Dear Dr. Petrone, were your spectra recorded in H2O or in D2O? I would suggest subtracting the IR absorption spectrum of your buffer so that you get a sufficiently flat baseline encompassing your Amide I band. For spectra in D2O, the baseline should stretch as far as 1250 cm-1 (near the absorption of D2O). After subtraction, you may draw a straight baseline at around 1800 cm-1. This should be sufficient, I would not make any other manipulation, because the Amide I band often overlaps with amino acid sidechain absorption especially on its low-frequency side.

I would start placing peaks according to the second derivative spectrum. I usually start inserting peaks at fixed positions, with initial 50:50 Lorentzian:Gaussian lineshapes, and let the software adjust the intensity, width, and Lorentzian-Gaussian character. After the fitting, you might want to compare the second derivative spectrum of your fit with that of your original absorption spectrum.

Hope it helps!

Hi Maurizio!

thanks for your reply.

I am going to use ATR-IR both in H2O and D2O as the protein native structure might change upon H/D replacement.

Then, I'll obtain the 2nd derivative spectrum and multiply it by -1.

Now the issues I still have are 2:

1) shall I perform ATR correction of the Amide I peak before fitting?

2) how to chose the right fitting procedure. Sometimes the width of the fixed peaks can vary greatly (from 8 up to 20 or 30 cm-1). This implies that the likelihood of having a certain component is much higher than others. So, I figured I might just fix the peak width, say to 10-12 cm-1.

Tricky!

Ciao Maurizio e grazie ancora

Luigi

Luigi,

Please take a look at the attached document. Hope it can be of help. Good luck!

There are numerous approaches to extract secondary structure information in terms of percentage fractions from the position and shape of the Amide I band.

- Curve fitting on original protein spectrum

- Curve fitting on Fourier Self Deconvolution (FSD) treated protein spectrum

- Factor analysis (Pezolet Biochemistry 1990, Chapman Biochemistry 1990) on original spectrum (using linear combinations of protein spectra with known conformations)

- 2nd derivative of protein spectrum/combination with curve fitting.

A commercial software and apparatus is available from BRUKER Optics GmbH, Ettlingen termed "Confocheck", where several pitfalls

-Correct water subtraction

-Using D2O instead of H2O

-Baseline

...

are addressed.

Also Search for author "Rahmelow", e.g. K. Rahmelow, W. Hübner, Anal. Biochem. 1996, 241(1), 5-13.) for additional informations.

Personally, I recommend a cross check of the FTIR results (a-helix, b-sheet, random, turn fractions) by circular dichroism.

Best regards,

Martin Müller