RopGAP4-Dependent Rop GTPase Rheostat Control of Arabidopsis Oxygen Deprivation Tolerance

See allHide authors and affiliations

Science  14 Jun 2002:
Vol. 296, Issue 5575, pp. 2026-2028
DOI: 10.1126/science.1071505


Transient soil flooding limits cellular oxygen to roots and reduces crop yield. Plant response to oxygen deprivation involves increased expression of the alcohol dehydrogenase gene (ADH) and ethanolic fermentation. Disruption of the Arabidopsis gene that encodes Rop (RHO-like small G protein of plants) guanosine triphosphatase (GTPase) activating protein 4 (ROPGAP4), a Rop deactivator, elevates ADH expression in response to oxygen deprivation but decreases tolerance to stress. Rop-dependent production of hydrogen peroxide via a diphenylene iodonium chloride–sensitive calcium-dependent reduced nicotinamide adenine dinucleotide phosphate (NADPH) oxidase is necessary for induction of both ADH and RopGAP4 expression. Tolerance to oxygen deprivation requires Rop activation andRopGAP4-dependent negative feedback regulation. This Rop signal transduction rheostat balances the ability to increase ethanolic fermentation with survival.

Plant endurance of transient flooding requires increased production of adenosine triphosphate through glycolysis and regeneration of nicotinamide adenine dinucleotide through ethanolic fermentation (1,2). Signal transduction processes that control changes in gene expression in O2-deprived cells involve oscillations in cytosolic free Ca2+(3–6). To identify the genes involved in regulating the expression of the sole alcohol dehydrogenase gene (ADH) of Arabidopsis thaliana, we screened lines carrying a gene-trap transposon (DsG) (7) for increased β-glucoronidase (GUS) histochemical staining and altered induction of ADH specific activity in response to O2 deprivation under low light (8) (see supplementary methods). We identified a line that displayed elevated GUS staining throughout the seedling vasculature in response to low O2 (Fig. 1A) but with no apparent abnormalities under control conditions. This line contained a single DsG transposon inserted into the first exon ofRopGAP4 (GTPase activating protein; 49 kD) (Fig. 1B) [GenBank accession number AC008153; MIPS At3g11490 (Munich Information Center for Protein Sequences identifier for Arabidopsis ROPGAP4 on chromosome 3); BAC F24K9.16 (Bacterial Artificial Chromosome number F29 and gene identifier #16), position 61811], resulting in a translational fusion within the CRIB (Cdc42/Rac-interactive binding) motif at the amino terminus of RopGAP4 (Fig. 1B). We designated this mutant alleleropgap4-1.

Figure 1

Characterization ofropgap4-1 and mRNA levels after O2deprivation. (A) Histochemical staining of 7-day-old ropgap4-1 seedlings with 5- bromo-4-chloro-3-indolyl-beta-d-glucuronic acid after O2 deprivation or 24-hour exposure to 5 mM caffeine. (B) Site and orientation of DsGinsertion within the CRIB motif of the first exon ofRopGAP4 in ropgap4-1. A, Ala; D, Asp; F, Phe; H, His; I, Ile; P, Pro; R, Arg; V, Val. (C) RT-PCR detection of ADH, RopGAP4, and actin (ACT2) mRNA in WT, ropgap4-1,CA-rop2, and DN-rop2seedlings after O2 deprivation.

RopGAPs were identified in a yeast two-hybrid system based on interaction with the RHO-like small G-protein of plants, Rop (9). RopGAPs possess a conserved GAP-like domain and a CRIB motif that enhances Rop interaction, allowing for efficient GTP hydrolysis (9). Rop signaling controls intracellular Ca2+ gradients and actin cytoskeletal dynamics required for tip growth of pollen (10–16) and polar growth of root hairs (17). Activation of Rop signaling is implicated in defense responses and developmental processes involving hydrogen peroxide (H2O2) (18–20), whereas inactivation of Rop signaling is necessary for abscisic acid–induced closure of leaf stomata (21).

RopGAP4 mRNA accumulation increased dramatically in response to O2 deprivation in wild-type (WT) seedlings, as detected by reverse transcriptase–polymerase chain reaction (RT-PCR) (Fig. 1C). RopGAP4 mRNA was not detectable inropgap4-1 seedlings, which indicates that theDsG insertion resulted in a loss-of-function mutation.

ropgap4-1 allowed us to consider whether Rop signaling is involved in regulating ADHexpression in response to O2 deprivation.ropgap4-1 seedlings showed a more rapid and dramatic increase in ADH mRNA accumulation and ADH specific activity threefold higher than WT after 12 hours of O2deprivation; paradoxically, they were more sensitive to the stress (Figs. 1C and 2A; Table 1). After 24 hours of O2deprivation, ADH mRNA and specific activity levels dropped dramatically and ropgap4-1 seedlings were unable to recover upon reoxygenation. Seedlings of a line expressing a dominant negative form of Rop2 [35S::DN-rop2 (T20N)] (22) showed no detectable induction of ADH mRNA or specific activity after O2 deprivation and increased stress sensitivity. This confirms that signaling through the Rop GTPase is mandatory for activation of ADH expression, a prerequisite for low O2 tolerance (8,23). In a line expressing a constitutive active form of Rop2 [35S::CA-rop2 (G15V)] (22), ADH specific activity was higher under control conditions and inducible by O2 deprivation. The limited induction of ADH in CA-rop2 versus the excessive induction in ropgap4-1 can be explained by negative feedback regulation of Rop signaling by ROPGAP4 (see below).

Figure 2

Rop signaling and H2O2production regulate ADH expression. (A) ADH specific activity in seedlings after O2 deprivation in the absence or presence of 30 μM DPI. (B) Rop-RIC1 interaction assay on extracts from WT and ropgap4-1 seedlings after O2 deprivation. Immunoblot shows detection of levels of total Rop (Rop-GTP and Rop-GDP) in crude extracts or Rop-GTP obtained by pull-down through interaction with RIC1-maltose binding protein. Data are representative of three independent experiments. (C) H2O2 levels after O2deprivation. In (A) and (C), values are mean ± SE of three independent experiments. Asterisk indicates a significant difference from WT at the same time point (P < 0.01; Student'st test).

Table 1

Effect of O2 deprivation and DPI treatment on seedling survival. +, Addition of 30 μM DPI in 3% dimethyl sulfoxide solvent; –, addition of solvent. Data are mean ± SE from three independent experiments.

View this table:

We confirmed transient activation of Rop signaling by O2deprivation with an assay that detects Rop-GTP by interaction with Rop-interacting CRIB motif-containing protein (RIC1) (24). Figure 2B compares the level of Rop in total cell extracts (Rop-GTP and Rop-GDP) with RIC1-interacting Rop (Rop-GTP) over 36 hours of O2 deprivation. Rop-GTP levels increased in WT seedlings after 1.5 hours, increased through 12 hours, and then decreased. Rop-GTP levels were constitutively high inropgap4-1 seedlings and increased in response to low O2 but showed no decrease even after 36 hours. O2 deprivation promotes activation of Rop, and RopGAP4 appears to negatively regulate this activation in WT seedlings.

Cotyledons of ropgap4-1 seedlings turned brown upon reoxygenation, whereas those ofCA-rop2, DN-rop2, and WT remained green, which led us to suspect that ropgap4-1seedlings succumb to O2 deprivation and reoxygenation as a result of oxidative stress. ropgap4-1 seedlings may fail to control the production of reactive oxygen species, because overexpression of a constitutive active form of Rop results in increased production of H2O2 via a reduced nicotinamide adenine dinucleotide phosphate (NADPH) oxidase in several plant species (18–20), and a GAP negatively regulates Rac GTPase activation of NADPH oxidase in mammals (25). We tested whether the response to O2deprivation was affected by treatment of seedlings with diphenylene iodonium chloride (DPI), which inhibits production of superoxide by flavin-containing NADPH oxidases and the resultant accumulation of H2O2. In all four genotypes, DPI reduced ADH activity under control and low O2 conditions, which indicates that ADH induction requires a DPI-sensitive NADPH oxidase (Fig. 2A). DPI also reduced the duration of stress that WT seedlings survived from >24 to <12 hours (Table 1). DPI treatment reduced ADH induction inropgap4-1 seedlings and increased their survival after O2 deprivation, which reveals that the inability to down-regulate DPI-sensitive NADPH oxidase reduces stress tolerance. Consistently, survival after O2 deprivation was improved inCA-rop2 and impaired inDN-rop2 seedlings in the presence of DPI.

H2O2 levels increased in response to O2 deprivation in WT, ropgap4-1, andCA-rop2 seedlings but did not change significantly in DN-rop2 seedlings (Fig. 2C), which supports a role of Rop signaling in H2O2production. In WT seedlings, H2O2 level and ADH specific activity increased coordinately over 24 hours of stress. H2O2 levels in ropgap4-1seedlings under control conditions and after 6 and 12 hours of O2 deprivation were significantly higher than in WT seedlings, consistent with ADH specific activity data. High H2O2 in the mutant may contribute to reduced stress tolerance. In CA-rop2 seedlings, H2O2 levels correlated with constitutively high ADH specific activity under control conditions but were not clearly responsible for intolerance of low O2.

ropgap4-1 seedlings have constitutively high levels of Rop-GTP but near normal levels of ADH specific activity until they are deprived of O2, which indicates that accumulation of Rop-GTP is insufficient for induction of ADH. An increase in cytosolic free Ca2+ due to organellar efflux or apoplastic influx is necessary for activation of ADHexpression in Arabidopsis (3). Treating maize cells with low levels of caffeine stimulates ADH1 expression and promotes an increase in cytosolic free Ca2+, similar to that observed in response to anoxia (4, 5). Caffeine treatment under nonstress conditions induced ADH specific activity to significantly higher levels than the maximal level observed in response to low O2 in all four genotypes (Fig. 3A). DPI effectively blocked the caffeine-stimulated increase in ADH specific activity and the concomitant increase in H2O2 (Fig. 3, A and B). The caffeine-promoted increase in ADH specific activity, consistent with O2 deprivation, was dramatic inropgap4-1 and limited inCA-rop2 seedlings. InDN-rop2 seedlings, the caffeine-stimulated induction may result from a Rop-independent mechanism or interaction between a Ca2+ signal and the residual activity of endogenous Rops. Topical application of a H2O2regenerating system, glucose and glucose oxidase, resulted in a rapid and efficient increase in ADH specific activity in WT seedlings (Fig. 4A), which confirms that H2O2 is a second messenger in ADHregulation.

Figure 3

ADH activity is stimulated by caffeine treatment via Rop-induced and DPI-sensitive H2O2 production. (A) ADH specific activity in seedlings treated with caffeine and/or DPI for 24 hours. (B) H2O2 levels in seedlings analyzed in (A). Values are mean ± SE of three independent experiments. Asterisk indicates significant difference from the maximal level detected after O2 deprivation (P < 0.01; Student's t test). (C) GUS specific activity in ropgap4-1 seedlings after O2deprivation, caffeine treatment, and DPI treatment. Values are mean ± SE of three independent experiments.

Figure 4

ADH activity and ROPGAP4 expression is induced by a H2O2-regenerating system. ADH specific activity in WT (A) and GUS specific activity inropgap4-1 seedlings (B) treated with glucose and glucose oxidase for up to 3 hours.

These results reveal that O2 deprivation stimulates a Rop signal transduction pathway, activating a DPI-sensitive NADPH oxidase that results in increased H2O2 production, which acts as a second messenger in the induction of ADHexpression (fig. S1). An increase in cytosolic free Ca2+appears to be necessary to complete this Rop-mediated signal. This could be due to the binding of Ca2+ by the plasma membrane DPI-sensitive NADPH oxidase gp91phox subunit (26) or to a Ca2+-dependent DPI-sensitive NAD(P)H dehydrogenase/oxidase of the inner mitochondrial membrane (27).

The attenuation of Rop signal transduction is also necessary for tolerance of O2 deprivation. Several lines of evidence indicate that Rop signaling drives this attenuation by activatingRopGAP4 expression. (i) Low O2 promotedRopGAP4 mRNA accumulation in WT but notDN-rop2 seedlings (Fig. 1C). (ii) GUS activity increased in ropgap4-1 seedlings in response to low O2 and caffeine, but it was blocked by DPI (Figs. 1A and 3C). (iii) Application of a H2O2regenerating system elevated GUS activity inropgap4-1 seedlings (Fig. 4B). (iv)RopGAP4 mRNA levels were constitutively elevated inCA-rop2 seedlings (Fig. 1C).

Thus, a Rop rheostat regulates the production of H2O2 that is required to trigger the expression of beneficial genes (for example, ADH) and the avoidance of H2O2-induced cell death. Rop signaling is controlled by negative feedback regulation through the stimulation ofRopGAP4 transcription by H2O2. The termination of Rop signaling by RopGAP4 would alleviate oxidative stress and limit consumption of carbohydrate reserves via glycolysis and ethanolic fermentation. The reduced O2 deprivation tolerance of the DN-rop2,CA-rop2, and ropgap4-1seedlings underscores the requirement for a fully functional Rop rheostat. We propose that a Rop rheostat is critical to developmental processes and environmental stress responses that use H2O2 as a second messenger or enhance H2O2 accumulation, including the response to abscisic acid, auxin, pathogen infection, and numerous abiotic stresses. Manipulation of the Rop signal transduction rheostat may enhance the productivity of crops that undergo transient submergence or soil waterlogging.

  • * To whom correspondence should be addressed. E-mail: serres{at}


View Abstract

Stay Connected to Science

Navigate This Article