It is possible the PCR is amplifying free inserts that are present on the plate, you should use one primer from inside the insert and one from the vector for colony PCR.

Usually false positives in colony PCR are from two things:

1. Amplification of left over "naked" DNA from the assembly reaction which is present on the plate but failed to find its way into an E. coli cell. This can be mitigated to some extent by being very careful not to poke the agar underneath a colony, but this doesn't always help. Other ways to get rid of this include:

• A. Obtaining a good amount of colony on a pipette (without gouging the agar), resuspending it in sterile PBS or saline in a centrifuge tube, centrifuging it, then discarding the supernatant to obtain a washed cell pellet which can be used for colony PCR (most of the naked DNA should go out with the discarded supernatant). This can be repeated a couple times if a single wash isn't enough.
• B. Simply poking a colony with a sterile pipette tip or tooth pick and patching it on a new plate then doing colony PCR from the patch plate the next day.

2. Colony PCR primers are amplifying something approximately the correct size in the E. coli genome. This is less common, but it happens. There's ~5 million base pairs in the E. coli chromosome any given primer could potentially match. Off target amplification can be prevented to a decent extent by designing primers using a program that checks for primer specificity. I like to use NCBI Primer Blast to design primers. For colony PCR primers, I usually tell it to check for off-target amplification on the whole genome sequence of the E. coli strain I'm using (or a closely related strain if that strain isn't sequenced).

Alexandra Johnson Yoav Lubelsky The problem is that I observe this situation even when I perform PCR with my gene primers on the plasmid extracted with miniprep. Moreover, I have never had this problem with my classical cloning attempts, where I have often used much larger amounts of insert compared to those used with the Gibson method, and I have never observed this issue of "false positives". This is why it seems very strange to me, but I really don’t know how to proceed since the sequencing gave this unusual result.

Laura Dionisi That is strange - A couple questions to help narrow down what the problem might be:

1. When you say sequencing did not reveal presence of your gene, do you mean the sequencing data of all your plasmids aligned to the sequence of the original vector without an insert? Or did it give you something else?

2. For sequencing, did you use classical primer based sequencing (ie. some version of Sanger sequencing) or was it long read whole plasmid sequencing?

Alexandra Johnson I tried performing two sequencing runs for two different cloning products. In the first attempt, Oxford Nanopore was used, and in this case the sequence aligned perfectly with that of the plasmid, but there was no trace of the gene. In the second case, Illumina was used (I know it’s not ideal, but that’s the platform available in my lab), and again the sequence assemblies showed very high coverage for my plasmid sequence, but not a single trace of the gene.

In the past few days, I also tried performing PCR again on my thawed colonies, and none of them were positive. Moreover, the analysis using a single restriction enzyme — attempting to linearize the plasmids and compare the sizes between those “with insert” and those without — also showed no significant differences. Considering that my genes to be inserted are around 750–1000 bp, on a 1% agarose gel I was expecting to observe a clear difference..

My problem probably lies in the PCR amplifications; I’m not sure how, but I need to optimize this step. It’s possible that some mismatch is occurring, preventing the correct annealing of the sequences?

Did you run a no template control in the PCR and was it clean? It is possible you have some contamination of your insert on your work area or in one of the reagents which is the source for the false positive result.

Yoav Lubelsky Yes, I always perform also a positive (with cDNA) and a negative control (with water) and it was everything good.

@Laura Dionisi

I would just try designing completely new primers for screening, preferably ones outside the 750 bp region of the current primers.

If that doesn't work I would try getting a new miniprep kit.

At this point it seems like the only things I could see it being are off target primer binding to some part of the E. coli genome (even in a good miniprep there is still a small amount of chromosomal DNA) or some part of the miniprep kit itself is contaminated with DNA.

Never use colony PCR from the plate you used to spread your transformed cells.

You need to streak potentially interesting colonies on a new plate (with selection), grow a liquid culture (also with selection) or do both.

It takes more time, but it's worth it to avoid the very avoidable problem of false positives due to unincorporated plasmid.

I'd recommend that you do a liquid culture, then a plasmid prep, digest to linearize*, and then PCR.

*Choose an enzyme that will cut at least once in the backbone and does NOT make any cuts in your insert. Double check to be darn sure, don't make assumptions about the enzyme being a "rare cutter". Trust me on this one.

As my advisor would say, "You can't rush quality"

Good luck!