õ10,000 MW (low-molecular-weight, LMW) fraction [C. C. Chao, L. G. Lund, K. R. Zinn, and A. E. Aust (1994) Arch. Biochem. Biophys. 314, 384–391]. The data here show that iron mobilization from crocidolite began immediately after treatment of the A549 cells and increased linearly with time. However, the synthesis of ferritin, an iron storage protein, did not begin until after 4 h of treatment, reaching a sustained maximum after 12 h. Mobilized iron was preferentially incorporated into the nonferritin-protein fraction up to 7 h after treatment, when the amount of iron mobilized was low and before significant accumulation of newly synthesized ferritin had occurred. This suggested that these cultured cells needed additional iron for synthesis of iron-requiring proteins and that iron mobilized from crocidolite could be utilized directly for this purpose. Subsequent to this, additional mobilized iron was incorporated into newly synthesized ferritin. Even though iron from crocidolite was incorporated into newly synthesized ferritin or into other proteins, the amount of iron from crocidolite in the LMW fraction remained constant during the 24 h. Thus, it appeared that synthesis of ferritin may not have fully protected the cells from the toxic effects of iron mobilized from crocidolite. q 1997 Academic Press Key Words: asbestos; crocidolite; ferritin induction; iron incorporation into ferritin; A549 cells; human lung epithelial cells; low-molecular-weight iron.

Some forms of asbestos, a family of naturally occurring hydrated silicate minerals, can cause pulmo1 To whom correspondence should be addressed. Fax: (801) 7973390. E-mail: [email protected].

nary interstitial fibrosis, mesothelioma of the pleura, pericardium, and peritoneum, and carcinoma of the lungs, esophagus, and stomach (1 – 5). Despite intensive investigation over the past 40 years, the mechanism by which asbestos causes its biological effects is still not known. The most carcinogenic forms of asbestos, the amphiboles, crocidolite and amosite, contain iron as an integral component of the crystalline structure to amounts as high as 27% by weight. Crocidolite can catalyze many of the same reactions that iron can in vitro, such as the formation of the hydroxyl radical (HO•) (6), DNA damage (7, 8), and lipid peroxidation (9, 10). The iron associated with asbestos could be mobilized in vitro by iron chelators, such as citrate and EDTA, and mobilization resulted in increased oxygen consumption (11, 12), HO• formation (12), and DNA damage (13). Iron was also mobilized from crocidolite in cultured human lung epithelial cells (A549)2 and was found to be associated with proteins and a õ10,000 MW, low-molecular-weight (LMW), fraction (14). At the highest exposure level in this study, 6 mg crocidolite/cm2, a total, intracellular concentration of 1.4 mM iron was mobilized in 24 h. The amount of iron in the LMW fraction directly correlated with the cytotoxicity of crocidolite to the cells. As a result of these studies, it was proposed that iron in the LMW fraction may be responsible for some of the pathological effects of crocidolite. Iron has been implicated in many diseases, such as cancer, hemochromatosis, liver cirrhosis, ischemia-reperfusion damage, inflammatory-immune injury, neurological disorders, and atherosclerosis (15). Iron can be toxic due to its ability to catalyze the formation of highly reactive oxygen species, such as the HO•, via the following reactions: 2 Abbreviations used: A549, human lung epithelial cells; LMW, low-molecular-weight; F12–Fe medium, Ham’s F12 complete growth medium without added iron; Fe–NTA, Fe–nitrilotriacetate; [55Fe]crocidolite, neutron-activated crocidolite; iNOS, inducible nitric oxide synthase; FBS, fetal bovine serum; NO, nitric oxide.

369

0003-9861/97 $25.00 Copyright q 1997 by Academic Press All rights of reproduction in any form reserved.

AID

ABB 9892

/

6b31$$$521

03-12-97 07:12:56

arcal

370

FANG AND AUST

Reductant(n) / Fe(III) r Reductant(n/1) / Fe(II) Fe(II) / O2 r Fe(III) / O • 2

•0 2

HO / O

•0 2

/ H r O2 / H2O2 /

Fe(II) / H2O2 r Fe(III) / HO• / HO0 (Fenton reaction). The ability of iron to catalyze these reactions depends upon how iron is complexed. It has been shown that many LMW iron complexes, such as Fe-citrate and Fe– nitrilotriacetate (Fe–NTA), can catalyze the Fenton reaction (16, 17). Normally, iron, bound to proteins and enzymes, does not catalyze these reactions (18). Under physiological conditions, the homeostasis of iron is controlled by transferrin, an iron transport protein, and ferritin, an iron storage protein. The amount of iron in the LMW fraction is negligible (15). However, under some pathological conditions, such as those mentioned earlier, the amount of iron in the LMW fraction increases dramatically and is thought to be responsible for the pathological symptoms (15). Treatment of animals or cells with iron compounds invariably induces ferritin synthesis (19, 20). It has been suggested that ferritin may protect cells from exposure to excess iron. However, little is known about how cells respond to iron mobilized from solids, such as asbestos. The phagocytosis of asbestos by cells in the lung results in bypass of iron uptake by transferrin and may represent an uncontrolled entry of iron into the cell. Therefore, the following studies were designed to investigate the intracellular distribution of iron mobilized from crocidolite and the induction of ferritin synthesis in A549 cells treated with crocidolite. Iron was mobilized immediately from crocidolite, yet ferritin synthesis did not begin until after 4 h of treatment. However, the amount of iron in the LMW fraction remained relatively constant throughout the 24-h exposure period. If the iron mobilized into the LMW fraction was redox active, it may have generated reactive oxygen species that were responsible for cellular damage. MATERIALS AND METHODS Materials. Crocidolite was obtained from Dr. Richard Griesemer, NIEHS/NTP (Research Triangle Park, NC). Standard human liver ferritin was from Calbiochem (San Diego, CA). The antibody to a mixture of human spleen and liver ferritin was obtained from Boehringer-Mannheim Biochemicals (Indianapolis, IN). Agarose-immobilized Protein A beads and ImmunoPure activated peroxidase were from Pierce Chemical Co. (Rockford, IL). [35S]Methionine was from ICN Pharmaceuticals, Inc. (Costa Mesa, CA). Protein molecular weight markers were obtained from Bio-Rad Laboratories (Hercules, CA). Microtiter plates were obtained from Corning Glass Works (Corning, NY). Fetal bovine serum was obtained from Summit Biotechnology (Ft. Collins, CO). Sodium dodecyl sulfate (SDS) was obtained from Aldrich Chemical Co., Inc. (Milwaukee, WI). RPMI 1640 medium, without methionine and cysteine, and Ham’s F12 complete growth medium, without added FeSO4 (F12–Fe medium), were obtained from Life Technologies, Inc. (Gaithersburg, MD). All the other

AID

ABB 9892

/

6b31$$$521

03-12-97 07:12:56

reagents were obtained from Sigma Chemical Co. (St. Louis, MO) or were the same as previously described (14). Cell culture. Human lung epithelial A549 cells were used for all of the experiments. These cells are thought to be of alveolar epithelial type II cell origin, which is one of the two target cell populations for the development of cancer caused by asbestos. The A549 cells were obtained from American Type Culture Collection (Rockville, MD) (ATCC No. CCL185). Conditions for culturing the cells were the same as described previously (14), except that A549 cells were cultured in F12–Fe medium. Neutron activation of crocidolite. Crocidolite was neutron activated at the University of Missouri Research Reactor, as previously described (14), resulting in crocidolite containing 55Fe and 59Fe. In the iron mobilization experiments, radioactivity was determined by liquid scintillation counting. The majority (ú99.6%) of the radioactivity in the crocidolite sample, expressed in disintegrations per minute (dpm), was due to 55Fe (14). Contributions from all other radionuclides were negligible as a result of their low abundance in the crocidolite sample or their low counting efficiency by scintillation counting. The neutron-activated crocidolite will be referred to as [55Fe]crocidolite. Preparation of crocidolite for treatment of cells. As previously described (14), crocidolite was suspended at the appropriate concentration in sterile NaHCO3 (1.176 g/L) before addition to cell cultures. Treatment of cells with crocidolite. Cells were treated, as previously described (14), with the following modifications. Briefly, 6 1 106 cells were plated in a 150-cm2 cell culture flask and incubated in F12–Fe medium containing 10% fetal bovine serum (FBS) for 24 h. The cells were treated in fresh medium containing 2.5% FBS with the indicated doses of crocidolite for ferritin-induction studies or of [55Fe]crocidolite for studies of the intracellular distribution of iron. The treatment doses of crocidolite are expressed as mg crocidolite/ cm2 of cell monolayer because the fibers rapidly settle onto the cells. Under the conditions used here, 1 mg crocidolite/cm2 was equivalent to 5 mg crocidolite/ml of treatment medium. At the indicated time intervals, cells were harvested and stored in 1 ml double-distilled H2O containing 0.1 mM phenylmethylsulfonyl fluoride at 0807C. Immunoprecipitation of ferritin from A549 cell lysate. The cells, frozen after harvest, were lysed, as described previously (14). Crocidolite and cell debris were removed by centrifugation at 10,000g for 30 min. The 10,000g supernatant was separated into a ú10,000 MW fraction and a õ10,000 MW fraction, using an Amicon ultrafilter with 10,000 MW cutoff. Appropriate volumes of 1 M Tris–HCl, pH 7.5, and 2 M NaCl were added to the ú10,000 MW fraction to make the final concentrations of 0.15 M NaCl in 10 mM Tris–HCl, pH 7.5. The antibody to human spleen and liver ferritin (25 mg) was added to the ú10,000 MW fraction to bind ferritin. Agarose-immobilized Protein A beads, 100 ml of a 20% suspension, were used to isolate the ferritin–antibody complex (21). The immunocomplex was removed from the agarose beads and dissolved by boiling for 20 min in 20 ml of the SDS sample buffer. The molecular weight of the proteins present in the solubilized, ferritin–antibody complex was determined by SDS–PAGE, using a 15% polyacrylamide gel (22). Proteins present in the gel were visualized by silver staining (23). Intracellular distribution of iron from crocidolite. The amount of Fe in the 10,000g supernatant from lysed, [55Fe]crocidolite-treated cells was determined by scintillation counting. The iron in this fraction was defined as total iron mobilized from crocidolite. The amounts of 55Fe in the ferritin fraction, ú10,000 MW nonferritin fraction, and õ10,000 MW fraction were also determined and were defined as ferritin associated iron, nonferritin-protein-associated iron, and LMW iron-containing fractions, respectively. The amount of iron adventitiously bound to proteins was determined by incubating the 10,000g supernatant with 1 mM EDTA for 1 h before the filtration. The additional iron present in the filtrate as a result of adding the EDTA was quantified by scintillation counting and was defined as adventitious iron. 55

arcal

FERRITIN INDUCTION IN ASBESTOS-TREATED HUMAN LUNG CELLS Determination of ferritin synthesis. The rate of ferritin synthesis was determined in control untreated cells or cells treated with 3 mg/ cm2 of crocidolite by measuring the incorporation of [35S]methionine into newly synthesized ferritin at the prescribed time intervals after crocidolite treatment. Control and treated cells were incubated in RPMI 1640 medium, without methionine or cysteine, for 30 min, before exposure to [35S]methionine (10 mCi/ml) in fresh RPMI 1640 medium for 2 h. Ferritin was removed by immunocomplex formation and subjected to SDS–PAGE, as described earlier. Proteins present in the gel were visualized by silver staining (23), the portion of the polyacrylamide gel containing ferritin was excised, and the amount of [35S]methionine in the ferritin, solubilized from the gel, was determined using scintillation counting. Total protein synthesis in control cells and treated cells was determined by incorporation of [35S]methionine into acid-precipitable protein in the 10,000g supernatant (21). The results are expressed as the percentage of the amount of ferritin synthesized in the treated cells relative to the control cells. Determination of ferritin concentration. The concentration of ferritin in the 10,000g supernatant was determined using a sandwich ELISA (24). The antibody to human spleen and liver ferritin was used as the capture antibody to coat the microtiter plate. The conjugate of peroxidase and antibody to human spleen and liver ferritin was used as the detector to determine the amount of ferritin bound to the capture antibody. The conjugate was prepared as described by Pierce Chemical Co., except that the molar ratio of antibody to activated peroxidase was 1:5. Tetramethylbenzidine was used as the substrate for the peroxidase. The absorbance of the tetramethylbenzidine at 450 nm was determined using a UVmax kinetic microplate reader (Molecular Devices, Menlo Park, CA). Human liver ferritin was used as the standard. There was a linear correlation between absorbance at 450 nm and the amount of standard human liver ferritin within the range of 0 to 7.5 ng. Total protein in the lysate was

371

FIG. 2. Effect of time of crocidolite exposure on ferritin synthesis in A549 cells. Control cells or cells treated with 3 mg/cm2 crocidolite for the indicated periods of time were pulse-labeled with [35S]methionine in RPMI 1640 medium, without methionine or cysteine, for 2 h. Immunoprecipitation of ferritin and SDS–PAGE of the solubilized ferritin were conducted at constant total protein levels, as described under Materials and Methods. The region of the ferritin band was excised from the gel, and the amount of radioactivity in the ferritin band was determined by scintillation counting. The results are expressed as the percentage of the amount of ferritin synthesized in the treated cells versus the control cells. The data are the average of two experiments. The bars associated with the data points represent the standard errors. The absence of a bar means that the error was smaller than the size of the symbol.

determined using bicinchoninic acid. The results are expressed as ng ferritin/mg total protein.

RESULTS

Effect of Crocidolite on Ferritin Synthesis in A549 Cells

FIG. 1. Induction of ferritin synthesis in A549 cells by crocidolite. Ferritin in control cells or cells treated with 3 mg/cm2 crocidolite for 24 h was immunoprecipitated, using an antibody to human spleen and liver ferritin, as described under Materials and Methods. The ferritin immunocomplex was solubilized and analyzed by SDS– PAGE. The proteins were visualized by silver staining. To the left of the gel, the ‘‘H’’ indicates where the heavy chain of human ferritin is located and the ‘‘L’’ where the light chain of human ferritin is located. The first lane on the left is standard human ferritin (Ferritin Std); the second lane is the solubilized immunocomplex from control cells (ImmPt/Con); the third lane is the solubilized immunocomplex from cells treated with 3 mg/cm2 crocidolite (ImmPt/Asb); the fourth lane is just the antibody, without any cell lysate or ferritin present (IgG); and the fifth lane is the molecular weight standards (MW Stds). The lines to the right of the gel show the molecular weights of the standard proteins.

AID

ABB 9892

/

6b31$$$521

03-12-97 07:12:56

Figure 1 shows the SDS–PAGE analysis of the ferritin–antibody complex. The sample from the cells treated with 3 mg/cm2 crocidolite for 24 h shows two distinct bands at 20,000 and 21,000 MW corresponding to the L- and H-subunits of the ferritin standard, respectively. The L-subunit was predominant in the treated sample. These bands were not observed in untreated cells (Fig. 1) or cells treated with crocidolite for only 4 h (data not shown). Figure 2 shows the time course of ferritin synthesis in cells treated with 3 mg/ cm2 crocidolite. Ferritin synthesis reached a maximum at 10 h, where it was five times greater than control levels. It returned to near control rate at 24 h. Total protein synthesis was the same in control cells and cells treated with crocidolite (data not shown). Effect of Crocidolite on the Amount of Ferritin The amount of ferritin in A549 cells increased linearly with increasing crocidolite concentration (Fig. 3A). The amount of ferritin remained constant in untreated cells for 24 h. However, in cells treated with 3 mg/cm2 crocidolite, the amount of ferritin began to increase after 4 h, reached a maximum after 12 h, and

arcal

372

FANG AND AUST

creased linearly with time up to 24 h. The incorporation of iron into ferritin occurred in two phases. From 1 to 4 h, the incorporation rate of iron into ferritin was 9 pmol Fe/106 cells/h. From 4 to 24 h, the incorporation rate increased dramatically to 58 pmol/106 cells/h. Approximately 69% of the iron mobilized in 24 h was incorporated into ferritin (Fig. 4B). The amount of mobilized iron incorporated into the nonferritin-protein fraction increased linearly with time, but with a slower rate than that for incorporation of iron into ferritin (Fig. 4A). Approximately 28% of the iron mobilized from crocidolite in 24 h was incorporated into the nonferritin-protein fraction (Fig. 4B). The amount of iron mobilized into the LMW fraction remained relatively constant from 1 to 24 h (Fig. 4A) and was approximately 3% of the total iron mobilized from crocidolite in 24 h (Fig. 4B). The mobilization of iron into each of

FIG. 3. Effect of dose of crocidolite or time of crocidolite exposure on the amount of ferritin synthesized in A549 cells. The amount of ferritin in treated or untreated A549 cells was determined using a sandwich ELISA, using human liver ferritin as the standard, as described under Materials and Methods. The results are expressed as means { SD (n Å 3). The absence of an error bar for any given data point means that the standard deviation was smaller than the size of the symbol. (A) A549 cells were treated with the indicated doses of crocidolite for 24 h before the determination of the amount of ferritin present in the cells (m). (B) The amount of ferritin present in A549 cells untreated (j) or treated with 3 mg/cm2 crocidolite for the indicated periods of time (l) is shown.

remained constant up to 24 h (Fig. 3B). We have observed that crocidolite treatment of A549 cells resulted in induction of inducible nitric oxide synthase (iNOS) and synthesis of NO (25). Since NO has been reported to inhibit ferritin synthesis (26), A549 cells were treated with crocidolite and an iNOS inhibitor, aminoguanidine (1 mM), to determine whether NO production affected the amount of ferritin produced after 12 or 24 h of exposure. There was no difference in the total amount of ferritin produced after 12 or 24 h of exposure to crocidolite in the presence of aminoguanidine (data not shown). Intracellular Distribution of Iron in A549 Cells Treated with Crocidolite The time course of mobilization of iron from crocidolite and the distribution of iron into subcellular fractions are shown in Fig. 4A. Total iron mobilized in-

AID

ABB 9892

/

6b31$$$521

03-12-97 07:12:56

FIG. 4. Effect of time of treatment with crocidolite on the intracellular distribution of iron from crocidolite in A549 cells. (A) A549 cells, treated with 3 mg/cm2 [55Fe]crocidolite for the time periods indicated, were harvested and the amount of mobilized iron in the 10,000g supernatant (j), nonferritin-protein fraction (l), ferritin fraction (l), and LMW fraction (h) was determined, as described under Materials and Methods. The inset shows the amount of iron mobilized into the LMW fraction. The results are expressed as means { SD (n Å 3). The absence of an error bar for any given data point means that the standard deviation was smaller than the size of the symbol. (B) The percentage of the total iron mobilized (55Fe in the 10,000g supernatant) into the nonferritin-protein fraction (l), ferritin fraction (l), and LMW fraction (h) is shown. The data for the calculations were taken from (A). The standard deviations were calculated according to the propagation of uncertainty (27).

arcal

FERRITIN INDUCTION IN ASBESTOS-TREATED HUMAN LUNG CELLS

373

tered by ferritin increased to 60% of the total iron mobilized from crocidolite. After 10 h, the proportion of iron in each fraction remained constant. The effect of crocidolite concentration on the percentage of iron in ferritin, nonferritin, or LMW fraction after 24 h is shown in Fig. 5B. When the concentration of crocidolite was varied from 1 to 3 mg/cm2, there was a transition from the majority of mobilized iron being in the nonferritin fraction to being in the ferritin fraction. DISCUSSION

FIG. 5. Effect of the dose of crocidolite on the intracellular distribution of iron from crocidolite in A549 cells. (A) A549 cells were treated with the indicated doses of crocidolite for 24 h. The total amount of 55 Fe mobilized from crocidolite (j) and the amount of 55Fe in the ferritin fraction (l), nonferritin-protein fraction (l), and LMW fraction (h) were determined, as described under Materials and Methods. The results are expressed as means { SD (n Å 3). The absence of an error bar for any given data point means that the standard deviation was smaller than the size of the symbol. (B) The percentage of the total 55Fe mobilized into each of the following fractions: ferritin (l), nonferritin protein (l), and LMW (h), was determined and plotted versus the treatment dose of crocidolite. The data for the calculations were taken from (A). The standard deviations were calculated according to the propagation of uncertainty (27).

the fractions up to 24 h increased linearly with increasing crocidolite concentration (Fig. 5A). As shown previously (14), the 55Fe present in the 10,000g supernatant was neither from residual [55Fe]crocidolite nor due to the release of 55Fe from [55Fe]crocidolite during the cell lysing process. The amount of iron adventitiously bound to proteins during the isolation process was negligible (data not shown). The percentage of iron incorporated into each subcellular fraction versus time was determined (Fig. 4B). From 1 to 7 h there was a transition in the compartmentalization of the mobilized iron from approximately equal percentages of iron in the LMW and nonferritin fractions and a very low percentage in ferritin to equal percentages of iron in ferritin and nonferritin fractions and a very low percentage in the LMW fraction. From 7 to 10 h, the percentage of iron in the LMW fraction remained constant, while the amount of iron seques-

AID

ABB 9892

/

6b31$$$521

03-12-97 07:12:56

The data presented here show that treatment of A549 cells with crocidolite, a particulate containing iron, resulted in the induction of ferritin synthesis. The rate of ferritin synthesis increased up to 10 h after treatment, which was the approximate time of maximum accumulation of ferritin protein in the cells. The decrease in ferritin synthesis after that time was not reflected in a decrease in ferritin protein in the cell. This suggested that the newly synthesized protein was not being degraded. There was a direct correlation between the amount of ferritin in the cells after 24 h of exposure and the treatment concentration of crocidolite. The amounts of ferritin induced by crocidolite in these experiments is within the range of ferritin levels observed in other cell types treated with soluble sources of iron. After exposure of mouse lymphocytes to Fe–NTA, 0.1 ng ferritin/mg protein was induced (28), while exposure of rat hepatoma cells to ferric ammonium citrate resulted in 25 ng ferritin/mg protein (29). Ferritin at levels of 1.5 ng/mg protein has been observed in rat lung tissue after treatment of the rats with methemoglobin (30). It appeared that L-subunit-dominant ferritin was induced in A549 cells treated with crocidolite for 24 h. Although the antibody used for these experiments resulted from treatment with human liver and spleen ferritin, both L-subunit-rich forms of ferritin, this antibody has been used successfully to complex H-subunitrich ferritin (31). Other investigators have reported that L-subunit-dominant ferritin was synthesized in response to iron treatment of cells (32). Since L-subunit-dominant ferritin can incorporate more iron than H-subunit-dominant ferritin (33), L-subunit-dominant ferritin may be synthesized after an iron challenge to more efficiently deal with the iron present in the cells. Therefore, it is not surprising that the challenge resulting from iron mobilization from crocidolite led to the induction of L-dominant ferritin. The phagocytosis of crocidolite (14) and mobilization of iron began immediately upon treatment of the cells. Significant ferritin synthesis began after 4 h of treatment with crocidolite, whereas ferritin synthesis started within 3 h after treatment with ferric ammonium citrate, a soluble iron form (data not shown). The

arcal

374

FANG AND AUST

delay in ferritin synthesis in the cells treated with crocidolite did not appear to be due to an inadequate amount of iron being mobilized from crocidolite during this period. In fact, the intracellular concentration of iron mobilized from crocidolite at 1 h was 44 mM. The concentration of iron mobilized into the LMW or nonferritin-protein fractions at 1 h was approximately 14 or 25 mM, respectively. While it is not known what form of iron is responsible for induction of ferritin synthesis, the total concentration of iron mobilized, as well as any of the subcellular fractions of iron determined, was higher than that of Fe–NTA (1 mM) reported to induce ferritin synthesis in rat skeletal myoblasts (20). Even though it is not known what form of iron is responsible for induction of ferritin, there is a direct correlation between the amount of ferritin in the cells after 24 h of treatment with crocidolite and the amount of iron in the LMW fraction. We have observed that the treatment of A549 cells with crocidolite resulted in induction of iNOS mRNA and synthesis of NO (25). It has been reported that NO can inhibit the synthesis of ferritin by removing Fe from the Fe–S cluster of the iron-responsive element binding protein (IREBP) (26). The presence of NO in the A549 cells does not appear to explain the lag in ferritin synthesis that we have observed, since iNOS mRNA did not appear until 4–6 h after treatment with crocidolite (unpublished observation), about the same time that ferritin synthesis began. The A549 cells appeared to utilize the iron from crocidolite directly for incorporation into iron-requiring proteins. When the amount of iron mobilized from crocidolite was low, as in cells treated with a low concentration of crocidolite (1 mg/cm2) or at early time points in cells treated with 3 mg/cm2, the majority of the iron was mobilized into the nonferritin-protein fraction and was likely bound to iron-requiring proteins. This suggests that the cells may have needed additional iron above the levels that they were able to acquire from the complete medium containing serum. When more iron was mobilized from crocidolite, the cells may have found the concentrations of iron to be beyond their metabolic needs, and the majority of the iron was incorporated into ferritin. The relative distribution among the various intracellular fractions remained constant after that time. Thus, it may be that the apparent lag in synthesis of ferritin was a reflection of the fact that the cells needed to use the iron for nonferritin-protein synthesis. We have previously reported that the toxicity of crocidolite correlated directly with the amount of iron mobilized into the intracellular LMW fraction (14). The amount of iron in the LMW fraction remained relatively constant from 1 to 24 h after addition of crocidolite to the treatment medium. Surprisingly, it did not decrease after synthesis of ferritin. The iron in the

AID

ABB 9892

/

6b31$$$521

03-12-97 07:12:56

LMW fraction was not an artifact of isolation and lysis of the cells, since the amount of iron in the LMW fraction remained constant, while the amount of crocidolite associated with the A549 cells increased during the first 9 h of treatment. It was likely that the iron mobilized from the fibers into the LMW fraction was in a redox-active form and was capable of catalyzing damage to the cells. Since iron was mobilized continuously over the 24 h of observation, this would mean that there was a relatively constant level of iron available for damage to the cells during this period of time. The chelator(s) responsible for mobilization of iron from crocidolite has not yet been identified, but it appeared to allow incorporation of iron into nonferritin proteins and ferritin. In conclusion, the phagocytosis of crocidolite and subsequent mobilization of iron may represent an iron challenge with which the cells are not prepared to cope. The results presented here show that crocidolite treatment of human lung epithelial cells in culture resulted in iron mobilization and ferritin induction. Ferritin synthesis may have offered some protection from the iron that was mobilized. However, since the toxicity of crocidolite was directly proportional to the amount of iron mobilized into the LMW fraction (14), it would appear that the mobilized iron was redox active and capable of causing damage to the cells. If this happens in vivo, this could result in DNA and other cellular damage leading to cancer. ACKNOWLEDGMENTS We thank Sun-Hee Park and Dr. Steven D. Aust for their helpful suggestions and Dr. Kurt Zinn of the University of Missouri Reactor Facility for the neutron activation of our crocidolite samples. This work was supported by a grant from the National Institute for Environmental Health Sciences (ES05782).

REFERENCES 1. Mossman, B. T. (1988) Environ. Carcinog. Rev. C4, 151–195. 2. Khan, G., Mahomood, N., Arif, J. M., and Rahman, Q. J. (1992) Sci. Ind. Res. 51, 507–514. 3. Seidman, H., Selikoff, I. J., and Hammond, E. C. (1979) Ann. N.Y. Acad. Sci. 330, 61–89. 4. Warnock, M. L., and Churg, A. M. (1975) Cancer 35, 1236–1242. 5. Hammond, E. C. (1966) Natl. Cancer Inst. Monogr. 19, 127–204. 6. Weitzman, S. A., and Graceffa, P. (1984) Arch. Biochem. Biophys. 228, 373–376. 7. Nejjari, A., Fournier, J., Pezerat, H., and Leanderson, P. (1993) Br. J. Ind. Med. 50, 501–504. 8. Libbus, B. L., Illenye, A. A., and Craighead, J. E. (1989) Cancer Res. 49, 5713–5718. 9. Jajte, J., Lao, I., and Wiseniewska-Knypl, J. M. (1988) Br. J. Ind. Med. 44, 180–186. 10. Kamal, A. A. M., Goma, A., Khafif, M. E., and Hammad, A. S. (1989) Environ. Res. 49, 173–180. 11. Lund, L. G., and Aust, A. E. (1991) Arch. Biochem. Biophys. 287, 91–96. 12. Aust, A. E., and Lund, L. G. (1991) in Mechanisms in Fiber Car-

arcal

FERRITIN INDUCTION IN ASBESTOS-TREATED HUMAN LUNG CELLS

13. 14. 15.

16. 17. 18. 19. 20. 21.

22. 23.

cinogenesis (Brown, R. C., Hoskins, J. A., and Johnson, N. F., Eds.), pp. 397–405, Plenum Press, New York. Lund, L. G., and Aust, A. E. (1992) Carcinogenesis 13, 4417– 4422. Chao, C. C., Lund, L. G., Zinn, K. R., and Aust, A. E. (1994) Arch. Biochem. Biophys. 314, 384–391. Haberland, M. E., and Smith, C. V. (1992) in Free Radical Mechanism of Tissue Injury (Moslen, M. T., and Smith, C. V., Eds.), pp. 45–63, CRC Press, Boca Raton, FL. Lund, L. G., and Aust, A. E. (1992) Carcinogenesis 13, 637–642. Kawabata, T., Awai, M., and Kohno, M. (1986) Acta Med. Okayama 40, 163–173. Miller, D. M., Buettner, G. R., and Aust, S. D. (1990) Free Radical Biol. Med. 8, 95–108. Drysdale, J. W., and Munro, H. N. (1966) J. Biol. Chem. 241, 3630–3637. Rittling, S. R., and Woodworth, R. C. (1983) Arch. Biochem. Biophys. 225, 390–397. Harlow, E., and Lane, D. (1988) in Antibodies: A Laboratory Manual, pp. 465–468, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. Laemmli, U. K. (1970) Nature 227, 680–685. Oakley, B. R., Kirsch, D. R., and Morris, N. R. (1980) Anal. Biochem. 105, 361–363.

AID

ABB 9892

/

6b31$$$521

03-12-97 07:12:56

375

24. Hornbeck, P. (1994) in Current Protocols in Immunology (Coico, R., Coligan, J. E., Kruisbeek, A. M., Margulies, D. H., Shevach, E. M., and Strober, W., Eds.), Vol. 1, pp. 2.1.9–2.1.11, Current Protocols, USA. 25. Chao, C.-C., Park, S.-H., and Aust, A. E. (1996) Arch. Biochem. Biophys. 326, 152–157. 26. Pantopoulos, K., and Hentze, M. W. (1995) Proc. Natl. Acad. Sci. USA 92, 1267–1271. 27. Harris, D. E. (1991) Quantitative Chemical Analysis, pp. 41–43, Freeman, New York. 28. Djeha, A., and Brock, J. H. (1992) Biochim. Biophys. Acta 1133, 147–152. 29. Goto, Y., Peterson, M., and Listowsky, I. (1983) J. Biol. Chem. 258, 5248–5255. 30. Balla, J., Nath, K. A., Balla, G., Juckett, M. B., Jacob, H. S., and Vercellotti, G. M. (1995) Am. J. Physiol. 268, L321–L327. 31. Eisenstein, R. S., Garcia-Mayol, D., Pettingell, W., and Munro, H. N. (1991) Proc. Natl. Acad. Sci. USA 88, 688–692. 32. Crichton, R. R. (1991) in Inorganic Biochemistry of Iron Metabolism, pp. 154–159, Ellis Horwood Limited, Chichester. 33. Harrison, P. M., Artymiuk, P. J., Ford, G. C., Lawson, D. M., Smith, J. M. A., Treffry, A., and White, J. L. (1989) in Biomineralisation (Mann, S., Webb, J., and Williams, R. J. P., Eds.), pp. 257–294, VCH, Weinheim.

arcal