INTRODUCTION

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

Carbonic anhydrase (CA, EC 4.2.1.1) is a zinc-containing metalloenzyme that catalyzes the reversible hydration of CO₂ to HCO₃. The widespread abundance of CA isoforms in plants, animals, and microorganisms suggest that this enzyme has many diverse roles in biological processes. CA plays a critical role in biological systems because CO₂ gas is the membrane permeable form of inorganic carbon for cells, and, in general, the uncatalyzed interconversion between HCO₃ and CO₂ is slow when compared with the required rate in living cells (Badger and Price, 1994).

In photosynthetic organisms, one generally accepted physiological role of CA is to provide sufficient levels of inorganic carbon as part of a CO₂-concentrating mechanism for improved photosynthetic efficiency. In Chlamydomonas reinhardtii, Badger and Price (1992) suggested that chloroplastic CA plays a role in photosynthetic carbon assimilation by converting accumulated pools of HCO₃ to CO₂, which is the substrate for Rubisco. Moroney et al. (1985) revealed that the reduction of periplasmic CA activity by using CA-specific inhibitors significantly reduced the efficiency of external inorganic carbon utilization for photosynthesis. Like green algae, CA in cyanobacteria plays an important role in the CO₂-concentrating mechanism and in photosynthesis (Badger and Price, 1994). Price et al. (1992) reported that there was an association of CA with Rubisco in the carboxysome of cyanobacteria (Synechococcus sp.). Also, Maeda et al. (2000) identified a gene (CmpA) that encodes a substrate-binding protein that can specifically bind to HCO₃ in cyanobacteria, which may further aid in the diffusion of HCO₃ and elevation of CO₂ in the carboxysome. In C₄ plants, CA is localized to the cytosol of mesophyll cells, where it supplies HCO₃ to phosphoenolpyruvate carboxylase (Burnell and Hatch, 1988; Hatch and Burnell, 1990), and calculations show that without CA, the rate of photosynthesis is reduced. In fact, a recent study with transgenic Flaveria bidentis clearly supported a role for CA in the assimilation of CO₂ in C₄ plants (Ludwig et al., 1998). CA in C₃ plants is distributed primarily in the stroma of chloroplasts (Poincelot, 1972) and is speculated to mediate in the diffusion of CO₂ from the cytosol to the site of carboxylation by Rubisco in the chloroplast stroma during photosynthesis (Reed and Graham, 1981). Thus, the principal function of CA in photosynthetic organisms is to support the efficient assimilation of inorganic carbon for the primary carboxylation reactions.

In animals, CAs have been shown to provide inorganic carbon to other important metabolic pathways such as pyrimidine biosynthesis, gluconeogensis, and lipogenesis (Sly and Hu, 1995), all requiring HCO₃ as the inorganic carbon substrate for initial carboxylation reactions. Cytological and biochemical evidence point to a metabolic role for CA in lipogenesis. For example, acetyl-coenzyme A (CoA) carboxylase (ACCase) was colocalized with CA in oligodendrocytes and fatty acid synthase was localized by immunostaining to be in the same cell type (Cammer, 1991). Lipogenesis was inhibited by acetazolamide, a CA-specific inhibitor, in human adipose tissue (Bray, 1977). The administration in vivo of acetazolamide in female mice resulted in decreased fatty acid synthesis (Cao and Rous, 1978). In addition, Herbert and Coulson (1983) demonstrated that de novo fatty acid synthesis (measured by [¹⁴C]acetate incorporation into total lipid) in liver of American chameleons (Anolis carolinensis) was inhibited by CA-specific inhibitors (ethoxyzolamide and acetazolamide). CA was suggested to play a role in de novo lipogenesis in hepatocytes by increasing the rate of CO₂ hydration to bicarbonate for ACCase (Dodgson et al., 1984), and Lynch et al. (1995) reported a reduction of [¹⁴C]acetate incorporation into total lipid in rat hepatocytes incubated with CA inhibitors trifluoromethylsulphonamide and ethoxyzolamide. Together, these results support the notion that CA assists in providing HCO₃ for lipid biosynthesis in animal systems.

Evidence in the literature from other eukaryotes and some preliminary results from our laboratory lead us to formulate a working hypothesis that plastidial CA in C₃ plants plays a role in plastidial lipid biosynthesis. cDNAs were identified recently in a dark-grown cotton (Gossypium hirsutum) seedling library that encoded functional CA enzymes with putative plastid-targeting sequences (Hoang et al., 1999). The expression of CA increased (estimated by relative mRNA abundance and specific enzyme activity) during the period of storage lipid accumulation in maturing embryos of cotton (Hoang et al., 1998; Fig. 1). In developing oilseeds, the majority of fatty acids that are synthesized in plastids are exported to the endoplasmic reticulum and converted to storage lipids (Ohlrogge and Browse, 1995). ACCase requires HCO₃ as a substrate and is considered to be the rate-limiting and committed step in fatty acid biosynthesis (Ohlrogge and Jaworski, 1997). Moreover, the ACCase K_m for HCO₃ is reported to be quite high from several plant species, in the millimolar range (Nikolau and Hawke, 1984). This suggests that in most plant systems, ACCase may be operating far below saturation because in higher plants the dissolved CO₂ concentration is in the micromolar range (Badger and Price, 1994). Here, we provide evidence to support our hypothesis in three different plant systems: (a) maturing embryos of developing cotton seeds, (b) cell suspensions of tobacco (Nicotiana tabacum), and (c) leaves of transgenic tobacco plants. Our results indicate that treatment of plant cells with CA-specific inhibitors reduced the rate of lipid synthesis (from [¹⁴C]acetate) in vivo and in vitro. In addition, molecular suppression of CA activity to 5% of wild-type (WT) levels (antisense-suppressed plants, Price et al., 1994) reduced lipid biosynthesis in chloroplasts from transgenic plants compared with chloroplasts from WT plants. We propose that CA is involved in lipid synthesis (and perhaps other HCO₃-requiring pathways in plastids) indirectly, serving to "concentrate" CO₂ in plastids as HCO₃ and reduce the rate of CO₂ diffusion out of plastids.

View larger version (23K): [in this window] [in a new window]   Figure 1.   Comparison of triacylglycerol content (A), total CA activity (B), and relative CA expression (C) in embryos excised from cotton bolls at 25 and 40 DPA. The relative amounts of TAG in cell-free homogenates were estimated by scanning densitometry (National Institutes of Health Image) of thin-layer chromatography (TLC)-fractionated lipid classes in comparison with triolein standards. Total CA activity was determined electrometrically in cell free homogenates. Poly(A+) RNA was isolated from cotton embryos and the ratios of CA to actin transcripts were evaluated by northern-blot analyses. The results depicted here are representative of replicate experiments.

	RESULTS

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

CA and Lipid Synthesis in Embryos

To examine whether CA could be involved in lipid biosynthesis in developing cotton embryos, we compared CA activities and expression before (25 DPA) and during (40 DPA) the maximum period of TAG accumulation (Fig. 1). Total CA activity and steady-state CA mRNA levels increased approximately 15- and 9- fold, respectively, during embryo maturation and oil accumulation. It is possible that this substantial increase in CA activity is required to increase the efficiency of inorganic carbon assimilation for lipid biosynthesis in developing embryos.

To further determine whether CA could play a role in lipid biosynthesis in cotton embryos (32 DPA), CA activity was reduced by pre-incubation of embryo extracts with CA specific inhibitors (at 1 and 10 mM sulfanilamide, acetazolamide, or ethoxyzolamide; Fig. 2). Both ethoxyzolamide and acetazolamide at 1 or 10 mM concentration inhibited more than 50% of CA activity in homogenates of cotton embryos; however, ethoxyzolamide was the most potent CA inhibitor. Embryos at 32 DPA were radiolabeled with [¹⁴C]acetate in vivo and the incorporation of [¹⁴C]acetate into total cottonseed lipids was quantified (Fig. 3). Cotton embryos (30-38 DPA) incubated with CA inhibitors at 1 and 10 mM ethoxyzolamide showed a significant reduction in the rate of [¹⁴C]acetate incorporation into total lipids compared with controls (no inhibitors). A linear rate of [¹⁴C]acetate incorporation into lipids was established within 10 min of incubating embryos with [¹⁴C]acetate (Fig. 3) and continued up to 30 min. A 50% reduction in the rate of radiolabeled acetate incorporation into total lipids was observed for embryos treated with 10 mM ethoxyzolamide when compared with controls (no inhibitors). Application of inhibitors appeared to be specific for reserve lipid accumulation because storage protein synthesis, measured by [³⁵S]-Met incorporation into total protein, was relatively unaffected (Fig. 4A). We confirmed that CA-specific inhibitors did not inhibit ACCase activity significantly in vitro (Fig. 4B) to reduce lipid synthesis in vivo.

View larger version (45K): [in this window] [in a new window]   Figure 2.   Effects of different CA inhibitors in homogenates of cotton embryos (34 DPA). The graph shows the percent remaining CA activity when inhibited with 1 and 10 mM sulfanilamide, acetazolamide, and ethoxyzolamide. Values shown represent the means and SDs from three separate experiments. CA activity with dimethyl sulfoxide (DMSO) alone in the amounts required to dissolve the above inhibitors was 98.6% ± 4.0% of that without.

View larger version (12K): [in this window] [in a new window]   Figure 3.   Time-dependent incorporation of [14C]acetate into cottonseed total lipids. Embryos (30-38 DPA) were pre-incubated with DMSO (control) or different concentrations of CA inhibitors. Both 1 and 10 mM concentrations of ethoxyzolamide were incubated for 30 min before the addition of radiolabeled [14C]acetate. Total lipids were extracted at 10, 15, 20, and 30 min after the addition of the radiolabeled acetate. Incorporation of [14C]acetate into lipids was not linear after 30 min. Aliquots were used to quantify the incorporation of [14C]acetate into embryos by liquid scintillation counting. Data points represent mean and SD of three independent experiments for 10 and 30 min. Lines are plotted from linear regression analyses (Prism version 3.02, GraphPad Software, San Diego) of the data with r2 = 0.97 for controls, r2 = 0.97 for 1 mM ethoxyzolamide, and r2 = 0.94 for 10 mM ethoxyzolamide. Rates estimated from linear regression analyses were 12.36 ± 0.71, 9.42 ± 0.52, and 6.15 ± 0.53 pmol acetate min1 mg1 protein, respectively. For reference, the rate of acetate incorporation into embryos not pre-incubated with DMSO was 12.44 ± 0.62 pmol acetate min1 mg1 protein.

View larger version (32K): [in this window] [in a new window]   Figure 4.   Comparison of embryo protein synthesis in vivo (A) and ACCase activities in vitro in the absence or presence of CA inhibitors. Embryos (34 DPA) were pre-incubated with ethoxyzolamide and acetazolamide 30 min before the addition of [35S]-Met. After 1 h of incubation, the incorporation of [35S]-Met into trichloroacetic acid (TCA)-precipitated protein was quantified by liquid scintillation counting. ACCase activity was assayed in cell free homogenates of 34-DPA embryos. The embryo homogenates were pre-incubated with DMSO (control), 1 mM ethoxyzolamide, 10 mM ethoxyzolamide, or 10 mM acetazolamide for 30 min before assays. Values shown represent the mean and SD from three independent experiments. With DMSO added (the controls shown above), total protein synthesis was 92% ± 9% and ACCase activity was 94% ± 3%, respectively, of that without DMSO.

The distribution of radioactivity in major lipid classes after incorporation of [¹⁴C]acetate for 30 min in cotton embryos was evaluated by TLC (summarized in Table I). About 82% of the radioactivity was in polar lipids (mostly phosopholipids) and 18% was in the nonpolar lipids (free fatty acids and triacylglyercol) of untreated cotton embryos. Embryos incubated with 10 mM ethoxyzolamide or acetazolamide had a distribution of radioactivity in polar and nonpolar lipids (albeit lower overall) similar to DMSO controls, suggesting that lipid synthesis overall was reduced, and that inhibition was not selective for extraplastidial fatty acid elongation.

View this table: [in this window] [in a new window]   Table I.   Distribution of radioactivity in major lipid classes after incorporation of [14C]acetate in cotton embryos (30 min) and tobacco cell suspensions (4 h) Values represent the mean and SD of three independent experiments. Disintegrations per minute × 103.

To examine specifically whether de novo fatty acid synthesis in cotton embryos was influenced by CA inhibitors, we isolated plastids from cotton embryos (34 DPA). Acetate is commonly used as a radioactive tracer in in vitro studies of fatty acid biosynthesis because it can be incorporated efficiently into fatty acids (Qi et al., 1995; Roughan and Ohlrogge, 1996). Plastids treated with 10 mM ethoxyzolamide revealed an approximately 67% reduction of the rate of lipid synthesis (mostly free fatty acid product, determined by radiometric scanning of TLC-fractionated lipid classes as above) when compared with controls (Table II). Incorporation of [¹⁴C]acetate into plastid lipids was inhibited to a lesser extent by acetazolamide. Hence, results from both in vivo and in vitro [¹⁴C]acetate labeling experiments indicated that application of CA inhibitors effectively reduced the rate of lipid synthesis from acetate in cotton embryos.

View this table: [in this window] [in a new window]   Table II.   Radiolabeling of total lipids with [14C]acetate in isolated plastids of cotton embryos (34 DPA) and tobacco cell suspensions and chloroplasts for 1 h Approximately 75 µg of embryo or tobacco plastid proteins or 0.167 mg of tobacco chloroplast chlorophylls were pre-incubated with 10 mM of ethoxyzolamide or acetazolamide for 30 min prior to the addition of [14C]acetate. Values represent the mean and SD of three independent experiments. Rates of lipid synthesis in embryo or tobacco cell plastids are represented as nmol acetate h1 mg1 protein and for tobacco chloroplasts as nmol acetate h1 mg1 chlorophyll.

CA and Lipid Synthesis in Tobacco Cell Suspensions

The influence of CA on lipid biosynthesis was examined with a similar approach in a different plant system (tobacco cell suspensions, Fig. 5). Tobacco cell suspensions (in log phase) were radiolabeled with [¹⁴C]acetate in vivo and the incorporation of [¹⁴C]acetate into total lipids was quantified (Fig. 5). A linear rate of [¹⁴C]acetate incorporation into total lipid was established after 1 h of incubating tobacco cells with radiolabeled acetate. An approximate 65% reduction in the rate of [¹⁴C]acetate incorporation was observed with 10 mM ethoxyzolamide (159.4 ± 41 pmol acetate h¹ mg¹ protein) when compared with controls (453.4 ± 42.4 pmol acetate h¹ mg¹ protein). The distribution of radioactivity in the major lipid classes of tobacco cell suspensions after 4 h was analyzed with TLC (Table I). Approximately 75% of radiolabeled acetate was incorporated into polar lipids (phospholipids and glycolipids) and 25% into nonpolar lipids (fatty acids and triacylglyercols) in control samples. Similar to results with cotton embryos, tobacco cell suspensions incubated with 10 mM ethoxyzolamide showed an overall reduction in radiolabel incorporated into both polar and nonpolar lipids compared with controls (no inhibitors).

View larger version (11K): [in this window] [in a new window]   Figure 5.   Incorporation of [14C]acetate into total lipids of tobacco cv xanthi cell suspensions. Cells were pre-incubated with DMSO (control), 1 mM ethoxyzolamide, or 10 mM ethoxyzolamide for 30 min before the addition of radiolabeled [14C]acetate. Total lipids were extracted at 1, 2, and 4 h after the addition of the radiolabeled acetate. Data points represent mean and SD of three independent experiments. Lines are drawn from linear regression analyses of the data with r2 = 0.98 for controls, r2 = 0.98 for 1 mM ethoxyzolamide, and r2 = 0.88 for 10 mM ethoxyzolamide. Rates estimated from linear regression analyses were 453.4 ± 42.4, 300.7 ± 31.4, and 159.4 ± 41.3 pmol acetate h1 mg1 protein, respectively. For reference, the rate of acetate incorporation into tobacco cell suspensions not pre-incubated with DMSO was 461.3 ± 11.5 pmol acetate h1 mg1 protein.

Plastids from tobacco cells were isolated and incubated with [¹⁴C]acetate in vitro. The rate of [¹⁴C]acetate incorporation into lipids was reduced by about 50% in plastids incubated with ethoxyzolamide when compared with controls (no ethoxyzolamide; Table II), consistent with results from in vivo studies with intact tobacco cells (Fig. 5), and also consistent with results of plastids isolated from other plant cell types (embryo plastids and leaf chloroplasts, Table II).

To examine if there was a relationship between the rate of lipid synthesis, CA activity, and the availability of intracellular inorganic carbon for lipid synthesis, several experiments were conducted in which external CO2 concentrations were measured and manipulated. First, the release of CO₂ was measured from tobacco cell suspensions incubated with 1,000 µM ethoxyzolamide. The data revealed about a 3-fold increase in the release of CO₂ from treated tobacco cell suspensions when compared with controls (Table III), suggesting that the reduction by CA-specific inhibitor of acetate incorporation into lipids may be due to the reduction of intracellular inorganic carbon pools. More directly, tobacco cells grown under limiting CO₂ (scrubbed air) conditions had about a 50% reduction in the rate of [¹⁴C]acetate incorporation into total lipids compared with cells grown under ambient CO₂ conditions (Table IV). The rates of lipid synthesis for tobacco cells grown under scrubbed CO₂ conditions were approximately equivalent to the rates of lipid synthesis for cells grown under ambient conditions, but treated with ethoxyzolamide. In both cases, incorporation of [³⁵S]-Met into total protein was similar between treatments and controls (not shown), indicating that like embryos (Fig. 4), inhibition of CA and acetate incorporation in tobacco cell suspensions did not affect other vital cellular processes such as general protein synthesis. Rates of acetate incorporation into total lipids were further reduced in tobacco cells grown under limiting CO₂ conditions and treated with ethoxyzolamide (Table IV). Taken together, these results revealed that inhibiting CA with ethoxyzolamide (or acetazolamide) in tobacco cell suspensions or incubating tobacco cells under limiting CO₂ conditions could effectively reduce lipid synthesis from [¹⁴C]acetate.

View this table: [in this window] [in a new window]   Table IV.   Radiolabeling of total lipids with [14C]acetate in tobacco cell suspensions incubated under ambient or limiting CO2 conditions and with or without 1 mM ethoxyozlamide Values represent the mean and SD of three experiments. Rates of lipid synthesis are represented as pmol acetate h1 mg1 protein.

CA and Lipid Synthesis in Tobacco Chloroplasts

Inhibition of CA at the molecular level also significantly reduced the rate of lipid synthesis from [¹⁴C]acetate (Fig. 6A). Chloroplasts of WT or antisense CA-suppressed transgenic tobacco plants were analyzed for [¹⁴C]acetate incorporation in vitro. Transformed plants refer to TOBCA 1.10 AS-A and AS-B (two separate antisense plants), which had total CA activity levels reduced to less than 5% and 10% of WT levels, respectively (Fig. 6B). Price et al. (1994) previously reported that chloroplasts isolated from these antisense plants (TOBCA 1.10) had 2% of CA activity when compared with WT (on a chlorophyll basis). Rates of [¹⁴C]acetate incorporation into lipids, equivalent to 4.2 ± 0.5 nmol acetate h¹ mg¹ protein, were routinely measured in chloroplasts isolated from expanding young tobacco leaves (WT). A linear rate of [¹⁴C]acetate incorporation into lipids was established after 5 min of incubating chloroplasts with [¹⁴C]acetate and continued up to 4 h. Approximately a 50% reduction in the rate of [¹⁴C]acetate incorporation into lipids was observed in chloroplasts from both transgenic (antisense CA-suppressed A and B) plants when compared with WT plants (Fig. 6A). For comparison, a substantial reduction in [¹⁴C]acetate incorporation into total lipids was observed in WT chloroplasts incubated with ethoxyzolamide (Table II).

View larger version (22K): [in this window] [in a new window]   Figure 6.   Incorporation of [14C]acetate into total lipids of chloroplasts isolated from tobacco leaves. Total lipids were extracted at 5 min, 10 min, 30 min, 1 h, 2 h, and 4 h after the addition of radiolabeled acetate (A). CA activity determined in leaves of antisense CA (A and B)-suppressed transgenic and WT plants (B). Data points represent mean and range of duplicate samples within a single experiment. Similar trends were observed in replicate experiments. Lines represent linear regression analyses of the data with r2 = 0.95 for controls, r2 = 0.98 for antisense plant A (AS-A), and r2 = 0.97 for antisense plant B (AS-B). Rates estimated from linear regression analyses were 70.3 ± 35.3, 39.9 ± 1.6, and 41.5 ± 2.3 pmol acetate min1 mg1 chlorophyll, respectively.

	DISCUSSION

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

We report here that biochemical and molecular inhibition of CA effectively reduced the rate of lipid synthesis in both cotton and tobacco. The data revealed that cotton embryos incubated with ethoxyzolamide or acetazolamide reduced CA enzyme activities in vitro by 60% to 90% and in turn could effectively reduced the rate of [¹⁴C]acetate incorporation into lipids in vivo and in vitro (Figs. 2 and 3; Table II). Also, tobacco cells treated with ethoxyzolamide had a significantly reduced rate of lipid synthesis when compared with controls (no inhibitors; Fig. 5). Likewise, chloroplasts of antisense CA-suppressed transgenic tobacco plants had a 50% reduction in the rate of [¹⁴C]acetate incorporation into lipids when compared with WT chloroplasts (Fig. 6). Collectively, the results presented in this paper clearly indicate that a reduction in plastidial CA activity leads to a reduction in the rate of plastidial lipid synthesis (from [¹⁴C]acetate) in developing cotton embryos, cell suspensions of tobacco, and leaves of tobacco plants.

Values of in vitro rates of fatty acid synthesis by isolated plastids and chloroplasts were significantly higher than in vivo rates (Figs. 3, 5, and 6; Table II). Low in vivo rates of lipid synthesis observed could be due to a limited uptake of [¹⁴C]acetate into plant tissues, thus limiting the availability of radiolabeled acetate for optimum rate of incorporation into total lipids. In addition, there has been much debate over the actual metabolic source of acetyl-CoA for de novo plant fatty acid synthesis and it appears to vary considerably depending upon cell type and physiological demand (Eastmond and Rawsthorne, 2000). Others showed that exogenous pyruvate and Glc-6-P supplied to oilseed rape embryos during the maximum period of lipid synthesis had the highest rate of incorporation into fatty acid, when compared with dihydroxyacetone phosphate, malate, or acetate as substrates. In any case, we used [¹⁴C]acetate as a general radiotracer for lipid synthesis to evaluate the impact of CA activity on the rates of lipid synthesis in cotton or tobacco tissues. Despite the differences between in vitro and in vivo rates of lipid synthesis from [¹⁴C]acetate, we consistently found a reduction of acetate incorporation into total lipids in both cotton and tobacco plants when CA was inhibited.

Although the precise mechanism of CA involvement in plant lipid synthesis is unclear at this point, there are a number of possibilities. One is that CA may be present to aid in the diffusion of CO₂ into the chloroplast as suggested by Badger and Price (1994). Because CO₂, but not HCO₃, can exit plastids by simple diffusion, movement of CO₂ out of plastids could result in a substantial loss of inorganic carbon and require additional energy for re-incorporation. The HCO₃ that is utilized by ACCase to form malonyl-CoA is released as CO₂ by subsequent reactions of the fatty acid synthase complex. It is possible that an enzymatic hydration of CO₂ at this point increases the efficiency of CO₂ utilization in plastids, a concept that is similar to the conservation of CO₂ in mesophyll cells of C₄ plants by the cooperative action of CA and phosphoenolpyruvate carboxylase (Hatch and Burnell, 1990). In fact, CA in plastids may interact specifically with ACCase and enzymes of the fatty acid synthase complex to efficiently "channel" carbon into fatty acid (Roughan and Ohlrogge, 1996), although a direct interaction remains to be shown.

A substantial increase in CO₂ loss was measured from tobacco cell suspensions treated with CA inhibitors (ethoxyzolamide), compared with untreated cells (Table III), which supports a "trapping" role for plastidial CA. In this case, the metabolic role for CA would be an indirect one, wherein CA improves the efficiency of fatty acid synthesis by rapidly cycling inorganic carbon for ACCase. As such, a long-term physiological consequence of CA reduction likely would be difficult to observe under optimal growth conditions (where CO₂ levels were not limited), especially because the hydration of CO₂ occurs spontaneously at appreciable rates. We detected no obvious growth differences in tobacco plants with less than 2% of WT CA activity consistent with previous results (Price et al., 1994). The physiological similarities between WT and antisense CA-suppressed tobacco plants were noted previously by Price and coworkers (1994). Rates of Rubisco activity and CO₂ assimilation were not different when CA was suppressed. However, there was a lower carbon isotope composition, ¹³C/¹²C, in leaf dry matter of antisense CA-suppressed plants compared with WT plants. This suggested that there was a greater loss of CO₂ in antisense CA-suppressed plants than WT plants (Price et al., 1994), which is consistent with our notion of a "trapping" function for CA.

Besides concentrating inorganic carbon for lipogenesis in plastids, CA may participate in other plastidial carboxylation reactions, such as carbamoyl phosphate synthetase (a plastid-localized enzyme in higher plants; Nara et al., 2000), which synthesizes the precursor for pyrimidine biosynthesis. Also, Kavroulakis et al. (2000) have suggested that CA facilitates the recycling of CO₂ in developing soybean (Glycine max) root nodules during early stages of development. The recycling and concentrating of CO₂ thus would provide adequate availability of inorganic carbon for lipid synthesis and other carboxylation reactions.

Another possible mechanism by which CA might act indirectly to influence fatty acid synthesis in plastids is to modulate plastidial pH. The optimal rate of fatty acid synthesis by fatty acid synthase complex could be influenced by changes in stromal pH. Jacobson et al. (1975) suggested that chloroplastic CA in spinach (Spinacia oleracea) could act to buffer against transient pH changes in the stroma during photosynthesis. This remains a possibility that warrants investigation, given the well-known role of pH regulation by CA in animal cells (Sly and Hu, 1995).

Although most research attention for plastidial CA has focused on its putative role in photosynthetic carbon fixation, it is becoming increasingly likely that CA has a variety of additional functions in non-photosynthetic tissues. cDNAs encoding CA proteins have been isolated from non-photosynthetic tissues such as cotton seedlings and alfalfa (Medicago sativa) nodules (Coba de la Pena et al., 1997; Hoang et al., 1999). In fact, it may be that non-photosynthetic plant systems, lacking the complicating carboxylating activity of Rubisco, are particularly well suited for evaluating various physiological roles of plastidial CA. In addition to our work here, others have implicated CA isoforms in two distinctly different roles in nitrogen metabolism in root nodules (Galvez et al., 2000). Multiple CA and CA-like genes are expressed in Arabidopsis (Arabidopsis Genome Initiative, 2000). It seems likely that as more CA isoforms are found, so will increase the number of physiological functions attributed to this evolutionarily conserved enzyme in plants.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

Plant Material

Cotton (Gossypium hirsutum L. cv Paymaster HS26) seeds were provided by Dr. John Burke (U.S. Department of Agriculture-Agricultural Research Service, Lubbock, TX). WT and transgenic tobacco (Nicotiana tabacum L. SR1) seeds (Price et al., 1994) were kindly provided by Dr. G. Dean Price (Australian National University, Canberra). Cotton and tobacco plants were grown in a greenhouse or growth room with 14-h photoperiod (supplemented with sodium lamps to extend day length when necessary) and temperatures of approximately 38°C during the day and 25°C at night. Plants were watered daily and fertilized biweekly with a dilute solution of fertilizer (Miracle Gro, Stern's Miracle Gro Products, Inc., Port Washington, NY). The age of cotton bolls and developing embryos was determined by tagging flowers at anthesis and by ovule morphology (Choinski and Trelease, 1978).

Tobacco leaves of similar sizes and developmental stages were harvested for chloroplast isolation. Tobacco cv xanthi cell suspensions were grown and maintained as described previously (Chapman et al., 1995). Cell suspensions in log phase (72 h after subculture) were used for plastid isolations and [¹⁴C]acetate labeling experiments.

Total Lipid Extraction/TLC

Total lipids were isolated from cotton embryos, tobacco cell suspensions, and plastids essentially as described previously (Chapman and Moore, 1993). The chloroform fraction was washed three times with 1 M KCl and then the chloroform/lipid mixture was evaporated to dryness with N₂. Total lipids were resuspended in 50 µL of chloroform and analyzed by fractionation on TLC. Total lipids were fractionated on TLC plates (silica gel G-60, Whatman, Clifton, NJ) for 40 min in hexane:diethylether:acetic acid (80:20:1 [v/v/v]) solvent system. Radiometric scanning (System 200 Imaging Scanner, Bioscan, Washington, DC) was used to quantify the radiolabeled lipids after TLC separation (Chapman and Sprinkle, 1996).

Enzyme Assays

Freshly harvested embryos were weighed and then frozen in liquid nitrogen. The frozen samples were ground with a mortar and pestle in a 1:1 (w/v) homogenization buffer containing 400 mM Suc, 100 mM sodium phosphate (pH 7.2), 10 mM KCL, 1 mM MgCl₂, and 1 mM EDTA. Homogenates were filtered through four layers of cheesecloth and the filtrate was analyzed immediately for CA activity. Also, embryo homogenates were pre-incubated with sulfanilamide, acetazolamide, and ethoxyzolamide inhibitors for 30 min before assaying for CA activity. Protein content in cell-free extracts was determined according to Bradford (1976) using bovine serum albumin for standard curve calibration. Total CA activity was determined electrometrically (Wilbur and Anderson, 1948) as described previously (Hoang et al., 1999). One unit of activity (Wilbur-Anderson unit) was defined as 10[(T_o/T)  1], where T_o and T are equal to the rate of pH change of the reaction without (control) and with cell homogenates, respectively.

Embryos were homogenized in 1:1 (w/v) buffer (same buffer used as in preparation of fractions for carbonic anhyrase assays) with a polytron (PT10/35, speed at 8, Brinkmann Instruments, Westbury, NY). Embryo homogenates were pre-incubated with different inhibitors for 30 min before assaying for ACCase activity. ACCase activity was based on acetyl-CoA-dependent H¹⁴CO₃ (0.05 uCi per sample; 8.4 mCi mmol¹) incorporation into acid-stable product (malonyl-CoA; Roesler et al., 1996). The production of [¹⁴C]malonyl-CoA was quantified by liquid scintillation counting (same conditions as above).

mRNA Isolation and Northern-Blot Analyses

Total cellular RNA was isolated from embryos of cotton plants by a hot borate procedure developed by Wan and Wilkins (1994). RNA yield and quality were evaluated spectrophotometrically and by analytical gel electrophoresis according to Sagerström and Sive (1996). Poly(A⁺) RNA from cotton embryos was isolated by oglio-dT cellulose column chromatography according to Aviv and Leder (1972). Approximately 2 µg of mRNA was electrophoresed in a 1% (w/v) agarose gel containing 6% (v/v) formaldehyde and 1× MOPS buffer (20 mM MOPS-NaOH, pH 7.0; 5 mM sodium acetate, and 0.1 mM EDTA; Sagerström and Sive, 1996). RNA was transferred to nylon membranes by capillary transfer with 20× SSC (overnight) and probed with a random prime-labeled (Gene Images random prime-labeling module, Amersham, Buckinghamshire, UK) 1.16-kb CA and 539-bp actin probe (Hoang et al., 1999). Hybridization and washing were both carried out at 62°C and with a final wash at 0.2× SSC. Hybridized bands were identified by an alkaline phosphatase-catalyzed chemiluminescent reaction (Gene Images CDP-Star detection module, Amersham) and quantified by densitometric scanning (National Institutes of Health version 6.1 image software).

Incorporation of [³⁵S]-Met into Proteins in Vivo

Cotton embryos were excised from bolls and placed on moist filter paper in petri dishes. CA inhibitors in 5 µL at different concentrations (0.1-10 mM) were dispensed onto each embryo 30 min before the addition of [³⁵S]-Met (2 µCi per embryo; 1.175 mCi µmol¹). After radiolabeling for 1 h with [³⁵S]-Met, the embryos were homogenzied in a 1:1 (w/v) buffer (same buffer used above in CA assays) and the total proteins from cell-free homogenates were precipitated in TCA (4% [w/v] final concentration). TCA precipitated material was washed twice with 70% (v/v) ethanol, once with diethyl ether, and then resuspended in 0.4 M NaOH. Insoluble material was removed by centrifugation and supernatants were air dried overnight (Coligan et al., 1983). Liquid scintillation counting (LS 3861 counter [ Beckman Instruments, Fullerton, CA] and ScintiSafe Plus 50% LSC cocktail, Fisher Scientific, Houston) was used to quantify [³⁵S]-Met incorporation into total protein.

Enzyme-Specific Inhibitors

CA-specific sulfonamide inhibitors (Maren, 1967) were utilized for in vivo and in vitro [¹⁴C]acetate incorporation experiments into total lipids. These were 4-aminobenzene-sulfonamide (sulfanilamide), 5-acetamido-1,3,4-thiadiazole- 2-sulfonamide (acetazolamide), and 6-ethoxy-2-benzothiazolesulfonamides (ethoxyzolamide; Sigma Chemical Co., St. Louis). Sulfanilamide was dissolved in water, whereas acetazolamide and ethoxyzolamide inhibitors were dissolved in 1:2 [v/v] DMSO:water.

Measurements of CO₂ Released from Tobacco Cell Suspensions

Tobacco cell suspensions in log phase were incubated with 10, 100, and 1,000 µM ethoxyzolamide and subsequently CO₂-free air (scrubbed with soda lime) was flushed through the system. CO₂ released from tobacco cell suspensions either treated with DMSO (control) or ethoxyzolamide was monitored in line with a plant CO₂ analysis system (Qubit Systems Inc., Kingston, Ontario). These same conditions were utilized in combination with [¹⁴C]acetate radiolabeling to evaluate the influence of external CO₂ (without and with CA inhibitors) on lipid synthesis in vivo.

Incorporation of [¹⁴C]Acetate into Total Lipids of Cotton Embryos and Tobacco Cells in Vivo

Cotton bolls were harvested and immediately placed on ice. The embryos were excised from the ovules and then placed on wet filter paper. CA inhibitors in 5 µL at different concentrations (0.1-10 mM) were dispensed onto each embryo and then incubated for 30 min before the addition of radiolabeled [2-¹⁴C]acetate (54 mCi mmol¹). Total lipids were extracted at different time points after the addition of the 0.5 µCi of radiolabeled acetate (Chapman and Moore, 1993). Aliquots of total lipids dissolved in chloroform were evaporated to dryness, then the incorporation of [¹⁴C] acetate was quantified by liquid-scintillation counting (Beckman LS 3861) or by radiometric scanning after TLC.

Tobacco cell suspensions in log phase were used for the [¹⁴C]acetate incorporation experiments in vivo. Tobacco cells were collected, washed, and suspended in fresh medium. Two milliliters of the suspended cells were transferred to a 15-mL tube and pre-incubated with CA inhibitors for 30 min before the addition of radiolabeled acetate. Also, tobacco cell suspensions were incubated under zero CO₂ conditions and subsequently analyzed for the incorporation of [¹⁴C]acetate into total lipids. Aliquots of the total lipids were analyzed and quantified as described above.

Plastid Isolation

Tobacco plants were taken out of the growth room and placed in the dark for 48 h in preparation for chloroplast isolation to reduce the amount of starch and allow for higher yields of intact chloroplasts. Chloroplasts were isolated according to Yu and Woo (1988), but with the following modifications. The pellet at 800g for 5 min was layered over a Percoll gradient composed of 3 mL of 90% (w/v), 15 mL of 35% (w/v), and 10 mL of 15% (w/v) Percoll. Intact chloroplasts were recovered and the chlorophyll content was estimated according to Bruinsma (1961) in 80% (v/v) acetone.

Plastids were isolated from tobacco cell suspensions (in log phase) and cotton embryos by centrifugation through a 10% (w/v) Percoll gradient (Sparace and Mudd, 1982; Trimming and Emes, 1993). Plastid protein content was estimated according to Bradford (1976).

Incorporation of [¹⁴C]Acetate into Total Plastid Lipids in Vitro

The conditions used for assaying the incorporation of the [2-¹⁴C]acetate into fatty acids in isolated plastids are described by Stahl and Sparace (1991). The reaction mixture contained 0.02 µM Na-[2-¹⁴C]acetate, 0.04 mM cold sodium acetate, and 3 mM each of MgCl₂ and ATP and 50 to 100 µg of plastid protein or 50 to 200 µg of chlorophyll.