Calculate accurate protein concentrations

Video Transcript

• 00:01 - 00:05: Welcome to today’s podcast on calculating and evaluating
• 00:05 - 00:06: your ELISA data.
• 00:06 - 00:10: My name is Quentin, and we go through our recommendations
• 00:10 - 00:13: on how to calculate ELISA data and get
• 00:13 - 00:16: an accurate protein concentration for your samples.
• 00:16 - 00:18: To go through these calculations,
• 00:18 - 00:22: we’ll be using the HumanPD-L1 SimpleStep ELISA kits
• 00:22 - 00:26: that we have recently developed in our lab.
• 00:26 - 00:30: Let’s recap first on the ELISA protocol.
• 00:30 - 00:33: Most ELISA kits have sequential protocols,
• 00:33 - 00:37: where all components have to be incubated separately.
• 00:37 - 00:42: A standard ELISA generally takes up to five hours.
• 00:42 - 00:45: Abcam’s SimpleStep ELISA technology
• 00:45 - 00:48: allows detection in only 90 minutes.
• 00:48 - 00:52: Indeed, capture and detector antibodies
• 00:52 - 00:55: are incubated all together with your standard sample
• 00:55 - 00:57: for one hour only.
• 00:57 - 01:02: You wash, add the TMB substrate, and it’s done.
• 01:02 - 01:03: Mix, wash, read.
• 01:03 - 01:06: It’s that easy.
• 01:06 - 01:09: Apart from the substantial time reduction,
• 01:09 - 01:11: Abcam’s SimpleStep ELISA kits also
• 01:11 - 01:16: provide many advantages, such as extensive validation
• 01:16 - 01:20: with biological samples, a broad dynamic range,
• 01:20 - 01:21: and high specificity.
• 01:22 - 01:24: For the purpose of this podcast, we’ll
• 01:24 - 01:28: be using the Human PD-L1 SimpleStep ELISA kit.
• 01:28 - 01:31: We have combined our highly specific RabMAb PD-L1
• 01:31 - 01:36: antibody with a 90-minute SimpleStep ELISA protocol
• 01:36 - 01:39: to create a kit with nearly two-fold higher sensitivity
• 01:39 - 01:41: than our main competitors’ published results.
• 01:44 - 01:47: First step: choosing the standard
• 01:47 - 01:48: and preparing the plate.
• 01:52 - 01:55: The Human PD-L1 ELISA kit is designed
• 01:55 - 01:58: to be used with relevant biological samples, such as
• 01:58 - 02:03: plasma, serum, cell culture supernatant, urine, cell
• 02:03 - 02:06: pellets, and tissue homogenate.
• 02:06 - 02:09: Sample preparation should be performed differently
• 02:09 - 02:12: depending on the sample type.
• 02:12 - 02:16: Plasma, serum, and cell culture supernatant samples
• 02:16 - 02:19: should be diluted in the sample diluent and buffer
• 02:19 - 02:24: that is included in the kit.
• 02:24 - 02:26: The standard curve dilution will vary between 1,000 to 17.9
• 02:26 - 02:31: picogram/mL of PD-L1 recombinant protein.
• 02:31 - 02:36: Samples like urine, cell pellets, and tissue homogenate
• 02:36 - 02:42: should be diluted in the PTR cell extraction
• 02:42 - 02:46: buffer that is also included in the kit.
• 02:46 - 02:51: The standard will vary between 1,400 to 21.88
• 02:51 - 02:54: picogram/mL.
• 02:54 - 02:58: In this podcast, we are going to use Duocat cell pellets
• 02:58 - 03:02: as an example to measure PD-L1 concentrations.
• 03:02 - 03:06: Samples are diluted in the PTR cell extraction buffer.
• 03:06 - 03:12: The standard curve will have a 21.88 to 1,400 picogram
• 03:12 - 03:15: mL range.
• 03:15 - 03:19: Here is the plate layout we will use for this presentation.
• 03:19 - 03:23: The first two columns will be our standard in duplicates.
• 03:23 - 03:25: The use of at least two replicates
• 03:25 - 03:29: for every measurement, standards, samples, and blanks
• 03:29 - 03:32: is highly recommended due to statistical reasons,
• 03:32 - 03:35: as we will see later.
• 03:35 - 03:38: As we already stated, since we are
• 03:38 - 03:40: going to test cell pellet samples,
• 03:40 - 03:44: we need a standard curve ranging from 21.88
• 03:44 - 03:48: to 1,400 picogram/mL.
• 03:48 - 03:53: Columns 3 and 4 will be our samples, also in duplicates.
• 03:53 - 03:59: We use Duocat cells treated with LPS and interferon gamma.
• 03:59 - 04:03: In this example, we measure the total protein concentration
• 04:03 - 04:04: of our samples.
• 04:04 - 04:07: A two-fold dilution was prepared,
• 04:07 - 04:15: starting at 1,000 microgram/mL down to 15.625 microgram/mL.
• 04:15 - 04:19: Finally, we also need two replicates as a negative control.
• 04:19 - 04:25: In this case, the wells will only contain sample diluent.
• 04:25 - 04:31: Second step: verifying the validity of the measure.
• 04:31 - 04:33: Once the assay is completed and the OD
• 04:33 - 04:37: has been assessed at 450 nanometers,
• 04:37 - 04:39: we obtain the following raw data.
• 04:39 - 04:43: We use here an endpoint measurement.
• 04:43 - 04:45: At first sight, OD values are decreasing
• 04:45 - 04:49: from the top to the bottom wells in accordance
• 04:49 - 04:51: with our dilutions.
• 04:51 - 04:54: The first thing we need to do is to subtract background noise
• 04:54 - 04:57: from our sample and standard values.
• 04:57 - 05:00: To do that, we calculate the mean value
• 05:00 - 05:02: obtained by the blank control wells,
• 05:02 - 05:05: containing only the sample diluent.
• 05:05 - 05:11: From duplicates at 0.051 and 0.056,
• 05:11 - 05:17: we obtain a mean value of 0.053.
• 05:17 - 05:21: This value is then subtracted from each value in each well.
• 05:21 - 05:26: We obtain the corrected ODs for standards and our samples.
• 05:26 - 05:30: Next, we need to calculate the coefficient of variation
• 05:30 - 05:33: for every point of our experiment.
• 05:33 - 05:36: This value is calculated for each duplicate
• 05:36 - 05:38: by dividing the standard deviation
• 05:38 - 05:42: by the OD average value of the two replicates.
• 05:42 - 05:47: This value should always be below 20%.
• 05:47 - 05:49: A higher percentage would indicate
• 05:49 - 05:51: that the variation between the duplicates
• 05:51 - 05:54: is too high to be considered correct.
• 05:54 - 05:58: If this would be the case, we should leave those values out
• 05:58 - 06:02: of the calculation and repeat the assay.
• 06:02 - 06:07: In this slide, we can see the calculation for each duplicate.
• 06:07 - 06:10: First are the average OD values for each duplicate
• 06:10 - 06:12: of standards and samples.
• 06:12 - 06:16: Next are the values of standard deviations.
• 06:16 - 06:19: In the last columns are the coefficients
• 06:19 - 06:22: of variation for each duplicate.
• 06:22 - 06:26: For example, the highest point of our standard curves
• 06:26 - 06:30: has the mean OD value of 2.805.
• 06:30 - 06:37: Using Excel, we calculated a standard deviation of 0.022.
• 06:37 - 06:41: Hence, the coefficient of variation between our duplicates
• 06:41 - 06:44: is 1%.
• 06:44 - 06:46: In our experiments, every point of the standards
• 06:46 - 06:52: and the samples are valid for interpretations.
• 06:52 - 06:55: Third step: creating the standard curve.
• 07:00 - 07:05: Bioassays like ELISA normally follow a sigmoidal curve.
• 07:05 - 07:08: Therefore, it is advisable to use a four parameters
• 07:08 - 07:13: model, or 4PR, for interpreting the results.
• 07:13 - 07:17: 4PR models are available in the software of most readers
• 07:17 - 07:21: and in GraphPad Prism, for example.
• 07:21 - 07:23: Alternatively, a linear regression
• 07:23 - 07:25: is possible using Excel.
• 07:25 - 07:28: However, please bear in mind it is only
• 07:28 - 07:31: suitable in the linear part of the dilution.
• 07:31 - 07:34: In our example, if we use a linear regression
• 07:34 - 07:39: curve with Excel, we still obtain a very good R squared.
• 07:39 - 07:42: However, if we look at the log-log representation
• 07:42 - 07:46: of our data, we can observe that with the linear regressions,
• 07:46 - 07:48: values at the low end of the curve
• 07:48 - 07:50: will not be as reliable as values at the higher
• 07:50 - 07:52: end of the curve.
• 07:52 - 07:55: On the contrary, plotting the same data
• 07:55 - 07:59: with a log-log scale using a 4PR model, the fit
• 07:59 - 08:00: is more accurate.
• 08:03 - 08:09: Fourth step: calculating the concentration of your samples.
• 08:09 - 08:11: Once we have our standard curve, we
• 08:11 - 08:14: can use it to calculate the concentration of our samples
• 08:14 - 08:17: depending on their ODs.
• 08:17 - 08:21: Here, we are calculating how many picograms of PD-L1
• 08:21 - 08:25: we have per microgram of GEOCAD total proteins.
• 08:25 - 08:27: We will also compare the results
• 08:27 - 08:32: calculated with a 4PR model and a linear regression.
• 08:32 - 08:35: We will consider as reliable data
• 08:35 - 08:38: only the values that are in between the lowest
• 08:38 - 08:41: and the highest point of the standard curve.
• 08:41 - 08:44: The lowest value included in the standard curve
• 08:44 - 08:47: is 21.88 picogram/mL.
• 08:47 - 08:50: And as you can see, the two lowest values
• 08:50 - 08:53: are lower than this concentration.
• 08:53 - 08:58: Therefore, in this case, we will exclude those values.
• 08:58 - 09:01: Since we have used a two-fold dilution for GEOCAD cell
• 09:01 - 09:05: pellets, we can calculate the corrected values of PD-L1
• 09:05 - 09:10: in each well by multiplying by the respective dilution factor.
• 09:10 - 09:15: For example, in 500 microgram/mL of GEOCAD cells,
• 09:15 - 09:20: we obtain an OD of 0.526.
• 09:20 - 09:23: Using a 4PR regression of the standard curve,
• 09:23 - 09:29: we’ve calculated a value of 230 microgram/mL of PD-L1.
• 09:29 - 09:32: Multiplied by a dilution factor of 2,
• 09:32 - 09:38: the corrected value is 460 microgram/mL.
• 09:38 - 09:40: As you can see on the right-hand side,
• 09:40 - 09:45: these values are overestimated with a linear regression.
• 09:45 - 09:49: If we calculate the linearity between each dilution,
• 09:49 - 09:54: we can see that using a 4PR regression for calculation
• 09:54 - 09:59: provides a better interpretation of our results.
• 09:59 - 10:02: In the end, this is our result representing the PD-L1
• 10:02 - 10:06: concentration in your different dilutions of GEOCAD cells
• 10:06 - 10:08: pellets.
• 10:08 - 10:10: In our protocol booklet, you will
• 10:10 - 10:15: find the dynamic range of all tested samples.
• 10:15 - 10:17: I hope you enjoyed this presentation
• 10:17 - 10:21: about how to calculate and evaluate your ELISA data using
• 10:21 - 10:25: our PD-L1 SimpleStep ELISA kit.
• 10:25 - 10:28: Please feel free to contact our scientific support
• 10:28 - 10:30: if you have any questions.