Reading a PCR genotyping panel where the controls misbehave
A state seed-testing laboratory receives 40 cotton seed lots certified as non-GM. The routine conventional PCR panel uses primers for the Cauliflower mosaic virus 35S promoter (450 bp product), primers for part of Cry1Ac, and a cotton lectin primer pair as a plant-DNA control. Results: the extraction blank and the no-template control both show a faint 450 bp band. Lot 17 shows strong 35S and lectin bands but no Cry1Ac. Lot 22 shows all three bands. Lot 31 shows only a very weak lectin band. The laboratory must sign a certificate within a week. Nearby fields have been found with volunteer Bt cotton, and the buyer's specification is zero tolerance for GM seed. The analyst has access to a real-time PCR machine and to an event-specific primer set for the widely grown Bt cotton event.
1. Classify the three lots and the controls. Which results can be reported and which cannot?
Read the worked answer
Controls first. A 450 bp band in the extraction blank and no-template control is amplicon carry-over or reagent contamination. No 35S result from this run is trustworthy, so the whole run is invalid and must be repeated with fresh reagents, separated pre- and post-PCR rooms, aliquoted primers, UV/bleach treatment and uracil-DNA-glycosylase carry-over control. Faint positives in controls are especially dangerous because they mimic low-level GM contamination.
Lot 17. Lectin positive, so DNA is amplifiable, but 35S positive and Cry1Ac negative. If real, this is not a Cry1Ac Bt event; it could be another GM construct, or a mixture. It certainly cannot be certified non-GM on this evidence.
Lot 22. All three bands - compatible with Bt cotton, but conventional PCR gives presence, not amount or event identity, and contamination in this run undermines it too. Retest and identify the event.
Lot 31. A very weak lectin band means poor DNA quality, degraded template or PCR inhibition; the negative 35S/Cry1Ac results are uninterpretable and must be repeated with diluted template and an added internal amplification control.
Only after a clean re-run can any lot be reported, and lot 22 must additionally be quantified and identified.
2. Why is the 35S promoter a poor sole marker for "GM", and what design would you use instead?
Read the worked answer
Weaknesses of 35S
- It is not universal: many constructs use maize ubiquitin, rice actin, or FMV promoters, and some traits are not promoter-driven in the same way; a 35S-negative result does not prove absence of a transgene.
- It is not specific to any event; 35S appears in herbicide-tolerance constructs, reporter constructs and research lines, and is a viral sequence, so any CaMV-contaminated material is a hazard in brassica testing.
- It cannot distinguish an authorised event from an unapproved one, nor give quantity relative to a labelling threshold.
Better design
- Event-specific primers spanning the transgene–plant junction, where the sequence is unique to that insertion; the event-specific set available to the analyst is the right primary tool.
- Multiplex with a species-specific endogenous reference (lectin for cotton) to prove DNA quality and absence of inhibition, plus a construct element (promoter or terminator) for screening.
- Real-time PCR with certified reference material and a standard curve or comparative Ct method, reporting the GM fraction with a stated limit of detection and quantification.
- Confirm amplicon identity by sequencing, and run no-template and extraction blanks in every batch.
3. The buyer demands zero-tolerance certification. What else must the analyst know before rejecting lot 22?
Read the worked answer
Sampling, not just chemistry, decides this. Seed lots are heterogeneous; a 200 g test portion from a 20-tonne lot can easily miss a low level of GM seed, and the grinding and sub-sampling steps contribute most of the measurement variance. The analyst needs the sampling plan, the number of increments, the test-portion size and the accepted operating characteristic curve for the method.
Analytical limits. Every PCR has a limit of detection; "zero" is not a measurable quantity. Certification should state the limit of detection and the probability of missing a given contamination level, not assert absence.
Identity and origin. The event must be identified and quantified; contamination could come from pollen flow, volunteers, shared machinery or a seed-lot mix-up rather than a GM variety. Chain of custody and reference material matter.
Specification. If the buyer's contract is genuinely zero tolerance, the correct action is rejection pending a clean re-test, with the lot quarantined; if a threshold applies, quantification against certified standards decides. A single conventional PCR band cannot support either conclusion, and the contaminated run makes any report indefensible.