• 4-hydroxynonenal;
  • aging;
  • C. elegans;
  • longevity regulation;
  • molecular biology of aging;
  • oxidative stress;
  • stress resistance;
  • transgenes


  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgments
  8. References

Caenorhabditis elegans expresses a glutathione transferase (GST) belonging to the Pi class, for which we propose the name CeGSTP2-2. CeGSTP2-2 (the product of the gst-10 gene) has the ability to conjugate the lipid peroxidation product 4-hydroxynonenal (4-HNE). Transgenic C. elegans strains were generated in which the 5′-flanking region and promoter of gst-10 were placed upstream of gst-10 and mGsta4 cDNAs, respectively. mGsta4 encodes the murine mGSTA4-4, an enzyme with particularly high catalytic efficiency for 4-HNE. The localization of both transgenes was similar to that of native CeGSTP2-2. The 4-HNE-conjugating activity in worm lysates increased in the order: control < mGsta4 transgenic < gst-10 transgenic; and the amount of 4-HNE-protein adducts decreased in the same order, indicating that the transgenic enzymes were active and effective in limiting electrophilic damage by 4-HNE. Stress resistance and lifespan were measured in transgenic animals (five independent lines each) and were compared with two independent control lines. Resistance to paraquat, heat shock, ultraviolet irradiation and hydrogen peroxide was greater in transgenic strains. Median lifespan of mGsta4 and gst-10 transgenic strains vs. control strains was increased by 13% and 22%, respectively. In addition to the cause–effect relationship between GST expression and lifespan observed in the transgenic lines, correlative evidence was also obtained in a series of congenic lines of C. elegans in which lifespan paralleled the 4-HNE-conjugating activity in whole-animal lysates. We conclude that electrophilic damage by 4-HNE may contribute to organismal aging.


  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgments
  8. References

The complete oxidation of food using molecular oxygen as the ultimate electron acceptor, with the concomitant capture of free energy released in the process, was a major evolutionary innovation. It is likely that the switch from fermentative to oxidative metabolism permitted many salient features of extant life forms, including size, mobility, predation, nervous system function and perhaps even multicellularity (Lane, 2002). These advances came, however, at a price. A small but significant fraction of electrons flowing through the mitochondrial respiratory chain prematurely leaks to molecular oxygen and forms the superoxide radical and other ‘reactive oxygen species’ (ROS) (Cadenas & Davies, 2000). Other biochemical and physical processes also generate ROS, either as byproducts or as part of their function. Mitochondria are usually the predominant ROS producers. In contrast to the dioxygen molecule, ROS and their downstream products such as peroxynitrite (Crow, 2000) are highly reactive and readily damage cellular constituents. If unopposed by appropriate defense mechanisms, ROS create a state of oxidative stress that has been implicated in the etiology of a variety of diseases, most of which are degenerative (Scandalios, 1997; Mates et al., 1999; McEwen et al., 2005), and in aging (Harman, 1956, 2003; Beckman & Ames, 1998; Cadenas & Davies, 2000).

The considerable reactivity of some ROS leads to their low selectivity. Although all cellular constituents can be targets of ROS, radical-initiated reaction with polyunsaturated fatty acids is unique because it results in a chain reaction (Gutteridge & Halliwell, 1990). The resulting amplification of the original insult may be of biological significance. The products of the process, lipid hydroperoxides, change the properties of membranes, act as oxidants or can be converted (Schneider et al., 2001, 2004; Sun & Salomon, 2004) to a variety of α,β-unsaturated aldehydes of which 4-hydroxynon-2-enal (4-HNE) is the most abundant and prototypical example. At physiological concentrations, 4-HNE is a signaling molecule which modulates a variety of fundamental biological processes (reviewed by Petersen & Doorn, 2004). The compound becomes toxic at higher levels owing to its electrophilic character (Esterbauer et al., 1991). It forms Michael adducts with nucleophilic centers on proteins and nucleic acids (Petersen & Doorn, 2004), can cross-link proteins and contributes to the formation of lipofuscin (Xu & Sayre, 1998; Xu et al., 2000), among other properties.

The metabolism of 4-HNE can be oxidative (oxidation of the aldehyde function to a carboxylic acid), reductive (reduction of the aldehyde function, or reduction of the double bond in position 2) or conjugative (Michael adduct formation with glutathione) (reviewed by Petersen & Doorn, 2004). Conjugation with glutathione is probably the predominant pathway in most tissues (Petersen & Doorn, 2004) and, although catalysed by most glutathione transferases (GSTs), it is most efficiently carried out by specialized enzymes of this superfamily exemplified by the murine mGSTA4-4 (Zimniak et al., 1994). We have also identified GSTs capable of efficient conjugation of 4-HNE in invertebrates (Singh et al., 2001; Sawicki et al., 2003), including Caenorhabditis elegans (Engle et al., 2001). The mammalian enzymes with high catalytic efficiency for 4-HNE belong to the Alpha class of GSTs, whereas those in Drosophila melanogaster are classified as Sigma and Delta, and the nematode GST is of the Pi class (Engle et al., 2001). The emergence of this activity in evolutionarily distant GST classes indicates convergent evolution (Sawicki et al., 2003). This, together with the wide species distribution, suggests an important physiological role for 4-HNE conjugation, perhaps common to all aerobic organisms.

The chain reaction character of lipid peroxidation and the ensuing amplification of the initiating oxidative stimulus raise the intriguing question of whether this mechanism could contribute to organismal aging, as would be predicted from the free radical theory of the aging process (Harman, 1956). If the answer is yes, two aspects of the process may be biologically relevant. One is the above-mentioned increase in the number of potentially damaging molecules, inherent in the chain reaction, which could overwhelm the existing defense mechanisms. The other aspect is a shift in the chemical characteristics of the insult: an oxidative or radical-mediated stress would be converted into 4-HNE-mediated electrophilic stress. 4-HNE is relatively selective for nucleophilic centers, especially soft nucleophiles such as those present in proteins, whereas the most damaging ROS, the hydroxyl radical, is very short-lived and is thus non-selective. Formation of 4-HNE adducts frequently alters the function of proteins, and many known regulatory proteins are targets of 4-HNE modification (Dianzani, 2003; Forman et al., 2003; Petersen & Doorn, 2004). It is therefore conceivable that 4-HNE, although sharing with ROS the ability to inactivate cellular components stochastically, has an additional, perhaps more insidious effect. 4-HNE could modify key regulatory proteins and alter signaling cascades (Leonarduzzi et al., 2004) or inappropriately realign metabolism. This could cause cell loss or tissue dysfunction more effectively than accumulation of random damage.

Here we provide evidence in support of the hypothesis that electrophilic stress is indeed a contributor to aging in the model organism C. elegans. Mitigation of 4-HNE-mediated electrophilic damage by overexpression of 4-HNE-metabolizing enzymes led to an extension of both median and maximal lifespan of the nematode. This finding contributes to a better understanding of the molecular basis of the aging process and its relationship to stress resistance.


  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgments
  8. References

Isolation and characterization of CeGSTP2-2

A GST with glutathione-conjugating activity for 4-HNE was isolated from C. elegans by a combination of glutathione (GSH)-affinity chromatography and isoelectric focusing (Engle et al., 2001). The major N-terminal sequence of the protein obtained by Edman microsequencing matched a single conceptually translated protein among the 44 GSTs and seven GST-like sequences listed in release WS135 of WormBase ( The protein, denoted as the gst-10 gene product in WormBase, was identified as a Pi-class GST on the basis of conserved amino acid residues and an analysis of intron–exon boundaries (Engle et al., 2001). Because another GST (the gst-1 gene product) has been previously assigned to the Pi class (Weston et al., 1989), we propose the name CeGSTP2-2 for the gst-10 gene product, according to the convention developed for GST nomenclature (Mannervik et al., 1992; Chelvanayagam et al., 2001).

The kinetic parameters of CeGSTP2-2 with 4-HNE as substrate are listed in Table 1. The catalytic efficiency of CeGSTP2-2 toward this electrophilic lipid peroxidation product is approximately three-fold lower than that of a mammalian Alpha-class GST, mGSTA4-4 (an enzyme highly specialized for 4-HNE metabolism; see Zimniak et al., 1994). However, the catalytic efficiency of CeGSTP2-2 for 4-HNE is similar to that of other invertebrate GSTs that carry out this reaction (Singh et al., 2001; Sawicki et al., 2003). CeGSTP2-2 has low activity with acrolein but is relatively efficient in catalysing the conjugation of trans-non-2-enal (Table 1). For comparison, we measured the corresponding activities of the C. elegans gst-4 gene product, also denoted as GST-p24 (Leiers, 2002; Leiers et al., 2003) (Table 1). The specific activity with trans-non-2-enal was almost two orders of magnitude higher for CeGSTP2-2 than for the gst-4 gene product (Table 1). For unknown reasons, however, in our hands the activities of the gst-4 gene product for several substrates were significantly lower (Table 1) than previously reported (Leiers, 2002; Leiers et al., 2003).

Table 1.  Enzymatic activity of bacterially expressed CeGSTP2-2, gst-4 gene product and, for comparison, murine mGSTA4-4
Enzyme/substrateSpecific activity (µmol mg−1 min−1)Vmax (µmol mg−1 min−1)KMm)kcat/KM (s−1 m−1)
  • Specific activities are given as means ± SD, and kinetic parameters are given as estimates ± asymptotic SE determined by least-squares fitting of a Michaelis–Menten hyperbola to the experimental data.

  • *


  • Protein purified by Ni2+ chelation chromatography (see Experimental procedures).

 CDNB*5.5 ± 0.7   
 4-HNE7.3 ± 0.510.3 ± 0.350 ± 40.17 × 106
 trans-non-2-enal (100 µm)10.8 ± 1.0   
 acrolein1.7 ± 0.1   
 cumene hydroperoxide0.07 ± 0.01   
gst-4 gene product
 CDNB1.1 ± 0.1   
 4-HNE1.6 ± 0.03   
 trans-non-2-enal (100 µm)0.12 ± 0.01   
 trans-non-2-enal (1 µm)0.03 ± 0.01   
 cumene hydroperoxide0.87 ± 0.01   
 4-HNE260 ± 16427 ± 2470 ± 10 5.1 × 106

During purification of CeGSTP2-2 from C. elegans lysate, only 36% of the total 4-HNE-conjugating activity was retained on the GSH affinity column (data not shown). A similar fraction of activity was depleted from C. elegans lysate by anti-CeGSTP2-2 antibody (see below). By contrast, bacterially expressed CeGSTP2-2 was retained on the GSH column essentially quantitatively, attesting to a high binding affinity and capacity of GSH-agarose for the enzyme. This indicates that CeGSTP2-2 is responsible for approximately one-third of the total 4-HNE-conjugating activity in adult worms.

Generation of antibodies against CeGSTP2-2

Polyclonal antibodies against bacterially expressed CeGSTP2-2 were raised in chicken. On Western blots of crude C. elegans lysates, the antibodies, but not the pre-immune IgY fraction, recognized a band identical in size to CeGSTP2-2 protein heterologously expressed in Escherichia coli, and a second band at 13 kDa (Fig. 1). Under the same conditions, the antibodies did not recognize hGSTP1-1, hGSTA1-1 and hGSTA4-4 (data not shown). The specificity of the antibody was further checked by immunoprecipitation followed by mass spectrometry. C. elegans lysates were treated with the antibody, and the immune complexes were recovered using bead-bound secondary antibody and separated by SDS-PAGE. Mass spectrometric analysis of tryptic peptides derived from a broad band migrating at approximately 25 kDa revealed the expected presence of IgY light chain, and of the following peptides: LLFLDNGIK, AVPQLYYFTIR and VDGQEIVQTGAIMR. The three peptides give 16% sequence coverage and a Mascot score (see Experimental procedures) of 104. Each of the three peptides yields a perfect match to two C. elegans proteins: CeGSTP2-2 and predicted protein F56A4.3. The latter is a 105-amino-acid protein with a calculated molecular weight of 12.2 kDa. Its existence is confirmed by the presence of an expressed sequence tag (yk773h08.5 in WormBase). CeGSTP2-2 and F56A4.3 share the sequence of 78 N-terminal amino acids, which include the three peptides listed above. The antibody may thus recognize F56A4.3 in addition to CeGSTP2-2, and the 13-kDa band observed on Western blots of C. elegans lysates (Fig. 1) could represent F56A4.3. Possible functions of F56A4.3 were not investigated. However, a protein that corresponds only to the N-terminal domain of a typical GST would not be expected to have GST activity as it would lack the electrophile binding site (Armstrong, 1997). This, together with the lack of other GSTs detectable by mass spectrometry sequencing, indicates that the antibody is specific for a single functional C. elegans GST, CeGSTP2-2.


Figure 1. Characterization of anti-CeGSTP2-2 antibody. Antibodies against bacterially expressed CeGSTP2-2 were raised in chicken and tested on Western blots (12% SDS-PAGE, Tris–glycine buffer system). Lanes 1 and 3: 25 ng CeGSTP2-2 expressed in E. coli and purified by GSH affinity chromatography. Lanes 2 and 4: 40 µg C. elegans (Bristol-N2) lysate. Lanes 1 and 2: pre-immune chicken IgY. Lanes 3 and 4: anti-CeGSTP2-2 IgY.

Download figure to PowerPoint

Immunodepletion of CeGSTP2-2 in C. elegans lysates

Approximately two-thirds of the total 4-HNE-conjugating activity of wild-type N2 C. elegans lysates were not retained on GSH affinity columns (see above). This suggests that one or more of the C. elegans GSTs other than CeGSTP2-2 may have 4-HNE-conjugating activity. It was recently reported that the gst-4 gene product has activity toward trans-non-2-enal (Leiers et al., 2003), and we found that this enzyme also conjugates 4-HNE, albeit with an almost five-fold lower activity than CeGSTP2-2 (Table 1). To estimate what fraction of the worm's 4-HNE-conjugating activity is due to CeGSTP2-2, C. elegans lysates were immunodepleted with anti-CeGSTP2-2 antibody. As shown in Fig. 2, approximately 20–25% of the 4-HNE-conjugating activity could be depleted by immune IgY raised against bacterially expressed, purified CeGSTP2-2, but not by pre-immune IgY. This result is in reasonable agreement with the fraction of activity that binds to a GSH affinity column, and indicates that the balance of 4-HNE-conjugating activity can be attributed to GST(s) which immunologically do not cross-react with CeGSTP2-2, and which probably belong to a class that fails to bind effectively to GSH-agarose.


Figure 2. Immunodepletion of 4-HNE-conjugating activity in C. elegans lysate by anti-CeGSTP2-2 antibody. Bristol-N2 worm lysates were incubated with various amounts of pre-immune (○) and anti-CeGSTP2-2 (•) chicken IgY, and the immune complexes were pelleted with agarose-bound anti-IgY (see Experimental procedures for details). 4-HNE-conjugating activity was measured in the supernatants.

Download figure to PowerPoint

Expression of CeGSTP2-2 and mGSTA4-4 transgenes

Western blot analysis demonstrated the absence in pha-1 control worms of material immunologically cross-reactive with mGSTA4-4. By contrast, mGSTA4-4 was expressed at similar levels in several independent mGsta4 transgenic lines (Fig. 3A). CeGSTP2-2 was detectable by Western blotting in pha-1 worms, but was present at higher levels in gst-10 transgenic animals (Fig. 3B). Quantitation of CeGSTP2-2 levels has shown that, in the five transgenic lines represented in Fig. 3(B), the protein was overexpressed by a factor of 2.4 ± 0.6 relative to control worms.


Figure 3. Expression of mGSTA4-4 and CeGSTP2-2 in mGsta4 and gst-10 transgenic worms. (A) Western blot (probed with anti-mGSTA4-4 antibody). Lane 1: 25 ng purified mGSTA4-4; lane 2: pha-1 control; lanes 3–5: three independent lines of mGsta4 transgenic C. elegans. For lanes 3–5, total worm lysates (2 mg protein each) were applied to a GSH affinity column, the column was eluted with GSH (Simons & Vander Jagt, 1977), and half of the eluate (i.e. GSTs isolated from 1 mg lysate protein) was loaded on the gel (SDS−12% polyacrylamide gel, Tris–glycine buffer system). (B) Western blot (probed with anti-CeGSTP2-2 antibody). Lane 1: pha-1 control; lanes 2–6: five independent lines of gst-10 transgenic C. elegans. In lanes 1–6, 25 µg each of worm lysate was loaded on the gel (precast SDS/4–12% polyacrylamide gradient gel in Bis–Tris buffer system).

Download figure to PowerPoint

4-HNE-conjugating activity in gst-10 and mGsta4 transgenic C. elegans

The expression or overexpression of mGSTA4-4 and CeGSTP2-2, respectively, in transgenic worms (Fig. 3) was accompanied by an increase of 4-HNE-conjugating activity over that measured in pha-1 controls (Table 2). The specific activity increased by 21% and 36% in mGsta4 and gst-10 transgenic animals, respectively. The increase was statistically significant in both cases (Table 2). The 36% increase in specific activity in gst-10 transgenic worms is consistent with the 2.4-fold increase in CeGSTP2-2 protein, given the fact that this enzyme accounts for approximately one-quarter to one-third of the total 4-HNE-conjugating activity (see Discussion).

Table 2.  Glutathione-dependent conjugation activity for 4-HNE in lysates of control and transgenic C. elegans
C. elegans genotype4-HNE-conjugating activity* (µmol mg−1 min−1)
  • *

    By anova, followed by multiple comparisons with the Tukey–Kramer correction, pha-1 differs from mGsta4 (P = 0.03), pha-1 differs from gst-10 (P < 0.0005), and the difference between mGsta4 and gst-10 does not reach statistical significance (P = 0.12).

pha-1 control0.217 ± 0.020
mGsta40.263 ± 0.052
gst-100.295 ± 0.043

Level of 4-HNE-protein adducts in gst-10 and mgsta4 transgenic worms

The level of 4-HNE adducts varied considerably between individual measurements (Fig. 4). However, after adjusting for interexperimental variability, the level of 4-HNE-protein adducts was significantly higher in pha-1 controls than in either of the transgenic lines. Moreover, the 4-HNE-conjugating activity in pha-1, mGsta4 and gst-10 worms inversely correlated with the amount of accumulated 4-HNE-protein adducts in these animals (Fig. 4). This correlation demonstrates that the overexpressed enzymes are functional in situ, and is consistent with the expected biochemical function of both enzymes in limiting 4-HNE-mediated electrophilic modifications and presumably damage to cellular proteins.


Figure 4. 4-HNE–protein adducts in control and transgenic C. elegans. Adducts were determined in pha-1 controls and in mGsta4 and gst-10 transgenic worms by ELISA (see Experimental procedures) in six separate experiments. Each individual experiment is denoted by a different plotting symbol, and measurements for the three genotypes obtained within each experiment are connected. The three groups are positioned on the abscissa according to their average 4-HNE-conjugating activity taken from Table 2. After adjusting for variability between experiments by General Linear Model anova or by multiple regression, the amount of 4-HNE adducts was found to be unequal for the three groups of worms (P = 0.0006). Multiple comparisons with Tukey–Kramer correction have further shown that pha-1 differs in the level of 4-HNE adducts from both mGsta4 (P = 0.013) and gst-10 (P = 0.0006), but that the difference between mGsta4 and gst-10 does not reach statistical significance (P = 0.15).

Download figure to PowerPoint

Localization of CeGSTP2-2

Transgenic worms were generated which express green fluorescent protein (GFP) under the control of the autologous gst-10 promoter. A punctate distribution of the reporter protein was observed that was most prominent in a small set of cells at the mouth and in the head (adjacent to the pharyngeal bulbs) of the worm; intense fluorescence was also observed at the tail (Fig. 5). Two promoter fragments, 3.2 and 2.4 kb in length, yielded an identical pattern of GFP expression. After these experiments were completed, the tissue distribution of GFP expressed under control of the gst-10 promoter, reported in an online database (R. Johnsen et al., British Columbia C. elegans Gene Expression Consortium, funded by Genome Canada and Genome British Columbia and accessible at, included head neurons among the expression sites. In the latter experiments, GFP was expressed under the control of a 2.9-kb promoter fragment, the 3′ end of which abuts the translation start codon of gst-10.


Figure 5. Tissue localization of autologous CeGSTP2-2 and heterologous GSTs driven by the gst-10 promoter. (A) expression of gst-10::GFP fusion (composite from two different fields of view). (B–E) Indirect immunofluorescence. (B) gst-10 transgenic animals with pre-immune antibody. (C) pha-1 animals with anti-CeGSTP2-2 antibody. (D) gst-10 transgenic animals with anti-CeGSTP2-2 antibody. (E) mGsta4 transgenic animals with anti-mGSTA4-4 antibody. In all cases, the fluorescence image (green) is overlaid on a bright-field image. For all fluorescence images, the exposure time was 1000 ms. Arrows labeled ‘H’ denote head localization, and those labeled ‘T’ denote tail localization.

Download figure to PowerPoint

An expression pattern similar to that of the reporter protein GFP was also obtained by indirect immunostaining of wild-type or gst-10 transgenic worms with anti-CeGSTP2-2 antibody (Fig. 5C,D, respectively), as well as mGsta4 transgenic worms with anti-mGSTA4-4 antibody (Fig. 5E). Both transgenes were driven by a fragment of the gst-10 promoter (see Experimental procedures). These results cross-verify the methods used, and confirm the selectivity of the gst-10 promoter for a relatively small subset of cells, which probably includes head and tail neurons.

Stress resistance of GST-expressing transgenic animals

C. elegans expressing either the gst-10 or the mGsta4 transgene had an increased resistance to hydrogen peroxide, paraquat, heat shock and ultraviolet (UV) irradiation at 254 nm as compared with pha-1 controls (Fig. 6). The increase in stress resistance was relatively small, typically 10–30% gain in median survival time, but was robust across independent transgenic lines (Fig. 6) and between experiments; and was statistically significant, except for H2O2 treatment of animals expressing mGSTA4-4 for which significance was marginal (Fig. 6A). With the exception of heat shock (Fig. 6E,F), the increase in stress resistance was more pronounced in gst-10 than in mGsta4 transgenic animals. This differential effect on stress resistance correlates with a greater conjugating activity for 4-HNE and lower level of 4-HNE-protein adducts in gst-10 as compared with mGsta4 transgenic animals (Table 2 and Fig. 4, respectively).


Figure 6. Enhanced stress resistance of mGsta4 and gst-10 transgenic C. elegans relative to controls. (A,C,E,G) mGsta4 transgenic animals; (B,D,F,H) gst-10 transgenic animals. Two independent lines of control pha-1 worms (closed symbols: •, ▪) and five independent lines each of mGsta4 and gst-10 transgenic animals (open symbols: ○, ▵, ▿, ◊, □) were exposed to H2O2 (A and B), paraquat (C and D), heat shock (E and F) and UV (G and H) as described in the Experimental procedures, and surviving worms were scored as a function of time. Not all C. elegans lines were used in all experiments. Because of the superimposition of the curves, individual symbols are not well resolved in the graph. Therefore, the number of transgenic and control lines used in each experiment is listed in the corresponding panels. The P-value given in each panel is the result of a log rank test to compare the stress resistance of pha-1 controls and transgenic worms shown in that panel.

Download figure to PowerPoint

Longevity of GST-expressing transgenic animals

The lifespan of C. elegans was extended by expression of mGSTA4-4 as well as by overexpression of the endogenous CeGSTP2-2 (Figs 7 and 8). The median lifespan of pha-1 control worms was 10.8 days at 25.5 °C, and was extended by 13% and 22% in C. elegans expressing mGSTA4-4 and overexpressing CeGSTP2-2, respectively (Table 3). Survival curves conformed approximately to the Gompertz function (r2 = 0.993, 0.984 and 0.992 for pha-1, mGsta4 and gst-10, respectively). The Gompertz parameters are listed in Table 3. Lifespan extension was statistically highly significant for both types of transgenic animals vs. controls, and gst-10 worms were longer-lived than mGsta4 worms at marginal statistical significance (see legend to Fig. 7 for further details).


Figure 7. Lifespan extension of mGsta4 and gst-10 transgenic C. elegans relative to control. Lifespan of pha-1 control (•), mGsta4 (▵), and gst-10 (○) transgenic animals was measured at 25.5 °C as described in the Experimental procedures. For illustrative purposes, non-linear fitting of the Gompertz function to the data points is represented as solid lines; maximum likelihood estimates of Gompertz parameters are listed in Table 3. The survival curves of pha-1 controls and of mGsta4 transgenic animals are different at P = 0.0024 (log rank test) and at P = 0.0027 (randomization log rank test using 106 Monte Carlo samples), and the survival curves of pha-1 controls and gst-10 transgenics are different at P < 10−9 and P < 10−5, respectively, in the two tests. Survival curves of gst-10 and mGsta4 transgenics are different from each other at P = 0.072 and P = 0.075, respectively, in the log rank and randomization log rank tests. All P-values include the Bonferroni correction for multiple comparisons.

Download figure to PowerPoint


Figure 8. (A,B) Pairwise comparison of lifespans of pha-1 controls (closed symbols) and gst-10 transgenic C. elegans (five independent lines; open symbols). A and B represent two independent experiments. (C,D) Pairwise comparison of lifespans of pha-1 controls (closed symbols) and mGsta4 transgenic C. elegans (five independent lines; open symbols). C and D represent two independent experiments.

Download figure to PowerPoint

Table 3.  Lifespan parameters of control and transgenic C. elegans
C. elegans genotypeGompertz parametersMedian lifespan (days)
  1. Maximum likelihood estimates of Gompertz parameters (listed with their lower and upper 95% confidence limits) and of median lifespans (± SE) were obtained from the data points shown in Fig. 7 using the program WinModest (Pletcher, 1999; Promislow et al., 1999).

pha-1 control0.0078 (0.0053, 0.0114)0.292 (0.261, 0.326)10.8 ± 0.2
mGsta40.0091 (0.0053, 0.0156)0.229 (0.193, 0.273)12.2 ± 0.1
gst-100.0059 (0.0040, 0.0088)0.248 (0.222, 0.276)13.2 ± 0.3

In addition to the experimental data shown in Fig. 7, in which all three genotypes (pha-1 control and the gst-10 and mGsta4 transgenic strains) were compared side by side, survival of the five available gst-10 lines was compared with that of the pha-1 control in two independent experiments (Fig. 8A,B), and the five available mGsta4 lines were compared with the control strain in two separate independent experiments (Fig. 8C,D). Because slight shifts in absolute survival times precluded direct averaging of data from the five independent experiments, all results were analysed together using a bivariate Cox regression model containing Experiment as a covariate in addition to Genotype as the treatment of interest. After covariate-adjusting for variability between experiments, the difference in survival curves between pha-1 controls and gst-10 transgenic animals was significant at P = 10−30, between pha-1 and mGsta4 at P = 10−8, and gst-10 and mGsta4 transgenic worms differed from each other in terms of survival curves at P = 10−5.

Correlation between lifespan and ability to conjugate 4-HNE in a series of C. elegans congenic lines

Two series of congenic lines were previously constructed in C. elegans, in order to pursue quantitative trait loci mapped initially for effects on longevity (Ayyadevara et al., 2001, 2003). These congenic lines demonstrated allelic effects on lifespan and resistance to specific stresses. The parental strain RC301 was used to derive lines SR101 and SR102, and the parental strain CL2a gave rise to lines SR704 and SR708. Each parental line with its derived congenic lines represents a set of naturally occurring polymorphic variants of genes within a region ranging approximately from 0.3 to 5 Mb, in an otherwise identical genetic background. Combined data for both sets of congenic lines showed a highly significant linear correlation (r2 = 0.96, P = 6 × 10−4) between the median lifespans and the specific activity of glutathione-dependent conjugation of 4-HNE in the corresponding worm lysates (Fig. 9).


Figure 9. Correlation between median lifespan and 4-HNE-conjugating activity in two series of C. elegans congenic lines. Two sets of congenic lines differing in stress resistance and longevity were selected as previously described (Ayyadevara et al., 2001, 2003). Specific activity of 4-HNE conjugation was measured in worm lysates as described in the Experimental procedures. Lines SR704 and SR708 were constructed in a CL2a strain background, whereas SR101 and SR102 have an RC301genetic background.

Download figure to PowerPoint


  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgments
  8. References

C. elegans has the ability to synthesize ω-6 polyunsaturated acids including arachidonic and linoleic acids (Watts & Browse, 2002), which are the biological precursors of 4-HNE (Esterbauer et al., 1991). In the present work, we have shown the existence of protein−4-HNE adducts (Fig. 4), thus demonstrating directly that 4-HNE is indeed formed in C. elegans. The manifold and complex physiological effects of 4-HNE (Dianzani et al., 1999; Nakashima et al., 2003; Carini et al., 2004; Petersen & Doorn, 2004) and the toxicity of the compound at high levels (Esterbauer et al., 1991) imply the existence of pathways to metabolize 4-HNE, both for signal termination and for detoxification. Indeed, we have previously identified a glutathione transferase in C. elegans, for which we propose the name CeGSTP2-2 for reasons outlined in the Results section, which is able to catalyse glutathione conjugation of 4-HNE (Engle et al., 2001). We now present evidence that this GST accounts for one-quarter to one-third of the 4-HNE-conjugating capacity of the worm, with the balance of activity probably attributable to GST(s) of other classes which are not retained on GSH affinity columns. However, measurable expression of CeGSTP2-2 is limited to a small subset of cells (Fig. 5). The resulting high local concentration of CeGSTP2-2 could play a significant role in the physiology of these cells and in their protection from oxidative or electrophilic damage.

To investigate the biological role of 4-HNE-conjugating activity in C. elegans, and specifically in the cells that express CeGSTP2-2, we created transgenic lines in which either the overexpression of the autologous CeGSTP2-2 or heterologous expression of murine mGSTA4-4, an enzyme with particularly high catalytic efficiency for 4-HNE, was driven by the autologous gst-10 promoter. We have verified, both by the use of the GFP reporter protein and by detection of the two transgenic GSTs through indirect immunofluorescence, that the length of the 5′ flanking fragments of the gst-10 gene used in the transgenic constructs was sufficient to ensure cell-specific transgene expression indistinguishable from that of the native gene. Thus, mGSTA4-4 and the additional CeGSTP2-2 in mGsta4 and gst-10 transgenic lines, respectively, retained the expression pattern of autologous CeGSTP2-2. It should be noted that all transgenes used herein were extrachromosomal episomes, so that variation in expression between worms could be attributed to somatic segregation of the transgene.

As described in the Results section, the anti-CeGSTP2-2 antibody used in the present work is likely to recognize, in addition to CeGSTP2-2 itself, the C. elegans protein F56A4.3, which consists of the N-terminal domain of CeGSTP2-2 fused to a short C-terminal peptide of 27 amino acids. The functions and expression pattern of this protein are unknown. However, the genomic sequences of gst-10 and f56a4.3 are identical for more than 25 kb upstream of their respective translation start codons, as determined by alignment of sequences deposited in WormBase, implying that f56a4.3 arose by a partial duplication of the gst-10 gene. As we verified that a 5′ flanking region of gst-10 much shorter than 25 kb is sufficient to restrict transgene expression to the native localization of CeGSTP2-2, we conclude that F56A4.3 and CeGSTP2-2 should be expressed in the same cells. Thus, F56A4.3 is unlikely to be a confounding factor in the immunolocalization experiments.

The overexpression of CeGSTP2-2 in gst-10 animals was 2.4 ± 0.6-fold greater than in controls (see Results). From this factor and from the estimate, based on retention on GSH columns and on immunodepletion, that CeGSTP2-2 accounts for 25–33% of 4-HNE-conjugating activity in wild-type worms, it can be calculated that the gst-10 transgenic animals should have a 4-HNE-conjugating activity in the range of 0.26–0.36 µmol mg−1 min−1, in good agreement with the experimental value of 0.295 ± 0.043 µmol mg−1 min−1 (Table 2). This indicates that the overexpressed CeGSTP2-2 is enzymatically fully active. Interestingly, lysates of mGsta4 transgenic worms have a 4-HNE-conjugating activity which is similar to, or perhaps slightly lower than, that of gst-10 transgenic animals (Table 2), in spite of a catalytic efficiency of mGSTA4-4 that is 30-fold higher than that of CeGSTP2-2 (Table 1). This suggests that the heterologous mGSTA4-4 protein may be poorly expressed or rapidly degraded in C. elegans. In fact, CeGSTP2-2 was readily detected in 25 µg of crude lysate from wild-type or transgenic animals, whereas 40 times more protein (i.e. an enriched GST fraction equivalent to 1 mg of lysate) was necessary to visualize mGSTA4-4 on Western blots of mGsta4 transgenic worms (Fig. 3), even though the sensitivity of the two anti-GST antibodies for detection of their respective antigens is similar (data not shown). This result is consistent with a lower steady-state expression level of the murine GST as compared with the autologous CeGSTP2-2. The resulting smaller increment in enzyme activity in mGsta4 as compared with gst-10 transgenics is also reflected in the level of 4-HNE–protein adducts, which is inversely proportional to the 4-HNE-conjugating activity: it is lowest in gst-10 worms, intermediate in mGsta4 transgenics and highest in pha-1 controls (Fig. 4).

It is noteworthy that the amount of 4-HNE–protein adducts decreased almost two-fold in gst-10 transgenic vs. control C. elegans in each of the individual measurements (Fig. 4), even though the absolute levels of the adducts were quite variable, perhaps due to uncontrolled day-to-day differences in unavoidable environmental stresses. This stands in sharp contrast to the limited number of cells in which CeGSTP2-2 was detected (Fig. 5 and the British Columbia C. elegans Gene Expression Consortium; see Results). Several possibilities could be invoked to reconcile the low percentage of cells expressing CeGSTP2-2 with the large effect of CeGSTP2-2 overexpression on whole-body 4-HNE–protein adducts. For example, a significant proportion of all 4-HNE-modified proteins could be present in the few neurons that express CeGSTP2-2, or the enzyme could be expressed below the detection threshold by cells other than the putatively identified neurons. Alternatively, cells expressing CeGSTP2-2, even if low in number, could constitute a ‘sink’ for 4-HNE diffusing from surrounding cells.

The biochemical parameters discussed so far indicate that the three groups of C. elegans lines used in this work, namely pha-1 controls, mGsta4 transgenics and gst-10 transgenics, constitute a graded series with progressively greater 4-HNE-conjugating activity and gradually decreasing steady-state level of 4-HNE-mediated electrophilic damage. Although free 4-HNE was not measured, 4-HNE–protein adducts are thought to reflect levels of 4-HNE integrated over time (but limited by the average half-life of the modified proteins).

Although the extrachromosomal maintenance of transgenes in C. elegans eliminates genetic positional effects, the level of transgene expression may still vary between individual sublines. Therefore, two independent pha-1 controls, five independent mGsta4 transgenic lines and five independent gst-10 transgenic lines were generated. Within each group, the lines behaved similarly in all experiments that were conducted. The availability of a graded series of three groups of lines differing in 4-HNE levels (pha-1 control, and mGsta4 and gst-10 transgenics) permitted a more conclusive evaluation of the physiological role of 4-HNE than would be possible in the more usual binary situation (i.e. control vs. transgenic strain). Indeed, with only one exception (heat shock), the stress resistances correlated with the 4-HNE-conjugating activity of the animals (Fig. 6). Similarly, the median lifespans ranked in the same order as 4-HNE-conjugating activity, i.e. pha-1 control < mGsta4 < gst-10 transgenic worms (Fig. 7 and Table 3). Thus, experimental manipulations which alter the ability of the organism to metabolize 4-HNE affect stress resistance and lifespan in a dose-dependent manner. The autologous CeGSTP2-2 and the heterologous murine mGSTA4-4 appeared to exert the same effect, differentiated only by their respective activities. This indicates that lifespan and stress resistance are influenced by 4-HNE rather than by other potential consequences of GST overexpression such as direct protein–protein interactions. Such regulatory interactions, which are independent of catalytic activity, have been described for GSTs (Adler et al., 1999; Dorion et al., 2002; Ryoo et al., 2004), including C. elegans GSTs (Greetham et al., 2004), but are unlikely for a murine GST in the context of a C. elegans cell.

Evidence obtained by perturbing the system (expression of GST transgenes) and measuring the outcome establishes a cause–effect relationship between 4-HNE metabolism and lifespan, although neither the underlying mechanism nor the number of steps between cause and effect can be derived from the data. The conclusion that there is a cause–effect relationship between the overexpression of some but not all (see next paragraph) GSTs and lifespan is corroborated by independent correlative evidence. We have demonstrated that lifespan positively correlates with 4-HNE-conjugating activity in a series of congenic lines of C. elegans (Fig. 9). In these lines, DNA regions 0.3–5 Mb in length are replaced by the corresponding regions from worms of a distinct genetic background. Regions of such length cover approximately 30–500 genes, out of the estimated 17 500 genes in C. elegans. The regions that have been identified as relevant to lifespan (Ayyadevara et al., 2001, 2003) may include the gst-10 locus (located on chromosome V, physical map position 2 561 521–2 564 525 bp, genetic map position −13.68 cM). However, at least part of the observed differences in 4-HNE-conjugating activity between the congenic lines could be due to changes in GSTs other than CeGSTP2-2, to indirect effects on 4-HNE conjugation by CeGSTP2-2 or to regulation of GST transcription by trans-acting factors encoded by genes in the replaced regions. Additional work will be needed to identify the gene(s) responsible for modulation of 4-HNE-conjugating activity, and to test directly their effects on lifespan.

In contrast to the effect observed by us for CeGSTP2-2 and mGSTA4-4, not all GSTs extend lifespan of C. elegans when overexpressed. Strains either overexpressing the gst-4 gene product or carrying a deletion of that gene did not differ in their lifespans from the wild-type control (Leiers et al., 2003). The gst-4 gene product has low activity for α,β-unsaturated aldehydes, including 4-HNE (Table 1). Overall, our findings strongly suggest that 4-HNE, or other similar α,β-unsaturated carbonyl compounds, affect lifespan. The difference between the gst-4 gene product and CeGSTP2-2 in the effect on lifespan could be due to disparate substrate specificities or differences in tissue distribution.

Our finding that an increased capacity for 4-HNE metabolism in a subset of cells (probably including neurons) extends lifespan in C. elegans raises the broader question of a possible mechanism underlying this effect, and its relationship to other known genetic and metabolic factors which modulate aging. The best-studied mechanism that controls lifespan is the insulin/insulin-like growth factor-1 (IGF-1) signaling pathway (reviewed in Kenyon, 2001; Warner, 2005). Mutations in that pathway can extend lifespan several-fold, i.e. to a significantly greater extent than changes in 4-HNE conjugation. Caloric restriction (Weindruch & Sohal, 1997) or mutations that reduce food intake or food utilization (Lakowski & Hekimi, 1998; Apfeld & Kenyon, 1999; Rogina et al., 2000) extend lifespan in a variety of organisms. Caloric restriction is thought to shift the NAD+/NADH ratio, which in turn may regulate the histone/protein deacetylase of the Sir-2 family (Blander & Guarente, 2004). The IGF-1 and Sir-2 pathways are interdependent, at least in mammalian cells (Nemoto et al., 2004). The specific target proteins whose expression and/or activity are modulated by the two pathways are not fully known but include stress-resistance gene products. Strikingly, neuronal cells, some of which appear to express CeGSTP2-2, are of central importance in the aging process, either as originators of IGF-1 and other signals (Wolkow, 2002; Alcedo & Kenyon, 2004), as targets of pharmacological or genetic interventions (Evason et al., 2005; this work), or as a ‘weak link’ which could determine organismal aging (Parkes et al., 1998; Wolkow et al., 2000).

The aging process is often considered within the conceptual framework of the free radical theory of aging (Harman, 1956, 2003) or its variants. Our finding that products of lipid peroxidation such as 4-HNE are relevant to aging fits within that framework. In particular, our results corroborate the recently proposed ‘green theory of aging’ (McElwee et al., 2004; Gems & McElwee, 2005). According to this theory, initiating events, including but not limited to ROS and encompassing a variety of stochastic perturbations, lead to the formation of a wide spectrum of secondary deleterious and toxic metabolites which are the actual biochemical effectors of damage. The damage accumulates and eventually becomes manifest as aging. In this view, the aging process would be retarded most effectively by detoxification reactions (hence ‘green’, by analogy with environmental concerns for clean waste disposal). The significance of pathways such as IGF-1 signaling would lie in their roles in regulating a wide range of detoxification enzymes. GSTs constitute a subset of these enzymes, explaining why the effect of CeGSTP2-2 on lifespan, although robust, is smaller than that of changes in the IGF-1 pathway. If correct, the above hypothesis predicts that CeGSTP2-2 should be overexpressed in long-lived C. elegans strains carrying hypomorphic mutations of daf-2 (which encodes an insulin/IGF-1 receptor), and that lifespan extension in such strains should be partially but not fully reversed by abrogating CeGSTP2-2 expression. Work is in progress in our laboratories to test these hypotheses; preliminary data indicate that the predictions are indeed correct.

Experimental procedures

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgments
  8. References

C.elegans culture conditions and DNA preparation

Worms were cultured in nematode growth medium [NGM: 25 mm potassium phosphate, pH 6.0, 50 mm NaCl, 0.25% (w/v) peptone, 0.5% (w/v) cholesterol, 1 mm MgCl2, 1 mm CaCl2] and fed with E. coli strain OP50 (2 × 109 mL−1) grown in 3XD medium (10.5 g L−1 Na2HPO4, 4.5 g L−1 KH2PO4, 0.6 g L−1 NH4Cl, 15 g L−1 casein hydrolysate, 24 g L−1 glycerol and 3 mL L−1m MgSO4). Animals were grown at 15 °C [LGIII, pha-1 (e2123) mutant], at 25 or 25.5 °C (transgenic strains in which the pha-1 mutation has been rescued by microinjection of the pBX plasmid; see below), or at 20 °C (wild-type Bristol-N2). For high yields, large liquid cultures were grown in bulk, followed by removal of bacteria by washing and flotation on a sucrose gradient. Genomic DNA was prepared from worms by proteinase K digestion in the presence of 1% SDS, followed by standard phenol/chloroform extraction and ethanol precipitation.

Determination of C. elegans lifespan

Worms were grown for two generations without depletion of bacteria, and were harvested by rinsing with S buffer (0.1 m NaCl, 0.05 m potassium phosphate, pH 6.0) (Brenner, 1974) into 15-mL plastic Falcon tubes. Adults were allowed to settle, and eggs were recovered from hermaphrodites by alkaline hypochlorite lysis (5 min at room temperature in 0.5 n NaOH, 1.05% hypochlorite) (Emmons et al., 1979). The eggs were rinsed with S buffer and transferred to fresh agar plates containing E. coli OP50. Survival cultures (in 60-mm dishes) were set up 1 day after the L4/adult molt and were incubated at 25.5 °C. Worms were counted and transferred daily to fresh 60-mm plates seeded with E. coli OP50. Worms were considered dead when they failed to move, either spontaneously or in response to touch, and showed no pharyngeal pumping. Worms that crawled away from the plate were censored (considered to have escaped). The time of death or escape was assumed to be at the midpoint of the interval between the time the worm was found dead or missing, respectively, and the time of the preceding observation. Between 100 and 150 worms per strain were used in each lifespan experiment.

Determination of stress resistance

To assess resistance to H2O2, age-synchronized worms were grown in liquid culture at 25.5 °C. Fifty adult worms (day 4 post-hatch) were added to each 35-mm dish containing 3 mL S buffer supplemented with cholesterol (5 µg mL−1), E. coli OP50 (2 × 109 cells mL−1), and freshly added H2O2 (5 mm). The total number of live worms in each dish was counted at regular time intervals. Worms that did not swim and showed no pharyngeal pumping were considered dead. Paraquat resistance was determined by transferring 4-day-old adult worms to plates containing 150 mm paraquat, and scoring as described for lifespan determination. For assessing heat resistance, young adult worms were incubated in 60-mm plates at 36 ± 0.4 °C, and were scored every 2 h for motility, provoked movement and pharyngeal pumping, as described for lifespan determination. Resistance to UV radiation (254 nm) was measured by placing 50 adult (4 days post-hatch) hermaphrodites on NGM agar plates (without E. coli) and delivering an irradiation dose of 5000 J m−2 in a UV Stratalinker 1800 (Stratagene, La Jolla, CA, USA), followed by transfer to NGM plates with E. coli and incubation at 25.5 °C. Worms were scored as dead or censored as described for lifespan measurements.


Germline transformation was carried out using C. elegans LGIII, pha-1 (e2123) mutants. The pha-1/pBX system (a kind gift from R. Schnabel, Braunschweig, Germany) is based on the temperature-sensitive embryonic lethal mutation pha-1. Constructs carrying Gfp, gst-10 or mGsta4 transgenes under the control of the gst-10 promoter (see below) were adjusted to 80 µg mL−1 and were microinjected into the distal arm of the hermaphrodite gonad as previously described (Mello et al., 1991). The pBX plasmid that contained a wild-type copy of the pha-1 gene was co-injected with the fusion constructs (Granato et al., 1994; Mello & Fire, 1995). Following microinjection, the animals were raised to 25 °C. Embryonal lethality caused by the pha-1 mutation is complemented in transgenic animals carrying the pBX plasmid. Thus, transgenic animals can be selected by shifting the F1 larvae of injected hermaphrodites from 15 to 25 °C. Only transformed progeny survive this selection and can be maintained by cultivation at 25 °C. For each construct, visible phenotypes of multiple animals from at least three independent lines were examined using Nomarski optics.

Isolation of 4-HNE-conjugating GST from C. elegans

The Bristol-N2 strain of C. elegans was grown in liquid culture, in NGM containing E. coli OP50 (2 × 109 mL−1). Worms were collected and stored at −70 °C until use. For purification of 4-HNE-metabolizing GSTs, 5 g of frozen C. elegans pellet was ground in a mortar on dry ice. The frozen powder was transferred to a glass-Teflon homogenizer containing 35 mL of 20 mm potassium phosphate, pH 7.0, 1.4 mm 2-mercaptoethanol (buffer A) supplemented with 1 mm PMSF. Following homogenization for 5 min, the lysate was sonicated for 2 min (power setting 3) using a Branson Sonifier 450 (VWR, West Chester, PA, USA) equipped with a tapered microtip, 3 mm in diameter, and was centrifuged for 15 min at 25 000 g. The supernatant was dialysed for 4 h against buffer A, and was used for isolation of a GST fraction by GSH affinity chromatography according to the method of Simons & Vander Jagt (1977). The GST pool was further fractionated by isoelectric focusing as described in Ahmad et al. (1988). Fractions from isoelectric focusing were assayed for GST activity with both 4-HNE and 1-chloro-2,4-dinitrobenzene as substrates (Engle et al., 2001).

Cloning, bacterial expression and enzymatic characterization of CeGSTP2-2

The coding sequence of gst-10 was amplified from total C. elegans cDNA using the sense primer 5′-AGAGCTCGTCATATGGCCGTGCCACAACTTTACTAC and the antisense primer 5′-TCATTTGCATTTATTCCATGGTCAGTTCATTCCCTTCTCAATC. The primers were modified (underlined) to create an NdeI restriction site that includes the initiation codon, and an NcoI site immediately downstream of the translation termination codon, in the sense and antisense primers, respectively. The amplification product was digested with NdeI and NcoI, and was subcloned between the same sites of the pET-30a(+) vector (Novagen, Madison, WI, USA). E. coli BL21(DE3)pLysS was transformed with the expression vector, and the CeGSTP2-2 protein was expressed upon induction with 1 mm isopropyl-β-D-thiogalactopyranoside (IPTC) and purified by GSH affinity chromatography (Simons & Vander Jagt, 1977).

Cloning, bacterial expression and enzymatic characterization of the gst-4 gene product

The coding sequence of gst-4 was amplified from total C. elegans cDNA using the sense primer 5′-AGCCATATGCCAAACTATAAGCTATTG and the antisense primer 5′-AGCGGATCCTTAAACAATACTATCCTTTCTTG (the introduced NdeI and BamHI sites, respectively, are underlined). The amplification product was digested with NdeI and BamHI, and was subcloned between the same sites of the pET-19b vector (Novagen). This added an (His)10 tag to the N-terminus of the GST. The entire open reading frame was sequenced and was found to be identical to the gst-4 sequence reported previously (Leiers, 2002), and to that deposited in WormBase. The protein was expressed in E. coli BL21 Star (DE3)pLysS (Invitrogen, Carlsbad, CA, USA) after induction with 1 mm IPTG, and was purified either by Ni2+ chelation chromatography utilizing the added His tag (Leiers et al., 2003), or by glutathione affinity chromatography (Simons & Vander Jagt, 1977). The two preparations had very similar enzymatic properties with 1-chloro-2,4-dinitrobenzene, 4-HNE, trans-non-2-enal and cumene hydroperoxide as substrates (Table 1).

Enzyme activity assays

Catalytic activity with 4-HNE as substrate was measured according to Alin et al. (1985). For determination of kinetic parameters for 4-HNE, the concentration of the aldehyde was varied at a constant GSH concentration (0.5 mm), and the kinetic parameters were obtained from non-linear fitting of a Michaelis–Menten hyperbola to the data. Catalytic activities with trans-non-2-enal and acrolein were determined according to Pal et al. (2000) and Nilsson et al. (2000), respectively, and activity with 1-chloro-2,4-dinitrobenzene was measured according to Habig et al. (1974). Glutathione peroxidase activity with cumene hydroperoxide as substrate was determined as described by Awasthi et al. (1975).

Generation of polyclonal antibodies against CeGSTP2-2

An antibody against affinity-purified bacterially expressed CeGSTP2-2 was raised in chicken commercially (Aves Laboratories, Tigard, OR, USA). The IgY fraction was obtained from egg yolks by Aves Laboratories using a proprietary protocol involving solvent extraction and salt fractionation. Pre-immune IgY was isolated from eggs of the same chicken collected prior to antigen injection.

Western blot analysis

Samples were separated by SDS-PAGE and electroblotted onto nitrocellulose membranes (Schleicher & Schuell, Keene, NH, USA). The membrane was blocked and incubated with the primary antibody against mGSTA4-4 (Zimniak et al., 1994) or CeGSTP2-2 at a 1 : 20 000 dilution at 4 °C overnight, then incubated with the secondary antibody (peroxidase-coupled anti-IgY; Sigma, St Louis, MO, USA) at 1 : 10 000 dilution at room temperature for 1 h, followed by chemiluminescence detection using the ECL Plus kit (Amersham Biosciences, Piscataway, NJ, USA) on a Storm 860 phosphoimager (Molecular Dynamics, Sunnyvale, CA, USA). Signal quantitation was carried out with the ImageQuant software (Molecular Dynamics).

Immunodepletion of CeGSTP2-2 in worm lysates

Bristol-N2 worms were grown in liquid culture at 20 °C, were harvested, and a lysate was then prepared. Worm lysate (300 µg) was incubated in a total volume of 0.5 mL of immunoprecipitation buffer [0.14 m NaCl, 0.01 m KCl, 8 mm Na2HPO4, 2 mm KH2PO4, pH 7.4, supplemented with Calbiochem protease inhibitor cocktail (La Jolla, CA, USA), set I, at 0.5× strength] with varying amounts (up to 1000 µg) of pre-immune or anti-CeGSTP2-2 IgY overnight at 4 °C, followed by an addition of agarose-coupled goat anti-chicken IgY (Aves Laboratories) for 2 h at room temperature. The beads were pelleted, and the supernatant was analysed for 4-HNE-conjugating activity.

Mass spectrometric analysis of proteins recognized by anti-CeGSTP2-2 antibody

C. elegans lysate was obtained as described in the previous section on immunodepletion, and was pre-adsorbed with agarose-coupled goat anti-chicken IgY antibody for 90 min at room temperature to remove non-specifically reacting proteins. The beads were pelleted by centrifugation and discarded, and anti-CeGSTP2-2 antibody was added to the supernatant at a ratio of 500 µg of IgY per 300 µg of lysate. The immunoprecipitate was recovered using fresh goat anti-chicken IgY antibody coupled to agarose beads. The beads were washed seven times with immunoprecipitation buffer, were extracted by heating with SDS sample buffer, and the solubilized proteins were separated on an SDS−12% polyacrylamide one-dimensional gel (Tris–glycine system). The gel was stained overnight in SYPRO Ruby solution (Molecular Probes, Eugene, OR, USA). Gel images were acquired using the ProPic imaging and spot-picking robot from Genomic Solutions (Ann Arbor, MI, USA). A broad band migrating at 25–29 kDa was excised into a 96-well plate with the ProPic, and trypsin digestion was performed using the ProGest in-gel enzymatic digestion robot from Genomic Solutions. The resulting peptide mixture was separated by nanoflow capillary HPLC-MS/MS using a CapLC XE pump (Waters Corp., Beverly, MA, USA) operating at 200 nL min−1 and equipped with a C18 (75 µm × 15 cm) column. The eluate was analysed in-line by ESI-MS/MS using a Micromass Q-Tof Micro (Waters) tandem mass spectrometer operating in the positive ion mode. Data acquisition was performed in a data-dependent fashion. The resulting data were processed using ProteinLynx 2.1 (Waters), and the resulting peak list was submitted for database searching on Mascot (

Determination of 4-HNE–protein adducts by ELISA

The level of 4-HNE adducts on proteins was measured in C. elegans lysates by competitive ELISA (Satoh et al., 1999) using a polyclonal antibody against 4-HNE-modified keyhole limpet hemocyanin. The antibody was a generous gift of Dr D. R. Petersen, University of Colorado, Denver.

Tissue specificity of CeGSTP2-2 expression as determined by a reporter gene

In order to investigate the cell-specific expression of CeGSTP2-2, transgenic nematodes were created. The basic strategy involved the insertion of fragments of the 5′ region of gst-10 into the multicloning site of the vector pPD95.77 (provided by A. Fire, Carnegie Institute, Baltimore). The inserted promoter sequence drives the expression of the GFP reporter gene. The GFP coding region is followed in the vector by translation termination and poly(A) addition signals. The putative promoter region of gst-10 was amplified by PCR using a proof-reading polymerase mix (Expand Long Template PCR System, Roche Diagnostics, Indianapolis, IN, USA) with C. elegans genomic DNA as template. Two promoter fragments were used. For both, the antisense primer 5′-CGCGGATCCGAACAGAAGACGGATGTATTCTCCGAA was used; the primer covers positions 89–115 downstream of the translation start site in exon 2 of the gst-10 gene, and contains an added BamH I restriction site (underlined). The sense primers were 5′-CGCGGATCCTGTCATCATTTGTGTATATTACCTTTGTTC and 5′-CGCGGATCCATTTCCTGACGAGAACCTATAGTATAAAAC, located 3.1 and 2.3 kb, respectively, upstream of the translation start site of gst-10 (added BamHI sites are underlined). For microinjection (see above), the plasmid DNA was prepared using the Endo Free Plasmid Maxi Kit (Qiagen, Hilden, Germany). Both promoter fragments yielded the same localization pattern of the GFP reporter.

CeGSTP2-2 and mGSTA4-4 transgene expression in C. elegans

A 1.0-kb putative gst-10 promoter fragment comprising the intergenic sequence between gst-10 and the upstream adjacent str-154 gene was amplified using the Expand Long Template PCR System (Roche Diagnostics). The sense primer 5′-AAAACTGCAGTTTCCAAATGTGAAATTTAACCGGAGA was modified (underlined) to insert a PstI restriction site. The antisense primer 5′-GGCACGGGGATCCTACGAGCTCTTATCGTTTA was used to ablate the translation initiation codon of gst-10 by inserting a BamHI restriction site (underlined). The PCR product (promoter insert) was digested with PstI and BamHI and subcloned into the pPD49.26 vector (provided by A. Fire) digested with the same restriction enzymes, yielding plasmid pPD49.26/CeP2pro.

The coding sequence of gst-10 was amplified from its cDNA using the sense primer 5′-TTAAGTTGAGCTAGCATGGCCGTGCCACAACTT (NheI restriction site, introduced immediately upstream of the translation initiation codon, is underlined) and the antisense primer 5′-TCATTTGCATTTATTCCATGGTCAGTTCATTCCCTTCTCAATC (NcoI restriction site, added immediately downstream of the translation termination codon, is underlined). Similarly, mGsta4 cDNA was amplified using the sense primer 5′-AGCCGTAACGCTAGCATGGCAGCCAAACCTAAG (NheI restriction site, introduced immediately upstream of the translation initiation codon, is underlined) and the antisense primer 5′-CTAGCATTGTCCATGGCTAGAACTTCAGGACAA (NcoI restriction site, added immediately downstream of the translation termination codon, is underlined). For neither gene did the introduction of the NheI site compromise the Kozak consensus sequence for the translational start site as utilized in C. elegans. The two amplification products were digested with NheI and NcoI, and were subcloned between the same sites in pPD49.26/CeP2pro. This yielded expression vectors pPD49.26/CeP2pro/gst-10 and pPD49.26/CeP2pro/mGsta4, respectively, in which the expression of each GST gene was placed under the control of the gst-10 promoter. The two constructs were used for microinjection. Single F1 larvae were transferred to new agar plates and the introduction of the expression vectors was confirmed in their progeny by single-worm PCR according to Jansen et al. (1997).

Tissue specificity of CeGSTP2-2 expression as determined by indirect immunostaining

Cells expressing CeGSTP2-2 were identified by staining wild-type adult C. elegans or worms overexpressing CeGSTP2-2 with anti-CeGSTP2-2 antibody according to Finney & Ruvkun (1990). Briefly, worms were fixed and permeabilized by freezing in 1% buffered formaldehyde, thawing, and exposure to 1% Triton X-100 and 1% 2-mercaptoethanol, followed by 10 mm dithiothreitol. After washing, the worms were incubated in 0.3% H2O2 in 25 mm sodium borate, pH 9.2, washed, blocked consecutively with 0.2% bovine serum albumin and 1% goat serum, and exposed to anti-CeGSTP2-2 antibodies (1 : 300 dilution) in the presence of 2% bovine serum albumin and 0.5% Triton X-100. The worms were washed, stained with an anti-chicken secondary antibody labeled with Alexa 488 (1 : 200; Molecular Probes) and photographed under a Nikon Eclipse E1000 fluorescence microscope. The expression pattern of mGSTA4-4 in mGsta4 transgenic worms was determined as described above, except that the primary antibody was anti-mGSTA4-4 raised in chicken as described previously (Singh et al., 2002).


Values of biochemical parameters were compared by anova followed by pairwise t-tests with a correction for multiple comparisons (Tukey–Kramer). Maximum likelihood estimates of Gompertz survival parameters (Arking, 1998) and of median lifespans were obtained using the program WinModest (Pletcher, 1999; Promislow et al., 1999). Survival curves were compared by the log rank test and by randomization analysis (using 106 Monte Carlo samples). Data from multiple lifespan experiments, each involving control and transgenic strains, were evaluated in aggregate by Cox regression. The Bonferroni correction was applied to multiple comparisons of lifespans. All statistical procedures were carried out in NCSS software (Number Cruncher Statistical Systems, Kaysville, UT, USA).


  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgments
  8. References

We thank Dr Dennis R. Petersen, University of Colorado, Denver, for the generous gift of an antibody against 4-HNE–protein adducts, and Eric R. Siegel, M.S., Division of Biostatistics, University of Arkansas for Medical Sciences, for statistical advice. This work was supported in part by National Institutes of Health grants R01 AG18845 and R01 ES07804 (to P.Z.), and Program Project grant P01 AG20641 (to R.J.S.R.). P.Z. and R.J.S.R. are recipients of VA Research Career Scientist Awards. M.E. was a recipient of an institutional grant awarded by the University of Arkansas for Medical Sciences Committee for Allocation of Graduate Student Research Funds (CAGSRF). Some nematode strains used in this work were provided by the Caenorhabditis Genetics Center, which is funded by the NIH National Center for Research Resources (NSRR). The Arkansas Cancer Research Center Proteomics Core Facility (C.F.L., co-director) is supported in part by NIH Grant Number P20 RR-16460 from the BRIN Program of the National Center for Research Resources and by the Arkansas Biosciences Institute (funded by the Tobacco Settlement Proceeds Act).


  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgments
  8. References
  • Adler V, Yin Z, Fuchs SY, Benezra M, Rosario M, Tew KD, Pincus MR, Sardana RM, Henderson CJ, Wolf CR, Davis RJ, Ronai Z (1999) Regulation of JNK signaling by GSTp. EMBO J. 18, 13211334.
  • Ahmad H, Singh SV, Medh RD, Ansari GAS, Kurosky A, Awasthi YC (1988) Differential expression of alpha, mu and pi classes of isoenzymes of glutathione S-transferase in bovine lens, cornea, and retina. Arch. Biochem. Biophys. 266, 416426.
  • Alcedo J, Kenyon C (2004) Regulation of C. elegans longevity by specific gustatory and olfactory neurons. Neuron 41, 4555.
  • Alin P, Danielson UH, Mannervik B (1985) 4-Hydroxyalk-2-enals are substrates for glutathione transferase. FEBS Lett. 179, 267270.
  • Apfeld J, Kenyon C (1999) Regulation of lifespan by sensory perception in Caenorhabditis elegans. Nature 402, 804809.
  • Arking R (1998) Measuring age-related changes in populations. In Biology of Aging: Observations and Principles (ed. ArkingR), pp. 2759. Sunderland, MA: Sinauer Associates.
  • Armstrong RN (1997) Structure, catalytic mechanism, and evolution of the glutathione transferases. Chem. Res. Toxicol. 10, 218.
  • Awasthi YC, Beutler E, Srivastava SK (1975) Purification and properties of human erythrocyte glutathione peroxidase. J. Biol. Chem. 250, 51445149.
  • Ayyadevara S, Ayyadevara R, Hou S, Thaden JJ, Shmookler Reis RJ (2001) Genetic mapping of quantitative trait loci governing longevity of Caenorhabditis elegans in recombinant-inbred progeny of a Bergerac-BO x RC301 interstrain cross. Genetics 157, 655666.
  • Ayyadevara S, Ayyadevara R, Vertino A, Galecki A, Thaden JJ, Shmookler Reis RJ (2003) Genetic loci modulating fitness and life span in Caenorhabditis elegans: categorical trait interval mapping in CL2a x Bergerac-BO recombinant-inbred worms. Genetics 163, 557570.
  • Beckman KB, Ames BN (1998) The free radical theory of aging matures. Physiol. Rev. 78, 547581.
  • Blander G, Guarente L (2004) The Sir2 family of protein deacetylases. Annu. Rev. Bioch. 73, 417435.
  • Brenner S (1974) The genetics of Caenorhabditis elegans. Genetics 77, 7194.
  • Cadenas E, Davies KJA (2000) Mitochondrial free radical generation, oxidative stress, and aging. Free Radic. Biol. Med. 29, 222230.
  • Carini M, Aldini G, Facino RM (2004) Mass spectrometry for detection of 4-hydroxy-trans-2-nonenal (HNE) adducts with peptides and proteins. Mass Spectr. Rev. 23, 281305.
  • Chelvanayagam G, Parker MW, Board PG (2001) Fly fishing for GSTs: a unified nomenclature for mammalian and insect glutathione transferases. Chem. Biol. Interact. 133, 256260.
  • Crow JP (2000) Peroxynitrite scavenging by metalloporphyrins and thiolates. Free Radic. Biol. Med. 28, 14871494.
  • Dianzani MU (2003) 4-Hydroxynonenal from pathology to physiology. Mol. Aspects Med. 24, 263272.
  • Dianzani MU, Barrera G, Parola M (1999) 4-hydroxy-2,3-nonenal as a signal for cell function and differentiation. Acta Biochim. Pol. 46, 6175.
  • Dorion S, Lambert H, Landry J (2002) Activation of the p38 signaling pathway by heat shock involves the dissociation of glutathione S-transferase Mu from Ask1. J. Biol. Chem. 277, 3079230797.
  • Emmons SW, Klass MR, Hirsh D (1979) Analysis of the constancy of DNA sequences during development and evolution of the nematode Caenorhabditis elegans. Proc. Natl Acad. Sci. USA 76, 13331337.
  • Engle MR, Singh SP, Nanduri B, Ji X, Zimniak P (2001) Invertebrate glutathione transferases conjugating 4-hydroxynonenal: CeGST 5.4 from Caenorhabditis elegans. Chem. Biol. Interact. 133, 244248.
  • Esterbauer H, Schaur RJ, Zollner H (1991) Chemistry and biochemistry of 4-hydroxynonenal, malonaldehyde and related aldehydes. Free Radic. Biol. Med. 11, 81128.
  • Evason K, Huang C, Yamben I, Covey DF, Kornfeld K (2005) Anticonvulsant medications extend worm life-span. Science 307, 258262.
  • Finney M, Ruvkun G (1990) The unc-86 gene product couples cell lineage and cell identity in C. elegans. Cell 63, 895905.
  • Forman HJ, Dickinson DA, Iles KE (2003) HNE – signaling pathways leading to its elimination. Mol. Aspects Med. 24, 189194.
  • Gems D, McElwee JJ (2005) Broad spectrum detoxification: the major longevity assurance process regulated by insulin/IGF-1 signaling? Mech. Ageing Dev. 126, 381387.
  • Granato M, Schnabel H, Schnabel R (1994) pha-1, a selectable marker for gene transfer in C. elegans. Nucleic Acids Res. 22, 17621763.
  • Greetham D, Morgan C, Campbell AM, Van Rossum AJ, Barrett J, Brophy PM (2004) Evidence of glutathione transferase complexing and signaling in the model nematode Caenorhabditis elegans using a pull-down proteomic assay. Proteomics 4, 19891995.
  • Gutteridge JMC, Halliwell B (1990) The measurement and mechanism of lipid peroxidation in biological systems. Trends Biochem. Sci. (TIBS) 15, 129135.
  • Habig WH, Pabst MJ, Jakoby WB (1974) Glutathione S-transferases. The first enzymatic step in mercapturic acid formation. J. Biol. Chem. 249, 71307139.
  • Harman D (1956) Aging: a theory based on free radical and radiation chemistry. J. Gerontol. 11, 298300.
  • Harman D (2003) The free radical theory of aging. Antioxid. Redox Signal. 5, 557561.
  • Jansen G, Hazendonk E, Thijssen KL, Plasterk RH (1997) Reverse genetics by chemical mutagenesis in Caenorhabditis elegans. Nat. Genet. 17, 119121.
  • Kenyon C (2001) A conserved regulatory system for aging. Cell 105, 165168.
  • Lakowski B, Hekimi S (1998) The genetics of caloric restriction in Caenorhabditis elegans. Proc. Natl Acad. Sci. USA 95, 1309113096.
  • Lane N (2002) Oxygen: the Molecule That Made the World. New York: Oxford University Press. .
  • Leiers B (2002) Funktionelle Charakterisierung von zwei stressinduzierbaren Glutathion S-Transferasen in Caenorhabditis elegans. Doctoral Thesis, Heinrich-Heine-Universität, Düsseldorf.
  • Leiers BR, Kampkotter A, Grevelding CG, Link CD, Johnson TE, Henkle-Duhrsen K (2003) A stress-responsive glutathione S-transferase confers resistance to oxidative stress in Caenorhabditis elegans. Free Radic. Biol. Med. 34, 14051415.
  • Leonarduzzi G, Robbesyn F, Poli G (2004) Signaling kinases modulated by 4-hydroxynonenal. Free Radic. Biol. Med. 37, 16941702.
  • Mannervik B, Awasthi YC, Board PG, Hayes JD, Di Ilio C, Ketterer B, Listowsky I, Morgenstern R, Muramatsu M, Pearson WR, Pickett CB, Sato K, Widersten M, Wolf CR (1992) Nomenclature for human glutathione transferases. Biochem. J. 282, 305308.
  • Mates JM, Perez-Gomez C, Nunez de Castro I (1999) Antioxidant enzymes and human diseases. Clin. Biochem. 32, 595603.
  • McElwee JJ, Schuster E, Blanc E, Thomas JH, Gems D (2004) Shared transcriptional signature in Caenorhabditis elegans dauer larvae and long-lived daf-2 mutants implicates detoxification system in longevity assurance. J. Biol. Chem. 279, 4453344543.
  • McEwen JE, Zimniak P, Mehta JL, Shmookler Reis RJ (2005) Molecular pathology of aging and its implications for senescent coronary atherosclerosis. Curr. Opin. Cardiol. 20, 399406.
  • Mello C, Fire A (1995) DNA transformation. Meth. Cell Biol. 48, 451482.
  • Mello CC, Kramer JM, Stinchcomb D, Ambros V (1991) Efficient gene transfer in C.elegans: extrachromosomal maintenance and integration of transforming sequences. EMBO J. 20, 39593970.
  • Nakashima I, Liu W, Akhand AA, Takeda K, Kawamoto Y, Kato M, Suzuki H (2003) 4-hydroxynonenal triggers multistep signal transduction cascades for suppression of cellular functions. Mol. Aspects Med. 24, 231238.
  • Nemoto S, Fergusson MM, Finkel T (2004) Nutrient availability regulates SIRT1 through a forkhead-dependent pathway. Science 306, 21052108.
  • Nilsson LO, Gustafsson A, Mannervik B (2000) Redesign of substrate-selectivity determining modules of glutathione transferase A1–1 installs high catalytic efficiency with toxic alkenal products of lipid peroxidation. Proc. Natl Acad. Sci. USA 97, 94089412.
  • Pal A, Hu X, Zimniak P, Singh SV (2000) Catalytic efficiencies of allelic variants of human glutathione S-transferase Pi in the glutathione conjugation of α,β-unsaturated aldehydes. Cancer Lett. 154, 3943.
  • Parkes TL, Elia AJ, Dickinson D, Hilliker AJ, Phillips JP, Boulianne GL (1998) Extension of Drosophila lifespan by overexpression of human SOD1 in motorneurons. Nat. Genetics. 19, 171174.
  • Petersen DR, Doorn JA (2004) Reactions of 4-hydroxynonenal with proteins and cellular targets. Free Radic. Biol. Med. 37, 937945.
  • Pletcher SD (1999) Model fitting and hypothesis testing for age-specific mortality data. J. Evol. Biol. 12, 430439.
  • Promislow DEL, Tatar M, Pletcher SD, Carey JR (1999) Below-threshold mortality: implications for studies in evolution, ecology and demography. J. Evol. Biol. 12, 314328.
  • Rogina B, Reenan RA, Nilsen SP, Helfand SL (2000) Extended life-span conferred by cotransporter gene mutations in Drosophila. Science 290, 21372140.
  • Ryoo K, Huh SH, Lee YH, Yoon KW, Cho SG, Choi EJ (2004) Negative regulation of MEKK1-induced signaling by glutathione S-transferase Mu. J. Biol. Chem. 279, 4358943594.
  • Satoh K, Yamada S, Koike Y, Igarashi Y, Toyokuni S, Kumano T, Takahata T, Hayakari M, Tsuchida S, Uchida K (1999) A 1-hour enzyme-linked immunosorbent assay for quantitation of acrolein- and hydroxynonenal-modified proteins by epitope-bound casein matrix method. Anal. Biochem. 270, 323328.
  • Sawicki R, Singh SP, Mondal AK, Benes H, Zimniak P (2003) Cloning, expression, and biochemical characterization of one Epsilon-class (GST-3) and ten Delta-class (GST-1) glutathione S-transferases from Drosophila melanogaster, and identification of additional nine members of the Epsilon class. Biochem. J. 370, 661669.
  • Scandalios JG (1997) Oxidative Stress and the Molecular Biology of Antioxidant Defenses. Plainview, NY: Cold Spring. Harbor Laboratory Press.
  • Schneider C, Porter NA, Brash AR (2004) Autoxidative transformation of chiral ω6 hydroxy linoleic and arachidonic acids to chiral 4-hydroxy-2E-nonenal. Chem. Res. Toxicol. 17, 937941.
  • Schneider C, Tallman KA, Porter NA, Brash AR (2001) Two distinct pathways of formation of 4-hydroxynonenal: mechanisms of nonenzymatic transformation of the 9-and 13-hydroperoxides of linoleic acid to 4-hydroxyalkenals. J. Biol. Chem. 276, 2083120838.
  • Simons PC, Vander Jagt DL (1977) Purification of glutathione S-transferases from human liver by glutathione-affinity chromatography. Anal. Biochem. 82, 334341.
  • Singh SP, Coronella JA, Benes H, Cochrane BJ, Zimniak P (2001) Catalytic function of Drosophila melanogaster glutathione S-transferase DmGSTS1–1 (GST-2) in conjugation of lipid peroxidation end products. Eur. J. Biochem. 268, 29122923.
  • Singh SP, Janecki AJ, Srivastava SK, Awasthi S, Awasthi YC, Xia SJ, Zimniak P (2002) Membrane association of glutathione S-transferase mGSTA4–4, an enzyme that metabolizes lipid peroxidation products. J. Biol. Chem. 277, 42324239.
  • Sun M, Salomon RG (2004) Oxidative fragmentation of hydroxy octadecadienoates generates biologically active γ-hydroxyalkenals. J. Am. Chem. Soc. 126, 56995708.
  • Warner HR (2005) Longevity genes: from primitive organisms to humans. Mech. Ageing Dev. 126, 235242.
  • Watts JL, Browse J (2002) Genetic dissection of polyunsaturated fatty acid synthesis in Caenorhabditis elegans. Proc. Natl Acad. Sci. USA 99, 58545859.
  • Weindruch R, Sohal RS (1997) Caloric intake and aging. N. Engl. J. Med. 337, 986994.
  • Weston K, Yochem. J, Greenwald I (1989) A Caenorhabditis elegans cDNA that encodes a product resembling the rat glutathione S-transferase P subunit. Nucleic Acids Res. 17, 2138.
  • Wolkow CA (2002) Life span: getting the signal from the nervous system. Trends Neurosci. 25, 212216.
  • Wolkow CA, Kimura KD, Lee M-S, Ruvkun G (2000) Regulation of C. elegans life-span by insulinlike signaling in the nervous system. Science 290, 147150.
  • Xu GH, Liu YH, Sayre LM (2000) Polyclonal antibodies to a fluorescent 4-hydroxy-2-nonenal (HNE)-derived lysine–lysine cross-link: characterization and application to HNE-treated protein and in vitro oxidized low-density lipoprotein. Chem. Res. Toxicol. 13, 406413.
  • Xu G, Sayre LM (1998) Structural characterization of a 4-hydroxy-2-alkenal-derived fluorophore that contributes to lipoperoxidation-dependent protein cross-linking in aging and degenerative disease. Chem. Res. Toxicol. 11, 247251.
  • Zimniak P, Singhal SS, Srivastava SK, Awasthi S, Sharma R, Hayden JB, Awasthi YC (1994) Estimation of genomic complexity, heterologous expression, and enzymatic characterization of mouse glutathione S-transferase mGSTA4–4 (GST 5.7). J. Biol. Chem. 269, 9921000.