thebench.bio #1 Ecosystem for the Bio-Tech Industry
All Molecular Biology questions

Question 16 of 20 Intermediate Troubleshooting

A cloning reaction gives no colonies. How would you find the step that failed?

Short answer

Work backwards with controls that isolate one step each. Transform an intact control plasmid to test the cells and the plate, transform the uncut vector to test the backbone and antibiotic, and run vector-only ligation or assembly to see whether the problem is the insert or the join.

Why

No colonies has only a few causes: the cells did not take up DNA, the plate or antibiotic was wrong, the assembly or ligation did not work, or the construct is toxic to the bacteria. Each has a control. Addgene's transformation protocol recommends a positive control plasmid supplied with the competent cells, and notes that efficiency with ligated inserts is roughly ten times lower than with intact plasmid, so a weak batch of cells can succeed on the control and fail on the real sample.

It also points out that more DNA is not always better. Checking that the resistance gene on the plasmid matches the antibiotic on the plate sounds trivial, and it is a frequent cause. Toxicity shows up as colonies that appear only after long incubation, or tiny ones, and a lower copy backbone or a tighter promoter usually fixes it.

Heat shock temperature and timing matter, as does recovery in antibiotic-free medium before plating, especially for ampicillin alternatives.

How to work it through

  1. Plate a known control plasmid with the same cells and plates to test competence and selection.
  2. Transform uncut vector to confirm the backbone and antibiotic match.
  3. Run the ligation or assembly without insert to measure background from re-circularized vector.
  4. Check the insert and vector on a gel for correct size, quantity and complete digestion.
  5. Repeat the join with fresh enzyme and correct molar ratios, and verify the DNA concentration with a fluorescent assay rather than absorbance.
  6. If the control works and the construct does not, suspect toxicity and move to a lower copy backbone, a tighter promoter or a different strain.

An example

A three-fragment assembly gives no colonies. The control plasmid gives thousands, so the cells are fine. Uncut vector grows on kanamycin, confirming the plate.

The vector-only assembly gives two colonies, so background is low. A gel shows the middle PCR fragment is faint and the wrong size. Remaking that fragment with a higher annealing temperature and gel-purifying it gives 40 colonies, and 6 of 6 sequenced clones are correct.

Where the answer stops holding

Controls localize the failing step but do not always name the cause within it. Toxicity can be intermittent and strain-dependent, so one failed attempt does not prove a construct cannot be built. Some failures are upstream, in primer design or a sequence with repeats, which no amount of transformation troubleshooting will fix.

A strong answer covers

  • One control per step: cells, plate, vector, join
  • Positive control plasmid tests competence
  • Vector-only reaction measures background
  • Check digestion, size and quantity on a gel
  • Consider insert toxicity and change backbone or strain

Common mistakes

  • Repeating the whole experiment unchanged instead of running controls
  • Adding far more DNA to the transformation to compensate
  • Assuming the antibiotic on the plate matches the plasmid without checking