20.109(S08):Assay protein behavior (Day6)
I promised that we would revisit the topic of scale today. You might remember that back on Day 3, you used a Nanodrop fluorimeter to obtain the titration curve for wild-type inverse pericam against calcium. You painstakingly pipetted each sample onto the platform and clicked a button to measure its fluorescence. Now imagine if you wanted to test three different samples, or ten, each at a dozen calcium concentrations. You might end up with a repetitive strain injury if you had to do research like that for too long!
Today you will obtain titration curves for your wild-type and mutant proteins using an automated fluorescence plate reader. This machine reads multiple samples in a standard format – in our case, a 96-well microtiter plate. The output is a grid of up to 96 fluorescence values, for rows A-G and columns 1-12, which is amenable to analysis with a program like Excel.
One improvement upon testing single samples is mechanical rather than digital, namely your new friend the multichannel pipet. This tool allows you to suck up and expel equivalent volumes of multiple identical samples (usually 8-12 at a time) with just one stroke. You will use this type of pipet to fill each row of a microtiter plate with one type of protein sample, and each column with a different concentration of calcium. Although a multichannel pipet can be sufficient for a typical research lab, in pharmaceutical companies that may be assaying thousands of samples a day, yet more steps of automation and scaling up are required, such as robotic pipet arms that obviate the need for manual pipetting at all. The degree of automation commercially available, or developed ‘in-house’ in a certain lab or corporation, depends in part on the frequency with which a certain assay is used. Assays used by many different labs and companies (such as fluorescence or absorbance spectrophotometry) are likely to breed commercially available high-throughput machines.
While the concept of scale is a pragmatic concern, a perhaps more substantive topic of interest to us today is that of confidence in our results. As you are probably well-aware, every manipulation and measurement you make in the lab has an error associated with it. For example, consider the ubiquitous P200 pipetman. According to one pipet manufacturer, its accuracy is 1% (slightly worse at the lowest volumes). So an attempt to pipet 100 μL would result in an actual volume of 99-101 μL from the error of the instrument alone, which could be further compounded by a sleepy pipet operator, say. The precision of a pipet is typically better than its accuracy, 0.25% for the example given above. Precision refers to the reproducibility of a given measurement, not its absolute accuracy. This simple example demonstrates the general principles applicable to other types of error.
You will attempt to get a sense of the overall error of today’s experiment by running your protein samples in duplicate. That is, for each protein-calcium combination, you will perform two independent measurements. These measurements can then be averaged to smooth out your data, and hopefully improve the signal to noise ratio (where signal here refers to true differences between samples mixed with different amounts of calcium, and noise means inherent fluctuations in the system due to error). Noise in this experiment can also refer to background fluorescence of the sample buffer. Thus, another way that maintain a reasonable signal:noise ratio is by keeping our protein fairly concentrated, so that the absolute fluorescence values we obtain are high compared to the background. The figure at right demonstrates the above concepts. Scatter in the data (not all of the circles are at the same height) is one kind of noise. The level of background is another kind of noise: the left-hand data has a relatively low signal and thus poor signal:noise ratio, while the right-hand data has a relatively high absolute signal and improved signal:noise ratio.
Only two groups at a time will work in lab today. Last time you should have signed up to arrive at 1:05, 2:20 or 3:35 pm.
Part 1: Protein gel observation
Take a look at your Coomassie-stained gel if you would like to. Scanned photos of the gels are available on the "talk" page of Day 5.
Part 2: Prepare samples for titration curve
- Take a 96-well plate, and label the first six rows as follows: top two rows are wild-type, next two rows are your first mutant, and the final two are your second mutant.
- Transfer an aliquot of wild-type protein to a plastic reservoir. Use the multichannel pipet to add 30 μL of protein (per well) to the top two rows of your plate.
- Rinse the reservoir with distilled water, and dry with chemical wipes.
- Repeated steps 2 and 3 for your mutant proteins, adding each one to the appropriately labeled rows.
- Finally, add plain EB buffer (no protein) to the seventh row of the plate. (Why do you think we are including a protein-free row of solutions?)
- Take a new reservoir. Fill the left-hand side with 500 μL of the lowest concentration calcium solution. Now add 30 μL to the top seven rows in the first column of the plate. Discard the pipet tips.
- Empty the reservoir and add 500 μL of the next lowest calcium solution to the right-hand side. This time add 30 μL to the (top seven rows of the) second column of the plate.
- Continue with each of the twelve calcium solutions, alternating the reservoir side each time to minimize propagated error.
- Finally, replace the plate’s cover and then wrap the plate in aluminum foil.
Part 3: Fluorescence assay
- BPEC (the Biological Process Engineering Center) has graciously agreed to let us use their plate reader. Walk over to the BPEC instrument room with a member of the teaching staff.
- You will be shown how to set the excitation (485 nm) and emission (515 nm) wavelength on the plate reader to assay your protein.
- Your raw data will be posted or emailed to you; alternatively, you can bring your own flash drive to recover the data immediately.
For next time
- Update: no longer mandatory, due to shift in Part 2 due date. Prepare a figure and caption for your SDS-PAGE results. Look up the expected molecular weight using the IPC sequence document and this or a similar website. Be sure to add ~ 3 KDa for the size of the N-terminus of pRSET (His tag, etc). If you see two strong bands, what do you think the second one is?
- If you haven't already, start thinking about what journal article you would like to present to the class a week from today. Suitable articles for presenting are here but you should not feel restricted to this list. If you have another article in mind please email me the citation for approval. Keep in mind that there is a written component to the journal club as well (Part 3 of the portfolio).
- Remember that your Module 1 lab report re-write is due next time.