Báo cáo Y học: The mechanism of nitrogen monoxide (NO)-mediated iron mobilization from cells NO intercepts iron before incorporation into ferritin and indirectly mobilizes iron from ferritin in a glutathione-dependent manner pot

The mechanism of nitrogen monoxide NO-mediated iron mobilization from cells NO intercepts iron before incorporation into ferritin and indirectly mobilizes iron from ferritin in a glutath

The mechanism of nitrogen monoxide (NO)-mediated iron mobilization from cells

NO intercepts iron before incorporation into ferritin and indirectly mobilizes iron from ferritin in a glutathione-dependent manner

Ralph N Watts and Des R Richardson

The Iron Metabolism and Chelation Group, The Heart Research Institute, Camperdown, Sydney, New South Wales, Australia

Nitrogen monoxide (NO) is a cytotoxic effector molecule

produced by macrophages that results in Fe mobilization

from tumour target cells which inhibits DNA synthesis and

mitochondrial respiration It is well known that NO has a

high affinity for Fe, and we showed that NO-mediated Fe

mobilization is markedly potentiated by glutathione (GSH)

generated by the hexose monophosphate shunt [Watts, R.N

& Richardson, D.R (2001) J.Biol.Chem 276, 4724–4732]

We hypothesized that GSH completes the coordination shell

of an NO–Fe complex that is released from the cell In this

report we have extended our studies to further characterize

the mechanism of NO-mediated Fe mobilization Native

PAGE 59Fe-autoradiography shows that NO decreased

ferritin-59Fe levels in cells prelabelled with [59Fe]transferrin

In prelabelled cells, ferritin-59Fe levels increased 3.5)fold

when cells were reincubated with control media between 30

and 240 min In contrast, when cells were reincubated with

NO, ferritin-59Fe levels decreased 10-fold compared with control cells after a 240-min reincubation However, NO could not remove Fe from ferritin in cell lysates Our data suggest that NO intercepts 59Fe on route to ferritin, and indirectly facilitates removal of 59Fe from the protein Studies using the GSH-depleting agent,L -buthionine-(S,R)-sulphoximine, indicated that the reduction in ferritin-59Fe levels via NO was GSH-dependent Competition experi-ments with NO and permeable chelators demonstrated that both bind a similar Fe pool We suggest that NO requires cellular metabolism in order to effect Fe mobilization and this does not occur via passive diffusion down a concentra-tion gradient Based on our results, we propose a model of glucose-dependent NO-mediated Fe mobilization

Keywords: chelators; ferritin; glutathione; iron; nitrogen monoxide

Many of the diverse biological effects of nitrogen monoxide

(NO) are mediated through its binding to iron (Fe) in the

haem prosthetic group of soluble guanylate cyclase [1–3]

Indeed, the high affinity of NO for Fe is a well-characterized

branch of coordination chemistry [2] Apart from the

regulatory role of NO, its cytotoxic actions are found when

it is produced in large quantities by cells such as activated

macrophages [3] Interestingly, NO produced by such

systems inhibits the proliferation of intracellular pathogens

and tumour cells [3–5] These effects can be explained by the

reactivity of NO with Fe in the [Fe–S] centres of critical

proteins, including aconitase and complex Iand IIof the electron transport chain [4–6] The high affinity of NO for

Fe probably results in both the removal of Fe from [Fe–S] centres and the formation of dinitrosyl Fe species within [Fe–S] proteins (reviewed in [7])

It has already been shown that NO forms complexes with

a range of Fe-containing proteins including ferritin [8], ribonucleotide reductase [9], haem-containing proteins [10–12], and ferrochelatase [13] Further, it has been suggested that ferritin can act as a store of NO [8], and NO-mediated Fe release from isolated and purified ferritin has been demonstrated [14] When activated macrophages are cocultured with tumour cells, this inhibits target cell DNA synthesis and results in the release of 64% of cellular

59Fe within 24 h [15] This loss of Fe may be due to the NO-mediated release of Fe from enzymes such as mito-chondrial aconitase [4,16–18] Others have suggested that

NO can also target loosely bound pools of nonhaem Fe [19] Nonetheless, the identification of Fe–nitrosyl complexes (Fe–dithiol dinitrosyl complexes and haem–nitrosyl com-plexes) by EPR spectroscopy in activated macrophages and their tumour cell targets show the importance of the Fe–NO interaction [17–23]

Apart from the above effects, NO can also increase the RNA-binding of iron-regulatory protein 1 (IRP1), that plays

an important role in regulating intracellular Fe homeo-stasis (reviewed in [3,24]) The effect of NO on IRP1-RNA binding activity occurs via two main mechanisms, a direct effect on the [4Fe)4S] cluster and Fe mobilization from

145 Missenden Road, Camperdown, Sydney, New South Wales,

2050 Australia.

Fax: + 61 2 9550 3302, Tel.: + 61 2 9550 3560,

E-mail: d.richardson@hri.org.au

balanced salt solution; DFO, desferrioxamine; GSH, reduced

glutathione; GSNO, S-nitrosoglutathione; HMPS, hexose

monophosphate shunt; IRP1, iron-regulatory protein 1; MEM,

minimum essential medium; NAP, N-acetylpenicillamine; PIH,

pyridoxal isonicotinoyl hydrazone; SNAP,

S-nitroso-N-acetylpenic-illamine; Sper, spermine; SperNO, Spermine-NONOate; Tf,

transferrin; 311, 2-hydroxy-1-naphthylaldehyde isonicotinoyl

hydrazone; DTPA, diethylenetriaminepentaacetic acid.

(Received 14 February 2002, revised 29 April 2002,

accepted 6 May 2002)

cells [25–29] Our previous studies have shown that a range

of NO-generators [e.g S-nitroso-N-acetylpenicillamine

(SNAP), S-nitrosoglutathione (GSNO), and spermine

NONOate (SperNO)], could mobilize59Fe from prelabelled

cells as, or more, effectively than the clinically used Fe

chelator desferrioxamine (DFO) [29] In contrast, the

precursor compounds of these latter NO-generators,

namely N-acetylpenicillamine (NAP), glutathione (GSH),

and spermine (Sper), respectively, had no effect [29]

Previous studies have suggested that NO may be released

from cells as a complex composed of NO, Fe, and

thiol-containing ligands such as cysteine or GSH [23,30,31]

Considering this and the other data described above, we

recently examined the energy-dependency of NO-mediated

Fe release from cells [32] Our investigation showed that

metabolism ofD-glucose potentiates NO-mediated Fe efflux

from a variety of cell types Further, we demonstrated that

the metabolism ofD-glucose by the hexose monophosphate

shunt (HMPS) and the maintenance of GSH levels was

essential for NO-mediated Fe mobilization [32] However,

we are not proposing a direct coupling between glucose

import/metabolism and NO metabolism Rather, our

ex-periments suggested that the generation of GSH after

incubation withD-glucose could result in GSH acting as a

ligand which together with NO would complete the

coordi-nation shell of Fe [32] Such a
mixed Fe complex with both

NO and GSH ligands bound to Fe may provide an

appro-priate lipophilic balance to allow diffusion through the

membrane and/or transport by a carrier In fact, we showed

that NO-mediated59Fe release was both temperature- and

energy-dependent, suggesting a membrane transport

mech-anism could be involved [32] However, the intracellular site

of NO-mediated Fe release was not established

In this investigation we have extended our knowledge of

NO-mediated Fe mobilization For the first time, we

demonstrate using a cellular system that NO intercepts Fe

before being incorporated into ferritin in a similar manner to

Fe chelators Further, NO facilitates removal of59Fe from

ferritin probably by an indirect mechanism This process of

depleting ferritin-bound 59Fe was dependent on cellular

GSH Our studies also indicate that cellular metabolism was

required for NO-mediated Fe mobilization which appears to

be an active rather than a passive process These results may

be important in understanding the cytotoxic actions of NO

produced by activated macrophages

E X P E R I M E N T A L P R O C E D U R E S

Cell treatments and reagents

The NO-generator SNAP was synthesized by established

techniques [33] from the precursor compound NAP (Sigma

Chemical Co.) Apotransferrin (apoTf),L

-buthionine-(S,R)-sulphoximine (BSO), diethylenetriaminepentaacetic acid

(DTPA), E`DTA, GSH, GSNO, horse spleen ferritin and

Sper were obtained from Sigma SperNO was obtained from

Cayman Chemicals Eagle’s minimum essential medium

(MEM) was obtained from Gibco BRL DFO was obtained

from Novartis Pharmaceutical Co Pyridoxal isonicotinoyl

hydrazone (PIH) and its analogue,

2-hydroxy-1-naphthylal-dehyde isonicotinoyl hydrazone (311), were synthesized by

standard techniques [34] Both PIH and 311 are strong

Fe-binding ligands [34] and were used as positive Fe chelation

controls Apolactoferrin was from Calbiochem A polyclonal rabbit anti-(human ferritin) Ig was from Roche Diagnostics All other chemicals were of analytical reagent quality The NO-generators and other compounds were dissolved in media immediately prior to an experiment [29,35]

Cell culture Human SK-N-MC neuroepithelioma cells, SK-Mel-28 melanoma cells, and MCF-7 breast cancer cells were from the American Type Culture Collection The mouse LMTK– fibroblast cell line was from the European Collection of Cell Cultures The BE-2 neuroblastoma cell line was a gift from

G Anderson, Queensland Institute of Medical Research (Brisbane, Australia) All cell lines were grown in MEM containing 10% foetal calf serum (Gibco), 1% (v/v) non-essential amino acids (Gibco), 100 lgÆmL)1streptomycin (Gibco), 100 UÆmL)1penicillin (Gibco), and 0.28 lgÆmL)1 fungizone (Squibb Pharmaceuticals) Cells were grown in an incubator (Forma Scientific) at 37°C in a humidified atmosphere of 5% CO2/95% air and subcultured as described previously [36] Cellular growth and viability were assessed by phase contrast microscopy, cell adherence to the culture substratum, and Trypan blue staining

Nitrite determination The accumulation of nitrite in cell culture supernatants is commonly used as a relative measure of NO production [25,29,35] Nitrite was assayed using the Griess reagent that gives a characteristic spectral peak at 550 nm [37] Protein preparation and labelling

Apotransferrin was labelled with59Fe (Dupont NEN) or

56Fe to produce Fe2-Tf using established procedures [36]

Efflux assay of59Fe from prelabelled cells Standard techniques were used to examine the effect of NO and other agents on the efflux of59Fe from prelabelled cells [29,32,34] Briefly, cells were labelled with59Fe-Tf (0.75 lM) for 3 h at 37°C in MEM After this incubation, the cell culture dishes were placed on a tray of ice, the medium aspirated, and the cell monolayer washed four times with ice-cold balanced salt solution (BSS) The cells were then reincubated for various incubation times up to 240 min at

37°C After this incubation, the overlying supernatant (efflux medium) was transferred to c-counting tubes The cells were removed from the petri dishes after adding 1 mL BSS and by using a plastic spatula to detach them Radioactivity was measured in both the cell pellet and supernatant using a c-scintillation counter (LKB Wallace

1282 Compugamma, Finland)

Determination of intracellular iron distribution using native-PAGE-59Fe-autoradiography Native-PAGE-59Fe-autoradiography was performed using standard techniques in our laboratory [38] Bands on X-ray film were quantified by scanning densitometry using a Laser Densitometer and analysed byBIOMAX Isoftware (Kodak Ltd)

Glutathione assay

GSH was measured as described previously [39] Cellular

GSH levels were reduced using the GSH synthesis inhibitor,

BSO (0.1 mM) This latter agent is a potent and selective

inhibitor of the enzyme c-glutamylcysteine synthetase that is

involved in GSH synthesis [40] A 20-h incubation with BSO

at a concentration of 0.1 mMwas used, as these conditions

were shown in our previous studies to markedly deplete

GSH levels without affecting cellular viability [32]

Assay for examining the ability of NO or iron chelators

to bind59Fe from cell lysates

Cells grown to near confluence in T75 culture flasks were

labelled with59Fe-Tf (0.75 lM) for 3 h at 37°C, placed on a

tray of ice, the medium decanted and the cell monolayer

washed six times with ice-cold BSS The cells were lysed by

one freeze-thaw cycle and then detached from the flask

using a Teflon spatula in the presence of the nonionic

detergent Triton X-100 (1.5%) at 4°C These samples were

then centrifuged at 21 300 g for 30 min at 4°C and the

cytosol removed and assessed for radioactivity using the

c-counter described above The cytosolic samples were then

incubated for 3 h at 37°C with DFO (0.5 mM) or GSNO

(0.5 mM) The generation of nitrite by GSNO was used as a

control to ensure that the NO-donor was producing NO in

the lysate After this incubation, the samples were then subjected to centrifugation at 4°C through a 5-kDa Mr exclusion filter (Vivaspin 500, Sartorius AG) After centri-fugation, the eluent, eluate, and membrane were taken to estimate 59Fe levels Examination of 59Fe levels on the membrane were considered important to assess the possi-bility of adsorption of the59Fe-complex

Statistics Experimental data were compared using Student’s t-test Results were considered statistically significant when

P< 0.05

R E S U L T S The effects of NO on intracellular iron distribution:

NO decreases ferritin-59Fe levels Considering our previous studies demonstrating that incu-bation with NO results in intracellular Fe mobilization [29,32], it was important to determine the source of the

Fe mobilized For these experiments we used native PAGE-59Fe-autoradiography that has proved useful in examining the intracellular distribution of 59Fe in our previous studies [38,41] Cells were prelabelled with59Fe-Tf for 3 h at 37°C, washed on ice, and then reincubated for up

0 10 20 30 40 50 60

Control GSNO

SNAP NAP

Sper DFO 311

Control GSNO

0 25 50 75 100 125 150

Control GSNO

Ferritin- 59 Fe Low Mr59 Fe

Supershifted Band

densitometric results of the autoradiograph The results in (A) are the mean ± SD of three replicates in a typical experiment of three performed The data shown in (B) are a representative experiment of three performed.

to 240 min at 37°C in the presence or absence of the agents

to be tested The cells were then lysed and subjected to

native PAGE-59Fe autoradiography

Cells prelabelled with59Fe were incubated with a variety

of NO-generating agents, including GSNO, SNAP and

SperNO at a concentration of 0.5 mM(Fig 1A) The effects

of these NO donors were compared to their respective

control compounds without the NO group, namely GSH,

NAP, and Sper We also compared NO-mediated 59Fe

mobilization to the efficacy of the well-characterized Fe

chelators, DFO and 311 [34,41] Each of the NO-generators

resulted in the release of 18–24% of total cellular 59Fe,

whereas the relevant control compounds were no more

effective than media alone which released 3 ± 1% of59Fe

(Fig 1A) The NO-generators were more effective at

mobilizing cellular 59Fe than DFO, but less active than

the potent Fe chelator 311 [34,41] (Fig 1A)

Examining intracellular59Fe distribution in SK-N-MC

cells (Fig 1B), the most pronounced band identified

comi-grated with purified horse spleen ferritin (data not shown)

Experiments incubating the lysate with an anti-ferritin

antibody demonstrated that only this band could be

super-shifted (Fig 1B), again indicating that it was ferritin As

described previously in neoplastic cells [41], a faint and very

diffuse band below ferritin was present which comigrated

with low Mr Fe complexes (59Fe-citrate) (Fig 1B) We

previously showed that this low MrFe appeared to be59Fe

bound from the lysate by the low Mrchelators in the gel

running buffer e.g Tris [41] In the present study, the low Mr

band will not be considered in detail as this component

remains undefined and its relevance uncertain

In each case, GSNO, SNAP and SperNO, decreased

ferritin-59Fe levels to 55–63% of the control while their

respective control compounds (GSH, NAP, and Sper) had

either little effect or increased ferritin-59Fe levels (Fig 1B)

The effect of GSH or NAP at increasing ferritin-59Fe was not a consistent finding in repeat experiments In addition, the chelators DFO and 311 decreased ferritin-59Fe levels (Fig 1B) to 69% and 51% of the control, respectively (Fig 1B) As all NO generators had a similar effect, subsequent studies examining the effect of NO on cellular

Fe metabolism were performed using GSNO because of its potential physiological importance

The effect of GSNO concentration and reincubation time

on ferritin-59Fe levels: NO intercepts59Fe before it reaches ferritin

To determine the efficacy of NO on59Fe mobilization, the effect of GSNO concentration (0.01–1 mM) on59Fe release from prelabelled cells (Fig 2A) and ferritin-59Fe levels (Fig 2B) was examined These experiments showed that GSNO appreciably increased59Fe mobilization from pre-labelled cells at a GSNO concentration of 0.05 mM, and then plateaued at 0.5 mM (Fig 2A) When assessing the effect of NO on intracellular 59Fe distribution, a GSNO concentration of 0.05 mM decreased ferritin-59Fe levels to 26% of the control value (Fig 2B) Higher concentrations

of the NO-donor were no more effective at reducing ferritin-59Fe levels (Fig 2B)

Studies were performed to determine the effect of reincubation time in the presence and absence of GSNO

on 59Fe mobilization (Fig 3A) and ferritin-59Fe levels (Fig 3B) As in the studies above, cells were prelabelled with

59Fe-Tf for 3 h at 37°C, washed on ice, and then reincubated in the presence of control media or GSNO (0.5 mM) for 30–240 min at 37°C In the control samples

< 3% of total cellular59Fe was released, while in GSNO-treated cells59Fe mobilization increased linearly from 30 to

180 min and then plateaued at 240 min (Fig 3A)

GSNO Concentration (mM)

0 3 6 9 12 15 18

GSNO Concentration (mM)

0 20 40 60 80 100 120

59 Fe-Ferritin Levels

Control 0.01 m

GSNO Concentration

cellular cytosols treated as described above and densitometric results of the autoradiograph The results in (A) are the mean ± SD of three replicates in a typical experiment of four performed The data shown in (B) are from a representative experiment of three performed.

Examining the intracellular distribution of59Fe in control

cells, ferritin-59Fe increased as a function of reincubation

time (Fig 3B) In fact, after a reincubation time of 240 min

in control media, ferritin-59Fe levels were 3.5-fold that

found after 30 min (Fig 3B) These results in control cells

demonstrated that there was some redistribution of 59Fe

between compartments, with 59Fe incorporation into

ferritin gradually increasing as a function of reincubation

time (Fig 3B) In contrast, in cells treated with GSNO,

ferritin-59Fe levels decreased after 90 min of reincubation

In fact, after 240 min, ferritin-59Fe levels were 10-fold less

than those of control cells at the same reincubation time

(Fig 3B) These data suggest that NO directly or indirectly

results in59Fe mobilization from ferritin and also intercepts

59Fe before it is deposited within this protein

NO reduces ferritin-59Fe levels in a variety of cell types

The effect of NO at reducing ferritin-59Fe levels was found

in a number of cell types including SK-N-MC

neuroepi-thelioma cells, MCF-7 breast cancer cells, LMTK–

fibro-blasts, BE-2 neuroblastoma cells and SK-Mel-28 melanoma

cells However, there was marked variation in the effect of

NO between cell types, with a 15–74% decrease in

ferritin-59Fe being observed (data not shown)

Depletion of intracellular GSH prevents the NO-mediated

decrease in ferritin-59Fe levels

Our previous studies showed that NO-mediated59Fe efflux

was GSH-dependent [32] Considering this, we examined

the effect of a 20-h incubation of both LMTK–and

SK-N-MC cells with the specific GSH synthesis inhibitor, BSO

(0.1 m ) [40], on the ability of GSNO to increase 59Fe

mobilization (Fig 4A) and reduce ferritin-59Fe levels (Fig 4B) As shown previously, preincubation with BSO markedly decreased NO-mediated 59Fe mobilization from both cell types (Fig 4A) Examining intracellular 59Fe distribution, NO decreased ferritin-59Fe levels, while BSO treatment totally prevented this decrease (Fig 4B) These results indicate that GSH is required for the effect of NO at decreasing ferritin-59Fe levels

It is of interest that incubation of BSO-treated LMTK– and SK-N-MC cells with GSNO resulted in an increase in the amount of ferritin-59Fe (Fig 4B) These data were in contrast to the decrease observed after treatment of control cells with GSNO (Fig 4B)

Cell membrane-impermeable or -permeable iron chelators do not increase NO-mediated59Fe efflux

It was possible that passive diffusion may be involved in NO-mediated59Fe release from cells Previous studies have shown that Fe mobilization in the absence of NO is increased by incubation with apoTf and extracellular chelators, presumably due to the ability of these agents to act as an extracellular Fe
sink to increase the concentration gradient across the cell membrane [42–44] Considering this, and the fact that a NO–Fe–GSH complex may be released from cells [32], experiments were designed to investigate the effects of 0.1 mgÆmL)1of apoTf, apolactoferrin, or BSA (as

a protein control) on59Fe mobilization during incubation with increasing concentrations of GSNO (0.025–0.5 mM) However, apoTf, apolactoferrin and BSA only very slightly increased 59Fe release from prelabelled cells at a GSNO concentration of 0.5 mM (Fig 5A) However, compared with the BSA protein control, there was no significant effect

of apoTf or apolactoferrin on 59Fe mobilization from

Control GSNO Control GSNO Control GSNO Control GSNO

Time (min):

Time (min)

0 2 4 6 8 10 12

14

Control GSNO

Time (min)

0 2 4 6 8

10

Control GSNO

Fe

the mean ± SD of three replicates in a typical experiment of two performed The data shown in (B) are a representative experiment of three performed.

SK-N-MC cells (Fig 5A) Studies combining GSNO

(0.5 mM) with increasing concentrations (0.03–1 mM) of

the extracellular chelators, EDTA or DTPA, also

demon-strated no potentiation of 59Fe mobilization from

prela-belled cells (Fig 5B)

As NO acted like an Fe chelator to mobilize59Fe from

prelabelled cells [29,32], studies were performed to

deter-mine if the same Fe pool bound by the permeable chelators,

DFO (Fig 6A) or PIH (Fig 6B), was bound by NO In these studies, cells were prelabelled with59Fe-Tf for 3 h at

37°C, washed, and then reincubated for 3 h at 37 °C with either increasing concentrations of DFO (0.05–1 mM) or PIH (1–50 lM) or these chelators combined with GSNO (0.5 mM) The addition of increasing concentrations of PIH

or DFO to GSNO had no significant effect on cellular59Fe mobilization (Fig 6A and B) These results suggested that

0 5 10 15 20 25

Control

BSO

LMTK–

Control

BSO

SK-N-MC

SK-N-MC

59 Fe-Ferritin Levels Relative Density (% Control) 0

40 80 120 160

in (A) are the mean ± SD of three replicates in a typical experiment of seven performed The data shown in (B) are a representative experiment of five performed.

Fig 5 The extracellular high affinity Fe-binding proteins, apoTf and apolactoferrin, and the extracellular Fe chelators, DTPA and EDTA, do not

representative experiment of three performed.

GSNO and the chelators were acting on the same

intracel-lular compartment of59Fe In contrast, when added without

GSNO, increasing concentrations of DFO or particularly

PIH, resulted in enhanced 59Fe mobilization from cells

(Fig 6A,B)

Examination of the direct effect of NO

and iron chelators on iron pools in cytosolic lysates:

comparison with intact cells

As NO could form an intracellular low Mr Fe–dithiol

dinitrosyl complex [23], experiments were performed to

determine if NO or DFO could mobilize59Fe from lysates

prepared from cells labelled with59Fe-Tf for 3 h at 37°C

(Fig 7A) The lysates were centrifuged to obtain cytosols and then incubated for 3 h at 37°C with DFO (0.5 mM) or GSNO (0.5 mM) The cytosol was then subjected to centrifugation at 4°C through a 5-kDa Mrexclusion filter

In five experiments, only DFO significantly (P < 0.009) increased the amount of59Fe that was passing through the membrane, while GSNO had no significant effect (Fig 7A)

In contrast with the lysates, when cells were prelabelled with

59Fe-Tf and reincubated with DFO or GSNO under the same conditions, GSNO was significantly (P < 0.00001) more effective than DFO or media at mobilizing cellular

59Fe (Fig 7B) These results suggest that intact cellular metabolism was required for NO-mediated59Fe mobiliza-tion

Fe levels in each assessed (see Materials and methods) These results are mean ± SD (three replicates) in a representative experiment of three performed.

are mean ± SD (three replicates) in a representative experiment of three performed.

Further experiments examined the effect of incubation of

lysates derived from59Fe-labelled cells with either DFO,

311, SNAP, NAP, GSNO, GSH, or GSH in the presence

of GSNO The lysates were then subjected to native

PAGE-59Fe-autoradiography (Fig 8) In these studies

examining the direct effect of the chelators and NO on the

lysate, no significant effect was observed on ferritin-59Fe

levels In addition, in the presence of both GSH and GSNO

no effect was apparent (Fig 8) This indicated that intact

cellular metabolism was required for the NO-mediated

effect on this molecule, and that NO did not directly remove

substantial59Fe from the protein (Fig 8)

D I S C U S S I O N Previous studies have clearly demonstrated that NO has a marked effect on cellular Fe metabolism [25–27,32] Indeed, NO-mediated Fe depletion of tumour target cells by activated macrophages could play an important role in immune surveillance [3–5,15,16,45] Our previous studies have shown that NO-mediated Fe mobilization is potenti-ated by incubating cells with D-glucose due to the subsequent generation of GSH [32] In the present study

we have significantly extended our knowledge of this process For the first time, we demonstrate in a cellular system that NO intercepts Fe before it is incorporated into ferritin and appeared to indirectly mobilize Fe from this protein

NO could remove Fe from ferritin by two possible mechanisms: (a) by directly chelating ferritin-bound Fe, or (b) by chelating a cellular Fe pool which leads to ferritin releasing its Fe Of these two possibilities our evidence favours the second mechanism, as NO could not remove

59Fe from ferritin in cellular lysates (Fig 8) Furthermore, the processes resulting in cellular Fe mobilization and Fe release from ferritin were dependent on cellular metabolism (Fig 7) and the generation of GSH (Fig 4 and [32]) These latter observations indicate that active cellular metabolism was required for Fe mobilization rather than direct chela-tion of ferritin-Fe by NO

A previous in vitro study by Reif & Simmons [14] using isolated horse spleen ferritin showed that NO generated by sodium nitroprusside could remove some Fe from this protein Our current data are obviously different, and these inconsistent results may relate to the very different experi-mental systems being used Lee and colleagues [8] have reported, using isolated ferritin, that NO forms a complex with Fe in its core, and have suggested that ferritin could act

as a store of NO Again, it is difficult to compare this latter study to our present experiments, as we have examined the effect of NO using intact cells or cellular lysates It is significant that we have shown that NO not only releases Fe from ferritin indirectly (Figs 1B, 2B, 4), but can also intercept Fe on route to this molecule (Fig 3B) At present the precise molecular mechanism(s) involved in the intra-cellular trafficking and delivery of Fe to ferritin remain unclear, although intermediates of low Mr[46] or high Mr

(e.g metal-binding chaperones) [47] could be involved Nevertheless, our results demonstrate that both permeable

Fe chelators (e.g PIH and DFO) and NO can intercept the same intermediary pool of Fe (Fig 6)

It is of interest that the ability of NO to induce Fe mobilization is dependent on GSH while that for chelators is independent of GSH [32] This suggests that NO by itself does not have the capacity to remove Fe from intermediates and could require the reducing capacity of GSH Alternat-ively, or in combination with this latter mechanism, GSH may form a mixed Fe complex with NO that has the appropriate lipophilicity and charge to diffuse or be transported from the cell Previous studies using EPR spectroscopy have demonstrated the presence of dithiol dinitrosyl–Fe complexes within cells [23,30] Further, Rogers and Ding [48] have shown thatL-cysteine is necessary for the removal of dinitrosyl–Fe complexes from [Fe–S]-containing proteins in Escherichia coli Interestingly, these authors showed that GSH was able to perform the same function but

Fig 8 The direct effect of incubating GSNO and Fe chelators on

59

Fe-containing molecules in cytosolic lysates derived from prelabelled

Fe-autoradio-graphy (see Materials and methods) The results are a representative

experiment of three performed.

Cell Membrane

ADP

GS-Fe-NO

Transporter ?

?

GS-Fe-NO

Protein

?

ATP

TCA

HMPS

Glucose

G-6-P

GSH

Glucose

Transporter

Fe-Protein

NO

NO NO-Fe-Protein

Transporter

Diffusion

by the tricarboxylic acid cycle for the production of ATP and by the

HMPS for the generation of reduced GSH Nitrogen monoxide (NO)

either diffuses or is transported into cells where it intercepts and binds

Fe bound to proteins or Fe on route to ferritin The high affinity of NO

for Fe results in the formation of an NO–Fe complex and GSH may

either be involved as a reductant to remove Fe from endogenous

lig-ands or may complete the Fe coordination shell along with the NO

ligand(s) This complex may then be released from the cell by an active

process requiring a transporter (see text for details).

not as efficiently as L-cysteine [48] Hence, these results

provide support for the possible mechanism of action of

GSH in our experimental system However, at present, we

cannot exclude that GSH has other roles Indeed, recently

the S-nitrosylated form of GSH has been suggested to act

as a transport molecule for NO which increases its half-life

and allows effective biological activity [49,50]

Considering that a low MrFe–NO–GSH complex may

be released from cells and that a concentration gradient

across a membrane can facilitate diffusion [44], we examined

whether NO-mediated Fe release could be potentiated by

strong extracellular Fe chelators (Fig 5) In these studies,

high concentrations of both physiological chelators

(apolactoferrin and apoTf) and synthetic chelators (EDTA

and DTPA) were used, and for all ligands no significant

potentiation of 59Fe mobilization was observed upon

combination with NO These studies suggest that

enhance-ment of the concentration gradient across the cell membrane

did not alter NO-mediated Fe release Considering this, it is

of note that intact cellular metabolism was required for Fe

mobilization by NO (Fig 7), and NO-mediated Fe release

was prevented by metabolic inhibitors and at 4°C [32]

Collectively, these data suggest that an energy-dependent

mechanism was required to enable efflux of the NO–Fe

complex

Based upon the results presented in this and our previous

study [32], we suggest in Fig 9 a hypothetical model of

D-glucose-dependent NO-mediated Fe mobilization from

cells.D-Glucose is transported into cells and is used by the

tricarboxylic acid cycle (TCA) for the production of ATP

and by the HMPS for the generation of GSH NO diffuses or

is transported [51] into cells where it intercepts Fe on route to

ferritin and binds Fe bound to proteins (Fig 9) The high

affinity of NO for Fe [2] results in the formation of an NO-Fe

complex and GSH may either be involved as a reductant to

remove Fe from endogenous ligands [48] or may complete

the Fe coordination shell along with NO [17,20,23,30] This

complex may then be transported out of the cell by an

energy-dependent transporter such as ferroportin 1 [52], or

alternatively, the ATP-binding cassette (ABC) transporter

family (e.g glutathione-S-conjugate export pump), which

are known to mediate the efflux of glutathione-conjugates

[53,54] (Fig 9) Further studies aimed at identifying the exact

molecular nature of the Fe released by NO and the

transporter involved are underway Finally, out current

results may be important in understanding the cytotoxic

actions of NO produced by activated macrophages

A C K N O W L E D G E M E N T S

The authors thank J Kwok for her excellent suggestions on this

manuscript prior to submission This work was supported by an

Australian Research Council Large Grant and Grants 970360 and

981826 from the National Health and Medical Research Council of

Australia.

R E F E R E N C E S

1 Moncada, S., Palmer, R.M.J & Higgs, E.A (1991) Nitric oxide:

physiology, pathophysiology and pharmacology Pharmacol.Rev.

43, 109–142.

2 Stamler, J., Singel, D.J & Loscalzo, J (1992) Biochemistry of

nitric oxide and its redox-activated forms Science 258, 1898–1902.

3 Richardson, D.R & Ponka, P (1997) The molecular mechanisms

of the metabolism and transport of iron in normal and neoplastic cells Biochim.Biophys.Acta 1331, 1–40.

4 Drapier, J.-C & Hibbs, J.B Jr (1986) Murine cytotoxic activated macrophages inhibit aconitase in tumor cells Inhibition involves the iron-sulfur prosthetic group and is reversible J.Clin.Invest.

78, 790–797.

5 Drapier, J.-C & Hibbs, J.B Jr (1988) Differentiation of murine macrophages to express non-specific cytotoxicity for tumor cells

iron-sulfur enzymes in the macrophage effector cells J.Immunol.140, 2829–2838.

6 Henry, Y., Ducrocq, C., Drapier, J.-C., Pellat, C & Guissani, A (1991) Nitric oxide as a biological effector molecule Electron paramagnetic resonance detection of nitrosyl-iron-protein com-plexes in whole cells Eur.Biophys.J.20, 1–15.

7 Drapier, J.-C (1997) Interplay between NO and [Fe-S] clusters: relevance to biological systems Methods 11, 319–329.

8 Lee, M., Arosio, P., Cozzi, A & Chasteen, N.D (1994) Identifi-cation of the EPR-active iron-nitrosyl complexes in mammalian ferritins Biochemistry 33, 3679–3687.

9 Lepoivre, M., Fieschi, F., Coves, J., Thelander, L & Fontecave, M (1991) Inactivation of ribonucleotide reductase by nitric oxide Biochem.Biophys.Res.Commun.179, 442–448.

10 Ignarro, L.J (1991) Heme-dependent activation of guanylate cyclase by nitric oxide: a novel signal transduction mechanism Blood Vessels 28, 67–73.

11 Khatsenko, O.G., Gross, S.S., Rifkind, R.R & Vane, J.R (1993) Nitric oxide is a mediator of the decrease in cytochrome P450-dependent metabolism caused by immunostimulants Proc.Natl Acad.Sci.USA 90, 11147–11151.

12 Griscavage, J.M., Fukuto, J.M., Komori, Y & Ignarro, L.J (1994) Nitric oxide inhibits neuronal nitric oxide synthase by interacting with the heme prosthetic group Role of tetra-hydrobiopterin in modulating the inhibitory action of nitric oxide J.Biol.Chem.269, 21644–21649.

13 Kim, Y.-M., Bergonia, H.A., Muller, C., Pitt, B.R., Watkins, W.D & Lancaster, J.R Jr (1995) Loss and degradation of enzyme-bound heme induced by cellular nitric oxide synthesis J.Biol.Chem.270, 5710–5713.

14 Reif, D.W & Simmons, R.D (1990) Nitric oxide mediates iron release from ferritin Arch.Biochem.Biophys.283, 537–541.

15 Hibbs, J.B Jr, Taintor, R.R & Vavrin, Z (1984) Iron depletion: possible cause of tumor cell cytotoxicity induced by activated macrophages Biochem.Biophys.Res.Commun.123, 716–723.

16 Hibbs, J.B Jr, Taintor, R.R., Vavrin, Z & Rachlin, E.M (1988) Nitric oxide: a cytotoxic activated macrophage effector molecule Biochem.Biophys.Res.Commun.157, 87–94.

17 Lancaster, J.R & Hibbs, J.B Jr (1990) EPR demonstration of iron-nitrosyl complex formation by cytotoxic activated macro-phages Proc.Natl Acad.Sci.USA 87, 1223–1229.

18 Pellat, C., Henry, Y & Drapier, J.-C (1990) IFN-gamma-activated macrophages: detection by electron paramagnetic

and non-heme iron proteins Biochem.Biophys.Res.Commun.

166, 119–125.

19 Vanin, A.F., Men’shikov, G.B., Moroz, I.A., Mordvintcev, P.I., Serezhenkov, V.A & Burbaev, D.S (1992) The source of non-heme iron that binds nitric oxide in cultivated macrophages Biochim.Biophys.Acta 1135, 275–279.

20 Drapier, J.-C., Pellat, C & Henry, Y (1991) Generation of EPR-detectable nitrosyl-iron complexes in tumor target cells co-cultured with activated macrophages J.Biol.Chem.266, 10162–10167.

21 Bastian, N.R., Yim, C.Y., Hibbs, J.B Jr & Samlowski, W.E (1994) Induction of iron-derived EPR signals in murine cancers by

nitric oxide Evidence for multiple intracellular targets J.Biol.

Chem 269, 5127–5131.

22 Commoner, B., Woolum, J.C., Senturia, B.H Jr & Ternberg, J.L.

(1970) The effects of 2-acetylaminofluorene and nitrite on free

radicals and carcinogenesis in rat liver Cancer Res 30, 2091–2097.

23 Vanin, A.F (1991) Endothelium-derived relaxing factor is a

nitrosyl iron complex with thiol ligands FEBS Lett 289, 1–3.

24 Theil, E.C (1998) The iron responsive element (IRE) family of

mRNA regulators Regulation of iron transport and uptake

compared in animals, plants and microorganisms Met.Ions Biol.

Syst 35, 403–434.

25 Drapier, J.-C., Hirling, H., Wietzerbin, J., Kaldy, P & Ku¨hn, L.C.

(1993) Biosynthesis of nitric oxide activates iron regulatory factor

in macrophages EMBO J 12, 3643–3649.

26 Weiss, G., Goossen, B., Doppler, W., Fuchs, D., Pantopoulos, K.,

Werner-Felmayer, G., Wachter, H & Hentze, M.W (1993)

Translational regulation via iron-responsive elements by the nitric

oxide/NO-synthase pathway EMBO J 12, 3651–3657.

27 Richardson, D.R., Neumannova, V., Nagy, E & Ponka, P (1995)

The effect of redox-related species of nitrogen monoxide on

transferrin and iron uptake and cellular proliferation of

erythro-leukemia (K562) cells Blood 86, 3211–3219.

28 Kennedy, M.C., Antholine, W.E & Beinert, H (1997) An EPR

investigation of the products of the reaction of cytosolic and

mitochondrial aconitases with nitric oxide J.Biol.Chem.272,

20340–20347.

29 Wardrop, S.L., Watts, R.N & Richardson, D.R (2000)

Nitrogen monoxide activates iron regulatory protein 1

RNA-binding activity by two possible mechanisms: effect on the [4Fe-4S]

cluster and iron mobilization from cells Biochemistry 39,

2748–2758.

30 Vanin, A.F., Bliumenfel’d, L.A & Chetverikov, A.G (1967) EPR

study of non-heme iron complexes in cells and tissues Biofizika

[Russian] 12, 829–841.

31 Woolum, J.C., Tiezzi, E & Commoner, B (1968) Electron spin

resonance of iron-nitric oxide complexes with amino acids,

pep-tides and proteins Biochim.Biophys.Acta 160, 311–320.

32 Watts, R.N & Richardson, D.R (2001) Nitrogen monoxide (NO)

and glucose: unexpected links between energy metabolism and

NO-mediated iron mobilization from cells J.Biol.Chem.276,

4724–4732.

33 Field, L., Dilts, R.V., Ravichandran, R., Lenhert, P.G &

Carnahan, G.E (1978) An unusually stable thionitrite from

of 2-(acetylamino)-2-carboxy-1,1-dimethylethyl thionitrite J.Chem.

Soc.Chem.Commun.249, 249–250.

34 Richardson, D.R., Tran, E & Ponka, P (1995) The potential of

iron chelators of the pyridoxal isonicotinoyl hydrazone class as

effective antiproliferative agents Blood 86, 4295–4306.

35 Pantopoulos, K., Weiss, G & Hentze, M.W (1996) Nitric oxide

by different pathways Mol.Cell.Biol.16, 3781–3788.

36 Richardson, D.R & Baker, E (1992) Two mechanisms of iron

uptake from transferrin by melanoma cells The effect of

desfer-rioxamine and ferric ammonium citrate J.Biol.Chem.267,

13972–13979.

37 Green, L.C., Wagner, D.A., Glogowski, J., Skipper, P.L.,

Wishnok, J.S & Tannenbaum, S.R (1982) Analysis of nitrate,

131–138.

38 Richardson, D.R., Ponka, P & Vyoral, D (1996) Distribution of iron in reticulocytes after inhibition of heme synthesis with suc-cinylacetone Examination of the intermediates involved in iron metabolism Blood 87, 3477–3488.

39 Sedlack, J & Lindsay, R (1968) Estimation of total protein-bound and nonprotein sulfhydryl groups in tissue with Ellman’s reagent Anal.Biochem.25, 192–205.

40 Griffith, O.W & Meister, A (1979) Potent and specific inhibition

of glutathione synthesis by buthionine sulfoximine (S-n-butyl homocysteine sulfoximine) J.Biol.Chem.254, 7558–7560.

41 Richardson, D.R & Milnes, K (1997) The potential of iron chelators of the pyridoxal isonicotinoyl hydrazone class as effective antiproliferative agents II The mechanism of action of ligands derived from salicylaldehyde benzoyl hydrazone and 2-hydroxy-1-naphthylaldehyde benzoyl hydrazone Blood 89, 3025–3038.

42 Baker, E., Page, M & Morgan, E.H (1985) Transferrin and iron release from rat hepatocytes in culture Am.J.Physiol.248, G93–G97.

43 Baker, E., Morton, A.G & Tavill, A.S (1980) The regulation of iron release from the perfused rat liver Br.J.Haematol.45, 607–620.

44 Young, S.P., Fahmy, M & Golding, S (1997) Ceruloplasmin, transferrin and apotransferrin facilitate iron release from human liver cells FEBS Lett 411, 93–96.

45 Nestle, F.P., Greene, R.N., Kichian, K., Ponka, P & Lapp, W.S (2000) Activation of macrophage cytostatic effector mechanisms during acute graft-versus-host disease: release of intracellular iron and nitric oxide-mediated cytostasis Blood 96, 1836–1843.

46 Jacobs, A (1977) Low molecular weight intracellular iron trans-port compounds Blood 50, 433–439.

47 Harrison, M.D., Jones, C.E & Dameron, C.T (1999) Copper chaperones: function, structure and copper-binding properties J.Biol.Inorg.Chem.4, 145–153.

destabiliza-tion of dinitrosyl iron complexes in proteins J.Biol.Chem.276, 30980–30986.

49 Lipton, A.J., Johnson, M.A., Macdonald, T., Lieberman, M.W., Gozal, D & Gaston, B (2001) S-Nitrosothiols signal the venti-latory response to hypoxia Nature 413, 171–174.

50 Lipton, S.A (2001) Nitric oxide and respiration Nature 413, 118–121.

51 Zai, A., Rudd, M.A., Scribner, A.W & Loscalzo, J (1999) Cell-surface protein disulphide isomerase catalyzes transnitrosation and regulates intracellular transfer of nitric oxide J.Clin.Invest.

103, 393–399.

52 Donovan, A., Brownlie, A., Zhou, Y., Shepard, J., Pratt, S.J., Moynihan, J., Paw, B.H., Drejer, A., Barut, B., Zapata, A et al (2000) Positional cloning of zebrafish ferroportin 1 identifies a conserved vertebrate iron exporter Nature 403, 776–781.

53 Suzuki, T., Nishio, K & Tanabe, S (2001) The MRP family and anticancer drug metabolism Curr.Drug Metab.2, 367– 377.

54 Dietrich, C.G., Ottenhoff, R., de Waart, D.R & Oude Elferink, R.P (2001) Role of MRP2 and GSH in intrahepatic cycling of toxins Toxicology 167, 73–81.