Maize Genetics Cooperation Newsletter vol 87 2013



Waksman Institute, Rutgers University


Montclair State University


A sequence-indexed single gene knockout resource for maize


Li, Y1; Huang, J1; He, L1; Wang, Q1; Xiong, W2; Segal, G1; Du, C2; Dooner, HK1


The purpose of this note is to apprise the maize genetics community of progress in our NSF-funded project to develop a single-gene-knockout reverse genetics resource based on the transposon Ds.  We have three objectives:  (1) To construct a set of 120 roughly equidistant transgenic Ds launching platforms that will allow simple visual selection of element transposition from any region of the genome and, thus, enable researchers to generate regional gene knock-out collections, (2) To isolate several thousand Ds insertion sites from model platforms and sequence-index them using a combination of next-generation sequencing (NGS) technology and computational tools that should make the method generalizable to any collection of insertions produced in a common background; and (3) To develop a web-searchable database of insertion site sequences cross-referenced to stocks available from the Maize Genetics Stock Center.

The transposons Ac and Ds tend to insert in or close to genes and are, therefore, excellent gene-searching engines in the highly repetitive maize genome (Cowperthwaite et al., Plant Cell 14:713-726, 2002; Vollbrecht et al., Plant Cell 22:1667-1685, 2010).  Both elements show a distinct preference to transpose to linked sites: about one-third of all transpositions are within 7 cM on either side of the donor site (Greenblatt, Genetics 108:471-485, 1984; Dooner and Belachew, Genetics, 122:447-457, 1989; Cowperthwaite et al., Plant Cell 14:713-726, 2002).  Therefore, Ac/Ds elements are excellent gene tagging tools for localized transposon mutagenesis and complement Mutator elements, which show a more random pattern of insertion across the genome.

In our project, we transform the c1 HiII hybrid with Agrobacterium using a standard binary vector system (Frame et al., Plant Physiology 129:13-22, 2002).  Our engineered construct is based on the c1-m2 mutable allele originally described by McClintock (Carnegie Inst Wash Yrbk 47, 155-169, 1948), in which a Ds element is inserted in the third exon of the c1 gene (Cone et al., Proc Natl Acad Sci USA 83: 9631-9635, 1986).  In the presence of Ac, c1-m2 shows a spotted aleurone phenotype. We have modified the Ds element to include a GFP marker expressed behind a 22-kD zein promoter that allows us to trace the movement of the element in the genome (Figure 1).

As outlined in Figure 2, when test crossed with a colorless Ac donor line, most transgenotes showed kernel spotting.  The c-m spotted phenotype resembles that of the native c1-m2 allele and segregates 1:1, suggesting transgene integration at a single locus.  Furthermore, all spotted kernels are green fluorescent, confirming that the spots are due to transpositions of Ds:GFP in response to Ac.  Southern-blot and progeny analyses confirmed the integration, expression, and inheritance of the transgenes in the T1 and T2 generations.  Sequences adjacent to the T-DNA launching platforms were isolated by inverse PCR, sequenced, and mapped to the B73 reference genome.  So far, we have generated 160 active transgenic lines and mapped 82 platforms to the maize genome.  Their location and those of any new mapped platforms can be found in our project website,  These lines are being deposited in the Maize Stock Center and will serve as starting materials for the generation of gene knockouts by community researchers.

We have measured the reversion frequency of each platform by crossing to a colorless c1 tester and selecting for C revertants.  The average reversion frequency was 2.97 x 10-2 on the male side and 1.64 x 10-2 on the female side, a bit higher than that of the native c1-m2 allele (2.0 x 10-2 as male, and 0.7 x 10-2 as female). The higher reversion frequency makes these lines more efficient for generating mutations.

To date, more than 12,000 purple C revertants have been selected from nine platforms and those from three platforms have been further characterized genetically.  The GFP marker in the c1-m2 (Ds) allele serves to track Ds* reinsertion after excision from the c1 gene.  The C revertant kernels carrying a trDs* are readily distinguished from those without Ds under blue light illumination in a fluorescence dissecting microscope.  The average reinsertion frequency is close to 50%, similar to what has been observed in nontransgenic systems.

We use the c1-m2 transgenic lines as pollen donors because about 93% of C selections from the male side are concordant (C endosperm and embryo), whereas only 1/3 of C selections from the female side are concordant, i.e., heritable.  The genetic distance between the trDs* and the C allele in the T-DNA can be readily obtained by scoring the fraction of green fluorescent colorless kernels in the testcross progeny.  As expected, the majority (74.1%) of the genetically linked reinsertions are tightly linked (within 7 cM) to the original T-DNA platform.  80% of the tightly linked sites fall between 0 and 3.5 cM.

The trDs* elements are mapped to the reference genome by isolating and sequencing the reinsertion sites.  To do so, we adopted the splinkerette-PCR method (Uren AG, et al., Nature Protocols 4:789-798, 2009) and constructed libraries for next generation sequencing (NGS) of pooled trDs* insertion sites.  In brief, genomic DNA of seedling tissue from C GFP revertants arranged in 3-D pools is sheared to 2-3 kb and ligated with a double stranded splinkerette oligonucleotide.  The ligation products are amplified by nested PCR, in which the first reaction is carried out with a GFP primer and an adaptor primer that can only anneal to a template synthesized by linear amplification with the GFP primer.  The nested reaction is carried out with a primer from the end of Ds and a barcoded adaptor primer.  The amplified insertion junctions are sequenced in-house by a SOLiD 5500xl system.

At present, we have used the above protocol on two sets of 960 C revertants arranged in 3-D pools of 10 plates x 8 rows x 12 columns. . A new pipeline, InsertionMapper, was specifically developed for the project.  Using this pipeline, we have been able to assign trDs* junction sequences to 1320 individual C revertants and anchor them to the reference genome.  Among them, 1183 are inserted in single copy DNA, and 809 are in genes.  The results of the physical mapping are generally consistent with those of the genetic mapping.  All the above information is provided in our website,, and updated periodically.  The website offers BLAST search capabilities for researchers to identify the stock(s) of interest.  The stocks will be available from the Maize Genetics Stock Center, upon advance APHIS notification of interstate movement.

Acknowledgment.  This project is supported by NSF-PGRP grant DBI 0929350.



Figure 1. T-DNA construct used in Agrobacterium transformation of Hi-II (c1) embryos.  The gus HindIII (H) fragment of pTF102 was replaced with the c1-m Ds* excision reporter shown in TAG 21 (LB, left border; RB, right border). 




Figure 2. Genetic scheme to generate transgenic c1-m lines and isolate C revertants carrying a trDs* element


Please Note: Notes submitted to the Maize Genetics Cooperation Newsletter may be cited only with consent of authors.