INTRODUCTION

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

To survive prolonged periods of oxygen deficiency, all aerobic organisms have had to evolve mechanisms for sensing oxygen availability and to adjust their cellular metabolism accordingly. Upon transfer from aerobic to hypoxic/anoxic conditions, animal and plant cells switch from aerobic respiration to lactic fermentation (Roberts et al., 1984a, 1984b). Continued lactic fermentation throughout hypoxia leads to the acidification of cytoplasm and rapid cell death in animal tissues. In contrast, after a transient period of lactic fermentation, maize root tip cells will further switch to alcoholic fermentation and allow glycolysis to continue for a longer period (Roberts et al., 1984a, 1984b). Comparative studies of cytoplasmic acidosis indicate that cytoplasmic pH regulation is an important factor in survival under hypoxia (Roberts et al., 1984a, 1984b; Xia and Saglio, 1992).

Anaerobic treatment of maize seedlings causes repression of pre-existing protein synthesis and induces the synthesis of about 20 anaerobic proteins (ANP) after approximately 90 min (Sachs et al., 1980). Most of the ANPs are enzymes involved in glycolysis and fermentation (for review, see Sachs et al., 1996). It was shown recently that most hypoxia-induced proteins in maize root tip cells are also enzymes involved in glycolysis and primary carbohydrate metabolism (Chang et al., 2000). Transcriptional, post-transcriptional, and translational controls have been shown to regulate synthesis of ANPs under low-oxygen stress (Fennoy and Bailey-Serres, 1995; Bailey-Serres and Dawe, 1996; Drew, 1997). Several cis-acting elements and trans-regulatory factors involved in anoxic and hypoxic inductions of the alcohol dehydrogenase (ADH) genes in maize and Arabidopsis have been identified (Ferl and Laughner, 1989; Yang et al., 1993; Dolferus et al., 1994; Kyozuka et al., 1994; Hoeren et al., 1998).

Lysogenic aerenchyma formation, which is characterized by continuous gas spaces in roots and shoots, occurs in the root cortex of several plant species during hypoxia (Campbell and Drew, 1983; Justin and Armstrong, 1987; Drew et al., 2000) and may correlate with tolerance to flooding (Justin and Armstrong, 1987). Lysogenic aerenchyma formation results from the lysis of cells in the cortical tissues of hypoxic-treated plants (He et al., 1994) and is associated with an increased cellulase activity, as well as the induction of a gene encoding a homolog of xyloglucan endo-transglycosylase, a putative cell wall loosening enzyme (He et al., 1994; Saab and Sachs, 1996).

An ethylene signal is required for aerenchyma formation in hypoxic maize roots (for review see, Drew et al., 2000; He et al., 1994, 1996). In contrast, no aerenchyma formation could be observed in maize roots under anoxic conditions in which ethylene biosynthesis is inhibited because the conversion of 1-aminocyclopropane-1-carboxylic acid (ACC) to ethylene by ACC oxidase requires the presence of oxygen (Yang and Hoffman, 1984; Kende, 1993). A series of studies using various signal transduction antagonists showed that an increase in intracellular Ca²⁺ is involved in the transduction of an ethylene signal, leading to the formation of aerenchyma in roots of maize under hypoxia (He et al., 1996). Ca²⁺ may also be involved in the signaling pathway leading to the activation of ADH and glycolytic genes. There is a transient increase in cytosolic Ca²⁺ concentration in the early stage of the flooding of maize roots (Subbaiah et al., 1994a, 1994b). Inhibition of this transient cytosolic Ca²⁺ increase blocked the induction of the ADH1 gene. A similar anoxic/hypoxic-inducible Ca²⁺ increase was observed in Arabidopsis (Sedbrook et al., 1996) and Ca²⁺ signaling is required for the activation of the Arabidopsis ADH gene (Chung and Ferl, 1999).

Although ethylene was shown to be involved in aerenchyma formation, its functional role in the hypoxic induction of ADH remains to be determined. In this report we examined the effect of aminooxy acetic acid (AOA), an inhibitor of ethylene biosynthesis, on the hypoxic induction of ADH in Arabidopsis. In addition, we also examined the hypoxic induction of ADH in mutants that are defective in ethylene responses. Our results suggested that an ethylene signal is required, but not solely responsible, for the induction of ADH during hypoxia.

	RESULTS

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

Effects of AOA on the Hypoxic Induction of ADH::-glucuronidase (GUS) Transgene

Ethylene is synthesized from S-adenosyl-Met (SAM) via ACC (Adams and Yang, 1979). AOA is a competitive inhibitor of ACC synthase, which catalyzes the conversion of SAM to ACC (Yang and Hoffman, 1984; Abeles et al., 1992). We used the AG2 Arabidopsis transgenic line, which contains an ADH promoter and GUS-coding region fusion (Conley et al., 1999), to examine the effect of AOA on the hypoxic induction of ADH. Arabidopsis plants were subjected to hypoxic treatment for 24 h in the presence of different concentrations of AOA. The data show that there was a dosage-dependent inhibition of hypoxic induction of the ADH::GUS transgene by AOA (Fig. 1). Hypoxic treatment resulted in an 8- to 10-fold increase in GUS activity as compared with the normoxic controls (Fig. 1, columns 1 and 2). The addition of 100 µM AOA resulted in nearly a 50% reduction in the accumulation of GUS activity during hypoxia (Fig. 1). However, further increases in concentrations of AOA resulted in no further reduction in levels of GUS activity. The partial inhibitory effect of AOA could be reversed by the addition of 10 µM ACC (Fig. 1, column 8) and completely reversed by 50 to 100 µM ACC (Fig. 1). These results suggest that an ethylene signal may be involved in the hypoxic induction of ADH and that AOA exerted its effect by blocking the biosynthesis of ethylene.

View larger version (60K): [in this window] [in a new window]   Figure 1.   Dosage effect of AOA on the hypoxic induction of the ADH::GUS transgene in AG2 plants. AG2 plants were subjected to hypoxic treatment for 28 h in the presence of different concentrations of AOA (columns 2-7 on the left) or of 100 µM AOA plus various concentrations of ACC (right) and were harvested for GUS enzyme activity assays. Column 1 is the GUS activity from AG2 plants grown under normoxic conditions. GUS activity is expressed as pmol of 4-methylumbelliferone (4-MU) min1 mg1 protein. The data presented are the average of the determinations from three separate hypoxic treatments done on separate occasions. Plants grown at different times were used for replicated treatments. Bars represent SD.

Temporal Expression Patterns of the ADH Gene during Hypoxia

We next examined the effect of AOA on the temporal expression of the ADH::GUS transgene during hypoxia. Figure 2 shows that there was a gradual increase in GUS activity during the hypoxic treatment of AG2 plants, reaching a maximum at 28 h of hypoxic treatment. The addition of 100 µM AOA resulted in a 30% to 50% reduction in the accumulation of GUS activity in later stages of hypoxia. When ACC was included in the medium, the inhibitory effect of AOA was mostly reversed (Fig. 2). However, there was no significance difference in GUS activity between the controls and AOA-treated plants at early stages of hypoxia.

View larger version (65K): [in this window] [in a new window]   Figure 2.   Effects of AOA and ACC on temporal expression of the ADH::GUS transgene in AG2. AG2 plants were subjected to hypoxic treatment in different media. At different times, samples were harvested and assayed for GUS activity. Bar graphs at each time point (from left to right) represent activities from hypoxic treatment in Murashige and Skoog medium, Murashige and Skoog medium containing 100 µM AOA, or 100 µM AOA plus 100 µM ACC. GUS activity is expressed as pmol 4-MU min1 mg1 protein. The data presented are the average of six independent hypoxic treatments. Bars represent SD.

Northern-blot analysis was used to examine the effect of AOA and ACC on the temporal expression pattern of the endogenous ADH gene during hypoxia. The results from one set of representative northern blots are illustrated in Figure 3. The nuclear gene ACT2 that encodes actin from Arabidopsis, the expression of which was not affected by growth conditions (An et al., 1996; M.-C. Shih, unpublished data), was used as an RNA loading standard. Quantification of northern blots indicated that ADH mRNA levels increased gradually during hypoxia, reaching a maximal level after 24 to 28 h of hypoxic treatment (Fig. 3A). This pattern of mRNA accumulation for the endogenous ADH gene during hypoxia is very similar to that of the ADH::GUS transgene shown in Figure 2. As with GUS activity, the addition of AOA and ACC had no apparent effect on ADH mRNA levels during early stages of hypoxia (Fig. 3, B and C). At later stages of hypoxia, the addition of AOA resulted in 30% to 50% reduction in levels of ADH mRNA (Fig. 3B). However, when ACC was added, the inhibitory effect of AOA was mostly reversed (Fig. 3C). Taken together, these results suggest that an ethylene signal contributes to the induction of the ADH gene at later stages during hypoxia.

View larger version (76K): [in this window] [in a new window]   Figure 3.   Effects of AOA and ACC on the accumulation of ADH mRNA during hypoxia. AG2 plants were subjected to hypoxic treatment in Murashige and Skoog medium (A), Murashige and Skoog medium containing 100 µM AOA (B), or 100 µM AOA plus 100 µM ACC (C). Total RNA (10 µg) samples from these plants were fractionated by agarose gel electrophoresis and were hybridized to the ADH or ACT2 probes. The numbers on top of each lane represent the time (in hours) under hypoxia. Each northern-blot analysis was repeated three times using RNAs prepared from three independent hypoxic treatments.

Production of Ethylene during Hypoxia

If ethylene is involved in the hypoxic induction of ADH, one would expect an increase in ethylene production in hypoxic-treated Arabidopsis. To investigate this possibility, AG2 plants were subjected to different lengths of hypoxic treatment and were harvested for measurement of ethylene production. The production of ethylene increased rapidly in the first 4 h of hypoxic treatment (Fig. 4). Rates of ethylene production remained roughly constant between 8 and 16 h. Between 20 and 24 h of hypoxia, there was a second increase in the rate of ethylene production. However, ethylene production started to decrease after 28 h of hypoxia. This pattern of hypoxia-induced ethylene production is similar to that of flooded tomato plants (Olson et al., 1995; Shiu, et al., 1998).

View larger version (21K): [in this window] [in a new window]   Figure 4.   Production of ethylene in AG2 during hypoxia. AG2 plants subjected to hypoxic treatment () or normoxic treatment () for different time periods were harvested and assayed for ethylene production as described in "Materials and Methods." The mean of two independent hypoxic treatments is plotted. Bars represent SD.

Ethylene Alone Is Not Sufficient to Induce ADH under Normoxia

Two experiments were performed to determine whether ethylene alone is sufficient to activate ADH gene without a hypoxic signal. First, we investigated whether applying exogenous ethylene can induce ADH gene expression under normoxic conditions. Two- to 3-week-old plants were transferred to liquid Murashige and Skoog media containing 10 µM ethephon, which is an ethylene-generating compound (Abeles et al., 1992), and bubbled continuously with air. The data showed that ethephon alone could not induce the ADH::GUS transgene (Fig. 5A) or the endogenous ADH gene (Fig. 5B) in AG2 plants under normoxia. Second, we found that eto1-1, a mutant that overproduces ethylene in etiolated seedlings (Woeste et al., 1999), also produces a higher level of ethylene under growth conditions used in our laboratory. This ethylene level is comparable with that of AG2 under hypoxia (data not shown). When eto1-1 plants were subjected to normoxic treatment, there was no induction of ADH activity (Fig. 5B). In a similar manner, there was no detectable ADH mRNA level in eto1-1 plants grown under normoxic conditions. In addition, we found that ADH is induced by hypoxia in eto1-1 plants to the same extent as in wild-type plants (data not shown). Taken together, these results show that an addition of exogenous ethylene or an increase in cellular ethylene alone is not enough to activate the transcription of ADH. Therefore, we conclude that ethylene is required, but not sufficient for, the induction of ADH during hypoxia.

View larger version (32K): [in this window] [in a new window]   Figure 5.   Effects of ethylene on the expression of ADH in plants under hypoxia, normoxia, or normoxia plus ethephon treatment. GUS and ADH activities of AG2 and eto1-1 plants subjected to various treatments were determined. Bar graphs at each time point (from left to right) represent activities for AG2 under hypoxia, AG2 under normoxia, AG2 under normoxia with 10 µM ethephon added, and (B only) eto1-1 under normoxia. GUS activity is expressed as pmol 4-MU min1 mg1 protein. A unit of ADH enzyme is defined as an increase in the production of 1 nmol NADH min1 mg1 protein. The data presented are the average of three independent experiments done on separate occasions. Error bars indicate SD.

Hypoxic Induction of ADH Is Affected in etr1-1 and ein2-1

Several different classes of mutants that fail to display the triple response in the presence of saturating levels of exogenously applied ethylene have been isolated (Guzman and Ecker, 1990; Roman et al., 1995). We chose to examine temporal expression patterns of ADH during hypoxia in two of these mutants, etr1-1 and ein2-1. The ETR1 gene was identified and found to encode a receptor protein with homology to two-component regulators (Chang et al., 1993). Although ETR is present as a small gene family in Arabidopsis, mutations in one of the ETR genes result in a dominant phenotype and cause defects in many ethylene responses. EIN2 was shown to be a bifunctional transducer and may mediate crosstalk between ethylene and stress responses (Alonso et al., 1999).

Northern-blot analysis shows that ADH mRNA levels increased during hypoxia in etr1-1 (Fig. 6A) and ein2-1 (Fig. 6B). These blots were quantified using the ADH mRNA level from AG2 plants treated with 24 h of hypoxia (Fig. 6, A and B, lane 1) as 100%. Levels of ADH mRNA were similar among etr1-1, ein2-1, and AG2 in the first 4 to 8 h of hypoxic treatment (Fig. 6C). However, ADH mRNA levels in etr1-1 and ein2-1were about 30% to 50% lower than those of AG2 between 12 and 36 h of hypoxic treatment (Fig. 6C). These results indicated that mutations affecting ethylene responses could also affect the induction of ADH gene at later stages of hypoxia.

View larger version (40K): [in this window] [in a new window]   Figure 6.   Hypoxic induction of ADH in etr1-1 and ein2-1. RNA samples from etr1-1 (A) and ein2-1 (B) subjected to hypoxic treatment were analyzed by northern-blot analysis. Digitized images of the ADH bands were quantified and normalized to the ACT2 band in each lane using the National Institutes of Health Image Analysis Program 1.62f. The normalized ADH mRNA level from 24-h hypoxic-treated AG2 plants (lane 1) was used as the 100% level. The quantification data presented in C are the average of three independent hypoxic treatments. Bars indicate SD.

	DISCUSSION

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

Two cellular changes are known to occur in plants under oxygen deficiency: switching from aerobic respiration to anaerobic fermentation and the formation of aerenchyma tissues (Drew, 1997). Switching from aerobic respiration to anaerobic fermentation involves the induction of glycolytic and fermentative genes. Although much progress had been made in recent years in the identification of cis- and trans-acting regulatory elements of the hypoxic inducible genes, how the hypoxic signal is transduced in plant cells to trigger these cellular changes remains largely unknown.

Our studies indicated that ethylene, which is known to be involved in various stress responses in different plant species, is involved in the hypoxic induction of the ADH gene in Arabidopsis. We showed that AOA, which is an inhibitor of ACC synthase and hence an inhibitor of ethylene biosynthesis, could reduce the hypoxic induction of ADH::GUS transgene in a dosage-dependent manner (Fig. 1). However, AOA is also known to inhibit other processes such as Gly oxidation (Dry and Wiskich, 1986). The inhibitory effect of AOA on the hypoxic induction of ADH, therefore, could be due to its effect on ethylene production or on other cellular metabolism. If the response is mediated by an ethylene signal, an addition of ACC to the medium should reverse the inhibitory effect of AOA. Our data showed that when 50 to 100 µM of ACC is added, the inhibition of AOA on the induction of the ADH::GUS transgene during hypoxia was mostly reversed (Fig. 1). The amounts of ACC required to reverse the inhibitory effect of AOA on ADH induction is greater than the amounts required to elicit the triple response in etiolated Arabidopsis seedlings. It has been shown that ACC at concentrations between 10 and 100 µM has a saturating effect on the triggering of triple responses in etiolated Arabidopsis seedlings (Luschnig et al., 1998). For most plant species, the conversion of SAM to ACC, which is catalyzed by ACC synthase, is the rate-limiting step during ethylene biosynthesis (Yang and Hoffman, 1984). However, the conversion of ACC to ethylene, which is catalyzed by ACC oxidase, requires oxygen. It is likely that the conversion of ACC to ethylene would become rate limiting under very low oxygen concentration. If this were the case, higher cellular levels of ACC will be needed as a substrate to synthesize sufficient amounts of ethylene in hypoxic-treated plants. It was found that the conversion of ACC to ethylene becomes the rate-limiting step for ethylene synthesis during submergence of Rumex palustris and that higher cellular ACC levels were observed in submerged R. palustris plants (Banga et al., 1996; Vriezen et al., 1999).

There are two major classes of ethylene response mutants in Arabidopsis. One involves mutants that display constitutive triple ethylene responses, which result from either ethylene overproduction (eto1, eto2, and eto3) or constitutive activation of the pathway (ctr1), and the other involves mutants that are insensitive to ethylene, which can be due to defects in their ability to perceive (etrt1, etr2, ein4, ers, and other receptor mutants) or respond (ein2, ein3, and ein5) to ethylene (Guzman and Ecker, 1990; Roman et al., 1995). The analysis of these mutants has allowed much progress in elucidating the mechanisms of ethylene perception and signal transduction (for review, see Kieber, 1997; Bleecker et al., 1998; Johnson and Ecker, 1998; Theologis, 1998). Since we found that ethylene may contribute to the signaling pathways leading to the induction of ADH during hypoxia, we expect that hypoxic induction of ADH will be affected by mutations in the ethylene-insensitive class. Our studies showed that ADH mRNA levels in etr1-1 and ein2-1 were about 30% to 50% lower than those of AG2 during hypoxia (Fig. 6). In a similar manner, we found that levels of ADH activity in both mutants were lower than those of AG2 during hypoxia (data not shown). These results provide supporting evidence for the involvement of ethylene in the hypoxic induction of ADH in Arabidopsis.

Our observation that AOA could not completely block the hypoxic induction of ADH (Fig. 1) suggests that an ethylene-independent pathway is also involved. Consistent with this hypothesis, we found that AOA is effective in reducing the expression of ADH::GUS transgene (Fig. 2) and the endogenous ADH (Fig. 3) only in later stages of hypoxia. In contrast, there is no apparent difference in levels of GUS activity and ADH mRNA in early stages of hypoxia between hypoxic-treated AG2 plants in the absence or presence of AOA. In a similar manner, we found that ADH mRNA levels in etr1-1 and ein2-1were reduced in later stages, but not in earlier stages, during hypoxia (Fig. 6). These results can best be interpreted as that two signaling pathways, one ethylene-independent and one ethylene-dependent, are involved in the hypoxic induction of ADH in Arabidopsis and that AOA and mutations in etr1-1 and ein2-1 affect only the ethylene-dependent pathway.

Ethylene is involved in many physiological and developmental processes in plants (Yang and Hoffman, 1984; Kende, 1993). In some instances, ethylene function requires a concomitant contribution of other signaling molecules (Penninckx et al., 1998). In fact, it was shown that an addition of exogenous ethylene could not induce the formation of aerenchyma in anoxic roots in maize, although ethylene is required for the hypoxia-induced aerenchyma formation (He et al., 1994, 1996; Drew, 1997). It was reported that an addition of AOA completely inhibits the induction by flooding of a xyloglucan endo-transglycosylase gene in maize roots (Saab and Sachs, 1996). Under the same condition, the induction of the ADH1 gene decreased slightly. These results suggest that an ethylene-signaling pathway is sufficient for the induction of the xyloglucan endo-transglycosylase gene and that an ethylene-independent pathway is mainly responsible for the induction of ADH1 in flooded maize roots. In Arabidopsis, we found that an application of exogenous ethylene or an increased cellular ethylene in eto1-1 is not capable of inducing the expression of ADH during normoxia. These results suggest that ethylene is necessary, but not solely responsible, for the induction of ADH during hypoxia.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

Growth Conditions and Stress Treatment

Seeds of Arabidopsis AG2 were surface sterilized and treated with 15 µM gibberellin at 4°C overnight. Seeds were sown onto plates with Murashige and Skoog medium containing 1% (w/v) Suc and 0.8% (w/v) agar and were grown at 20°C with 16-h light/8-h dark cycles. After 1 week, seedlings were transferred to fresh Murashige and Skoog plates containing 2% (w/v) agar and were grown for additional 7 to 10 d with plates in vertical positions. For hypoxic treatments, plants were submerged in liquid Murashige and Skoog medium through which gas containing 4.5% to 5% (w/v) oxygen and balanced with nitrogen was bubbled continuously.

GUS and ADH Enzymatic Assays

GUS enzyme activity assays were performed essentially as described by Jefferson et al. (1987). Fluorescence of the 4-methylumbelliferyl product was quantified using a minifluorometer (model TKO-100, Hoefer Scientific Instruments, San Francisco). ADH activity assay was performed according to the procedures described in Xie and Wu (1989). The assay uses ethanol as the substrate and measures the production of NADH. Measurement of NADH formation was performed in a spectrophotometer (DU 64, Beckman Instruments, Fullerton, CA). A unit of ADH is defined as the production of 1 nmol of NADH min¹ mg¹ protein.

RNA Isolation and Northern-Blot Analysis

Total RNA was isolated by an acidic phenol protocol adapted from the procedures described by Chomczynski and Sacchi (1987). RNA samples (10 µg) were denatured in 6.5% (w/v) formaldehyde/50% (w/v) formamide at 65°C and electrophoresed through 1.2% (w/v) agarose gels with 6.5% (w/v) formaldehyde/1× MOPS [3-(N-morpholino)-propanesulfonic acid] buffer as described in Sambrook et al. (1989). RNA was transferred to a Magnacharge 0.45-µm nylon membrane (Micron Separations, Westborough, MA) overnight in 10× SSC and 0.1% (w/v) SDS. Filters were hybridized with random primer-labeled cDNA probes (Feinberg and Vogelstein, 1983). Final post-hybridization washings were performed at 65°C in 0.1× SSC/0.1% (w/v) SDS. The hybridization probes were as follows: ADH, a 525-bp cDNA fragment generated from reverse transcriptase-PCR based on the sequence from Chang and Meyerowitz (1986); and ACT2, a 800-bp cDNA fragment for Arabidopsis Actin2 gene generated by reverse transcriptase-PCR based on the sequence from An et al. (1996). Membranes were exposed to film (XAR-5, Eastman-Kodak, Rochester, NY) with intensifying screens at 70°C. Quantification was performed by scanning autoradiograms and analyzing the images using the National Institutes of Health Image Analysis Program 1.62f.

Measurement of Ethylene Production

AG2 plants were grown and subjected to hypoxic treatment exactly as described in prior sections. At different time intervals, 10 plants were collected in a 13 × 100-mm test tube and capped for 1 h at room temperature. The amounts of ethylene produced were measured as described by Jackson and Campbell (1976).

Chemicals and Seeds

AOA, gibberellin A₃, antibiotics, and other chemicals were purchased from Sigma Chemical (St. Louis). 5-Bromo-4-chloro-3-indolyl--D-GlcUA cyclohexylammonium salt was purchased from Gold Biotechnology (St. Louis). Restriction and modification enzymes were from New England BioLabs (Boston) and Promega (Madison, WI). Seed stocks for etr1-1 and ein2-1 were obtained from the Arabidopsis Biological Resources Center at the Ohio State University.