Maize Tilling Project

In this section we have tried to anticipate questions and problems that prospective Maize TILLING users might have and how to resolve them. The Maize TILLING project was started as a colloboration between the Seattle TILLING project and Cliff Weil's lab at Purdue University to establish a Maize TILLING facility modeled after the Arabidopsis TILLING Project (ATP). To learn more about TILLING you might want to look the ATP FAQ page.

Should I TILL my gene?

Most of the EMS-generated TILLING mutations will be missense mutations or silent mutations that occur in introns or do not change the amino acid due to the degeneracy of the genetic code. A small percentage of the mutations will be nonsense mutations, or those that affect splicing junctions.

If you have a gene knockout that results in a phenotype, then YES! TILLING can provide you with sub-lethal alleles. However, if the gene knockout does not have a phenotype neither will the TILLING mutants.

If you are not sure what phenotype the mutant will have, you should have a way to assay gene activity. The TILLING mutants will harbor many other mutations besides the one in the target gene, thus you will need to confirm that the phenotype is due to the mutation in the target gene and not one of the other background mutations. Please see the "mutant seed" section for more details on how our mutants were generated.

How long will my TILLING request take?

It varies and depends in part on the number of other TILLING requests received at the same time. Ideally, we would like to deliver results in ~ 4 months. Please see "TILLING order status" for a more detailed explanation.

What if I do not have the complete genomic DNA sequence for my gene?

TILLING primers are designed against the genomic sequence, and CODDLE needs to know the exon-intron structure in order select primers that will TILL regions containing the most protein coding sequence rather than intronic sequence. There are several ways you can obtain more gDNA sequence. If you are lucky, it is just a BLAST search away. Once you have found more sequence for your gene, you can use GeneSeqer to construct the gene model.

1. BLAST Search Zea mays in PlantGDB

PlantGDB contains contig assemblies of the various EST and gDNA fragments that have been deposited into various databases. Use either the genomic DNA you already have in hand or your cDNA sequence/protein sequence to BLAST (it is under the tools menu) against Zea mays.

2. BLAST search against TIGRAZM4

3. BLAST search against the trace archives at NCBI and select "Zea mays - WGS" or "Zea mays - other" as your search database.

4. BLAST against all plants (or various subsets) using PlantGDB. We suggest starting with rice and then other grasses and plants as needed.

Often times the position (though not necessarily the size or sequence) of introns within a gene is conserved between maize and rice, and maize and other grasses. If the gene is from a different plant, you will need to confirm the exon-intron organization in maize by designing primers to amplify and sequence the corresponding intron regions from maize genomic DNA.

As a last resort, you can create a maize-rice fusion gene model by inserting the rice intron sequences into the corresponding location in the maize cDNA. If you decide to CODDLe with such a fusion gene, be sure to check the primer output! Please make sure that the primers are not located in introns. If the primer is designed against a rice intron sequence, chances are the PCR will not work efficiently.

Why do my gDNA and cDNA/protein sequences need to be from the same genetic background?

The first step of TILLING is entering your gDNA and cDNA (or protein) sequences into CODDLe for analysis. If CODDLe detects any inconsistencies, or non co-linearity, you will get an error message and it will not proceed any further. A single nucleotide mismatch between the gDNA and cDNA will give an error. There can be slight variations in a gene's sequence depending on its genetic background. This error can be corrected by changing the mismatched base(s) in either one of the sequences.

What if my gene sequence is not from B73 or W22?

Your sequences do not have to be from B73 or W22 in order to CODDLe and submit a TILLING request. You should, however, verify that the primers selected by CODDLe work on both B73 and W22 gDNA and amplify a robust product of the expected size. If the PCR product is a little bit larger or smaller than expected, it is likely due to differences in intron size, which can be confirmed by sequencing the PCR product. As long as the PCR product is less than ~ 1550 bp, we will be able to TILL it. Our PCR conditions do not work well on targets > 1.6 kb.

What if my gene is part of a gene family, or is duplicated?

If you know that your target gene is part of a gene family or is duplicated you may have to do some manipulation of CODDLe and Primer3. You should try to select primers that will preferentially amplify one of the targets but not the other and you may already know where the primers should be located to accomplish this. You may need to define the region that CODDLe/Primer3 should use to select primers. We are able to TILL 1.5 kb targets, and that is the default parameter on CODDLe, however, TILLING a smaller fragment may allow you to obtain amplicon-specific primers. The CODDLe Tutorial describes how to change the TILLING window.

The default setting on Primer3/CODDLe is to return 5 primer sets, but you can increase that number and sort through the list until you find the primers in the desired location. You may opt to order several sets of primers for your gene target and test them in your own lab first using our PCR conditions. By doing the PCRs and sequencing the products, you can determine which primers are (and are not) suitable for TILLING before submitting your TILLING request through CODDLe.

We are able to TILL targets that contain a few heterozygous bases and these do not cause a problem later during mutation analysis because the heterozygous base(s) will be found in the W22 and B73 sequences as well as all the putative mutants we sequence for that target. Thus if there are only a few base differences between duplicated genes we may not be able to catch it. We did try to use thermal denaturation kinetics as a way to detect multiple amplicons, however the results were inconsistent and often would indicate that multiple amplicons were present when there were not based on the sequence and agarose gel data.

BLOCKS did not find any homologous proteins corresponding to my gene product. Why and what can I do about it?

BLOCKS uses InterPro 12.0 protein database to find proteins related to your target sequence. If the protein encoded by the gene is classified as "hypothetical" or "uncharacterized", the BLOCKS program may not find any matches or good hits because the database queried only has known proteins. If you are working with a hypothetical protein, we strongly recommend that you upload your own alignment of related proteins so that the TILLING mutations can be analyzed by SIFT to determine if they would be damaging. The protein alignment should be in either a .msf file or in FASTA format and then just follow the steps below.

1. After CODDLE displays the PSI-BLAST results, leave the results unchecked and click on "proceed with CODDLE". On the next screen, there will be a field for "User defined BLOCKS" that provides directions on how to input your own alignment.

2. Click on "User defined BLOCKS" to get to the block maker.

3. If your alignment is saved as .msf (one of the save as options in CLUSTAL), go to multiple alignment processor. Your sequence should be the top one in the alignment.

If your alignment is in FASTA, click on the FASTA format link to double check how your sequence should be formatted if you are using an alignment in FASTA.

4. Scroll down to the window and cut/paste or upload your file and click on the "submit" button.

5. After the BLOCKS have been made, click on "BLOCKS formatted". You should get a text file with the information arranged in blocks of text according to regions of similarity.

6. Save as an .html file. This html file will be accepted by CODDLE.