T7E1 and Surveyor assays: how they work, their limits, and how to interpret results

This is Arc 1, Part 11 of the CRISPR from Bench to Analysis series.

You’ve designed your guides, transfected your cells, and waited 48 hours. Now comes the moment of truth: did it cut?

Before you spend hundreds of dollars on Next-Generation Sequencing (NGS), you need a quick "yes/no" answer. That’s where the mismatch-cleavage assay comes in. Whether you call it a T7E1 assay or a Surveyor assay, the goal is the same: use a "mismatch-hungry" enzyme to find the small errors (indels) Cas9 left behind and cut them into smaller pieces you can see on a gel.

This post explains how these assays work, which enzyme to choose, and—crucially—why they often lie to you.

What you'll learn

• The molecular mechanism: Denaturation, Re-annealing, and Cleavage
• T7E1 vs. Surveyor: A head-to-head comparison
• The Protocol: From cell pellet to agarose gel in 4 hours
• Math time: How to calculate % editing from gel band intensity
• The "Lies": Why these assays consistently underestimate your editing efficiency

How it works: The "Mismatch" Principle

When Cas9 creates a double-strand break and the cell repairs it via Non-Homologous End Joining (NHEJ), it often leaves behind small insertions or deletions (indels). If you PCR-amplify the region around the target site, you get a mixture of two types of DNA:

1. Wild-type (WT): Perfectly intact sequences.
2. Edited: Sequences with 1-10 bp changes.

If you just run this PCR product on a gel, you won't see anything. A 1 bp difference on a 500 bp fragment is invisible.

To find them, we follow three steps:

1. Denaturation & Re-annealing

We heat the PCR product to 95°C to melt the double strands, then cool it down slowly. During cooling, the strands find partners. Most will find their original match, but some WT strands will pair with Edited strands. This creates a heteroduplex—a double helix with a "bubble" or "kink" where the sequences don't match.

2. Enzymatic Cleavage

We add a nuclease (T7E1 or Surveyor) that specifically recognizes these kinks and cuts both strands of the DNA at that spot.

3. Gel Electrophoresis

If the enzyme found a mismatch and cut it, your original 500 bp band will disappear and two smaller bands (e.g., 200 bp and 300 bp) will appear. The presence of these "cleavage products" is your "Yes, it worked" signal.

T7E1 vs. Surveyor: Which one to use?

Most labs have a preference, but they aren't identical.

The consensus: T7E1 is the workhorse of most CRISPR labs. It’s faster, cleaner, and handles the "random" nature of NHEJ repairs well. Surveyor is often reserved for when you are looking for single-nucleotide polymorphisms (SNPs) or very precise 1bp base edits.

Interpreting the Results

When you look at your gel, you'll see the large parental band (uncut) and two smaller bands (cleaved). To estimate the Percentage Gene Modification (% Indel), we use the following formula:

$\% \text{Indel} = 100 \times \left( 1 - \sqrt{1 - \frac{\text{fraction cleaved}}{1}} \right)$

Where fraction cleaved is the sum of the intensities of the two small bands divided by the total intensity of all three bands.

Wait, why the square root? Because a single indel can form multiple heteroduplexes. The math accounts for the probability of a mutant strand pairing with a wild-type strand.

The Real Talk: Why these assays lie

Mismatch assays are great for a "quick look," but they have three major limitations:

1. They are insensitive: If your editing efficiency is below 5%, you likely won't see anything on the gel.
2. They underestimate high efficiency: As your editing gets better, you start forming "homoduplex mutants" (mutant + mutant). If the sequences match each other, the enzyme won't cut, and you'll think your efficiency is lower than it actually is.
3. Primer Dimers and SNPs: Natural variations (SNPs) in your cell line will show up as "edits" because they create mismatches. Always run a non-transfected control!

My Take

Use T7E1 for your initial screen. It’s the fastest way to know if your guide is alive. But never publish a final editing percentage based solely on a gel. Once you know it works, move to TIDE/ICE (Sanger-based analysis) or Amplicon NGS for the "real" numbers. We'll cover those in the next two posts.

Staring at a gel and can't tell if that's a band or a smudge? Upload a photo or describe it below—let's troubleshoot your T7E1.

Resources