Functional, c-myc-tagged Cf-9 resistance gene products are plasma-membrane localized and glycosylated

Two c-myc-tagged Cf-9 alleles confer Avr9 recognition in tobacco

A triple c-myc sequence was inserted in frame in Cf-9, either behind the putative signal peptide cleavage site (domain B) near the N-terminus of the predicted extracellular part of the protein, or in the putative cytoplasmic part of the protein (domain G) (see Experimental procedures). We introduced three c-myc sequences (EQKLISEEDL) in tandem to improve sensitivity for the detection of the chimeric protein with the anti-myc antibodies. These constructs were fused to the 35S promoter and were cloned into an Agrobacterium binary plasmid. The resulting clones, SLJ9161 and SLJ9171, are represented in Fig. 1. Tobacco plants were transformed, and at least 10 independent kanamycin-resistant plants were analysed for each construct. Transgenic tobacco expressing the tomato Cf-9 gene can be used to study Cf-9- and Avr9-dependent responses ( Hammond-Kosack et al. 1998 ; Piedras et al. 1998 ; Romeis et al. 1999 ).

To assess whether these transgenic plants still retain Cf-9 function, IF(Avr9⁺) and IF(Avr9^–) was injected into the leaves. Most transformants exhibited either a grey necrosis or chlorosis in response to the IF(Avr9⁺) challenges, and no macroscopic response to the IF(Avr9^–) challenges ( Fig. 2a), indicating that these plants were still able to initiate the Avr9-dependent defence response. The response was stronger in plants expressing c-myc : Cf-9 with the tag in domain B (9171 plants) than plants with the tag in domain G (9161 plants).

Expression of c-myc-tagged Cf-9 protein

The expression of c-myc-tagged protein was analysed by Western blots with the anti-myc antibodies in total protein extracts from 9161 and 9171 tobacco plants. A strong cross-reacting band was detected in plants transformed with either plasmid ( Fig. 2b, lanes marked as T), whereas signal was not observed in the control extracts obtained from non-transformed plants ( Fig. 2b, PH panel). The protein band observed in both 9161 and 9171 microsomes showed an estimated molecular mass of 160 kDa. The signal was stronger in plants with the tag in the G domain (9161 plants). The fact that 9161 plants showed higher c-myc : Cf-9 protein levels but were less responsive to Avr9 infiltration suggests either that the insertion of the triple c-myc in the G domain affects the functionality of Cf-9, or that the B domain tag is more prone to proteolysis.

c-myc-tagged Cf-9 is glycosylated

The Cf-9-predicted protein has a molecular mass of 94 kDa and contains 22 putative glycosylation sites distributed throughout its sequence. The protein recognized by the c-myc antibodies shows a molecular mass of 160 kDa, higher than predicted from the Cf-9 amino-acid sequence, even considering the 5 kDa added by the inserted triple c-myc. This difference between the observed and estimated molecular masses led us to test whether the potential glycosylation sites in Cf-9 are used. Solubilized microsomal fractions from c-myc : Cf-9 plants were treated with PNGase F. This enzyme is a glycoamidase that cleaves the bond between the asparagine residue of the protein and the N-acetylglucosamine residue that joins the carbohydrate to the protein, liberating nearly all N-linked oligosaccharides from glycoproteins. After incubation of c-myc : Cf-9 microsomes with PNGase F, a shift in the mobility of the protein was observed. Within 5 min, two smaller bands, of 155 and 105 kDa, were detected ( Fig. 3). As the enzymatic reaction proceeded, the 105 kDa band increased in intensity concomitantly with the 155 kDa band becoming fainter. The 105 kDa size is close to the molecular mass of 100 kDa predicted for the c-myc : Cf-9 amino-acid sequence, indicating that Cf-9 is highly glycosylated.

c-myc-tagged Cf-9 is in the plasma membrane fraction

Domain F in Cf-9 is predicted to be a transmembrane domain, but this does not establish whether Cf-9 is a plasma membrane or an ER protein. Separation by ultracentrifugation of soluble and membrane-associated proteins confirms that c-myc : Cf-9 is associated with membranes; the c-myc : Cf-9 was detected in the microsomal pellet fraction and signal was not detected in the soluble fraction ( Fig. 2b) for both plants 9161 and 9171.

The presence of the KKRY sequence in the Cf-9 C-terminus prompted us to investigate whether Cf-9 is a plasma-membrane protein. We attempted to separate plasma membranes from ER using sucrose gradients, but clear resolution between ER and plasma membranes was not achieved (data not shown). Therefore we purified plasma membranes using the two-phase partitioning method of Kjellbom & Larsson (1984). Using this technique, plasma membranes are separated from other membranes based upon differences in their surface charge. The plasma-membrane vesicles obtained with this method are predominantly sealed and oriented right-side-out. The partitioning in dextran-polyethylene was repeated three times, generating U3 (upper phase re-extracted three times and enriched in plasma membrane) and L3 (lower phase re-extracted three times). Both fractions, U3 and L3, as well as the microsomal membranes, were assayed for the presence of H⁺-ATPase (plasma membrane marker), NADPH-cytochrome c reductase (ER marker), cytochrome c oxidase (mitochondrial marker), and chlorophyll (chloroplast marker). The results are summarized in Table 1. The values represent the specific activity compared with the microsomal preparation. The plasma membranes obtained from plants with tags in both positions were enriched seven-fold in vanadate-sensitive H⁺-ATPase, were chlorophyll-free and were also very pure with respect to contamination by mitochondria. However, a relatively high specific activity of cytochrome c reductase was found in the plasma membrane preparations. This NADPH-cyt c reductase activity is probably not due to contamination by ER, but is a true activity of the plasma membrane ( Askerlund et al. 1991 ; Kjellbom & Larsson 1984). Antibodies raised against the Arabidopsis plasma membrane aquaporin PIP1a were used as a second marker for plasma-membrane proteins ( Kammerloher et al. 1994 ). Western blots using those antibodies confirmed the data obtained using H⁺-ATPase activity. This aquaporin was enriched in our plasma membrane preparations. The same fractions were analysed for c-myc : Cf-9 protein using the myc antibodies. Enrichment for c-myc : Cf-9 was always observed in plasma membrane fractions ( Fig. 4).

Immunolocalization of c-myc : Cf-9 in protoplast plasma membranes

As an additional test of whether Cf-9 is plasma-membrane localized, protoplasts were obtained from suspension cultures derived from 9171 plants. In these plants c-myc is in the B domain and therefore should be extracellular. Protoplasts from these plants, as well as from control plants, were incubated with myc antibodies and with secondary gold-coupled antibodies. The gold particles on the surface of protoplasts were visualized after silver enhancement. A clear difference between 9171 and untransformed protoplasts was reproducibly obtained in many separate experiments ( Fig. 5). Although some antibody binds non-specifically to untransformed protoplasts, transformed cells show substantially enhanced specific binding of c-myc antibodies to the protoplast surface. These data provide independent verification of the plasma membrane localization of the c-myc : Cf-9, and confirm its anticipated orientation in the membrane.

Vector construction

The triple c-myc coding sequence was amplified by PCR from the plasmid CB2339 (a gift of Kim Arndt) to create EcoRI and HindIII cloning sites. The primers used were 5′ ATG GAA TTC GGT GAA CAA AAG TTG 3′ (creating an EcoRI site 5′) and 5′ ATG AAG CTT TCC GTT CAA GTC TTC TTC T 3′ (creating a HindIII site 3′). The amplified fragment was cloned in KS Bluescript after digestion with EcoRI and HindIII restriction enzymes. The EcoRI/HindIII fragment was inserted in two different positions in a Cf-9 clone from which the internal EcoRI and HindIII restriction sites were removed (SLJ8501).

Cf-9 clones with the myc inserted were created as follows. For clone SLJ9161, EcoRI and HindIII restriction sites were created in the G domain by oligomutagenesis (5′ TTT CAT TTT CGT AGT AAG CTT CAT GAA TTC AAT TAT GTG TTC CAA 3′). The underlined restriction sites were used to clone the HindIII–EcoRI fragment containing the triple c-myc. A ClaI restriction site was created by oligomutagenesis at the translational initiation ATG of Cf-9 to allow cloning of the resulting c-myc : Cf-9 behind the 35S promoter. The 35S promoter (as an EcoR–ClaI fragment from the clone SLJ4K1; Jones et al. 1992 ), and the final Cf-9 with the myc tag in the G domain (as a ClaI–BamHI fragment) were cloned in the binary vector SLJ7291 digested with EcoRI and BamHI to generate the vector SLJ9161. Clone SLJ9171 was constructed as for SLJ9161 but using the primer (5′ TTC GGG GCA CAA ATG AAG CTT CAT GAA TTC AGG CAA GGA TGA GGA 3′) to create the cloning sites EcoRI and HindIII in the B domain-encoding DNA of Cf-9.

Plant transformation and growth conditions

The binary clones SLJ9161 and SLJ9171 were mobilized into Agrobacterium tumefaciens LBA4404, and introduced into tobacco cultivar Petite Havana (Nicotiana tabacum) by the method of Horsch et al. (1985) . At least 10 independent transformants of each construct were selected for further analysis. To select the subset of primary tobacco transformants carrying single T-DNA integration sites, T₂ progeny seed was plated on Murashige & Skoog (1962) medium containing 300 μg ml⁻¹ kanamycin sulphate, and the ratio of kanamycin-resistant to sensitive seedlings was determined. Plants were grown as described in Hammond-Kosack et al. (1998) .

Generation of suspension cultures

Cultures were generated from leaf pieces taken from 9171 tobacco plants as described by Piedras et al. (1998) .

Treatment of tobacco plants

IFs from Petit Havana tobacco plants, either non-transgenic or expressing the Avr9 peptide (SLJ6201 A) ( Hammond-Kosack et al. 1994 ), grown under the same conditions, were extracted after infiltration with distilled water as described by De Wit & Spikman (1982) and frozen at −20°C until use. The IF from Avr9-expressing plants was designated IF(Avr9⁺) and the IF from control plants IF(Avr9^–). The interveinal regions of fully expanded leaves were injected with both IFs.

Isolation and purification of plasma-membrane vesicles

Leaves were cut into pieces and placed in a homogenizer fitted with razor blades and homogenized in lysis buffer (25 m m Tris–HCl pH 7.5, 0.5 m sucrose, 3 m m EDTA, 0.5 m m 4-(2-aminoethyl)-benzenesulfonyl fluoride, AEBSF) four times for 15 sec. The homogenate was filtered through four layers of miracloth and centrifuged at 6000 g for 10 min. The supernatant was recovered and centrifuged at 100 000 g for 1 h to pellet the microsomal membranes. Microsomes were resuspended in 5 m m potassium phosphate buffer, 0.33 m sucrose, 3 m m KCl, 1 m m AEBSF. Plasma-membrane vesicles were prepared by aqueous two-phase partitioning ( Kjellbom & Larsson 1984) with a final composition of 6.4% (w/w) dextran T500, 6.4% PEG 3350, 330 m m sucrose, 5 m m potassium phosphate pH 7.8, 3 m m KCl. Aliquots of microsomes were dissolved in the phase mixture to a final weight of 28 g. Partitioning was accelerated by 10 min centrifugation at 2000 g in a swinging-bucket rotor. The resulting lower and upper phases were repartitioned twice with fresh phases to produce U3 and L3. Those were then diluted, centrifuged at 120 000 g for 1 h, and resuspended in 5 m m potassium phosphate buffer, 0.33 m sucrose, 3 m m KCl, 1 m m AEBSF. All manipulations were conducted at 4°C.

Deglycosylation experiment

c-myc : Cf-9-containing microsomes were solubilized with 0.1% (octylphenoxy)-polyethoxyethanol (Nonidet P-40) before treatment with the peptide : N-glycosidase F (PNGaseF) (New England Biolabs, Hutchin, UK). The solubilized microsomal fraction (1 mg total protein) was centrifuged at 100 000 g for 1 h at 4°C, and the supernatant was incubated at 100°C for 10 min in denaturing buffer (0.5% SDS, 1% β-mercaptoehanol). The deglycosylation reaction was carried out at 37°C after the addition of 500 units of PNGaseF. Aliquots were taken at different times and the reaction was stopped by addition of loading buffer and boiling. The proteins were subjected to SDS–PAGE and Western blotting with the myc antibodies as described bellow.

Enzymatic activities

H⁺-ATPase, cytochrome c oxidase (EC 1.9.3.1) and NADPH-cytochrome c reductase (EC 1.6.2.5) were determined according to Schaller & De Wit (1995).

Protein determination

All protein concentrations were determined by BCA protein assay kit (Pierce, Chester, UK) using 1% (v/v) Triton X100. BSA was used as a standard.

SDS–PAGE and immunoblotting

Samples were mixed with loading buffer (with a final concentration of 100 m m SDS and 200 m m DTT) and incubated for 15 min at 60°C. After electrophoresis ( Laemmli 1970), proteins were transferred to nitrocellulose (Amersham, Little Chalfont, UK) by wet electroblotting (Mini-Protean II system; Bio-Rad, Hemel Hempsted, UK). Transfer buffer consisted of 20% (v/v) methanol, 38.64 m m glycine, 47.89 Tris, 1.28 m m SDS.

Western blots were blocked for at least 1 h in PBS-T (137 m m NaCl, 2.7 m m KCl, 4.3 m m Na₂HPO₄, 1.4 m m KH₂PO₄, 0.1% Tween 20 pH 7.3) with 5% (w/v) dried milk. Antibodies were diluted in PBS-T at the following concentrations: antiaquaporin (1 : 10 000) raised against the Arabidopsis plasma membrane aquaporin PIP1a ( Kammerloher et al. 1994 ), anti c-myc (1 : 2000) (Santa Cruz Biotechnology, Santa Cruz, CA, USA), goat anti-rabbit IgG (1 : 2000) (Sigma, Dorset, UK). Primary and secondary antibody incubations were for 1 h at room temperature and were followed by one wash for 15 min, and two washes for 5 min in PBS-T. Secondary antibodies were detected using alkaline phosphatase or peroxidase (Sigma, Dorset, UK).

Protoplast preparation

Suspension cultures were incubated at 27°C for 3 h in protoplasting medium (0.4 m mannitol, 10 m m ascorbate, 1 m m CaCl₂, 0.1% (w/v) Onozuka Cellulase RS, 0.05% (w/v) Onozuka Macerozyme R10, 0.05% (w/v) pectolyase Y23 (Seishin, Tokyo, Japan), 1% (w/v) BSA. Homogenate was filtered through a nylon mesh (80 μm pore size) and the protoplasts were washed three times in 0.55 m mannitol supplemented with 1 m m CaCl₂ by low-speed centrifugation (70 g; Labofuge 400 e, Heraeus, Hanau, Germany) and resuspension of the pellet.

Fixation, immunolabelling and silver enhancement

c-myc : Cf-9 protoplasts and protoplasts from untransformed plants were mildly prefixed for 12 h at 4°C (0.3% paraformaldehyde, 0.05% glutardialdehyde, 10 m m phosphate buffer pH 7, 0.47 m mannitol, 1 m m CaCl₂) before being post-fixed with 0.05% (w/v) OsO₄ for 20 min at room temperature. After washing for 10 min in 25 m m phosphate buffer pH 7, the protopasts were incubated for 30 min in PBS pH 7.4 supplemented with 1% BSA and 0.1% BSAc (Aurion, Wageningen, the Netherlands) as a blocking medium. Protoplasts were then incubated with primary and secondary gold coupled antibodies, respectively (GAM5; BioCell, Cardiff, UK) diluted in PBS pH 7.4 for 1 h at room temperature. Between the two incubations the protoplasts were washed three times with fresh PBS. To allow the visualization of 5 nm gold particles on the protoplast surface, protoplasts were washed twice with distilled water and subjected to silver enhancement for 5 min in the dark (Silverenhancement Kit, BioCell). Samples were then mounted on glass slides and observed and photographed with a Axivert 35 inverted microscope (Carl Zeiss, Oberkochen, Germany), equipped with a plan-neofluar 63×/1,25 oil Ph3 antiflex objective, using epipolarization or bright-field mode. Particle counting and determination of the labelling densities were according to Diekmann et al. (1994) .