1. A. Checks to run to make sure the instrument is in good working condition B. Decide which experiment type is most appropriate, equilibrium or velocity C. Optical system considerations D. Speed selection and length of an experiment E. Column height F. Concentration selection G. Temperature considerations H. Wavelength scans I. Instrument settings Session 2: Experimental Design and Data Collection

2. A. Checks to run to make sure the instrument is in good working condition: Absorbance optics: ● run an intensity wavelength scan with a windowless cell at 5.9 cm, at 6.5 and 7.1 cm. Overlay the scans and make sure that a) the peak intensity at 230 nm is at least 15,000 (or clean the lamp) and make sure that intensity plots from all three positions overlay (guarantees that the optical system is properly aligned). ● check radial alignment ● check the overlaid menisci of 40-50 scans of the same sample – they should form a single sharp peak, if the meniscus shifts back and forth, the slit assembly may need to be serviced. ● Take advantage of unique absorption peaks (i.e., heme, metal in the visible) to extend concentration range with absorbance optics. Match absorbance and emission peaks. Experimental Design and Data Collection

3. A. Checks to run to make sure the instrument is in good working condition: Interference optics: ● check to make sure the radial calibration is correct and matches the absorbance results from identical cells. ● Attempt to adjust the laser timing to produce a well-balanced contrast throughout both channels of a cell ● Once adjusted, the interference system does not require the counterbalance and the 4 th /8 th hole can be used for a sample position. Experimental Design and Data Collection

4. B. Decide which experiment type is most appropriate, equilibrium or velocity: Velocity: ● Sedimentation velocity experiments provide more information than equilibrium experiments and are suitable for 99% of all cases ● Net flow of solutes provides information on shape and kinetics ● Composition analysis in velocity is infinitely more sensitive than in equilibrium experiments ● Velocity experiments have longer columns which provide more datapoints for better statistics and more reliable data fits ● Best used for unknown sample ● Modulation in speed can be used to characterize a wide range of solution properties ● Global fitting is possible (multi-speed, multi-concentrations) Experimental Design and Data Collection

5. B. Decide which experiment type is most appropriate, equilibrium or velocity: Equilibrium: ● Use only with samples that are 95% pure or better based on SDS PAGE ● Measure Kd's for reversibly self-associating systems ● measure MW for discrete species or distributions (low resolution, and partial concentration unreliable for model-independent fit) ● Use only for MW's that are not too disparate (10% - 70% of each other) Experimental Design and Data Collection

6. C. Optical system considerations Absorbance optics: ● use 230 nm for optimal signal to noise ratio (largest emission peak), use lower wavelengths for small protein concentration. ● Protein extinction is usually 3-10 fold better when using 230 nm instead of 280 nm. Signal is often poor below 230 nm, do not measure at concentrations above 0.9-1.0 OD, for velocity experiments use 0.5-0.9 OD, regardless of wavelength ● In velocity experiments it is important not to change the wavelength mid-run. When measuring multiple concentrations at different wavelengths, perform multiple runs. This is not a concern when performing equilibrium experiments, wavelengths can be changed during the experiment, as long as globally fitted wavelength scans are made Experimental Design and Data Collection

7. C. Optical system considerations Intensity Measurements: ● Can be used to measure samples in the reference channel as long as the concentration is less than 0.3 OD to avoid resetting of photomultiplier gain setting ● Requires time invariant noise removal by fitting with the 2-dimensional spectrum analysis ● Reduces stochastic noise by a factor of square root of 2 ● Requires the use of one water channel for intensity reference Experimental Design and Data Collection

8. C. Optical system considerations Absorbance optics: ● Buffers may not absorb, or only very little (less than 0.1 OD). Always measure against a water blank, unless performing differential sedimentation experiments ● To reduce stress on centerpieces, try to closely match the fill volumes on both sides of the septum. When using 6-channel centerpieces, fill reference side completely and leave a small airbubble on the sample side to generate a meniscus. For 2-channel centerpieces, fill reference completely for velocity experiments, and also leave a small airbubble on the sample side to generate a meniscus. For equilibrium runs, fill 120-130 μl of sample and just a little more for the reference to generate a good meniscus. ● For velocity intensity measurements leave a small air bubble in both channels, but make column as large as possible. It is better to have a lower absorbance than a shorter column. Experimental Design and Data Collection

9. C. Optical system considerations Interference optics: ● Use for high protein concentrations (1 mg/ml and above) ● ALWAYS measure against water – if any buffer components sediment they can be fit with 2DSA as separate species. ● For all experiments fill both channels as full as possible, but leave enough room for a small air-to-air region, which is required to align the scans and correct integral fringe shifts during editing. ● Use for systems containing buffers that absorb too strongly Experimental Design and Data Collection

10. C. Optical system considerations Buffer considerations: ● Interference optics allows the use of absorbing buffers, since only concentration differences are measured, not absorbance. Use interference optics for experiments that involve nucleotides, reductants, absorbing buffers such as TRIS and other absorbing buffer components. If in doubt, scan the buffer against water in the desired wavelength range. ● Minimize or eliminate gradient forming materials such as glycerol, sucrose, etc. since they will introduce hard to model density and viscosity variations throughout the cell ● Absorbance experiments require use of non-absorbing buffers. TCEP may be used above 260 nm, other reductants can change extinction based on their oxidation state, causing shifts in the baseline absorbance during the run. Always run a buffer wavelength scan against water for an unknown buffer. Experimental Design and Data Collection

11. D. Speed selection and length of an experiment Velocity experiments: ● Attempt to obtain a minimum of 40 scans for each cell measured ● The faster the rotor speed, the better the s-value resolution. Whenever possible, measure at the maximum speed supported by the instrument to characterize composition ● For small and slow sedimenting samples, it is often possible to scan multiple samples (cells) and still obtain a sufficient number of scans on all cells ● For very large samples, diffusion will be small and will have less of an effect on the s-value resolution, and good resolution can be obtained even at lower speeds ● Use of interference optics is preferred, since it allows for faster scanning (seconds rather than minutes), and allows even multiple samples to be scanned at high speed before the sample is pelleted Experimental Design and Data Collection

12. D. Speed selection and length of an experiment Velocity experiments: ● ALWAYS collect scans from the first seconds of the early experiment all the way until most material is pelleted. Remember, you can always discard scans later, but repeating the experiment to obtain missed data is not desirable ● Use the finite element simulation routine to simulate all expected components in a system. You should model all components by shape, MW, s and D using the “Simulation:Model s, D and f from Molecular Weight for 4 basic shapes” and then use the “Simulation:Finite Element Simulation” module to predict how long to run the experiment and what speed should be selected. In order to guarantee that you will obtain enough scans, keep in mind that at the preferred setting for velocity experiments, you need to expect about 1.5 minutes for each absorbance scan of a properly filled cell (i.e, all the way full), and about 5 seconds for each interference scan. Experimental Design and Data Collection

13. Effect of Time on Resolution

14. Effect of Rotorspeed with constant ω 2 t on Resolution:

15. Resolution Comparison:

16. ● Use 3 different loading concentrations at the same wavelength ● Increase concentration range by measuring at different wavelengths such as 280 nm, 230 nm and ~210 nm, check absorbance spectrum! ● If interference optics are available, use them to extend concentration range. Always run several concentrations of your sample! Self -Associating Equilibrium Experimental Design:

17. In order to maximize the Resolution of your Analysis: ● Run at the fastest speed possible (simulate!) ● Always collect data until the end of the Run ● Collect early scans for best estimate of C 0 ● Recommended OD: 0.6 – 0.9 ● Fill cells as high as possible to get a long column ● Later scans provide better resolution than earlier scans DEMO – Finite Element Velocity Simulation Maximizing Resolution of the Sedimentation Coefficient:

18. D. Speed selection and length of an experiment Equilibrium Experiments: ● The speed in equilibrium experiments determines the steepness of the equilibrium gradient. By performing equilibrium experiments at multiple speeds, it is possible to generate multiple representations of the identical sample which can be fitted globally for better precision. ● Speed selection is dictated by the molecular weight of the sample, and the range of possible and reasonable speeds is given by the reduced molecular weight, sigma. The speeds should be chosen such that the sigma value is between 1 and 4 for the smallest component in the system (generally the monomer molecular weight, or the smallest associate in the system) Experimental Design and Data Collection

19. D. Speed selection and length of an experiment Equilibrium Experiments: ● Equally space 4-5 speeds between sigma limits of 1-4 ● Use the “Equilibrium:Suggest Best Speed” module to estimate the speed appropriate for an experiment DEMO Equilibrium Speed Selection Experimental Design and Data Collection

20. D. Speed selection and length of an experiment Equilibrium Experiments: ● The length of an equilibrium experiment depends on several factors: column height (discussed below), speed in the approach to equilibrium, and the diffusion properties of the molecule. To predict the approximate time for reaching equilibrium at multiple speeds use the “Simulation:Model s, D and f from Molecular Weight for 4 basic shapes” module to first estimate the molecular parameters for your sample, and then use the “Simulation:Estimate Equilibrium Times” module to estimate the time it takes to reach equilibrium at each speed. Always estimate the largest component in the system, since this component will have the smallest diffusion coefficient. You can also overestimate the axial ratio to be on the safe side. To be certain that equilibrium has been reached, take multiple scans spaced 4-6 hours apart and see if they remain unchanged. DEMO Equilibrium Time Simulation Experimental Design and Data Collection

21. E. Column height ● For velocity experiments, the full column length should always be used. Obviously, the longer the sample column, the more datapoints are available for fitting. Naturally, more data is always better than less data. Therefore, the maximum column height should always be used. When total sample is limiting, a 1.4 cm column experiment with a diluted sample at 0.45 OD is much more informative than a concentrated sample at 0.9 OD with a 0.7 cm column length. ● Column height has a large effect on the time it takes to reach equilibrium. The longer the column, the longer the experiment. Too short a column doesn't provide enough datapoints. A good compromise is to use a 3.0-3.1 mm column height, which translates into a loading volumne of approximately 120-130 μl. Equilibrium experiments should be performed at multiple concentrations to obtain additional equilibrium profiles and to span a better signal range for all associated multimers in the sample. Experimental Design and Data Collection

22. F. Concentration selection ● For velocity experiments, a loading concentration between 0.5-0.9 OD is recommended. ● For interference experiments, at least 1 mg/ml is desirable to reduce time- invariant noise from refractive index in-homogeneities in the cell windows. ● For equilibrium experiments, loading concentrations of 0.3, 0.5 and 0.7 OD at each wavelength are recommended. ● In order to maximize the signal from all species in a reversibly self- associating sample, it is recommended that the concentration range be made as large as possible to assure that sufficient signal from each species is present in the data. ● In order to provide maximum concentration spread in the analysis, it is recommended to combine data from 210, 230 and 280 nm, as well as interference in a global fit. Experimental Design and Data Collection

23. G. Temperature considerations Velocity experiments require a constant velocity. Therefore it is critically important to temperature-equilibrate the rotor before acceleration. This is best accomplished by letting the rotor with loaded sample cells sit in vacuum at the temperature at which the run is to be performed for at least 1 hour before the experiment is started. Other than thermal stability of a sample, there are no considerations for temperature. There is plenty of time for the rotor to temperature-equilibrate before the first equilibrium scan is taken. Experimental Design and Data Collection

24. H. Wavelength scans Whenever data at multiple wavelengths are measured, it is important to obtain extinction information for each wavelength, so appropriate corrections can be made when concentration dependent parameters like the association constant are determined. It is recommended to obtain wavelength scans at multiple concentrations spanning all wavelengths plus 40-50 nm on either side. The wavelength scans are then fitted globally to obtain an intrinsic extinction profile which can be normalized with the known extinction at 280 nm. Use the “Utilities:Global Extinction Fit” module to determine the extinction profile. The wavelength measurements should be performed as follows: measure 3 scans of each concentration with 1 nm resolution and zero averages. DEMO: Wavelength Fitter Experimental Design and Data Collection

25. I. Instrument settings For UV absorbance experiments it is important that optimal data acquisition settings are selected. Options include radial resolution, the number of repeat measurements, and for wavelength measurements the wavelength resolution. Settings for each experiment: ● Wavelength measurements: Wavelength measurements should be performed as follows: measure 3 scans of each concentration with 1 nm resolution and zero averages, continuous mode. Measurement should include 20 nm above and below the desired wavelengths. ● Velocity experiments: 0.003 cm resolution, no averaging, no scan delays, continuous mode ● Equilibrium experiments: 0.001 cm resolution, 20 averages, step mode. Optional: repeat scans 4-6 hours apart to determine if equilibrium has been reached. ● Interference data collection for velocity experiments should be performed in continuous mode without delay between scans. Superfluous scans can be discarded later. Experimental Design and Data Collection