Báo cáo khoa học: Transfection with 4-hydroxynonenal-metabolizing glutathione S-transferase isozymes leads to phenotypic transformation and immortalization of adherent cells pdf

The results showthat lowering intracellular levels of 4-HNE by incorporation of active hGSTA4-4, by either transfection or microinjection, led to phenotypic transformation of attached ce

Transfection with 4-hydroxynonenal-metabolizing glutathione

and immortalization of adherent cells

Rajendra Sharma1,*, David Brown1,*, Sanjay Awasthi2, Yusong Yang1, Abha Sharma1, Brad Patrick1, Manjit K Saini1, Sharda P Singh3, Piotr Zimniak3, Shivendra V Singh4and Yogesh C Awasthi1

1

Department of Human Biological Chemistry and Genetics, University of Texas Medical Branch, Galveston, TX, USA;

2

Department of Chemistry and Biochemistry, University of Texas at Arlington, TX, USA;3Department of Pharmacology

and Toxicology and Department of Biochemistry and Molecular Biology, University of Arkansas for Medical Sciences,

and Central Arkansas Veterans Healthcare System, Little Rock, AR, USA; 4 Department of Pharmacology,

University of Pittsburgh Cancer Center, PA, USA

4-Hydroxy-2-trans-nonenal (4-HNE), one of the major end

products of lipid peroxidation, has been shown to induce

apoptosis in a variety of cell lines It appears to modulate

signaling processes in more than one way because it has been

suggested to have a role in signaling for differentiation and

proliferation We showfor the first time that incorporation

of 4-HNE-metabolizing glutathione S-transferase (GST)

isozyme, hGSTA4-4, into adherent cell lines HLE B-3 and

CCL-75, by either cDNA transfection or microinjection of

active enzyme, leads to their transformation The dramatic

phenotypic changes due to the incorporation of hGSTA4-4

include rounding of cells and anchorage-independent rapid

proliferation of immortalized, rounded, and smaller cells

Incorporation of the inactive mutant of hGSTA4-4 (Y212F)

in cells by either microinjection or transfection does not

cause transformation, suggesting that the activity of

hGSTA4-4 toward 4-HNE is required for transformation This is further confirmed by the fact that mouse and Dro-sophila GST isozymes (mGSTA4-4 and DmGSTD1-1), which have high activity toward 4-HNE and subsequent depletion of 4-HNE, cause transformation whereas human GST isozymes hGSTP1-1 and hGSTA1-1, with minimal activity toward 4-HNE, do not cause transformation In cells overexpressing active hGSTA4-4, expression of trans-forming growth factor b1, cyclin-dependent kinase 2, pro-tein kinase C bII and extracellular signal regulated kinase is upregulated, whereas expression of p53 is downregulated These studies suggest that alterations in 4-HNE homeostasis can profoundly affect cell-cycle signaling events

Keywords: 4-hydroxy-2-trans-nonenal; glutathione S-trans-ferase; lipid peroxidation; oxidative stress; transformation

Oxidative stress causes generation of reactive oxygen

species, which leads to the onset of lipid peroxidation [1]

4-Hydroxynonenal (4-HNE) is one of the end products of

this process [2] In recent years there has been an increasing

interest in the role of 4-HNE in signaling mechanisms

[3–12] There are reports suggesting that 4-HNE can cause

cell cycle arrest [2], apoptosis [3,6,7,12], differentiation [12]

or proliferation [11,12] in different cell types in a concen-tration-dependent manner These seemingly opposite effects

of 4-HNE on cell cycle signaling (e.g cell cycle arrest and apoptosis vs proliferation) are intriguing If 4-HNE does indeed differentially affect signal transduction in a concen-tration-dependent manner, the regulation of 4-HNE homeo-stasis may be important for cell cycle signaling It is inherently difficult to characterize the functional conse-quences of changes in intracellular 4-HNE concentration because 4-HNE is formed by lipid peroxidation, mostly an uncontrolled nonenzymatic process In this study, we circumvented this problem by regulating 4-HNE concen-tration through its metabolism, and investigated the effect

of altered 4-HNE homeostasis on proliferation and cell cycle signaling in two different adherent cell lines

To test the hypothesis that 4-HNE may be a determinant

in cell cycle regulation, we stably transfected the human lens epithelial cell line (HLE B-3) with cDNA for human glutathione S-transferase (GST, EC 2.5.1.18) isozyme hGSTA4-4 This isozyme conjugates GSH to 4-HNE with high efficiency [13], and cells overexpressing it, or similar enzymes [14], have lower steady-state levels of 4-HNE [12]

In accordance with accepted convention [15], we refer to the gene and the dimeric enzyme as hGSTA4 and hGSTA4-4,

Correspondence to Y C Awasthi, 551 Basic Science Building,

Department of Human Biological Chemistry and Genetics, University

of Texas Medical Branch, Galveston, TX 77555-0647, USA.

Fax: + 1 409 772 6603, Tel.: + 1 409 772 2735,

E-mail: ycawasth@utmb.edu

Abbreviations: 4-HNE, 4-hydroxy-2-trans-nonenal; GST, glutathione

S-transferase; HLE B-3, human lens epithelial cell; CCL-75, human

lung fibroblast cell; JNK, c-Jun N-terminal kinase; OG-dextran,

Oregon green 488-dextran; GFP, green fluorescent protein; eGFP,

enhanced green fluorescent protein; GS-HNE, glutathione conjugate

of 4-HNE.

Enzyme: glutathione S-transferase (GST; EC 2.5.1.18).

*These authors contributed equally to this work.

(Received 22 January 2004, revised 24 February 2004,

accepted 2 March 2004)

respectively Surprisingly, the clonal lines of HLE B-3/

hGSTA4transfectants overexpressing enzymatically active

hGSTA4-4 acquired a transformed phenotype with time

We then examined whether an adherent cell line other than

HLE B-3 would also be affected by hGSTA4 transfection

and exhibit a similar transformed phenotype Furthermore,

to correlate specifically the effects of hGSTA4 transfection

with the increased metabolism and depletion of 4-HNE, we

investigated the effect of transfection with mutant hGSTA4

devoid of GST activity towards 4-HNE Finally, we

compared the effect of microinjection of different GST

isozymes from several species into HLE B-3 cells to rule out

nonspecific effects of GST overexpression of active or

mutant hGSTA4-4 protein The results showthat lowering

intracellular levels of 4-HNE by incorporation of active

hGSTA4-4, by either transfection or microinjection, led to

phenotypic transformation of attached cells into rounded,

smaller cells which acquired immortality and grew rapidly in

an anchorage-independent manner

Experimental procedures

Cell culture

HLE B-3 cells were a gift from U P Andley (Department

of Ophthalmology and Visual Sciences, Washington

Uni-versity at St Louis, MO, USA) The cells were received on

passage no 14 and were maintained in minimal essential

medium containing 20% fetal bovine serum and

50 lgÆmL)1gentamicin at 37C in 5% CO2 Human lung

fibroblast cell line, CCL-75, obtained from ATCC

(Man-assas, VA, USA) was maintained in minimal essential

medium containing 10% fetal bovine serum, 1 mMsodium

pyruvate and 10 mMnonessential amino acids

Antibodies

Polyclonal antibodies were developed against recombinant

hGSTA4-4 in chicken as described previously [16] All other

antibodies were from commercial sources

Preparation of recombinant hGSTA4-4 and other

GST isozymes

hGSTA4-4 was expressed in Escherichia coli and purified as

described previously [16] The purity of the enzyme was

confirmed by SDS/PAGE; a single band at 26 kDa was

recognized by hGSTA4-4 antibodies on Western blots

Activity of the purified enzyme using

1-chloro-2,4-dinitro-benzene and 4-HNE as substrates was measured as

described previously [6] Methods for preparation of

recombinant GST isozymes mGSTA4-4 [14], Drosophila

melanogaster DmGSTD1-1 [17], hGSTA1-1 [18] and

hGSTP1-1 [19] have been described previously

Preparation of hGSTA4-4 eukaryotic expression

constructs

The hGSTA4 ORF was amplified by PCR from the bacterial

expression vector pET-30a[+]/hGSTA4, and subcloned into

the pTarget-T mammalian expression vector (Promega)

The hGSTA4 insert was confirmed by sequencing

Transfection of HLE B-3 cells with p-Target-hGSTA4 expression vector

HLE B-3 cells (2· 105) at passage no 18 were plated in

60 mm dishes in complete growth medium When the cells reached nearly 80% confluency, the medium was changed, and the cells were transfected 3–4 h later with 6 lg plasmid using the Profection mammalian transfection kit (Promega) according to the manufacturer’s protocol After 4 h, the cells were treated with 10% dimethyl sulfamethoxazole in minimal essential medium for 30 s After dimethyl sulfa-methoxazole shock, the cells were allowed to recover in complete growth medium for 48 h Stable transfectants were selected in 200 lgÆmL)1G418 by the dilution method

in 96 well plates Wells containing single cells were marked, and growth in these wells was monitored daily Expression

of hGSTA4-4 protein was ascertained by Western blot analysis

Site-directed mutagenesis of hGSTA4-4 The Y212F mutation was introduced in both the bacterial and the mammalian hGSTA4-4 expression vectors using the Quickchange site-directed mutagenesis kit (Stratagene,

La Jolla, CA, USA) with the mutagenic sense primer 5¢-CCTGATGAATTTTCGTGAGAACCGT (mutation underlined) and the complementary antisense primer In this paper, hGSTA4-4(Y212F) is referred to as mut-hGSTA4-4

Immunohistochemical localization studies Immunofluorescence studies on adherent HLE B-3 and CCL-75 cells (wild-type, empty-vector-transfected and mut-hGSTA4-transfected) were carried out by seeding

1· 104cells on to coverslips Next day, the coverslips with attached cells were washed in NaCl/Pi(pH 7.0) three times (5 min each) and fixed in 4% paraformaldehyde solution prepared in NaCl/Pi (pH 7.4) for 15 min at room temperature The fixed cells were washed three times with NaCl/Pi, permeabilized in cold methanol ()20 C) for

30 s, treated with sodium borohydride (0.5 mgÆmL)1) for

15 min to reduce aldehyde groups, and washed three times with NaCl/Pi(5 min each) The cells were then incubated with blocking buffer (1% BSA + 1% goat serum in NaCl/Pi) for 2 h at room temperature in a humidified chamber, and incubated with primary antibodies against hGSTA4-4 developed in chicken (1 : 200 dilution pre-pared in 1% BSA in NaCl/Pi) overnight at 4C Cells were washed three times in NaCl/Pi and then incubated with Alexa fluor 488 fluorescein isothiocyanate-conjugated anti-chicken secondary IgG (Molecular Probes; 1 : 200, diluted in 1% BSA in NaCl/Pi) for 2 h at room temperature in a humidified chamber Cells were washed three times with NaCl/Pi, mounted on slides with 50% glycerol in NaCl/Pi, and visualized under a fluorescence microscope (Nikon Eclipse 600) The cells treated with preimmune chicken IgY were used as negative controls Slides for suspension culture of hGSTA4-transfected and transformed HLE B-3 cells were prepared by centrifu-ging the cells on polylysine-coated slides in a cytospin at

28 g

In situ detection of apoptosis

To detect cells undergoing apoptosis during the course of

microinjection experiments, we performed

immunolocali-zation of cleaved caspase-3 by using monoclonal

antibod-ies against cleaved caspase-3 After cytospinning the cells

at 28 g for 5 min, the cells w ere fixed in 4%

paraformal-dehyde (15 min) and washed three times in NaCl/Pi The

cells were permeabilized by incubation in 0.1% Triton

X-100 for 2 min, washed with NaCl/Pi, treated w ith

blocking buffer for 2 h at room temperature in a

humidified chamber, and then incubated with cleaved

caspase-3 IgG (1 : 100 dilution prepared in 1% BSA)

overnight at 4C Cells were washed three times in NaCl/

Piand then incubated with mouse tetramethyl rhodamine

isothiocyanate-conjugated secondary antibodies (1 : 500)

for 2 h After the cells had been washed and mounted as

described above, the expression of cleaved caspase-3, a

marker of apoptosis, was ascertained by observing the

cells under a fluorescence microscope

Determination of intracellular levels of malondialdehyde

and 4-HNE

Lipid peroxide levels as determined by malondialdehyde

and 4-HNE concentrations in hGSTA4-transfected and

control HLE B-3 cells were determined using the Biotech

LPO-586TMkit (Oxis International, Portland, OR, USA)

according to the manufacturer’s protocol as described

previously [6]

SDS/PAGE and Western blot analysis

For checking the expression of hGSTA4-4 by Western blot

analysis, cells (1· 106) were lysed in 10 mM potassium

phosphate buffer, pH 7.0, containing 1.4 mM

2-mercapto-ethanol, sonicated on ice for 30 s, and centrifuged at

28 000 g for 30 min Buffer-soluble proteins (25 lg) present

in the supernatants were mixed with Laemmeli’s sample

buffer [20] and loaded in the wells of gels containing 12%

polyacrylamide Proteins resolved on SDS/polyacrylamide

gels were transferred to nitrocellulose or poly(vinylidene

difluoride) membranes, and the blots probed by using

hGSTA4-4 antibodies developed in chicken as primary

antibodies, and secondary antibodies as horseradish

per-oxide-conjugated anti-chicken IgG developed in goat Blots

were developed by West Pico-chemiluminescence’s reagent

(Pierce) To check the expression of p53, transforming

growth factor b1, cyclin-dependent kinase 2 and protein

kinase C bII proteins in HLE B-3 cells, Western blot

analyses were performed by preparing whole cell extracts in

RIPA buffer [20 mMTris/HCl, pH 7.4, 150 mMNaCl, 1%

Nonidet P40, 1 mM EDTA, 1 mM NaF, 1 mM sodium

vanadate, 1 mM phenylmethanesulfonyl fluoride and

pro-tease inhibitor cocktail (Sigma Chemical Co)] For these

analyses extracts containing 100 lg protein were used for

each sample

Cell growth analysis

The growth kinetics of HLE B-3 cells and their

transfect-ants was measured both by manual cell count using a

hemocytometer (after trypsinization in the case of adherent cells) and by 3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyltetra-zolium bromide analysis as described previously [18] Assay of soft agar colony formation

This was performed as described previously [21] Briefly,

5 000 cells per dish, mixed in 0.35% agarose/complete medium, were seeded on to 0.7% agarose/complete medium bottom layer The Petri dishes were incubated at 37C and

a drop of medium was added every 3 days Four weeks later, cells were stained with 0.5% crystal violet (Sigma) in 20% methanol for 2 h, and colonies were counted under a microscope

Microinjection of cells Protein sample preparation Immediately before injection, the recombinant hGSTA4-4 protein (wild-type or mutant) was dialyzed against injection buffer (114 mMKCl, 0.5 mM

K2HPO4 and 5.5 mM KH2PO4, pH 7.4) for 10 min, and brought to a concentration of 2 mgÆmL)1 with injection buffer and a 5 mgÆmL)1stock solution of Oregon Green 488-dextran (OG-dextran; 70 kDa; Molecular Probes), bringing the injection samples to an OG-dextran concen-tration of 0.4 mgÆmL)1, a concentration used in previous studies that had no effect on cell viability and proliferation [22] The samples were then centrifuged at 10 000 g for

10 min to remove large aggregates All steps of the sample preparation were performed at 4C, and the samples kept

on ice until injected into cells Samples of recombinant mGSTA4-4, DmGSTD1-1, hGSTA1-1 and hGSTP1-1 used for microinjection were prepared in an identical manner

DNAsample preparation Both wild-type- and mut-hGSTA4-4 expression vectors were brought to a concentra-tion of 20 copies per 5 fL with injecconcentra-tion buffer when injected

as individual samples We had previously determined that optimal expression occurred with this concentration of DNA [22] For the experiments in which mGSTA4-4 expression vectors were coinjected with the green fluorescent protein (GFP) expression vector, all coinjected vectors were brought to a concentration of 40 copies per 5 fL with injection buffer Just before injection, coinjected samples were mixed 1 : 1 bringing the coinjected vectors to a concentration of 20 copies of each vector per 5 fL Just before injection of the vectors into cells, all samples were dialyzed against injection buffer for 10 min, and then centrifuged at 10 000 g for 10 min at room temperature Glass-needle-mediated microinjection of proteins and DNAexpression vectors HLE B-3 and CCL-75 cells were maintained as monolayer cultures as described above For the experiments performed in this study, HLE B-3 and CCL-75 cells were used at passage 18 On the day before each experiment, 2· 104cells were plated in 35 mm2tissue culture dishes (Corning) containing 1.5 mL medium Before plating of the cells, circles were etched into each of the dishes

to facilitate subsequent identification of injected cells Injection needles were pulled from borosilicate capillaries using a Flaming/Brown Micropipette Puller, model P-97

(Sutter Instrument Co., Novato, CA, USA) with a range of

outer tip diameters of 2.5–3 lm, as determined by scanning

electron microscopy [23] Phase contrast microscopy was

used to visualize the injection procedure using an Olympus

Corp (Melville, NY, USA) IX70 inverted microscope

equipped with a temperature-controlled stage kept at 37C

The cells were injected with either 5 fL sample containing

protein (cytoplasmic injections, with OG-dextran as the

marker of the injected cells) or sample containing DNA

(nuclear injections, with GFP as the marker of the injected

cells), using the electronically interfaced Eppendorf

Micro-manipulator (model 5171) and Transjector (model 5246) as

described previously [23] All injections were performed

manually, with each injection sample being injected into

75 cells per dish per experiment All experiments were

repeated two or more times Only cells within an etched

boundary were injected, to allow easy localization of the

injected cells

Single-cell assay of post-injection viability and

GS-induced cell rounding or apoptosis Fluorescence

micros-copy (IX70 inverted microscope) was used to identify

injected cells The percentage cytoplasmic and

post-nuclear injection viabilities were determined for both

HLE B-3 and CCL-75 cells by calculating: (number of

fluorescent cells 24 h after injection/75)· 100 Viabilities

were determined from cells coinjected with either

mut-hGSTA4-4 protein and OG-dextran or the mut-mut-hGSTA4-4

expression vector expressing the mutant form of hGSTA4

(Y212F) and the GFP expression vector, or with fluorescent

markers alone At 24 h after the injection, any cells killed by

the injection procedure were lifted off the dish leaving only

the injected cells that survived the injection Such cells were

flat and attached to the dish as shown in Fig 2 Stratagene

enhanced GFP (eGFP) and OG-dextran fluorescent

mark-ers of injected cells as well as mutant hGSTA4-4 protein had

no effect on post-injection viabilities The mean viability

after nuclear injection into HLE B-3 and CCL-75 cells

ranged from 40% to 70% To determine the effect of

wild-type and mutant forms of GST on injected cells, all

surviving HLE B-3 and CCL-75 cells were scored at 24 and

48 h and 24, 48, and 72 h, respectively, as being either flat,

round or apoptotic The mean percentage of the injected

cells showing the above morphologies was calculated with

data from three or more experiments for each injection

sample at each time point

Results

Effect of transfection of HLE B-3 cells withhGSTA4

The HLE B-3 cell line was originally developed after

infection with adenovirus (Ad12-SV40) [24] and is referred

to here as WT-HLE B-3 These cells have been reported to

be relatively resistant to oxidative stress [25], growin

monolayers (Fig 1A, a) with a population doubling time of

48–52 h, and become senescent after 76 population

dou-blings [24] Keeping this in view, we used WT-HLE B-3 cells

with low passage numbers (passages 18–20) for these

studies WT-HLE B-3 cells were transfected with a

eukary-otic expression vector containing hGSTA4 cDNA, and three

clones overexpressing hGSTA4-4, designated C4, D7 and

E1, were selected Initially, hGSTA4-transfected cells grew normally in monolayers (Fig 1A, b) with a doubling time identical with that of empty-vector-transfected cells How-ever, four weeks after transfection (two passages) during their clonal selection in medium containing G418, cells stopped proliferating and some began to enlarge (Fig 1A, c) Even though the growth medium was changed every

72 h, the cells remained in a quiescent state for the next four weeks Eight weeks after transfection, cells originating from clones C4, D7 and E1 started to transform their shape, as was apparent from the characteristic budding of round cells from giant cells A typical example of this transformation is show n in Fig 1A, d The transformed round cells becoming anchorage-independent (Fig 1A, e) continued to express higher levels of hGSTA4-4 (Fig 1B, a-p and a-f), and had lower levels of 4-HNE (Fig 1C) To date, these cells have undergone about 365 doublings in suspension cultures, with

no cells becoming senescent, a property characteristic of cancer-derived cell lines, e.g human erythroleukemic (K562) and small cell lung cancer (H69) cell lines The HLE B-3/hGSTA4 anchorage-independent cells had a sig-nificantly shorter doubling time than wild-type-transfected and empty-vector-transfected HLE B-3 cells (20 ± 3.4 h

vs 50 ± 4.3 h)

hGSTA4-4 expression and 4-HNE levels in transfected cells

The expression of hGSTA4-4 in stably transfected cells was confirmed by Western blots, which showed no detectable expression of hGSTA4-4 in the wild-type-transfected or empty-vector-transfected HLE B-3 cells, but a strong band

in hGSTA4-transfected cells (Fig 1D) All three clones (C4, D7 and E1) continued to express high levels of enzymat-ically active hGSTA4-4 and showed similar effects of hGSTA4transfection on their phenotype with a significant reduction in intracellular 4-HNE levels Most of the data presented here were obtained using the representative clone C4 Although there was detectable constitutive GST activity towards 4-HNE in WT- HLE B-3 cells, this activity was about sixfold higher in the transfected cells [1.5 vs 9.7 nmol 4-HNE consumedÆmin)1Æ(mg protein))1], indicating success-ful expression of enzymatically active hGSTA4-4 in trans-fected cells The 4-HNE level in clone C4 used for these studies was found to be 40 ± 8% of that observed in the wild-type-transfected or empty-vector-transfected HLE B-3 cells (Fig 1C) These results further confirm overexpression

of active hGSTA4-4 in the transfected cells

Anchorage-independent growth The anchorage-independent growth of phenotypically transformed cells was confirmed by assay of soft agar colonies [26] Clone C4 cells grewinto colonies within

3 weeks of plating, while WT-HLE B-3 cells did not form detectable colonies (data not presented) The colony-forming ratio of clone C4 (HLE B-3) cells to WT-K562 cells used as positive control in these experiments was found to be 3 : 1 Taken together, these results confirm the phenotypic transformation of WT-HLE B-3 cells to anchorage-independent growth on stable transfection with hGSTA4

Effect of transfection with enzymatically inactive

mutanthGSTA4

To establish whether the observed phenotypic changes were

specifically due to depletion of 4-HNE because of high

activity of hGSTA4-4 towards 4-HNE in the transfected

cells or to some unknown effect of transfection, we prepared

a mutant cDNA of hGSTA4-4 isozyme in which Tyr212

was replaced with phenylalanine Consistent with the

previous studies [13], recombinant mutant

hGSTA4-4(Y212F) had only 3% of the activity towards 4-HNE

compared with WT-hGSTA4-4 [1.9 vs 72 lmol 4-HNEÆ

min)1Æ(mg protein))1] There was no noticeable change in

morphology of the cells tarnsfected with mutant hGSTA4

(Y212F) even after six passages (Fig 1B, m-p) Despite high

expression of mutant protein as indicated by

immunolocal-ization (Fig 1B, m-f) and Western blot studies (Fig 1D,

lane 1) using hGSTA4-4 antibodies, there was no significant

change in either their GST activity towards 4-HNE or the steady-state levels of 4-HNE compared with those of WT-HLE B-3 cells (Fig 1C) These results strongly suggest that overexpression of enzymatically active hGSTA4-4 resulting

in accelerated metabolism of 4-HNE and thereby lowering

of the intracellular concentrations of 4-HNE leads to the observed phenotypic transformation and immortalization

of WT-HLE B-3 cells

Microinjection of the active hGSTA4-4 induces similar phenotypic changes

We also studied the effects of direct microinjection of the active hGSTA4-4, its inactive mutant, and their expression vector counterparts into WT-HLE B-3 cells To monitor the microinjection of active or inactive hGSTA4-4 recom-binant protein, the cells were coinjected with OG-dextran, a fluorescent marker, as detailed in the legend of Fig 2A The

Fig 1 Phenotypic transformation and biochemical characterization of hGSA4-transfected cells (A) Phenotypic transformation of hGSA4-trans-fected cells (a) Control WT-HLE B-3 cells; (b) HLE B-3 cells 2 weeks after hGSTA4 transfection; (c) growth arrest and enlargement of HLE B-3 cells 4 weeks after transfection; (d) budding of rounded cells from giant cells 8 weeks after transfection; (e) anchorage-independent growth of transformed rounded cells (B) Transfection of HLE B-3 with WT-hGSTA4 and Y212F mutant hGSTA4 (mut-hGSTA4) with no activity towards 4-HNE: (a-p) a typical phase contrast micrograph of transformed cells after transfection with WT-hGSAT4; (a-f) fluorescence micrograph showing expression of WT-hGSTA4-4 protein in transformed cells detected immunohistologically using hGSTA4-4 antibodies; (m-p) phase contrast micrograph of cells 8 weeks after transfection with mut-GSTA4; (m-f) fluorescence micrograph showing expression of mut-hGSTA4-4 protein in transfected cells (C) 4-HNE levels in HLE B-3 cells (D) Expression of hGSTA4-4 protein in transfected cells as detected by Western blots: lane 1, cells transfected with hGSTA4 Y212F mutant; lane 2, WT-HLE B-3 cells; lane 3, cells transfected with hGSTA4; lane 4, positive control of hGSTA4-4 Details for transfection, immunofluorescence studies, Western blots and 4-HNE determination are given in Experimental procedures.

cells were monitored from 12 to 48 h after injection After

24 h, cells injected with active protein began to round up

and detach (Fig 2A, a-p and a-f), whereas those injected

with OG-dextran and inactive protein remained flat and

attached (Fig 2A, m-p and m-p & f) There was a significant

increase in the percentage (52.5 ± 5%; mean ± SD) of the

round cells in active protein-injected cells over the first 48 h

after injection (data not presented) In contrast, in the cells

injected with the mutant hGSTA4-4 protein, only

8 ± 2.5% of the cells were rounded, which was similar to

the level observed in OG-dextran mock-injected cells (data

not presented)

In the experiments for microinjecting WT-hGSTA4 and

the inactive mut- hGSTA4 (Y212F) cDNA into HLEB-3

cells, the expression vector of eGFP was used as a marker

for successful microinjection As shown in Fig 2B, a-p and

a-f, microinjection of WT-hGSTA4 cDNA led to

charac-teristic rounding and anchorage-independent growth within

24 h Cells microinjected with mut-hGSTA4 cDNA

main-tained their original phenotype and did not undergo any

change (Fig 2B, m-p and m-f) A small but clearly

noticeable number of cells underwent apoptosis after

microinjection The apoptotic cells could be identified, as they showed activation of caspase-3 detected by staining the cells with antibodies to cleaved caspase-3 (data not shown) and loss of fluorescence due to extrusion of cytoplasm These cells could be easily distinguished from the rounded, transformed cells As shown in Fig 2C, these cells were not fully rounded and showed only minimal fluorescence of eGFP which was prominent in rounded, transformed cells The percentages of unchanged flat cells, transformed rounded cells, and apoptotic cells after 24 h and 48 h of microinjection of WT-hGSTA4 and mut-hGSTA4 cDNA in

a typical experiment are given in Fig 2D Together, these results further indicate that overexpression of active hGSTA4-4 is required for phenotypic transformation

Only GST isozymes that have high catalytic efficiency with 4-HNE have transforming activity

To further establish that high hGSTA4-4 activity was required for its transforming activity, we microinjected four different GST isozymes into HLE B-3 cells For these experiments, two GST isozymes with high activity and two

Fig 2 Microinjection of active WT-hGSTA4-4, inactive mutant hGSTA4-4 recombinant protein or the respective expression vector into HLE B-3 cells (A) (a-p) Phase contrast micrograph of a typical transformed HLE B-3 cell 24 h after cytosolic microinjection with WT-hGSTA4-4 protein; (a-f) fluorescence micrograph of same cell showing fluorescent marker, OG-dex, coinjected with WT-hGSTA4-4; (m-p) phase contrast micrograph

of a typical HLE B-3 cell after microinjection with inactive mut-hGSTA4-4 protein; (m-p&f) combined phase contrast micrograph and fluorescence micrograph of same cell indicating delivery of OG-dex marker Bar denotes 30 lm (B) (a-p&f) Combined phase contrast micrograph and fluorescence micrograph of a typical transformed HLE B-3 cell 24 h after nuclear microinjection with WT-hGSTA4 and the marker eGFP cDNAs; (a-f) fluorescence micrograph of same HLE B-3 cell (fluorescence represents expression of eGFP); (m-p) phase contrast micrograph of typical HLE B-3 cells 24 h after microinjection with inactive mut-hGSTA4 and eGFP cDNAs; (m-f) fluorescent micrograph of same cells (C) A small fraction of microinjected cells undergo apoptosis (p) phase contrast micrograph of cell undergoing apoptosis; (f) fluorescence micrograph of same cell These cells could be distinguished from the transformed cells as they were not fully rounded and expression of fluorescent marker eGFP was minimal (D) Quantitation of transformed (unfilled bars), nontransformed (grey bars) and apoptotic cells (black bars) after microinjection with WT-hGSTA4 or mut-hGSTA4 expression vectors Details of microinjection and immunofluorescence experiments are given in Experimental rocedures.

with minimal activity toward 4-HNE were selected Mouse

enzyme mGSTA4-4 [14] and Drosophila enzyme

DmG-STD1-1 [17] are known to have high activities for 4-HNE

(specific activities: 65 UÆmg)1and 32 UÆmg)1, respectively)

On the other hand, human enzymes hGSTA1-1 and

hGSTP1-1 have minimal activity towards 4-HNE [27]

These results showthat mGSTA4-4 and DmGSTD1-1

(Fig 3) trigger transformation and hGSTA1-1 and

hGSTP1-1 (Fig 4) do not A phase contrast micrograph

and fluorescent micrograph of a typical transformed cell

24 h after microinjection of mGSTA4-4 or DmGSTD1-1

are presented in Fig 3A and Fig 3B, respectively Results

presented in Fig 3C indicate that most microinjected cells

are transformed within 48 h In contrast, results presented

in Fig 4 showthat cells microinjected with either

hGSTA1-1 or hGSTPhGSTA1-1-hGSTA1-1 retain their original phenotype and do not

undergo transformation These results further support the

idea that the ability of the GST isozymes to induce

transformation is dependent on their ability to conjugate

4-HNE to GSH Furthermore, these results argue against

the possibility of a nonspecific effect of hGSTA4-4 causing

the transformation

Effect of hGSTA4-4 overexpression in the CCL-75 cell line

The effect of hGSTA4-4 overexpression was also examined

in a human lung fibroblast cell line, CCL-75, a nonviral

transformed adherent cell line with a finite lifetime of

50 ± 10 population doublings [28] In these experiments, when CCL-75 cells were microinjected with active and mutant hGSTA4-4 proteins in a manner similar to WT-HLE B-3 cells, comparable results were observed (Fig 5A– D) Interestingly, cell rounding was observed in CCL-75 cells 48 h after the microinjection of active protein, a delay

of nearly 24 h compared with WT-HLE B-3 cells The reasons for this time lag are not clear These results also showthat direct injection of active hGSTA4-4 protein or its cDNA into attached cells causes a characteristic trans-formed phenotype and further suggest that overexpression

of hGSTA4-4 leading to such transformation may be a generalized phenomenon

Effect of hGSTA4-4 on key cell-cycle genes

We also studied the effects of the hGSTA4-4 expression on some of the key genes involved in cell-cycle regulation and apoptosis In the hGSTA4-transfected and phenotypically transformed, anchorage-independent HLE B-3 cells, we found upregulation of transforming growth factor, cyclin-dependent kinase 2, protein kinase C bII, and extracellular regulatory stress kinase vs downregulation of p53 (Fig 6) These observations are consistent with the idea that lowering the intracellular concentrations of 4-HNE upreg-ulated genes involved in promotion of proliferation and

Fig 3 Cytoplasmic microinjection of recombinant mGSTA4-4 and DmGSTD1-1 in HLE B-3 cells Cells were microinjected with the respective protein as described in Experimental procedures and scored 24 h and 48 h after injection (A) (p) Phase contrast micrograph of a typical transformed cell 24 h after microinjection with mGSTA4-4; (f) fluorescence micrograph of same cell showing fluorescent marker OG-dex coinjected with mGSTA4-4 (B) (p) Phase contrast micrograph of a typical transformed cell 24 h after microinjection with DmGSTD1-1; (f) fluorescence micrograph of same cell showing fluorescent marker OG-dex coinjected with DmGSTD1-1; (C) Bar graph showing percentage of nontransformed (flat; black bars), transformed (rounded; light grey bars) and apoptotic (dark grey bars) cells 24 h and 48 h after microinjection.

downregulated genes (such as p53) that control the cell cycle

and are pro-apoptotic [29]

Discussion

Previous studies strongly suggest that intracellular 4-HNE

can influence signaling mechanisms and, depending on its

concentration, can promote apoptosis [3,6,7,12],

differenti-ation [12], or proliferdifferenti-ation [11,12] of cells GSTs in general,

and hGSTA4-4 and hGST5.8 in particular, are the major

4-HNE-metabolizing enzymes in humans [16] Dramatic

phenotypic transformation of attached cells on transfection

with hGSTA4-4 into smaller rounded immortalized cells

which grow rapidly in suspension is surprising but it seems

to be consistent with numerous previous studies suggesting

that 4-HNE is involved in cell-cycle signaling mechanisms

Our results showthat transfection or microinjection of cells

with enzymatically active hGSTA4-4 causes the emergence

of the transformed phenotype, whereas hGSTA4-4(Y212F),

a mutant with decreased activity for 4-HNE [13], is unable

to transform cells This result provides a reasonable basis for

proposing the hypothesis that the observed transformation

of HLE B-3 and CCL-75 cells is a consequence of its

conjugation of 4-HNE, rather than being linked to other

possible effects of hGSTA4-4, such as a hypothetical direct

binding to signaling kinases, as has been described for

Pi-class GSTs [30]

To test further the hypothesis that 4-HNE conjugation

is relevant to cellular transformation, we microinjected

cells with two additional GSTs which are known to metabolize 4-HNE but are structurally distinct and phylo-genetically distant from hGSTA4-4 The murine enzyme mGSTA4-4 has a catalytic efficiency for 4-HNE that is similar to that of hGSTA4-4 [14] However, antibodies against one enzyme do not cross-react with the other [16], indicating that at least parts of the surface of the two proteins differ substantially from each other The second microinjected protein was DmGSTD1-1 from D melano-gaster[31] This Delta-class insect GST also has a relatively high catalytic efficiency for 4-HNE conjugation [17] Insects and mammals diverged at least 600 million years ago [32], and hGSTA4-4 is only 22%/40% identical/similar to DmGSTD1-1 Thus, it is unlikely that DmGSTD1-1 could replace hGSTA4-4 in any putative regulatory protein– protein interactions in which hGSTA4-4 may be involved Our studies clearly showthat microinjection of hGSTA4-4, mGSTA4-4, and DmGSTD1-1 triggers cell transformation whereas microinjection of hGSTA4-4(Y212F), hGSTA1-1, and hGSTP1-1 does not The three proteins able to transform cells are structurally dissimilar but are all efficient

at conjugating 4-HNE, whereas those that lack 4-HNE-conjugating activity also fail to transform cells, even if they are structurally almost identical with an active enzyme, as in the case of hGSTA4-4(Y212F) Together, these results point to the conjugative metabolism of 4-HNE

as the common denominator and the causative principle

in the transformation process, and suggest that the level

of 4-HNE or its glutathione conjugate (GS-HNE) is the

Fig 4 Cytoplasmic microinjection of recombinant hGSTA1-1 and hGSTP1-1 in HLE B-3 cells (A) Typical cells 48 h after microinjection of hGSTA1-1 (p) Phase contrast micrograph showing that cells maintained their original phenotype; (f) fluorescence micrograph of same cells showing OG-dex fluorescent marker (B) Typical cells 48 h after microinjection of hGSTP1-1 (p) Phase contrast micrograph and (f) fluorescence micrograph of same cells (C) Bar graph showing percentage of nontransformed (flat; black bars) and transformed (rounded; light grey bars) cells

24 h and 48 h after microinjection.

most likely effector of the cell transformation we observed.

This is consistent with a lower 4-HNE level in cells

overexpressing hGSTA4-4 but not in cells overexpressing

the mutant Y212F, which is ineffective in triggering

transformation

Although a hypothetical substrate other than 4-HNE,

perhaps another Michael acceptor, cannot be ruled out at

present, 4-HNE is the only currently known common

physiological substrate of proteins as different as

hGSTA4-4 and DmGSTD1-1 However, the correlation of hGSTA4-

4-HNE-conjugating activity with the ability to transform cells which

holds for six different proteins [4,

hGSTA4-4(Y212F), mGSTA4-4, DmGSTD1-1, hGSTA1-1, and

hGSTP1-1] indicates that a causal involvement of 4-HNE

in the mechanism of hGSTA4-4-mediated transformation

of HLE B-3 and CCL-75 cells provides the simplest

explanation of all the available experimental data

Binding of GSTs, particularly hGSTP1-1 with c-Jun N-terminal kinase (JNK), modulates stress-mediated signa-ling for apoptosis In monomeric form, hGSTP1-1 binds to JNK and inhibits its activation, but under conditions of stress such as exposure to UV or H2O2 treatment, it oligomerizes and dissociates from the JNK complex leading

to abrogation of JNK inhibition [30] Such interactions of hGSTA4-4 with JNK or other key kinases may also be considered as the mechanistic basis for the observed transformation However, the inability of mutant hGSTA4-4(Y212F) to induce transformation argues against such a possibility because GSTP1-1 with a mutation in its active site is still able to prevent JNK activation An effector domain critical for its binding to JNK (residues 191–201) has been identified in hGSTP1-1 [33], and it seems unlikely that mutation of a single active-site residue (Y212F) would abrogate the binding of hGSTA4-4 to kinases

Fig 5 Microinjection of WT-hGSTA4-4, mut-hGSTA4-4 recombinant protein or respective expression vector in CCL-75 cells (A) (a-p) Phase contrast micrograph of a typical transformed CCL-75 cell 48 h after cytoplasmic microinjection with WT-hGSTA4-4 recombinant protein and OG-dex marker; (a-f) fluorescence micrograph of same cell (fluorescence indicates delivery of the marker OG-dex); (b-p) phase contrast micrograph

of a typical transformed CCL-75 cell 48 h after nuclear microinjection of expression vectors of WT-hGSTA4 and eGFP cDNAs; (b-f) fluorescence micrograph of same cell (fluorescence indicates expression of the marker eGFP); (c-p) phase contrast micrograph of a typical CCL-75 cell 48 h after cytoplasmic microinjection of mut-hGSTA4-4 protein and the marker OG-dex; (c-f) fluorescence micrograph of same cell (fluorescence indicates the marker OG-dex); (d-p) phase contrast micrograph of a typical CCL-75 cell 48 h after nuclear microinjection of mut-hGSTA4 and eGFP cDNAs; (d-f) fluorescence micrograph of same cell (fluorescence indicates expression of the marker eGFP) Details are provided in Experimental procedures Bar represents 30 lm (B–D) Bar graphs showing percentage of transformed (rounded; unfilled bars), unaffected (flat; dark grey bars) and apoptotic (black bars) cells after cytoplasmic microinjection of recombinant protein of active WT-hGSATA4-4 (B) or inactive mut-hGSTA4-4 (C) and nuclear microinjection of expression vectors (D).

Furthermore, GSTP1-1-mediated activation of JNK does

not appear to be applicable to all cell types because, in a

human lung fibroblast cell line, GSTP1-1 does not affect

JNK activation [34] We showthat neither hGSTP1-1 nor

hGSTA1-1 cause transformation but mouse and Drosophila

4-HNE-metabolizing GST isozymes (mGSTA4-4 and

DmGSTD1-1, respectively) showtransforming activity

comparable to that of hGSTA4-4 Thus, transformation

does not appear to be due to interaction of hGSTA4-4 with

signaling kinases but seems to be linked to its catalytic

ability to conjugate 4-HNE to GSH

GSTA4-4-catalyzed conjugation of 4-HNE to GSH

results in the formation of GS-HNE An increase in the

level of GS-HNE may also be a trigger of transformation

Previous studies have shown that overexpression of

4-HNE-metabolizing GST isozymes leads to accelerated formation

of GS-HNE in cells As confirmed by identification and

quantification of intact GS-HNE in the medium of cells

loaded with radioactive GS-HNE, most GS-HNE thus

formed is transported out of the cells through

ATP-dependent transport processes [6,7] However, a significant

proportion of intracellular GS-HNE can be metabolized to

the corresponding alcohol, glutathionyl dihydroxynonene

formed through NADPH-dependent reduction of GS-HNE

catalyzed by aldose reductase [35] In addition, GS-HNE

can also be converted into mercapturic acids, which can

then be x-hydroxylated by CYP-450 [36] to yield more

hydrophilic products The possibility of GS-HNE or its

metabolites being involved in the mechanisms of the

observed transformation phenomenon is not ruled out by

the present studies Aldose reductase, which can reduce

GS-HNE, has been shown to mediate mitogenic signaling in

vascular smooth muscle cells [37], and its inhibitors have

been shown to inhibit tumor necrosis factor-a-induced

apoptosis and caspase-3 activation [38] Channeling of 4-HNE towards accelerated formation of GS-HNE in hGSTA4-4-overexpressing cells may abruptly change the overall physiologic homeostasis of 4-HNE, GS-HNE, and its metabolites maintained by a co-ordinated action of 4-HNE-metabolizing enzymes including GSTs [16], aldose reductase [35], aldehyde dehydrogenase [39], and transport-ers of GS-HNE [6,7] According to this interpretation, it is not just the concentration of 4-HNE but also changes in the homeostasis of 4-HNE and its metabolites that provides the mechanistic basis for the transformation This possibility needs to be explored in future studies

Our results showthat some of the more prominent genes suggested to be involved in promoting proliferation are upregulated in hGSTA4-4-overexpressing HLE B-3 cells This, along with almost complete suppression of p53, may account for the observed threefold higher growth rate of the transformed cells Our studies are limited to evaluating the expression of only a fewkey genes known to be involved in cell-cycle events An assessment of the effect of transfection with hGSTA4-4 and 4-HNE depletion on global gene expression using cDNA microarrays is planned for the future Taken together with the results of previous studies showing that at higher concentrations, 4-HNE causes apoptosis [3,6,7,12] and differentiation [12,40], the present results suggest that the intracellular level of 4-HNE may be one of the determinants for leading cells towards pathways for transformation, differentiation, proliferation, or apop-tosis

The mechanisms through which 4-HNE affects signaling processes in a concentration-dependent manner are obscure and appear to be complex 4-HNE is a strong electrophile which reacts with nucleophilic groups of proteins [41,42], nucleic acids [43,44], and lipids [45] It interacts with thiols

Fig 6 Expression of genes known to be involved in cell-cycle regulation (A,B) HLE B-3 cells (1 · 10 6 ) transfected with wild-type, empty vector or hGSTA4 were lysed in RIPA buffer containing protease inhibitor cocktail, 1 m M phenylmethanesulfonyl fluoride and 2 m M sodium orthovanadate The cell extracts were centrifuged at 15 000 g at 4 C Supernatant containing 100 lg protein was loaded in each well and subjected to Western blot analysis using antibodies against proteins identified in the left hand margins of (A) and (B) Lanes 1, 2, and 3 in both panels represent extracts of HLE B-3 cells transfected with wild-type, empty vector and hGSTA4, respectively (C) For comparison of expression of extracellular signal regulated kinase 1/2, cells (1 · 10 6 ) from clone C4 (A4-4) and empty-vector-transfected (VT) HLE B-3 cells were serum starved for 24 h in separate Petri dishes and then treated with serum-containing medium (10%) for different times The cells were centrifuged at 654 g (5 min), their extracts were prepared in RIPA buffer as described in Experimental procedures, and Western blot analyses were performed using antibodies against extracellular signal regulated kinase 1/2 Lane 1, extract of cells before serum stimulation; lanes 2–6, extracts of the cells after treatment with 10% serum for 2, 5, 10, 15 and 30 min, respectively.