INTRODUCTION

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

The small genome of Arabidopsis simplifies in many ways the study of plant genes required for growth and development. Although the entire genomic sequence of Arabidopsis will be available in the near future, the in planta function of the vast majority of genes within that genome is unknown. In Arabidopsis, the current study of plant genetics is made more difficult by the seemingly abundant redundancy of gene function (i.e. Last et al., 1991). For many important metabolic functions in plants there seem to be any number of genes coding for enzymes with similar catalytic properties (Kopczak et al., 1992; Harper et al., 1994; Hrabak et al., 1996; Bevan et al., 1998). In the absence of genetic redundancy, i.e. when a plant has only a single isoform for a gene, one strategy used to investigate gene function is to disrupt the gene in question and observe the phenotype of the resulting plant. This can be accomplished by studying plant lines in which the gene of interest is altered by mutation (Hirsch et al., 1998) or by the expression of antisense or overexpression transgenes to alter gene function (Mizukami and Ma, 1995; Huang et al., 1998). However, these techniques are often insufficient to make apparent the function of a single gene in a family of genes of similar function. In antisense and overexpression studies the effect of the transgene on other members of the gene family is unpredictable (Flavell, 1994; Halliday et al., 1999). In the case of single gene mutations, the redundancy of gene function in the family may mask any phenotypic difference in plants in which the expression of only one isozyme is disrupted (Hua and Meyerowitz, 1998).

To alleviate this problem, an important goal would be to obtain mutant lines with disruptions in all of the isoforms within the gene family. To accomplish this, we have broadened the PCR-based approaches to identify T-DNA insertion mutants reported by McKinney et al. (1995) and Krysan et al. (1996). The major limitation in these initial studies was the number of T-DNA lines available. In McKinney et al. (1995), a total of 5,300 T-DNA lines were used and mutants in two of 10 known actin genes were recovered. In the Krysan et al. (1996) study, 9,100 T-DNA lines were used and, for example, only three of 12 calmodulin domain protein kinases (CPK) mutants were isolated.

In both studies, the techniques were adequate to identify most, if not all, of the available T-DNA inserts in the populations. However, because the T-DNA populations were relatively small, the number of PCR reactions required to perform the initial screens was also relatively small. McKinney et al. (1995) used highly degenerate primers and low annealing temperatures on 53 pools of 100 T-DNA lines each. Krysan et al. (1996) used pools of 1,300 T-DNA lines (seven pools) with gene-specific primers. The use of larger pool sizes was accomplished by making the PCR conditions for screening very stringent. This included using longer primers and higher annealing temperatures, and precluded the use of degenererate primers as described in McKinney et al. (1995).

In this report we describe techniques used to efficiently search the collection of 60,470 T-DNA lines described in Krysan et al. (1999) for individual members of large gene families. We have increased the pool size of primary screens to 2,025 lines and have optimized conditions for the use of gene-specific as well as degenerate primers. With this approach at least one mutant and as many as five additional alleles were identified for 10 of the 12 members of the Arabidopsis H⁺-ATPase (AHA) gene family.

	RESULTS

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

Using Degenerate PCR Primers

Degenerate primers provide the means to search for more than one gene isoform at a time in a reverse-genetic strategy. To design degenerate primers for the AHA gene family, the nucleotide sequence surrounding the conserved amino acid sequence (CSDK) region typical of plasma membrane P-type ATPases was aligned using the genomic sequences for AHAs1, 2, 3, 9, 10, and 11 (Fig. 1; see Table I for accession numbers of sequence corresponding to isoform designations as described by Harper et al. [1994]). Sequence for AHA11 was provided by Jeff Harper (Scripps Institute, San Diego). The degenerate PCR primers were designed based on the following criteria: oligo lengths greater than 27 bp, little or no divergence in the 3' ends, and limited degenerate positions. To limit degeneracy, several mismatch bases (1-3) were allowed in each isoform. In the case of the AHA genes, a 35-bp region provided adequate conservation for both forward and reverse primers (Fig. 1B). The AHA degenerate primer (DEG1) primes in the 5' to 3' direction of ATPase genes, and the DEG2 primer primes in the 3' to 5' direction (Fig. 1B). The DEG1 and DEG2 primers were tested with gene-specific primers for AHA1, 2, 3, 9,10, and 11 (Table I) and found to amplify the expected product in the PCR conditions required for subsequent pool searches (data not shown). Shorter primers and primers with higher degeneracy were found to provide higher background when using large pools (data not shown).

View larger version (32K): [in this window] [in a new window]   Figure 1.   Degenerate primer design and position. A, Degenerate primers were constructed from DNA sequence around the conserved CSDK region typical of plasma membrane P-type H+-ATPases. Degenerate bases (in bold font) are located in three positions in the AHA degenerate primers. The number of mismatched bases is indicated to the right of the gene sequence. The DEG2 primer complements the same conserved region. B, The degenerate primers DEG1 and DEG2 are located near the center of the AHA genes and prime in opposite directions, in the direction of the solid lines.

To identify T-DNA insertions within the AHA gene family using degenerate primers, four combinations of ATPase primer and T-DNA border primer are possible (DEG1 + T-DNA left border, DEG1 + T-DNA right border, DEG2 + T-DNA left border, and DEG2 + T-DNA right border). For illustration purposes, the results from the DEG2 + T-DNA right border primer PCR reactions are presented in Figure 2A. To determine whether the products visible on the ethidium-stained gel represent mutant alleles with T-DNA insertions inside AHA genes, a dot blot was performed. Rather than probe for each AHA gene separately, a mixture of probes comprised of AHA1, 3, and 10 was hybridized against a dot blot (Fig. 2B). To check for cross hybridization with other AHA isoforms, 2 ng of genomic DNA from AHAs 2, 6, 9, and 11 was also dotted and probed. Dots corresponding to lanes 5, 7, 17, 19, and 30 showed strong hybridizations (Fig. 2B). The PCR bands visible on the ethidium gel that did not hybridize to the probes are artifacts typical to this type of reverse genetic screen (McKinney et al., 1995; Krysan et al., 1996). Cycle sequencing of PCR products from those reactions with a T-DNA right border primer confirmed T-DNA right border/plant DNA junctions (data not shown). In similar experiments, summarized in Table I, DEG1 and the T-DNA left border primer produced five strongly hybridizing bands, and DEG2 and the T-DNA left border primer produced 11 strongly hybridizing bands. Of the positives, the size of five of the PCR products indicated that the T-DNA tag landed near, but outside of, the gene. Twelve of the remaining bands were sequenced, confirming insertions in all isoforms except AHAs 2, 5, 6, and 12. The presence of multiple mutant alleles was indicated for AHA1 (4) and AHA10 (2). A summary of the results from these PCR reactions and subsequent analysis is presented in Table I and Figure 3. The DEG1 and T-DNA right border primer combination was not tested, and thus additional inserts may be present in the population.

View larger version (93K): [in this window] [in a new window]   Figure 2.   PCR using AHA degenerate primers and a T-DNA right border primer. a, Ethidium gel of PCR on 30 super pools using DEG and XR2. b, Dot blot of PCR reactions probed with a mixture of probes comprised of AHAs 1, 3, and 10. Genomic fragments (2 ng) of AHAs 2, 6, 9, and 11 were dotted to test for probe cross hybridization. Positive controls for AHAs1, 3, and 10 not shown.

View larger version (24K): [in this window] [in a new window]   Figure 3.   Insertion sites for AHA gene family T-DNA mutants. Gene lengths have been normalized. Position of each T-DNA insertion is depicted by an inverted triangle. Insertions labeled (s) have been identified based on Southern analysis only.

Using Gene-Specific Primers to Search for Mutants in AHA Isoforms 2, 5, 6, and 12

Alignment of the AHA degenerate primers with AHA2 and 6 shows good homology; however, the matches to AHAs 5 and 12 are poor. Gene-specific primers were designed for these isoforms to determine the efficiency of the degenerate primers (Table II). Two mutant alleles for AHA2 were detected and three for AHA5 (Table I). For isoforms AHA6 and AHA12, no T-DNA insertions were detected (data not shown).

	DISCUSSION

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

This study demonstrates that degenerate primers provide an efficient means to isolate many mutant alleles in large gene families. With this approach we have identified at least 16 new Arabidopsis plasma membrane H⁺-ATPase mutants in an ordered population of 60,740 T-DNA insertion lines. This includes new mutants in eight of 12 isoforms. To augment the search with degenerate primers, we have also screened the ordered pools with gene-specific primers designed for isoforms for which no mutants were discovered with the degenerate primers. Searches with gene-specific primers indicated that AHA2 and AHA5 mutant alleles not detected with the degenerate primers were present in the collection. Searches for AHA6 and AHA12 with gene-specific primers were negative, indicating that a T-DNA insertional population of 60,740 is not yet saturating, even for large genes. The ATPases range between 4.5 and 6.5 kb. However, with a combined degenerate and gene-specific primer approach, we have quickly identified mutants in 10 of the 12 known isoforms and now have multiple mutant alleles for seven of the 12 AHA gene family members.

With these resources we will now be able to begin addressing questions of gene function and genetic redundancy in the AHA gene family. Once mutants homozygous for the T-DNA insertions are identified and backcrossed to establish a single insert line, the phenotype can be analyzed. With mutants in most AHA isoforms now available, tests for functional redundancy between isoforms can be performed by making genetic crosses between mutants. A directed approach can be taken in which crosses between gene family members with similar sequence or expression patterns can be made, or a more comprehensive approach can be taken in which double, triple, quadruple, etc. mutants are created. Once a variety of multiple-mutant lines are available they can in turn be crossed with each other to obtain complex segregating populations. At any point, the genotype for seedlings with observed phenotypes in these segregating populations can be determined using PCR genotyping (Krysan et al., 1996).

In the near future, increases in the size of T-DNA insertion populations will make it possible to obtain mutants in any gene of interest. However, the study of one particular gene often requires the analysis of many closely related genes. This study provides a framework for beginning the daunting task of describing the function and relationship between Arabidopsis gene family members. The results presented here indicate that this approach provides an efficient and effective start toward finding the genetic resources to bridge sequence information and plant biology.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

Organization of the T-DNA-Transformed Lines

This study uses 60,470 independently transformed lines established through the collaboration of Drs. Rick Amasino and Sandra Austin-Phillips (University of Wisconsin, Madison), Michael Sussman, and colleagues. Wild-type Arabidopsis (ecotype Wassilewskija) plants were transformed with the pD991 T-DNA vector derived from the binary plant transformation vector pCGN1547 (McBride and Summerfelt, 1990). Kanamycin-resistant seedlings (T1 generation seeds) were isolated and transferred to soil, nine seedlings per pot, and allowed to self-pollinate. Offspring of these plants (T2 generation seeds) were collected in pools of nine totaling 6,720 pools. Twenty-five pools of 9 were then combined (using approximately 180 seeds from each pool, i.e. approximately 20× sampling) to form seed pools of 225. These seeds were placed in 125 mL distilled water and gently shaken at 4°C on an orbital shaker for 2 d. The seeds were then moved to an orbital shaker under lights at room temperature, allowed to germinate, and harvested after 5 to 7 d. Harvested seedlings were rinsed with distilled water, patted dry, and then ground with a mortar and pestle in liquid nitrogen. DNA was extracted using the DNAeasy Plant Mini Kit (Qiagen, Valencia, CA). The DNA was suspended in a final volume of 500 µL of TE (10 mM Tris^.HCl/1 mM EDTA, pH 9). Super pools were comprised of equal volumes of DNA from nine pools of 225 DNA, resulting in 30 super pools of 2,025 lines each.

PCR Strategy and Conditions

Super Pool PCR (2,025 Lines)

Pools of 225 PCR (225 Lines)

Pool of Nine DNA Preps (Nine Lines)

J-Pool DNA PCR

Isolating Individual Lines

Cycle Sequencing

For the super pool PCR products that were the predominant product in the reaction (i.e. Fig. 2, lane 30), the reactions were cleaned using a QIAquick PCR Purification Kit (Qiagen) and subjected to cycle sequencing using the T-DNA border primers. The sequences were generated by the University of Wisconsin, Madison Biotechnology Center DNA Sequencing Facility. For reactions in which more than one band was present, individual bands were gel isolated, frozen in 500 µL water, thawed, and then spun at the highest speed in a microcentrifuge for 5 min. Fragments suitable for sequencing were then re-amplified with the primers used in the primary screen.