Q1 Can you profile GPCR's (membrane confined proteins) as well as purified proteins?
Q2 What changes in protein conformation can you see?
Q3 How sensitive is the assay?
Q4 Are absorbance changes in proteins or in the small molecules?
Q5 Is the data predictive or explanatory?
Q6 Can you see weak binding events?
Q7 Can you fingerprint combinations of proteins and lysates?
Q8 What is the concentration of the sample?
Q9 What are the necessary volumes in the channels and reservoirs?
Q10 How much sample is needed -- on cassette and in well-plate holder?
Q11 What secondary structures can you see in proteins?
Q12 How sensitive are the tests to changes in proteins caused by drugs?
Q13 What is the noise level in milliabsorbance units?
Q14 What is resolution of the measurement of wave numbers?
Q15 What range of wave numbers can be detected?
Q16 How long does it take to get the data?
Q17 Can you see effects of antigens on antibodies?
Q18 Is the cassette an open system or a closed system?
Q19 What do you do about the Joule heating from the isolectric focusing?
Q20 How many species can you see at one time?
Q21 What do you do about endoosmotic flow?
Q22 How do you avoid denaturing the protein on the spotting pins?
Q23 Can the cassettes be reused?
Q24 Can you profile peptides?
Q25 What is the risk of residues left from the cleaning of the pins?
Q26 Can I use my own ampholytes?
Q27 Can you see SNPs?
Q28 Can you measure on and off rates and at what time scale?
Q29 How long does the electrophoresis take?
Q30 Can you use polarized IR light?
Q1 Can you profile GPCR's (membrane confined proteins) and purified proteins?
Yes. The membrane absorbs at different frequencies so the protein absorbencies can be distinguished. Purified proteins are usually simple to profile. top
Q2 What changes in protein conformation can you see?
The spectra indicate relative changes in secondary structure by looking at the changes in areas under the peaks. For example, the beta rippled sheet, beta pleated sheet, alpha 1 helix, alpha 2 helix, 3-10 helix, beta turn I, beta turn II, and beta turn III all have absorbance peak areas that can be measured and compared to detect percentage changes in structure. The Solus100™ software will identify the substructures and their relative amounts. top
Q3 How sensitive is the assay?
The assay requires as little as 5 picomoles of protein per experiment with 24experiments possible on one chip. One experiment consumes about .25 micrograms of a 50,000 dalton molecule in 40 micromolar concentration. The percentage change in protein conformation is measured and a change as small as 4% has been detected. For example, if a protein has 30% alpha helix and a molecule changes the conformation to be 26% alpha helix, that protein alteration can be observed. The noise level is currently .25 milliabsorbance units. The platform can measure wave numbers from 950 to 340 wave lengths about 3 to 10 microns. Absorbance peaks separated by about 5 wave numbers can be distinguished from each other.
The absorbance A obeys Beers law:
A = el c
The quantity e is the molar absorptivity, l is the path length, and c is the molar concentration. top
Q4 Are absorbance changes in proteins or in the small molecules?
The response being measured is that of the system. However, the size of a protein is typically 60,000 daltons or more while the size of a drug is about 300 daltons. The absorbance by the proteins overwhelms that of the drug in most cases. With further analysis, one can measure the spectra of individual components of a mixture and by subtraction, isolate effects of interest. One can change the concentration of the small molecule to show there are no changes and it therefore is not a significant component in the spectrum of the mixture. In addition, proteins, lipids, and small molecules typically occupy different frequencies in the spectrum so in some experiments the changes can immediately be attributed to the sample source. In cases where it is important to isolate the changes of interacting proteins, one can use isotopic substitution to change the spectral peaks for secondary structure absorbance. For example, see Li T, Horan T, Osslund T, Stearns G, Arakawa T., "Conformational Changes in G-CSF/Receptor Complex as Investigated by Isotope-edited FTIR Spectroscopy," (1997) Biochemistry, 36, pp 8849-8857. top
Q5 Is the data predictive or explanatory?
The spectra from the Solus100™ system be used to guide future experimentation and to design research plans to be much more efficient because of the data. Taken literally, the data is explanatory and confirmatory, but in the sense that it guides further experiments it is predictive.
(1) Assume one has solved protein conformations. The spectrum can be used to confirm that a sample of protein has not denatured or been contaminated and is in the desired conformation. The Solus100™ system can verify that the protein is damaged and it will in that sense predict that experiments with the protein will be valid.
(2) When one wants to isolate one or a small number of small molecules that are useful drugs, the Solus100™ system can be used. For example, suppose a large number of drugs have affinity binding with a set of proteins. The changes in the spectra of the proteins caused by each of the drugs can be measured and clustered in groups that have equivalent effect on the protein conformation. Representative drugs from each cluster can be used in further experimentation. Those representatives that pass further tests then make the cluster data from the Solus100™ system predictive in the sense that drugs from the same cluster as the representative candidates are more likely to have the same properties.
(3) Suppose one mixes one or more proteins with one or more small molecules. By measuring the differences induced in the spectra, differences in conformation can be detected and the data is predictive in the sense it has identified the interesting candidates for further study such as crystallographic or in vivo.
4) The spectra for crystallized and in vitro proteins with the same conformations should be identical. This comparison is a good way to verify one has crystallized a protein with the desired conformation. For example, see "A comparison of infrared spectra of proteins in solution and crystalline forms", Biochim Biophys Acta. 1995 Apr 27;1248(2):115-22 top
Q6 Can you see weak binding events?
The assay measures conformational changes. If the weak binding causes sufficient conformational changes, it will be seen. The assay may be applicable to situations where standard binding assays are not effective or of sufficient sensitivity. top
Q7 Can you fingerprint combinations of proteins and lysates?
The spectral fingerprint of combinations of proteins can be taken. In theory, the Solus100™ system can fingerprint a lysate. (We have not conducted experiments in this area as of Dec 2005.) top
Q8 What sample concentration is required?
A concentration of 10 micromolar should be sufficient. A higher concentration, say 40 - 80 micromolar will give better signal to noise for measuring differences in some situations. The limit on higher concentrations is the risk of protein precipitation. top
Q9 How much sample is needed -- on the chip and in well-plate
holder?
The short answer is about 25 micrograms of sample in 384 well plate format are needed and about .25 micrograms of that will be consumed. In 1536 format, about 5 micrograms of sample are needed and .25 micrograms of that will be consumed. top
Q10 What secondary structures can you see in proteins?
The area under characteristic peaks in the spectra indicate the relative percentage or secondary structure elements of proteins -- beta rippled sheet, beta pleated sheet, alpha 1 helix, alpha 2 helix, 3-10 helix, beta turn I, beta turn II, and beta turn III. There is a substantial body of knowledge on absorbance at different frequencies. That knowledge will be incorporated into the Solus100™ software and the relative percentages of the secondary structures will be computed by the software. top
Q11 How sensitive are the tests to changes in proteins caused
by drugs?
The difference in the spectra between the protein before and after the drug is introduced measures the change in conformation. This difference is caused by a change in the bonds of the protein and in particular, the equipment is sensitive to the percentage change in bonds. This means for a small molecule, for example a peptide, a smaller change in numbers of bonds will be observable because it is a larger percentage of total bonds. The percentage changes in protein conformation is measured and a change as small as 4% has been observed. For example, if a protein has 30% alpha helix and a molecule changes the conformation to be 26%, that change should be observed. top
Q12 What is the noise level in milliabsorbance units?
The noise level is currently about .25 milliabsorbance units. If the signal to noise is a problem, then one can use a more concentrated sample or average over a greater number of spectra. The number of spectra that are averaged is set at 16 but can be changed by software. top
Q13 What is resolution of the measurement of wave numbers?
The resolution is about 5 wave numbers. That is if there are two absorbance peaks separated by 5 wave numbers, they can be distinguished by the platform. top
Q14 What range of wave numbers can be detected?
The focal plane array detector (cooled MCT - mercury cadmium telluride) and current optics have an effective range of wave numbers from 3400 (about 3 microns) to 950 (about 10 microns). top
Q15 How long does it take to get the data?
About 5 minutes for a 96-well plate. top
Q16 Can you see effects of antigens on antibodies?
Yes – though we haven’t done experiments in this area. top
Q17 Is the cassette an open system or a closed system?
Our chip is technically an "open system" (no top). However after the automated fluidics deposit the sample, the chip is moved under the optics which seals the system. top
Q18 How do you avoid denaturing the protein on the spotting pins?
The steel pins have covalent coatings to avoid denaturing the proteins. top
Q19 Can the cassettes be reused?
No. Proteins are viscous and difficult to remove from the channels. Any cleaning agent is likely to leave a film whose reflections will cause too much noise in subsequent spectra. top
Q20 Can you profile peptides?
Yes. Peptides are smaller molecules than proteins. As such, the change in small number of bonds will be a bigger percentage change to the spectrum of a peptide so small changes can be observed. top
Q21 What is the risk of residues left from the cleaning of the pins?
The pins are cleaned through a (programmable) cleaning cycle and our test routines include repeated spotting after cleaning to verify no detectable contamination. top
