During my undergraduate internship I was making different formulations of insulin nanoparticles that would, in theory, be delivered through an inhaler and into the lungs. My mentor told me about the lethal dose of insulin nanoparticles when they were delivered to the lungs of mice. My eyes must have widened, because he quickly followed up by telling me the lethal dose of water in the lungs, which was much less. Meaning, water in their lungs was more potent in mice compared to our insulin nanoparticles. The point was, it became understood that there are no lethal molecules only lethal doses.

A similar point applies to antibiotics and cell selection agents. At an appropriate concentration these tools will kill sensitive cells but will be innocuous to cells that harbor the resistance gene. However, at a low enough dose antibiotics and cell agents will have no impact on any cell, and at a high enough dose they will kill everything. That’s why finding the appropriate concentration for the tool at hand is the key step here.

A kill curve is a dose-response experiment using antibiotics or cell selection agents. It compares their potencies against sensitive and resistant cells to determine the optimal antibiotic or cell selection agent concentration.

In this article, we’ll discuss how kill curves are used to determine the appropriate concentration of antibiotic or cell selection agent to use for a given experimental setup.

 

In this article:

Antibiotics, cell selection agents, and resistance genes

What is a kill curve?

Comparing kill curves for sensitive and resistant cell lines

Do I need to make a kill curve?

References

 

Antibiotics, cell selection agents, and resistance genes

An antibiotic is a molecule that kills bacteria. A cell selection agent also kills cells. Usually when people use the term cell selection agent, they’re describing a molecule that also kills eukaryotic cells like fungi, plants, and mammalian cells.

Some, but not all, antibiotics are also cell selection agents that kill eukaryotic cells. For example, puromycin, blasticidin, hygromycin, and G418 are both antibiotics and cell selection agents.

---

---

 

Antibiotics and cell selection agents are used to isolate particular cells in a mixed population. Cells that carry the corresponding resistance gene for a given antibiotic or cell selection agent will grow just in the presence of that molecule (Table 1). That is because the resistance gene modifies and inactivates the antibiotic or cell selection agent (Figure 1).

Table 1. Antibiotics or cell selection agents with their corresponding resistance genes.

 

We’ve covered these terms really briefly here, but if you’re interested in learning more about antibiotics check out this article, and for cell selection agents this article is a good reference.

Figure 1. Cell selection agents are inactivated by resistance genes encoding enzymes that modify the selection agent.

 

 

What is a kill curve?

A kill curve is a dose-response experiment used to determine the concentration of an antibiotic needed to kill cells over a specific time period.

Usually the cells to perform the kill curve on are the sensitive cells that lack the resistance gene for the antibiotic or cell selection agent. But as we’ll discuss in the next section, there is also value in running a kill curve on the resistant cells to determine the optimal concentration for your experimental setup.

Conducting a kill curve is really easy.  You titrate your antibiotic or cell selection agent against the cells of interest and determine which concentration completely kills the cells over the specified time (Figure 2).

There are many assays you can use to measure live cells such as metabolic assays (MTT or ATP-based), trypan blue exclusion, and flow cytometry. By comparing to a control sample that wasn’t treated with an antibiotic or cell selection agent you can assign a percent viability to each of your samples, between 0 and 100%.

Additionally, for bacteria, yeast, and other microbes cell viability is often assessed simply by measuring the optical density at 600 nm (OD600) using a spectrophotometer.

For sensitive cells that lack a resistance plasmid, shown as the orange data points in Figure 2, you would be looking for the minimum concentration of antibiotic or cell selection agent that kills all untransfected cells.

Figure 2. Hypothetical example of a kill curve where each dot corresponds to a different concentration of an antibiotic or a cell selection agent.

 

What should your specified time period be? That depends on the cells and tools that you’re using. For bacteria and antibiotics like ampicillin or kanamycin, overnight incubation (~12-24 hours) is plenty. Conversely, for mammalian cell culture and cell selection agents this process usually takes a few weeks (Table 2).

GoldBio has a titration kill curve protocol available to help you set this up.

 

Table 2. Suggested concentration and duration for antibiotics and cell selection agents (VectorBuilder, 2026).

Selection Agent	Cell line	Recommended concentration	Recommended duration
Puromycin	293T	1-2 ug/mL	3-5 days
Blasticidin	293T	5-15 ug/mL	7-11 days
Hygromycin	293T	100-200 ug/mL	5-7 days
G418	HT1080	500-1,000 ug/mL	7-11 days

 

Comparing kill curves for sensitive and resistant cell lines

The overall idea with picking a concentration for your antibiotic or cell selection agent is that it will kill sensitive cells that lack the resistance gene, whereas cells that contain the resistance gene will grow just fine (Figure 3).

Figure 3. Left, the plasmid (purple) in the top left corner of the resistant cell allows it to grow with selection agent present. The right cell doesn’t have the plasmid with the resistance gene and so it is killed the selection agent.

 

The kill curve experiment described in the last section is usually performed for sensitive cells that lack a resistance gene, and it’s generally a safe bet that the minimum concentration to kill sensitive cells won’t perturb cells that harbor the resistance gene.

However, if it is important to determine the ideal concentration that kills sensitive cells but has the least impact on resistant cells, then you would want to run a kill curve on both cell types (Figure 4).

Figure 4. By performing kill curves against sensitive (orange) and resistant (purple) cells you can choose the concentration of antibiotic or cell selection agent that provides the largest difference between the two types of cells.

 

Cells that possess the resistance plasmid, shown as purple data-points in Figure 4, will have higher survival at most concentrations of antibiotic or cell selection agent. The difference between the two cell lines at any given concentration is called the selectivity (black arrow in Figure 4), and in general you would want to maximize this value. For your experiments, you’ll probably still want to choose the minimum concentration that kills all sensitive cells. Ideally your resistant line will still have a pretty high survival rate at this concentration.

 

Do I have to make a kill curve?

You don’t have to make a kill curve. Using concentrations of antibiotics or cell selection agents that have previously been described in the literature will usually work well for most experiments. For example, I have never optimized the antibiotic concentration to use in my standard molecular cloning and protein expression protocols. I just use standard concentrations and that has always worked well.

However, there may be times where you will want to perform kill curves to determine the optimal concentration of antibiotics or cell selection agents for your experimental setup. For example, cell selection agents have different killing potencies between different cell lines (Delrue et al, 2018). So, if the reference concentration you have for your cell selection agent was found using a different cell line, then you will probably want to optimize on your own.

Another way of thinking about this is the common phrase “is the juice worth the squeeze?” For the standard cloning and protein expression experiments that I described above, by using appropriate controls I would know within a day or two if my experiment had worked. And if the experiment didn’t work, I could go back and run a kill curve then to make sure that my antibiotic was performing at the concentration used in the experiment.

However, if you’re making a more substantial time and resource investment, it is probably worth the squeeze up front to make sure you’re using an optimal concentration in your experiments. For example, when establishing stable cell lines that will be routinely used, or running a CRISPR screen, it is likely worth the effort up front to setup the best possible experiment and get the best possible results out of your extensive efforts.

 

If you’re ready to kill the right kind of cells at an optimized concentration, then check out GoldBio’s wide ranging selection of antibiotics and cell selection agents using the links below or by searching on our website.

 

Tags

antibiotics Cell Selection Simon Currie

• Share :
• Deel
• X
• LinkedIn

---