INTRODUCTION

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

In plant and animal cells, the essential amino acid Lys is catabolized into acetyl CoA and several molecules of Glu (Fig. 1; Arruda et al., 2000; Galili et al., 2001). The first enzyme in the Lys catabolic pathway, Lys-ketoglutarate reductase (LKR), condenses Lys and -ketoglutarate into saccharopine, which is then converted by the second enzyme saccharopine dehydrogenase (SDH) into -amino adipic semialdehyde and Glu (Fig. 1). In animals, Lys catabolism via -amino adipic semialdehyde plays a significant physiological role in providing Glu for nervous signal transmission via Glu receptors (Arruda et al., 2000; Galili et al., 2001).

View larger version (25K): [in this window] [in a new window]   Figure 1.   The Lys catabolism pathway and metabolites derived from it. Broken arrows represent several nonspecified enzymatic reactions. Glu residues are situated inside large boxes.

The functional significance of Lys catabolism in plants is still unknown. In plants, the LKR level was shown to be significantly up-regulated in inflorescence tissues and developing seeds and in response to osmotic stress (Karchi et al., 1995; Kemper et al., 1999; Moulin et al., 2000; Tang et al., 1997). In addition, LKR activity in tobacco (Nicotiana tabacum) and maize (Zea mays) seeds was shown to be stimulated by excess cellular Lys, via an intracellular signaling cascade involving Ca²⁺ and protein phosphorylation (Karchi et al., 1995; Arruda et al., 2000).

In both plants and animals, LKR and SDH are linked on a single, bifunctional polypeptide encoded by a single gene (Arruda et al., 2000; Galili et al., 2001). Yet, the metabolic flux of Lys catabolism does not depend only on the bifunctional LKR/SDH enzyme. We have demonstrated that in Arabidopsis the LKR/SDH locus also encodes a monofunctional SDH enzyme, using an internal promoter (Tang et al., 1997, 2000). In the present report, we provide evidence showing that the LKR/SDH locus of plants expresses another novel enzymatic form, i.e. a highly active catabolic monofunctional LKR. We propose that this enzyme may provide a transient superefficient flux of Lys catabolism under specific developmental programs.

	RESULTS

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

An mRNA Encoding a Novel Catabolic Monofunctional LKR Is Expressed in Cotton (Gossypium hirsutum) Boll Abscission Zone

To obtain a deeper insight into the functional significance of Lys catabolism in plant growth and development, we searched for LKR/SDH related sequences in a number of expressed sequence tag (EST) databases. Of particular interest was the discovery of three LKR/SDH-related ESTs of only approximately 1,800 ESTs derived from a cotton boll abscission zone cDNA library (http://www.genome.clemson. edu/projects/cotton/). This relatively high frequency suggested that LKR/SDH gene expression might be significantly up-regulated during abscission. The three cotton ESTs were ordered and sequenced to completion. One of these ESTs encoded a bifunctional LKR/SDH (GhLKR/SDH; EST no. Cabc0005CA08x), which is similar to the LKR/SDH cDNA of Arabidopsis. Yet, the other two ESTs represented two novel catabolic monofunctional LKR mRNAs (EST nos. Cabc0001cE06x and Cabc0007af10x). These two clones were designated GhLKR-1 and GhLKR-2, respectively. As shown schematically in Figure 2, the cotton GhLKR-1 and GhLKR-2 cDNAs were identical in DNA sequence to the bifunctional GhLKR/SDH cDNA through the coding DNA sequence of the LKR domain and part of the intermediate domain. However, they contained additional 3'-DNA sequences that did not exist in the bifunctional LKR/SDH cDNA. These 3'-DNA sequences included codons for the five C-terminal amino acids (VSIHN) followed by a stop codon, 3'-non-coding sequences (varying in length between GhLKR-1 and GhLKR-2) and a poly(A) tail. The three original cotton cDNAs were all incomplete, lacking part of the 5' region of the LKR open reading frame (Fig. 2).

View larger version (29K): [in this window] [in a new window]   Figure 2.   Schematic diagram of the cotton LKR/SDH-related ESTs. 1, The Arabidopsis AtLKR/SDH cDNA, used as a reference; 2 through 4, the cotton GhLKR/SDH, GhLKR1, and GhLKR2, cDNAs, respectively. Boxes with identical filling represent the same domain. ATG and TAG represent initiator and terminator codons, respectively. The position of the five deduced new C-terminal amino acids VSIHN of the two cotton monofunctional LKR ESTs is indicated by an arrow at the bottom.

The Cotton LKR/SDH and Monofunctional LKR mRNAs Are Encoded by the Same Locus

The identical LKR-related DNA sequences between GhLKR/SDH, GhLKR-1, and GhLKR-2 suggested that the LKR/SDH and monofunctional LKR mRNAs are encoded by the same composite LKR/SDH locus. To study the genetic control of these two mRNAs, cotton DNA was digested with several restrictions enzymes, fractionated by agarose gel electrophoresis, and hybridized in a Southern blot with GhLKR-1 as a probe. As shown in Figure 3, hybridization with this probe hybridized to mostly two DNA bands. Because the G. hirsutum species of cotton is allotetraploid, it is likely that this species possesses two genes, one for each of its two diploid genomes. To clone the cotton LKR/SDH locus, we screened a cotton bacterial artificial chromosome (BAC) library by hybridization with GhLKR-1 as a probe. A positive BAC plasmid was identified and confirmed to possess the LKR/SDH locus by partial DNA sequence analysis (data not shown). To test whether the LKR/SDH locus inside this BAC plasmid encodes both the bifunctional LKR/SDH and monofunctional LKR mRNAs, it was hybridized with the unique 3' sequence of the GhLKR-1, which is missing in GhLKR/SDH cDNA. The BAC plasmid positively hybridized with this probe too (data not shown).

View larger version (100K): [in this window] [in a new window]   Figure 3.   Southern-blot analysis of LKR/SDH-related sequences in cotton. Cotton genomic DNA was digested with various restriction enzymes as indicated on the top of the panel. The digested DNA was reacted in a Southern blot with a probe derived from the cotton GhLKR-1 cDNA as a probe. The migration of DNA size markers is indicated on the left.

The LKR/SDH locus in the BAC clone was incomplete, lacking the promoter and a small part of the 5'-coding region. However, using inverse PCR of cotton genomic DNA, we were able to clone the 5'-coding region and to generate the full open reading frames of LKR/SDH and the two monofunctional LKR cDNAs (GenBank accession nos. AF264146, AF264147, and AF264148, respectively).

The Cotton Monofunctional LKR mRNA Is Produced by a Polyadenylation Site Located within a Large Intron in the Linker Region between LKR and SDH Coding Sequences

To test the origin of the unique 3' sequences of GhLKR-1 and GhLKR-2 mRNAs, we subcloned and sequenced the DNA region encoding the intermediate domain between the LKR and SDH domains in the gGhLKR/SDH locus. As shown schematically in Figure 4A, the 3' unique sequences of GhLKR-1 and GhLKR-2 were located inside a large intron (GenBank accession no. AF264630) located within this intermediate domain region. Furthermore, as shown in Figure 4B, analysis of this intron revealed at least two putative polyadenylation sites each containing an AT-rich PE plus additional elements upstream and downstream to it, which meet the criteria of plant polyadenylation sites as reported by Graber and associates (1999). Only one of these two putative polyadenylation sites was present in GhLKR-1, whereas both were present in GhLKR-2, slightly upstream of their poly(A) tails. None of these two polyadenylation sites resembled the complete consensus plant polyadenylation site (Graber et al., 1999), suggesting that they are rather low-affinity sites.

View larger version (37K): [in this window] [in a new window]   Figure 4.   Localization of the 3'-DNA sequence of the cotton monofunctional LKR cDNAs within an intron of the cotton LKR/SDH locus. A, Schematic diagram showing the position of the intron within the intermediate domain region. The location of the five deduced C-terminal amino acids of GhLKR1 and GhLKR2 followed by a stop codon (*) at the beginning of this intron (bolded capital letters), the position of the two poly(A) residues of these cDNAs, and the location of the two putative low-affinity polyadenylation sites (black boxes marked 1 and 2) are indicated. ATG and TAG represent initiator and terminator codons, respectively. B, The 5'-DNA sequence within the intron shown in A. The exon/intron junction is indicated by an arrow. The deduced seven C-terminal LEVSIHN amino acids (of which the last five encoded by intron sequences) are indicated in bold letters followed by a TAA stop codon (*). The positions of the two putative polyadenylation elements (PE) are indicated in bolded letters. The positions of sequences upstream and downstream to the PE, which resemble those reported by Graber and associates (1999), are indicated in bold underlined letters.

The Monofunctional LKR Is Abundantly Expressed in Parenchyma Cells of the Abscission Zone

To substantiate the presence of a monofunctional LKR in the cotton abscission zone and identify the cells expressing it, we induced abscission zone of cotton leaves, using the ethylene releasing compound ethephon (see "Materials and Methods"). Longitudinal sections of abscission zone at the base of the petiole were then subjected to in situ mRNA hybridization, using monofunctional LKR-specific antisense and sense RNA probes, derived from the 3' intron-located non-coding region of the monofunctional LKR. As shown in Figure 5, a nonspecific color was observed inside the vacuoles of epidermis cells treated with either the antisense or sense monofunctional LKR probes (Fig. 5, a-f). This color was also observed in untreated sections (data not shown). Nevertheless, no alkaline phosphatase staining was detected in the cytosol of these cells after hybridization with both the antisense (lanes a, c, and e) and sense (lanes b, d, and f) probes. In contrast, the parenchyma cells beneath the epidermis layers strongly hybridized with the antisense, but not the sense probes. Many of the parenchyma cells are unfortunately highly vacuolated with very little cytoplasm, and it is, hence, very difficult to see the alkaline phosphatase staining in a photograph. However, clear, intense alkaline phosphatase staining is seen in a number of cells where more elaborated cytoplasm is present (Fig. 5, a, c, and e, staining marked by arrowheads). No such staining is seen in sections treated with the control sense probe (Fig. 5, b, d, and f).

View larger version (172K): [in this window] [in a new window]   Figure 5.   In situ hybridization pattern of the cotton monofunctional LKR mRNA in cotton leaf abscission zone. Near median longitudinal sections (8 µm) through the base of the petiole were hybridized with Dig-labeled antisense (a, c, and e) and sense (b, d, and f) probes specific to the monofunctional LKR mRNA (see "Materials and Methods"). c and d are higher magnifications of a and b, respectively. e and f are higher magnification of neighboring cells. EP, Epidermis; NS, nonspecific color inside vacuoles. The location of the monofunctional LKR mRNA in the cytosol, surrounding the large vacuoles, in parenchyma cells is indicated by arrowheads.

Similar longitudinal sections as shown in Figure 5 where also probed with antisense and sense RNA probes derived the SDH domain of the cotton LKR/SDH cDNA to localize the LKR/SDH mRNA. The SDH antisense RNA probe, but not the sense probe, labeled the same type of parenchyma cells shown in Figure 5, although the labeling intensity did not seem as intense as that obtained with the monofunctional LKR probe (data not shown).

Expression of the monofunctional LKR in the leaf abscission zone was also studied by a quantitative reverse transcriptase (RT)-PCR, in comparison with normal leaf base (not treated with ethephon) and flower buds as controls (see "Materials and Methods"). Three sets of specific primers were used. One set was specific for the monofunctional LKR, the second specific for the bifunctional LKR/SDH, and the third specific for ubiquitin as an internal control. As shown in the top panel of Figure 6, the ubiquitin-specific primers amplified DNA bands with comparable intensities from the three different tissues, suggesting that the RT reaction operated at a comparable efficiency on RNAs extracted from these tissues. As shown in the middle and bottom panels of Figure 6, the relative intensity of the monofunctional LKR band was considerably higher in RT-PCR from abscission zone RNA than from the untreated leaf base and flower buds RNAs. The opposite was observed for the bifunctional LKR/SDH band, which was relatively more intense in RT-PCR from flower buds RNA than abscission zone and untreated leaf base RNAs. The intensity of the monofunctional LKR band derived from the untreated leaf base RNA was notably slightly higher than that derived from the flower buds RNA. This may indicate that leaf bases become committed to form abscission zones long before abscission can be detected morphologically. The quantitative RT-PCR was repeated several times with similar results.

View larger version (26K): [in this window] [in a new window]   Figure 6.   Quantitative RT-PCR analysis of the bifunctional LKR/SDH and monofunctional LKR mRNAs in cotton. Top, middle, and bottom, PCR products with primers specific to the control ubiquitin, monofunctional LKR, and LKR/SDH mRNAs, respectively. Lanes a through c, RNA taken from induced leaf abscission zone, noninduced leaf base, and flower buds, respectively.

The LKR/SDH Genes of Maize, Arabidopsis, and Tomato (Lycopersicon esculentum) Also Possess Putative Polyadenylation Sites within Introns for Production of a Monofunctional LKR mRNA

In two previous reports (Tang et al., 1997; Kemper et al., 1999), faint mRNA bands of smaller sizes than the LKR/SDH mRNA were detected in northern blots from maize and Arabidopsis plants, which were specifically hybridized with the LKR but not with the SDH domain of the LKR/SDH gene. The origins of these LKR-specific mRNA bands were not elucidated. To explore whether these low abundance mRNAs may have been produced by transcription termination within introns, we analyzed the DNA sequence of introns within the intermediate domain regions of the maize and Arabidopsis LKR/SDH genes (GenBank accession nos. AF271636 and U95758, respectively). Several low-affinity polyadenylation sites were identified within introns from both genes, which can potentially produce monofunctional LKR mRNAs. Examples of such polyadenylation sites within intron 12 of the maize LKR/SDH gene and intron 11 of the Arabidopsis LKR/SDH gene are illustrated in the supplementary Figure 1 and supplementary Figure 2, respectively (they can be viewed at www.plantphysiol.org). Using a similar DNA sequence analysis, we have also recently identified putative low-affinity polyadenylation sites within introns in the intermediate domain region of the tomato LKR/SDH gene (supplementary Fig. 3; it can be viewed at www.plantphysiol.org). These results suggest that a monofunctional LKR mRNA is produced in many plant species.

The Monofunctional LKR Is a More Efficient Enzyme Than Its Counterpart That Is Linked to SDH

To provide some insight into the physiological significance of the new catabolic monofunctional LKR, we wished to compare its biochemical properties with that of its counterpart that is linked to SDH. To address this, we used recombinant Arabidopsis LKR/SDH and monofunctional LKR, fused at their N termini to a tag of six histidines (His tag), which we have previously expressed in yeast (Saccharomyces cerevisiae) cells (Zhu et al., 2000a). We have also previously shown that yeast is a highly reliable system to study the activities of recombinant LKR/SDH enzymes from Arabidopsis (Zhu et al., 2000b). The recombinant LKR and LKR/SDH were purified on a nickel column and analyzed for LKR activity under increasing concentration of the three LKR substrates Lys, -ketoglutarate, and NADPH. Reactions were performed in a 0.1 M phosphate buffer at pH 7.5, which resembles the physiological pH of the cytosol where LKR/SDH and apparently also the monofunctional LKR are localized (Kemper et al., 1999; Zhu et al., 2000). As shown in Table I, the monofunctional LKR possessed a nearly 10-fold lower K_m to Lys than the LKR activity of LKR/SDH. No difference between these two enzymes was observed for the K_m of -ketoglutarate and NADPH. The average LKR specific activities of the monofunctional LKR and bifunctional LKR/SDH enzymes under conditions of excess substrate concentrations (V_max), as determined by units per microgram of protein ± SD, were also comparable being 20.59 ± 1.17 and 15.15 ± 1.11, respectively.

	DISCUSSION

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

The Cotton LKR/SDH Locus Encodes a Novel Form of a Catabolic Monofunctional LKR, Using Internal Low-Affinity Polyadenylation Sites

By complete sequencing of three ESTs derived from cotton boll abscission zone, we have identified a bifunctional LKR/SDH cDNA and two cDNAs encoding a novel form of a catabolic monofunctional LKR. The coding DNA sequences of the monofunctional LKR cDNAs were identical to each other and to that of the bifunctional LKR/SDH, implying that they are produced from the LKR/SDH locus. This was also confirmed by DNA sequence analysis of the LKR/SDH gene from a BAC clone showing that the monofunctional LKR-specific 3'-DNA sequences were derived from a large intron within the intermediate domain region between the LKR and SDH coding regions. Our Southern-blot analysis (Fig. 3) also supported this observation, indicating the presence of two LKR/SDH genes in tetraploid cotton, apparently a single gene per diploid genome. Because both of the cotton monofunctional LKR cDNAs ended with a 3'-poly(A) tail, we concluded that their transcription termination is regulated by two polyadenylation sites located inside the large intron within the intermediate domain region of the cotton LKR/SDH locus. Such putative elements (Graber et al., 1999) were identified within the DNA sequence of the intron (Fig. 4).

Production of the Cotton Monofunctional LKR mRNA Is Enhanced in the Parenchyma Cells of the Abscission Zone

Using in situ mRNA localization with RNA probes derived from the intron within the intermediate domain of the cotton LKR/SDH gene, we showed that the monofunctional LKR is predominantly expressed in parenchyma cells of the abscission zone. Moreover, our quantitative RT-PCR also suggested that the level of the monofunctional LKR in these cells is relatively higher than in nonabscised leaf bases and flower buds. This was also supported by the relative high frequency of LKR/SH-derived sequences in the cotton abscission zone library (three of only approximately 1, 800 ESTs) and by the fact that two of these three ESTs were derived from monofunctional LKR mRNAs.

The mechanism controlling the enhanced production of monofunctional LKR in abscission zones is not known but is likely related to the efficiency of transcription termination and mRNA polyadenylation at the two putative polyadenylation sites within the intron. The sequences of these two polyadenylation signals are notably much more diverged from the consensus plant polyadenylation signals (Graber et al., 1999) than the polyadenylation sites in the 3' end of the LKR/SDH gene. Thus, it is likely that in many tissues, such as flower buds, transcription termination and mRNA polyadenylation within the intron will be minor resulting in dominant production of the bifunctional LKR/SDH mRNA. Enhanced production of the monofunctional LKR mRNA in the abscission zone may be regulated by a mechanism that enhances the recognition of the two low-affinity polyadenylation sites within the intron. Such a mechanism was previously shown to control the synthesis variant proteins from a number of mammalian genes, with a good example being the production of a soluble and a membrane-bound form of IgM heavy chains from a single locus in human B cells (Takagaki et al., 1996). The switch to the synthesis of the smaller soluble IgM polypeptide occurs by preferred use of a low-affinity polyadenylation site within an intron. This preferred use is associated with enhanced synthesis of CstF64, a member of a protein complex that binds to the polyadenylation site (Takagaki et al., 1996). Whether a similar mechanism also operates in plants still remains to be elucidated.

A Monofunctional LKR mRNA Is Likely Also Produced from the LKR/SDH Loci of Other Plant Species

By analysis of the sequences of the intermediate domain regions of the maize, Arabidopsis, and tomato LKR/SDH loci, we identified putative low-affinity polyadenylation signals with introns that can direct the synthesis of monofunctional LKR in all of these plant species. Moreover, transcription termination at these low-affinity polyadenylation sites would have expected to produce transcripts that correspond in sizes to the faint LKR-specific mRNAs bands that were previously detected in maize and Arabidopsis (Tang et al., 1997; Kemper et al., 1999). Transcription termination and polyadenylation in sequences that diverge from the consensus plant polyadenylation sites were previously reported when yeast genes were expressed in plants (Grec et al., 2000), supporting the presence of a polyadenylation machinery that can recognize such low-affinity polyadenylation signals. Our results suggest further that such a machinery is quite dominant in abscission zones and is used to produce a novel functional gene product, i.e. a catabolic monofunctional LKR.

The Physiological Significance of the Catabolic Monofunctional LKR

Our biochemical studies showed that the Arabidopsis monofunctional LKR possesses a K_m of 0.328 mM Lys, whereas its counterpart that is linked to SDH possesses a significantly higher K_m of 5.18 mM Lys (Table I). We believe that these characteristics are also true for LKR/SDH and monofunctional LKR enzymes of other plant species because the LKR/SDH polypeptides of different plant species possess highly conserved sequences and biochemical properties (Arruda et al., 2000; Galili et al., 2001). Because the level of Lys in the plant cytosol is estimated to be around 1 mM (Winter et al., 1993), our results imply that the monofunctional LKR is much more efficient than its counterpart that is linked to SDH in vivo. Taken together, we, thus, hypothesize that plants possess two different fluxes of Lys catabolism. In some tissues, such as inflorescence and developing seeds, the LKR/SDH is apparently the major contributor to the total LKR activity, and the flux of Lys catabolism may therefore be relatively limited because of the inefficient nature of LKR/SDH. The major function of Lys catabolism in such tissues may be to regulate Lys homeostasis. A transient enhanced production of the more efficient monofunctional LKR (and in some plant species also a monofunctional SDH; Tang et al., 2000) in specific physiological programs, like abscission, may enable a temporary superefficient flux of Lys catabolism into Glu (Fig. 1). Glu is an important regulatory metabolite in plants. It serves as a precursor for the stress-related metabolites Pro, -amino butyric acid, and Arg, which is the donor for polyamines and nitric oxide and is also a modulator of Glu (for review, see Galili et al., 2001).

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

Materials

Cotton (Gossypium hirsutum) plants were grown in pots under a greenhouse condition (12-h photoperiod at 25°C ± 5°C). Ethephon (Dropp Ultra soluble concentrate, Agrevo, Cambridge, UK) was kindly provided by Dr. Alon Haberfeld (Hazera Quality Seeds Ltd., Kinyat Gat, Israel). Long-template PCR amplification Taq polymerase, PWO Taq polymerase, dNTPs, RNase-free DNase, and DIG RNA labeling kit were purchased from Roche Diagnostics (Mannheim, Germany). Moloney murine leukemia virus (M-MLV) RT was purchased from Promega (Madison, WI). Super-Therm DNA Polymerase was purchased from JMR Holdings (London).

Isolation of the 5' Region of the Cotton LKR/SDH Locus

The 5'-coding region of the cotton LKR/SDH (missing 5' regions in the three cotton ESTs) was cloned by inverse nested PCR as follows: Two micrograms of cotton genomic DNA was cleaved with XbaI for 2 h. After heat inactivating the enzyme at 65°C for 20 min, the DNA was allowed to self-ligate overnight. Two microliters of the overnight ligation reaction was used as a template for PCR analysis using the two oligonucleotides Cot-P9-R and Cot-P2-F (Table II). Reaction conditions included: annealing temp of 55°C, 8-min elongation, 40 cycles, and long-template PCR amplification Taq polymerase. Two microliters of the inverse PCR reaction was used as a template, for a second PCR reaction with the two oligonucleotides Cot-Inv-2-R and Cot-P7-F (Table II). Reaction conditions were as for the first PCR reaction, but the PWO Taq polymerase was used. The PCR products were cleaned with phenol/chloroform cut with XbaI and cloned into the XbaI and SmaI sites of the SK plasmid. The subcloned PCR products were sequenced and assembled into the 5' regions of the three cotton LKR/SDH-related ESTs to generate full open reading frames.

Induction of Leaf Abscission Zone

Blades of fully developed young cotton leaves were excised, and several drops of 20× dilution of the ethephon soluble concentrate were applied to the cutting. Control leaves were kept intact. About 5 d after addition of the ethephon, leaf bases were collected and either immediately frozen in liquid nitrogen and kept at 70°C or fixed for in situ analysis as previously described (Zhu-Shimoni et al., 1997).

Primer Use for Specific Amplification of cDNAs

The sequences of the primers used for PCR amplification are illustrated in Table II. The cotton ubiquitin mRNA (GenBank accession no. AI728891) was amplified by primers Cot-UBQ-5' and Cot-UBQ-3', generating a 312-bp-long DNA fragment. Specific amplification of the cotton LKR/SDH and monofunctional LKR were performed by two sets of primers, each containing the common forward Cot-P2-F primer derived from the LKR domain of GhLKR/SDH. For specific amplification of the bifunctional LKR/SDH, a second reverse primer derived from the SDH domain (Cot-P24-R) was used; whereas for specific detection of the monofunctional LKR, a second reverse primer (Cot-P5-R), located upstream the polyadenylation site within the intron situated within the intermediate domain was used. These two sets of primers amplified DNAs of 255 and 257 bp, respectively.

In Situ mRNA Hybridization

To construct specific sense and antisense probes for in situ detection of the cotton monofunctional LKR, the two primers Cot-P5-R and Cot-P16-F were used for PCR on cotton genomic DNA using PWO Taq polymerase. These primers amplify DNA sequences from the large intron within the intermediate domain of the cotton LKR/SDH gene. The PCR product was cloned into the XbaI and Sam sites of the Bluescript SK vector.

To construct specific sense and antisense probes for in situ detection of the cotton bifunctional LKR/SDH, the GhLKR/SDH cDNA was amplified with primers cot-P7-F and cot-P17-R from the SDH domain. The PCR product was digested with EcoRI and SmaI (inside the cot-P17-R primer) and subcloned into the Bluescript SK plasmid.

Digoxigenin-labeled sense and antisense RNA probes were obtained by in vitro transcription using the DIG RNA labeling kit. Tissue preparation and in situ hybridization were carried out as described (Drews, 1995). An antisense probe and the corresponding control sense probe were used in each experiment.

Extraction of Nucleic Acids and Quantitative RT-PCR Analysis

Cotton DNA was extracted as previously described (Bernatzky and Tanksley, 1986). Total RNA from various cotton tissues was isolated as previously described (Wan and Wilkins, 1994).

Total RNA (1 µg) was treated with 1 unit of RNase-free DNase. After enzyme inactivation at 65°C 20 min, the RNA was annealed with 10 nmol of oligo(dT) (17-mer) primer. Reverse transcription was performed by adding a mixture of 1× M-MLV buffer, 1 mM dNTPs, 20 units of RNase inhibitor, and 200 units of M-MLV RT in a 20-µL final volume. Reactions were incubated at 42°C for 1 h, and then the RT was inactivated at 95°C for 5 min. Two microliters of RT reaction was taken into a mixture of 1× PCR buffer, 0.25 mM dNTPs, 5 pmol of primers of the specific genes, 2.5 units of Super-Therm DNA Polymerase in a 50-µL final volume for PCR amplification. PCR was performed for 30 cycles each containing 45 s at 94°C, 1 min at 53°C, and 1 min at 72°C. The PCR products were then separated on a 1% (w/v) agarose gel. Amplification of the different RT reactions with the ubiquitin primers resulted in bands with similar intensities between different tissues, whereas their amplification with the LKR/SDH and monofunctional LKR primers resulted in bands with different intensities between tissues. This suggested that the PCR reactions were still in the linear range of amplification and did not reach the plateau level expected from deletion of the rate-limiting PCR components. To further corroborate the quantitative nature of the results, all first amplification reactions were diluted twice by 10- and 100-fold and subjected to a second amplification as specified above. Amplification of both dilutions resulted in very similar relative intensities of the different bands to that observed by the first amplification of the RT reaction, supporting the linearity of all amplification reactions.

Purification of His-Tagged Proteins and Analysis of LKR Activity

Expression in yeast (Saccharomyces cerevisiae) and purification of Arabidopsis bifunctional LKR/SDH and monofunctional LKR polypeptides, fused at their N termini to an epitope tag of six histidines (His tag), has been previously described (Zhu et al., 2000a, 2000b). LKR activity was performed in an incubation assay containing approximately 0.1 µg of protein from the nearly purified preparations in 0.3 mL of 0.1 M phosphate buffer (pH 7.5), 20 mM Lys, 14 mM -ketoglutarate, and 0.4 mM NADPH as previously described (Zhu et al., 2000b). Experiments elucidating the apparent K_m of the three LKR substrates used different concentration of these substrates. The apparent K_m values were determined graphically from double-reciprocal plots of activity against the variable substrate concentration.