ODsee is SparkleCell's OD600 spectrophotometer. It is used to estimate the cell density in the SparkleCell culture chamber in real time. All SparkleCell models except the tiny EasyGrow come with an ODsee. In this post we show you some of the functions and limitations of this convenient accessory.
ODsee is inserted through the bottom of SparkleCell's base via its built-in sanitary clamp connector. The white stalk points to the inside of the culture chamber.
The visible hole in the stalk contains a glass cuvette, similar to the ones used in desktop spectrophotometers. It is open on both ends, so that the surrounding culture can slosh through it freely. A tiny laser is housed in the outer tip of the stalk. It fires down through the glass of the cuvette and through the culture contained in it. A detector on the other side picks up how much light was scattered by the E. coli cells.
This information gets passed on to the electronics contained in ODsee's shiny steel body. An Arduino microcontroller then calculates the detector's raw result into a standard "OD600" value, that will be the same number any desktop spectrophotometer with 1 cm path length puts out. The OD600 value is passed out through the USB connector into your SparkleCell housing.
Potential issues to be aware of:
There are a few fundamental physical issues that can cause differences between the numbers seen by desktop spectrophotometers vs. in-chamber live photometers like ODsee:
- Temperature-dependence of the electronics
- Changes in solute solubility, and precipitation on the optical surfaces
- Changes in cell-specific scattering during protein expression ("puffing up")
- Differences in beam quality
- Saturation of the signal due to shadowing
ODsee is designed to operate while your air pump is running and the chamber is full of bubbles. It can see the cells, but sees right through the bubbles. Optimized beam quality, the angled cuvette and its low / sunken location in the chamber all work together to make this possible.
But some of the remaining physical issues do not have a perfect solution. For example, a desktop spectrophotometer has the luxury to hold its detector temperature constant, while ODsee's detector temperature adapts to the temperature of the culture chamber. Since the photoelectric effect is temperature-dependent, a change in culture temperature necessarily affects the sensor.
Therefore, there are small remaining differences between ODsee's number and the number reported by a desktop spectrophotometer. During engineering, we optimized ODsee's design to give you the best correlation early in the production run, when you need to make decisions based on it. Later in the run, the correlation can suffer, depending on circumstances.
Let's look at some data:
For the present experiment, we produced TEV protease, the enzyme used to remove purification tags from other proteins, from our open source plasmid pSparkleTEV. To express this enzyme in soluble, active form, a temperature drop to 28C is necessary at the time of induction. This is accomplished rapidly by dropping media ice cubes.
We took samples from the culture over time and measured them with a desktop spectrophotometer (Nanodrop ND-1000, cell cultures module), and compared them with the real-time readings produced by ODsee. Above OD600 = 5, we diluted the spectrophotometer samples 1:10, while ODsee of course does not have the luxury to use dilutions.
As can be seen, the correlation is initially very good. Steps in early culturing, like replenishing ampicillin or inducing protein expression can be decided with confidence, based on ODsee's reading alone. The correlation becomes unhinged during the temperature drop, and appears to recover somewhat thereafter. Plotting the two instruments against each other looks like this:
The data points highlighted in red are the three readings following the temperature drop. They were eliminated from the linear equation and R-squared calculations. Clearly, during this period, the correlation is worst. It could be related to precipitation of media components, such as the poorly soluble Antifoam 204 on ODsee's lens, in combination with temperature effects on the electro-optical sensor itself.
Let's look at another construct:
Here is another of our open source plasmids, pSparkleTaq, encoding Taq DNA Polymerase for PCR. This construct can achieve very high OD600 values during its 5 hour induction period. Again, the spectrophotometer measurements were again made using appropriate dilutions. But ODsee does not have the luxury to rely on dilutions, as it's pointed directly into the culture chamber.
As can be seen, ODsee's measurements become saturated above OD600 = 30. This is due to "shadowing": At these high densities, some E. coli cells are likely to sit in the shadow of other cells. The cells that were hit by the light first already scattered most of it. So the later-hit cells can't scatter as much, simply because there isn't as much light there. ODsee can't "tell" how many cells might be hiding in the shadows behind each other. We decided not to try to correct mathematically for shadowing, but rather show you the true physical data, along with this explanation of what's going on.
- ODsee is useful for estimating the cell density in SparkleCell's culture chamber within reasonable limits.
- Early in the culture period, when important decisions need to be made, ODsee tracks the OD600 values measured by an external spectrophotometer.
- The correlation becomes somewhat unhinged during induction of protein expression, especially in combination with a temperature drop.
- ODsee saturates above OD600 = 30 due to shadowing.