Serial Dilution: A Classic Technique used by Modern Robotics
Intro
In my last two posts, I discussed small volume dispensers and their usefulness in drug discovery and beyond. However, they do have some drawbacks. These instruments can be expensive; struggle with larger volumes; and samples need to be loaded by hand for some of them, leading to errors in sample tracking. Small volume dispensers have their advantages, but you can still get those small dilutions you need with serial dilution. I wanted to discuss automated serial dilution by starting with good practices for manual serial dilution, the benefits of switching to automation, some common errors with serial dilution on liquid handlers, and how to troubleshoot those problems.
Multichannel Best Practices
This is a blog about automation, but understanding how to optimize serial dilution on a robot requires us to start with the basics. As I discussed before, the technology behind many automated pipetting instruments is air displacement, so many best practices that apply to manual pipetting will also apply to automation.
Make Sure Pipettes are Calibrated and Maintained: Regular calibration is necessary to ensure that the volume you set on your pipette is close to the actual volume you are pipetting. It is suggested that pipettes be calibrated every 3-6 months, depending on usage. For best results, use a vendor that specializes in pipette calibration because they will have the most sensitive tools for the job and can repair your pipette if broken.
Sterilize Your Pipettes Regularly: To reduce cross contamination between samples, make sure you are spraying down your pipettes before and after each use, at the very least. Some liquids require special decontamination procedures, see a whole list here. Pipettes are also autoclavable!
Make Sure You’re Dispensing In Your Pipette’s Dispense Range: Pipettes have a dispense range, and oftentimes it’s printed right on the pipette itself. Your P200 pipette may be able to pipette less than 20uL, but doing so affects the accuracy of your pipetting. In serial dilution, those inaccuracies compound. Make sure you know the dispensing ranges of your pipettes before using them.
Pre-Wet Tips: Pre-wetting tips refers to aspirating and dispensing the full volume to be dispensed 2-3 times. This brings the temperature between the inside of the tip and the liquid to equilibrium and humidifies the air space between the pipette and tip. Without pre-wetting, the volume you dispense will be less than your desired delivery volume. This step also helps make the solution more heterogeneous, which brings me to my next point…
Mix Thoroughly Before Dispensing: Not only does this pre-wet the tip for you, but it makes the solution more heterogeneous, or uniform, so the dilution that you’re pipetting is closest to the concentration you want to transfer. I’ll go into how many mixes are necessary for the best results later in this post, but 3-5 mixes is a good amount to get good results without taking a ton of time.
Switch Tips: The biggest drawback of serial dilution is the compounding of errors over the course of the dilution. A big source of this is residual liquid on the outside of the tip, which can throw off the dilution. Imagine having 1uL of extra volume when performing a two fold dilution for ten dilutions. That adds up! Switching tips mitigates this. Just make sure you mix well with each set of new tips!
Excuse my lack of brevity on the subject. I probably could have written a whole post dedicated just to this. Since this is an automation blog, I’d be wrong to stop here. Serial dilution is better with robots, let me tell you why.
Automation is Better, Trust Me
If you’ve been following this blog, I’ve mentioned the benefits of automation many times. Let’s look into how those benefits impact serial dilution in particular.
Automation is Ergonomic: Performing serial dilutions in a 96 well plate involves a lot of wrist work. Between pipetting your diluent into each well, adding your starting dose, mixing, and transferring each dilution, you’re pressing down on that pipette plunger at least 100 times per plate. And that’s just to make your dilution plate. According to the journal of occupational health, carpel tunnel affects about 9.7% of scientists. Save your wrists and have the robot do the tedious work for you.
Time Saving: A well made protocol on a liquid handler will allow you to walk away from your robot and dedicate your time to another task. This effectively doubles your productivity or your rest time (I won’t judge).
Consistency: You can optimize the aspiration and dispensing steps of your protocol with laser precision. Things like tip height, tip location, tip retraction/extension, liquid classes and other more complex technologies control for different aspects of liquid handling that are almost impossible to execute manually. I’ll go into these a little later in this post. Every scientist has their own way of doing things that will make it hard to replicate data between operators. With a liquid handler, the technique will stay the same regardless of who presses the start button.
Less Human Error: How many times have you lost track of your dilution step and had to start over. Sure, there are techniques I used to use, like using tips to keep track of which step I was on (which is another reason to switch tips), but even this wasn’t foolproof. A liquid handler will keep track of all liquids being transferred, so if there ever is a mistake, you’ll find out by looking at the instrument process data rather than when you get your results.
Robots aren’t perfect, just better than humans at these kinds of tasks. Let’s look at some common dispensing errors that we see in automated liquid handlers.
Common Issues with Automated Liquid Handlers in Serial Dilution
Splashing and Cross Contamination: It takes a bit more work to set up a liquid class on a robot that will pipette with a speed appropriate for your liquid. With a manual pipette, you control the speed at which you pipette: more viscous liquid needs to be pipetted more slowly and less viscous liquids can be pipetted more quickly. If this isn’t optimized, you’ll see it in your results. Off the top of my head, I don’t know if a technology exists that can detect splashing on a robot
Wrong Tip Type Being Used: A liquid handler will often have only one head for different aspiration volumes, so the pipette range will vary by tip. Using the wrong tip type can cause similar inconsistencies that you’ll see with manual pipetting. Aspirating volumes that are larger than a tip type can handle can also damage your instrument if the liquid enters the system itself.
Specialized Training and Learning Curve to Program Robots: It takes time to learn how to program robots. For more complex platforms, you may even need to take training courses to become competent with the user interface. If you don’t have this training, you can get stuck and need to wait for help from a vendor before you can process.
Smaller Volumes Are Less Accurate: A calibrated manual multichannel pipette can pipette less than 1µl of liquid with a CV of <1%. When it comes to some liquid handlers, they claim as low as 4% for 1µl, but that requires a well-defined liquid class. This is where I would switch to a small volume dispenser if I had one.
If you noticed, a lot of these problems come from human error, so can we really blame the robots? Let’s look at ways we can mitigate these common liquid handling problems.
Solutions to Liquid Handling Issues
Doing optimization up front can save you a lot of time in the long run, so keep that in mind when reading over these solutions.
Proper Maintenance of Your Liquid Handler: Make sure you’re getting preventative maintenance done on your instrument by the vendor in regular intervals. Also make sure you’re running in-house maintenance protocols routinely. Before a large run, it wouldn’t hurt to QC your robot to make sure everything is in working order.
Well Defined Liquid Classes: Defining your liquid class takes many different factors into account. Work with a vendor or liquid handling specialist to make sure that the liquids you’re working with are being dispensed at the correct speed.
Mixing Well: Using the appropriate amount of mixing steps can improve CV%. In one study, 3 mixes yielded 11.8% CV while 20 mixes yielded 1.7% CV. There is a tradeoff of time, as mixing each step 20 times could take 15-20 minutes a plate. Use what works best for your needs and optimize the functions that I discuss in the next point to optimize your mixing steps.
Optimize Pipetting Functions: Let’s talk about the different settings you can optimize on your liquid handler when pipetting:
Tip height: Controlling the tip height can affect the amount of liquid that is circulated during a mix cycle. It can also mitigate splashing for certain liquids.
Tip Location: Aiming for the middle of the well will generally disperse the liquid more evenly.
Retraction/Extension: Extending tips closer to the bottom of the well during aspiration and retracting further from the well them during dispense to allow for a larger volume of the well to be affected when mixing.
Dispensing Technologies: Liquid Handlers can detect liquid levels with cLLD technologies, using capacitance change measurement to detect the surface of a conductive liquid. This is not so much something you can optimize, but rather something extra that ensures you’re dispensing the right volume every time.
Having an Automation Engineer Onsite: An automation engineer is an in-house expert that can program your protocols for you! They can save time and money versus contacting the vendor for your issues. It’s their job to make sure the robots are maintained and easy to use for any scientist, plus they may have a better idea of the scope of your project to offer a more precise solution.
Use the resources you have to make automation work for you. Science is a collaborative effort, so work with automation engineers and vendors to get things done!
Conclusion:
Serial dilution isn’t perfect, but it can still help you get valuable data to find the next best drugs! And with automation, the process is much less painful. Manual pipetting best practices set the standard for how robots can perform the same tasks, but the robot does it more efficiently and with consistency; furthermore, even though liquid handling isn’t perfect, you can make it work with some optimization and help from automation specialists. It may take a lot of time and money up front, but automation will save you time and money in the long run.
*I tried a longer format today, let me know what you think! Check out my website if you are looking to set up automation in your lab!