Common Plasmid Design Mistakes and How to Avoid Them
Why design mistakes are expensive
A single design error — a missing terminator, a frameshift, an incompatible origin of replication — can cost weeks of troubleshooting and hundreds of dollars in failed experiments. Most of these mistakes are detectable before you ever order a primer.
This guide covers the 10 most common plasmid design errors we see in PlasmidStudio's Design Health checks, explains why each one matters, and describes how to fix or prevent it.
1. Missing terminator
What happens: Read-through transcription past your gene of interest produces aberrant mRNAs, reduces expression levels, and can destabilize the plasmid.
Why it's missed: Terminators are less visually obvious than promoters in map editors. Many templates include them, so designers assume they're present without checking.
How to fix it: Add a strong terminator (BGH polyA for mammalian, rrnB T1/T2 for bacterial, CYC1 for yeast) downstream of your coding sequence. Match the terminator to your expression system.
2. Reading frame errors
What happens: If your insert is not in frame with the start codon (or with an upstream fusion tag), the ribosome translates the wrong protein — or hits a premature stop codon.
Why it's missed: Frame alignment is easy to verify manually for a single insert, but gets complicated with multi-fragment assemblies, especially when linkers or tags shift the frame.
How to fix it: Count nucleotides from the ATG to your insert junction. The total before the insert's first codon must be divisible by 3. Tools like PlasmidStudio flag this automatically.
3. Incompatible origins of replication
What happens: If you're co-transforming two plasmids with origins from the same incompatibility group (e.g., two ColE1-derived origins), one plasmid will be lost over generations.
Why it's missed: Origin incompatibility is a property of the replication machinery, not the sequence. You can't see it by looking at the map.
How to fix it: Use origins from different incompatibility groups. Common compatible pairs: ColE1 + p15A, pMB1 + pSC101. Check the incompatibility group in Addgene's vector database.
4. Internal restriction sites in the insert
What happens: If your cloning strategy uses restriction enzymes and your insert contains a site for one of those enzymes, the insert gets cut during digestion. You get fragments instead of your full-length insert.
Why it's missed: Designers check the backbone's MCS but forget to scan the insert sequence for internal sites.
How to fix it: Before choosing your restriction enzymes, run a restriction analysis on the complete insert sequence. The codon optimization tool includes restriction enzyme site detection and can avoid introducing new sites during optimization. Alternatively, use a scarless method (Gibson, Golden Gate) instead.
5. Missing or incorrect Kozak sequence (mammalian)
What happens: In mammalian expression systems, the ribosome scans for the first AUG in a favorable context (the Kozak sequence: GCCACCATGG). Without a strong Kozak, translation initiation is inefficient and expression drops significantly.
Why it's missed: The Kozak sequence is only 6–10 nucleotides. It's easy to overlook when focusing on the coding sequence.
How to fix it: Ensure your start codon is in a strong Kozak context: GCCACCATG (the G after ATG is the first codon's first position and should ideally be G). This is only relevant for mammalian/vertebrate expression systems.
6. Antibiotic resistance gene too close to the origin
What happens: If the resistance gene and origin overlap or are too close, deletions in one can affect the other, leading to plasmid instability or loss of selection.
Why it's missed: When assembling constructs from parts, spatial relationships between backbone elements get overlooked.
How to fix it: Maintain at least 100–200 bp between the resistance gene and the origin. Most well-characterized backbones (pUC, pET, pcDNA) already have proper spacing.
7. Wrong promoter for the expression system
What happens: A CMV promoter in E. coli gives zero expression. A T7 promoter in HEK293 cells gives zero expression. The promoter must match the host's transcription machinery.
Why it's missed: This seems obvious, but it happens frequently when adapting a construct from one system to another, or when using a backbone designed for a different host.
How to fix it: Verify promoter compatibility: CMV/EF1α/CAG for mammalian, T7/trc/lac for E. coli, GAL1/ADH1 for yeast, polh/p10 for insect. PlasmidStudio's Design Health checks flag mismatches when you set the expression system.
8. Repeated sequences causing recombination
What happens: Long repeated sequences (>20 bp direct repeats, >40 bp inverted repeats) are substrates for RecA-mediated recombination in E. coli. This leads to deletions, inversions, and plasmid instability during propagation.
Why it's missed: Repeats are invisible on a feature map. You need to analyze the raw sequence to find them.
How to fix it: Use recA⁻ strains (DH5α, TOP10, Stbl3) for cloning constructs with repeats. For lentiviral LTRs or other unavoidable repeats, use Stbl3 and grow at 30°C.
9. Codon usage mismatch
What happens: Rare codons in the host organism cause ribosome stalling, truncated proteins, and low expression. A gene optimized for mammalian expression may express poorly in E. coli.
Why it's missed: Codon usage bias is species-specific and not visible from the amino acid sequence alone.
How to fix it: Codon-optimize your coding sequence for the target organism. Check the Codon Adaptation Index (CAI): aim for >0.7. Use the free codon optimization tool to check CAI scores and optimize for E. coli, HEK293, CHO, yeast, or insect cells — with codon-by-codon comparison and restriction enzyme site checks.
10. Missing polyA signal (mammalian)
What happens: Without a polyadenylation signal, mRNA is not properly processed, leading to rapid degradation and virtually no protein expression in mammalian cells.
Why it's missed: Like terminators, polyA signals are easy to forget because they're small and don't show up prominently on maps.
How to fix it: Add BGH polyA, SV40 polyA, or hGH polyA downstream of your stop codon. BGH polyA is the most commonly used. This is separate from the terminator — some constructs need both.
Catching these mistakes automatically
All 10 of these mistakes can be detected by automated design checks before you order anything. PlasmidStudio's Design Health panel runs 14 checks in the background as you design, flagging issues with one-click fixes.
The checks include all the issues above plus: insert size warnings (>10 kb), GC content outliers, and missing start/stop codons. Each issue includes a brief explanation and a suggested fix that you can send directly to the AI chat.
Try PlasmidStudio
AI-assisted plasmid design with automated validation. Free during beta.
Join the beta