Hi,

We routinely measure below 180 nm, but use a 0.1mm cuvette with our Aviv spectrophotometer. As you mention, decreasing the phosphate should help, but so will a shorter path length. But, I suspect the problem is with oxygen removal.

Going to 200 and under is going to require a very good flow of nitrogen gas. I hope you are using liquid N2 boil off, as gas tubes go quickly. You can check purging on your system with your buffer blank (or no sample) set at a wavelength of 178 and recording a kinetic run while you increase the flow in steps. I suspect you need to purge more. Do not pay attention to the flow meter, simply increase the flow and watch the photomultiplier voltage. If it continues to drop, you have your answer.

Filtering itself should not affect pH, but degassing might. This is easily checked. A properly degassed sample will also help you achieve lower wavelengths. I would centrifuge the protein and not filter it. Some proteins do not fair well when pressed through a membrane.

Measuring high quality spectra below 180 nm will require considerable averaging. As you are attempting a very wide spectrum, it will be most efficient to break the spectrum up into segments with different averaging times. You will not need excessive purging or long averaging until you are nearing 200. This will save you nitrogen and time. Expect to average at least 12-16 seconds per point below 200 depending on your protein concentration which is likely to be on the order of 0.5-1 mg/ml (depending on how much absorbance your buffer has). If less, you will need to accumulate more time at each data point.

The bandwidth can also be chosen to increase your signal to noise. You should be able to use 1.5 nm without distorting your spectrum. It is usually good to check if you are not too familiar with your instrument.

Good luck, post a spectrum with your results.

Other than a careful preparation of the buffer (as mentioned by Berndt), I suggest you to be careful about two additional factors: 1) The purity of the Nitrogen gas (I assume that your CD machine is attached with an in-house supply of Nitrogen gas) and 2) The life-time of the Xe lamp of the two machine. Often these two factors also play important roles in affecting the signal/noise ratio of the CD spectrum. Also try to collect your spectrum at a lower scan-speed with suitable number of accumulations.

Thank you Saumya Dasgupta. 

Yes our machine has inhouse Nitrogen gas supply and the purity is about 99.99%. The machine is new and previously only two to three data were recorded. So the Xe arc lamp will be good i hope. Anyhow i will take consideration of our suggestions.

Thank you Berndt. 

You had given a very useful and valuable suggestion. Since I am a new user I am taking the guidance of the manufacturer's user manual and various literatures on CD. However I would be happy if you could explain about the breaking up of spectrum into segments with different average times and 12-16 secs per point below 200 nm. How to accumulate the data after doing that?

Hi again:

You will want to record spectra in segments. Segments are chosen so as to optimize the spectrum for the features you have. For example: 600-300, 300-220, 220-178. Think of these segments as separate spectra which will be combined when the experiment is done. Each segment should be recorded with optimized parameters and preferably without removing the sample or entering the sample compartment. I would strongly recommend using the "Step Scan" mode Jasco offers where you measure for a defined time after moving to a new wavelength. Remember that S/N goes as the square root of the acquisition time. 

Segment   wavelength    resolution  time/point  total time

      I           600-300nm   1nm/point       2 sec       602 sec

     II           300-230nm   1nm/point       8 sec       568 sec

    III           230-178nm   1nm/point     32 sec     1696 sec

total of 2866 sec = 48 min

The data can then be exported in ASCII format and combined to give one continuous spectrum. Note there is a 1nm overlap. This (or more substantial overlap) can be helpful in aligning segments if the instrument drifts.

Now you can see that this approach will depend on your spectropolarimeter being stable over time. A drifting instrument will result in collected segments that will be offset with respect to one another, making a later combination challenging. Stability of CD signal over time matters when you need to make longer measurements (to increase S/N). If the signal drifts, different regions of the spectrum will have offsets with respect to one another. I am not sure Jasco has a stability At some point you should make a stability test of your instrument. Check the manual for a procedure to reproduce the instrument specifications. It should not drift more than 0.05-0.1 mdeg/hr. If it drifts more, you will be limited to relatively short scans.

The approach you take will be somewhat different with different manufacturers of spectropolarimeters. As I said, it has been a very long time since I have used a Jasco (we have been using Aviv). A stable, low drift signal and control over these parameters are two reasons I prefer the Aviv spectropolarimeters.

What are your intentions with the spectra? For estimation of secondary structure content, we routinely measure from 260-178nm (see Johnson WC Jr, Proteins. 1990;7(3):205-14 for an excellent summary).You must have a reason for starting at 600nm. Depending on the longer wavelength CD signal you may even need to record the spectrum using different concentrations and/or different path length cuvettes. Think of it as maintaining a somewhat constant signal to noise (S/N). I would try to solve the problem using only one sample concentration and one cuvette if possible.

If you have more questions, do not be afraid to ask. Perhaps a test spectrum of your protein would also help.

Thats very very informative Berndt. Let me make note of all your points and workout.... Let me obtain the spectrum. If I have any further clarifications will sure ask you. Thank you so much for your detailed explanation

The suggestions given by different scientist, I completely agree with them. Apart from that,  some times the detector may have problem or the light source would have been a problem (life time of light would have over).

@Rajagopal is correct. To make high-quality measurements below 200 takes a properly functioning spectrometer. Lamp energy, drift, calibration, etc. should be performed regularly.

Thank you Berndt and I will check with those things suggested by you.

in addition to all the helpful contributions above:

http://structbio.vanderbilt.edu/wetlab/cd.sample.prep.php

http://olisweb.com/literature/pdf/cd_practical_guide.pdf

https://www.uic.edu/orgs/ctrstbio/manuals/kellybba.pdf

Thank you Ulrich.

The cd practcal guide and kellybba manuals are which i am already having with me and the structbio.vanderbilt.edu link is new to be and thank you for sharing it to me Ulrich

I think there is little else you can do regarding buffer and sample preparation. You should reduce the path length. To go down to 180 nm you probably need a 0.1 mm cuvette.

Yes you are right Francis. But unfortunately we have only 0.2cm cuvette... So i am supppsed to go upto 180nm with this. Thats why i am trying to minimize the all other risk factors. Still if the problem persists then I have no other option will try to buy 0.1cm cuvette.

Buffer is quite ok. The lower salt concentration the better, but you need some of it.

I would recommend a cuvette with smaller pathlenght. Max 1 mm.

Decreseing the pathlength twice will decrease the HT voltage also about twice.

There are also cuvettes with pathlength 0.5 mm or 0.2 mm.

I usually use 0.2 mm cuvette and I was able to obtain smooth spectra with 3 M TMAO which gives you very high signal at low wavelengths.

You just need higher protein concentration, but also smaller volume

@Magdalena... What is TMAO? 3M wouldnt be a high salt buffer? What concentration of protein and volume you use for 2mm cuvette? Could you explain?

Not 2 mm cuvette but 0,2 mm.

TMAO is osmolith, the point is that is absorbs around 190 so it cause high HT (as do salts).

For 0,2 mm cuvette I use 0.5-1.2 mg/ml of protein, sample volume 60 ul

thanks for your information Magdalena

You can lower the concentration of protein and optimize for signal quality (lack of saturation) below 200 nm