| | Abstract | Introduction | Materials and methods | Results | References
Abstract
Barley is an important cereal crop with a size of approx. 5300 Mb per haploid genome. This is too large to be considered for whole-genome sequencing. But barley genome resources including the Morex BAC library, abundant ESTs, and 22K microarray enable researchers to access the barley genome. We aim to couple these resources to accelerate a transition to comprehensive physical mapping and sequencing of the barley “gene-space”. We utilized unigene sequences to design more than 12,600 36-mer “overgo” probes to identify Morex barley BAC clones that carry expressed genes. These BAC clones will be fingerprinted to create BAC contigs, and a minimal set will be identified. In Phase I of this project, 21,161 BACs identified in our own work and that of A Kleinhofs, G Muehlbauer, R Wise, P Hayes, K Gill, N Stein, MA Saghai Maroof and co-workers were fingerprinted, with 13,067 BACs assembled into 2262 contigs comprising ca 9.4% (470 Mb) of the genome. These results are available through the “The Barley Genome” website http://phymap.ucdavis.edu:8080/barley/. In Phase II, more than 7700 abiotic stress related genes (drought, salinity, low temperature or ABA treatment) were identified using the Affymetrix Barley1 GeneChip. In total ~7000 overgos have been used as of June 2005. Of these, about 2149 overgo probes were related to an objective to genetically map 1000 genes associated to abiotic stress. For the purpose of anchoring these abiotic stress related regions on the genetic map, we investigated single feature polymorphisms (SFPs) using the Barley1 GeneChip data using Morex, Steptoe, Oregon Wolfe Barley (OWB) dominant and OWB recessive. We also developed a single nucleotide polymorphisms (SNPs) database from HarvEST:Barley EST sequences. A high throughput method for SNP mapping with R Waugh and N Rostoks (Scottish Crop Research Institute; SCRI) and N Stein, R Varshney and A Graner (Institute of Plant Genetics and Crop Plant Research; IPK) is in progress. Polymorphisms, and genetic and physical map data, will be added to HarvEST:Barley (http://harvest.ucr.edu). Phase III has a goal of probing the BAC library with the remaining ~5500 overgos to identify around 60,000 gene-bearing BACs in all, and to fingerprint and align them into contigs to derive a physical map of the overall minimal set.
Introduction
Triticeae genomes contain at least 80% of repetitive DNA (Bennet and Leitch, 1995), which has so far prevented the Triticeae from becoming the focus of large-scale genomic sequencing projects. In recent years, however, a number of barley genomic resources such as ~400,000 ESTs representing about 70% of all genes, the 6.4X Morex BAC library (Yu et al. 2000), cDNA libraries, several widely used mapping populations, a 22K microarray and its diploid nature have made barley a model Triticeae crop to access its genome.
The most commonly used probes for screening arrayed libraries have been sub-cloned DNA fragments, PCR amplified products (Xu et al. 1998) or DNA oligonucleotides (Klein et al. 2000). A novel approach for making probes, developed by Ross et al. (1999), has the advantage of oligonucleotides and also yields slightly larger probes with better hybridization kinetics and higher specific activity of labeling. These probes, termed overgos are made by annealing two 24-bp oligonucleotides with an 8-bp overlapping region at the 3’ end and filling in the overhanging bases with Klenow enzyme and radiolabeled nucleotides. Multiplexing of overgos enables the hybridization of large numbers of probes in a single experiment. For example, 10,642 overgos designed from ESTs were applied to 165,888 maize BACs in a 24×24×24 experimental design with an 88% success rate (Gardiner et al. 2004).
In our effort to isolate a large number of BAC clones from gene rich loci in the barley genome, we have developed a novel strategy that integrates the technical advantages of currently available library screening methods. We modified the labeling protocol and developed stringent criteria for the selection of sequences used for overgo probes. We developed software that can extract overgos from unique as well as popular sequences from the HarvEST:Barley database.
Following this approach, we designed a total of 12,661 “overgos”. The “OligoSpawn” website http://oligospawn.ucr.edu provides access to elements of our oligo design algorithms. We have been able to pool >200 overgo probes per hybridization for highly parallel hybridization-based screening of the Morex barley BAC library.
Materials and Methods
Barley BAC library. The library was derived from DNA of cultivar Morex using restriction endonuclease HindIII. This library consists of 313,344 individual clones stored in 816 384-well microtiter plates. This library provides about 6.3 haploid genome equivalents with an average insert size is 106 kb. The library is arrayed on 17 high-density DNA filters for screening by hybridization (http://www.genome.clemson.edu).
Oligonucleotide probe design. A computer program “OligoSpawn” was used to design the overgo probes used in this study (Zheng et al. 2004, http://www.oligospawn.ucr.edu). A total of 18,766 overgos were designed from the probesets used in 22K barley GeneChip (Close et al. 2004), and of these 9500 were selected for the present studies. The rest of the overgos (around 2600 for this project) were designed from unigenes that were not covered as probesets on the barley gene chip. These latter probes were chosen on the basis of functional categories of the unigenes from which they were derived.
Oligonucleotide probes and probe pairs. All barley oligonucleotide primers were purchased from Illumina Inc. (San Diego, CA). Each oligonucleotide was synthesized as a 22-mer at 25 nmol scale, dissolved in 250 µl of TE buffer, and diluted 50-fold in the final probe pair mix (final concentration 1 µM).
Oligonucleotide probe labeling. Ten microliters of each probe pair mix was labeled in a separate well of a 96-well PCR microplate with 10 µl of freshly prepared master mix composed of 4.0 µl of 2.5X Overgo Labeling Buffer, 1.0 µl 2 mg/ml acetylated bovine serum albumin (BSA) (Promega), 0.125 µl of all the four radioactive nucleotides ( [α-32P] dATP, [α-32P] dCTP, [α-32P] dGTP and [α-32P] dTTP) each at a concentration of 10 µCi/µl (~ 3000 Ci/mmol) (Perkin Elmer) and 1 unit of Klenow enzyme (New England Biolabs). A dNTP solution composed of 10 mM each of four non-radioactive dNTPs was used for cold chase. An oligonucleotide pair with sequences 5’-AACGGGCGAGTGATGAAAATA-3’ and 5’-TGATGGGATCGGGCTATTTTAC-3’ was used as background overgo to light up all of the bacterial clones. Labeling reactions were carried out at room temperature for 1 h followed by addition of 5 µl of the cold chase solution to each of the reaction tubes. Later, all the reactions were pooled and probes were denatured at 95°C for 5 min and immediately transferred to the hybridization tubes containing prehybridized high density BAC membranes.
High-density filter hybridization. Hybridization was performed in 40 ml of Church’s buffer at 60°C for two nights in a hybridization oven. After hybridization, membranes were extensively washed in solutions with increasing stringency starting with 2 liters of 4X SSC with 0.1% SDS followed by 2 liters of 1.5X SSC with 0.1% SDS and finally with 2 liters of 0.75X SSC with 0.1% SDS at 50°C. Membranes were then sealed in plastic wrap and exposed to Kodak X-ray films (Kodak BIOMAX MS Double Emulsion, 24 x 30 cm) at -80°C for 5-6 days.
Results
Phase I
As the initial step to compile all the barley resources, all of the available BAC addresses from major barley genomic researchers were collected. A total of 21,616 BAC addresses were compiled from seven sources including those identified from our own work and that of A Kleinhofs, G Muehlbauer, R Wise, P Hayes, K Gill, N Stein, MA Saghai Maroof and co-workers. The majority of these BAC clones were identified using mapped cDNA probes, while most of the others were recognized using EST-derived overgo probes. In Phase I of this project, an attempt was made to fingerprint all of these BACs, with 15,513 clones ultimately used for FPC assembly. Of these, 13,067 BACs assembled into 2262 contigs, while 2446 were singletons. These 2262 contigs account for 470 Mb which is about 9.4% of the barley genome. All data is publicly available at the “Barley Genome” website http://phymap.ucdavis.edu:8080/barley/ providing access to BAC contig data.
Phase II
The strategy that we have developed consists of identifiying gene-containing BAC clones through hybridization to pools of overgos designed from EST-derived unigene sequences in the HarvEST:Barley database. We design overgos using algorithms available through the OligoSpawn website (http://oligospawn.ucr.edu). OligoSpawn provides efficient selection of two types of oligos, namely unique and popular, from large unigene datasets. In the context of BAC library screening, unique oligos serve to unambiguously link one gene to BAC clones, while the purpose of popular oligos is to identify the largest possible list of gene-bearing BAC clones using the smallest number of probes. In order to obtain oligos for many genes of interest, and to probe selectively by functional category, we created a local information management software called oSearch. The majority of the overgos were derived from probesets on the Affymetrix barley GeneChip that were up- or down-regulated during abiotic stress including salinity, drought, low temperature or ABA treatment. This software recommends a 36-mer for each unigene or probeset and lists all unigenes represented by a given 36-mer.
Using oSearch we have generated a total of 12,661 overgos to be used in this work. Through early June 2005, a total of about 7000 overgos had been used, generally in pools of 96 to 300 simultaneous probes, most often 192 probes per pool. The reading of positive BAC addresses from all these hybridizations using Incogen’s High Density Filter Reader software (http://www.incogen.com) is in progress. Interestingly, results using 96 popular overgos detected about 4000 positive BAC clones, which is about 40 per overgo, about 6-7 times the expected frequency of probes representing unique genes. This result seems to validate our hypothesis that popular overgos provide economical screening of genomic libraries for gene-bearing clones that carry sequences found in numerous genes. Another result from screening the BAC library with 576 overgos, a mixture of unique and popular, identified more than 5000 positive BAC clones with an average of 9.2 clones per probe (Figure 1).
Figure 1. This chart shows the percent new BACs in each pool for this first 13 pools that we used, as well as the total number of BACs that were identified up to that point.
In order to tie the physical map of BAC clones to the genetic linkage map, around 1000 genes with unambiguous BAC address and contigs will be mapped. To identify and map these unigenes we developed a two-pronged approach based on single nucleotide polymorphisms (SNPs) and single feature polymorphisms (SFPs). SNP discovery was done “in silico” using a relaxed assembly (#32) from HarvEST:Barley and 36 pairwise comparisons between eight barley genotypes. This resulted in 12,615 eSNPs in 3509 unigenes, of which 29 of 32 (91%) randomly chosen eSNPs were validated by direct sequencing. Of these 3509 unigenes, only the subset in the abiotic stress list has been further considered for our mapping purposes. We combined our list of SNPs with others provided by collaborators N Rostocks and R Waugh at SCRI and N Stein, R Varshney and Andreas Graner at IPK. SNPs from 565 and 217 unigenes were provided by SCRI and IPK, respectively, the former list being mainly a subset of the abiotic stress unigene list that we previously shared with our SCRI colleagues. The collective list of SNPs was used to design of an Illumina Oligo Pool Assay (OPA). The OPA is a high throughput genotyping platform designed to genotype 1536 loci simultaneously. Among these, we hope to map as many as 1000 related to abiotic stress in order to accomplish our abiotic stress gene mapping objective. We plan to genotype 96 maplines each from Steptoe x Morex, Barke x Morex, and Oregon Wolfe Barley (OWB) dominant and recessive parents. Also 96 different cultivars, landraces and elite lines will be examined. Designs have been finalized for the Illumina OPA chip. To identify the BAC clones corresponding to these ~1000 abiotic stress genes, about 2149 corresponding overgo probes have been used to screen the BAC library filters.
For our second approach we investigated single feature polymorphisms (SFPs) using the Affymetrix Barley1 GeneChip hybridized with labeled cRNA from the parents of each of three barley mapping populations: OWB dominant x OWB recessive, Steptoe x Morex, Barke x Morex. We developed a detection method using the robustified projection pursuit (RPP) method in order to evaluate the overall differentiations of signal intensities of probe sets comparing two genotypes and to measure the individual contribution of each probe, from which the probes covering polymorphisms (SNPs or INDELs) can be identified (Cui et al., submitted). We randomly selected SFP-containing unigenes for sequence validation and found that 59 of 72 were validated (82%). A total of 2090 SFPs were detected of which 844 (722 probe sets) were abiotic stress responsive as defined by our expression data. A 12,000 probe Nimblegen array was designed to further test the performance of SFPs and optimize SFP representation. The results from the Nimblegen chip indicate that a 25-mer with the polymorphic nucleotide(s) positions within a central region of 6-18 nucleotides is best suited for obtaining higher signal intensity differences between the polymorphic parents.
References
Bennet MD, Leitch IJ. 1995. Nuclear DNA amounts in angiosperms. Ann Bot 76: 113-176.
Close TJ, Wanamaker S, Caldo RA, Turner SM, Ashlock DA, Dickerson JA, Wing RA, Muehlbauer GJ, Kleinhofs A, Wise RP. 2004. A new resource for cereal genomics: 22K barley GeneChip comes of age. Plant Physiology 134: 960-968.
Cui X, Xu J, Asghar R, Condamine P, Svensson JT, Wanamaker S, Stein, N, Roose M, Close TJ. 2005. Detecting single-feature polymorphismn through high density oligonucleotide microarray. (Submitted to Bioinformatics)
Gardiner J, Schroeder S, Polacco ML, Sanchez-Villeda H, Fang Z, Morgante M, Landewe T, Fengler K, Useche F, Hanafey M, Tingey S, Chou H, Wing R, Soderlund C, Coe EH Jr. 2004. Anchoring 9,371 maize expressed sequence tagged unigenes to the bacterial artificial chromosome contig map by two-dimensional overgo hybridization. Plant Physiol 134:1317–1326.
Klein PE, Klein RR, Cartinhour SW, Ulanch PE, Dong J, Obert JA, Morishige DT, Schlueter SD, Childs KL, Ale M, Mullet JE. 2000. A high-throughput AFLP-based method for constructing integrated genetic and physical maps: progress toward a sorghum genome map. Genome Res 10:789-807.
Ross MT, LaBrie T, McPherson J, Stanton VM. 1999. Screening large-insert libraries by hybridization. In Current protocols in human genetics (ed. Ann Boyl), pp.5.6.1-5.6.52. Wiley, New York.
Xu J, Daichang Y, Domingo J, Ni J, Huang N. 1998. Screening for overlapping bacterial artificial chromosome clones by PCR analysis with an arbitrary primer. Proc Natl Acad Sci USA 95:5661-5666.
Yu Y, Tomkins JP, Waugh R, Frisch DA, Kudrna D, Kleinhofs A, Brueggeman RS, Muehlbauer GJ, Wise RP, Wing RA. 2000. A bacterial artificial chromosome library for barley (Hordeum vulgare L.) and the identification of clones containing putative resistance genes. Theor Appl Genet 101:1093-1099.
Zheng J, Close TJ, Lonardi S, Jiang T. 2004. Efficient selection of unique and popular oligos for large EST databases. Bioinformatics 20:2101-2112.
Kavitha Madishetty (1), Jan T. Svensson (1), Pascal Condamine (1), Jie Zheng (2), Steve Wanamaker (1), Ming-Cheng Luo (3), Tao Jiang (2), Stefano Lonardi (2) and Timothy J. Close* (1)
* Corresponding author: timothy.close@ucr.edu
(1) Dept. of Botany & Plant Sciences;
(2) Dept. of Computer Sciences, University of California, Riverside, CA, 92521
(3) Dept. of Agronomy & Range Science, University of California, Davis, CA, 95616
Presented at the 18th North American Barley Researchers Workshop, July 17-20, 2005 |
|