This method generally follows that of standard large-interval screening (see the Barstead lab web site), with enhancements and changes as described below. It is not particularly fussy, but care should be taken to get good separation of products on the gels, as deletions as small as 78 bp have been detected by the method. A molecular weight marker that provides bands over the range of 100 bp to 3 kb is very helpful (we use Bio-Rad Molecular PCR Ruler, cat. # 170-8206).
Primer Design
Method 1: Use our Primer Design web site to design a matched high-quality set of two external primers, a poison primer, and two internal primers. This extended version of Steve Jones’ original program uses current ACeDB data, Primer3 and e-PCR to design primers targeting the gene and exons of your choice while minimizing the risk of non-specific amplification. It designs poisons specifically to target exons at the 5' end of the gene, if possible, and placed roughly in the center of the amplicon. Method 2: Design your own poison or poisons to match existing primers. Ideally, you should place the poison in the center of the amplicon and match its Tm to that of the external primers. For either method, the external amplicon for use with poisons should be between 1200 and 2600 bp.
Reaction Mix Formulation and PCR ParametersOne or two poison primers are included in the external round in amounts equimolar to those of the normal external primers. We typically use 5 or 10 pmoles of each primer per 25 l reaction. Internal round reactions are formulated as for normal screening. Use 61 C as the annealing temperature in the external/poison round, and 55 C in the internal round if your primers were designed using our web site. Such primers are predicted to have annealing optima of 55 C, but we found in temperature gradient tests that most have a broad range of usable temperatures, and most seem to prefer higher temperatures. 61 C is a good middle-of-the-road temperature, and in external-round gradient tests on wild-type DNA it most often resulted in a good band from the poison and its productive external partner, and an attenuated full-length product. Fine adjustment of extension time is not critical for success of this technique. Longer extension times may appear to override the effect of the poison on production of the full-length external product (measured in multiple side-by-side reactions on identical wild-type DNA), but this has not been shown to interfere in deletion detection. In fact, minor attenuation of the full-length wild-type product has resulted in recovery of otherwise undetectable deletions (see Figure 3 in the method description).
Use the following rules of thumb to determine extension time:
| Internal products up to 1200 bp: |
Extend 30 seconds |
| Internal products of 1200 2600 bp: |
Extend 1 minute |
| Internal products over 2600 bp: |
Extend 1.5 or 2 minutes |
Products, Gels and ImagingIdeally, addition of a poison to the external PCR round will limit production of the full-length external product to the point where it is only weakly amplified in the internal round. This is not always the case, so you may see a fair amount of the full-size product on your gels. You may also see smaller products resulting from the presence of the poison primer. In practice, such bands on the gels usually don’t interfere with seeing deletions. Our screening gels are 2% agarose in 1X TBE. We add 10 l loading buffer (15% Ficoll, bromophenol blue and xylene cyanol) to each 10 or 25 l reaction, and load 1 l of the mix on the gel. Gels are stained with SYBR Green and imaged on a Molecular Dynamics Fluorimager. This system is extremely sensitive, and has allowed us to see pretty faint bands that turned out to be real deletions, but most often deletion bands seen with poisons are fairly robust. Imaging can be accomplished with an ethidium bromide / UV lightbox system, but the lower sensitivity will require more of the reaction to be loaded on the gel. The complete recipe and cycling parameters follows at the end of this page. For 10 l reactions in 384-well PCR plates, divide all amounts by 2.5.
Library Screening Results to DateIn the most recent library for which we have final data, we screened 100,000 genomes for deletions in 210 genes (235 different primer intervals), accounting for 275,000 bp using poisons and 212,000 bp with normal primers. We obtained 59 hits in primary screening, which were put into row test (pooled and individual library DNAs each in duplicate reactions to confirm presence of an amplifiable deletion). 30 of the apparent deletions passed this test to give a worm well address. The worms were resampled and digested, and PCRs were done in quadruplicate on a positive control (library DNA), the resample, and a negative control (N2 DNA). 17 of the 30 passed this test and were put into sib selection. We obtained 15 of the 17 as viable worm strains. Of these 15, four were identified without poisons in the external reactions, and 11 with poisons. Poison plus/minus reactions were done for those 11 on pooled library DNA, and demonstrated that the poisons were required in nine cases. (One of the remaining two did not delete the poison, and the other did but was large enough that the poison was not necessary for detection.) The sizes of those for which the poison was required range from 368 bp to 850 bp, averaging 608 bp. In summary, the poison method is allowing us to detect small deletions in relatively complex libraries (1 chromosome in 1200) that were undetectable using normal screening methods, accounting for over half the deletions obtained in a single screen. Also, the targeting feature of this method makes it possible to detect deletions in specific regions of genes.
