Precision Spectroscopy


Introduction



The hemes of the mitochondrial electron transport chain have very strong absorption spectra in the α-band (520-630nm) and dominate the absorption spectra of mammalian cells. However, they are present at very low concentrations and so the signal is very weak. In fact, the signal is weaker than when an exoplanet passes in front of its star.

CellSpex has developed a precision spectroscopy system that captures and utilizes every available photon. We use a 0.3 metre imaging spectrograph from Horiba that has excellent stray light rejection, spectral resolution and throughput. This is combined with our own cooled CCD imaging camera that has been optimized for this application.

We collect light from the cells with a 1mm optical fiber and F-number match it onto the slits of the spectrograph. The image on the CCD sensor fills the whole height of the sensor and the spectrum fills the whole width so we are using every pixel to maximize signal. We then employ full spectra fitting to use every available photon and precisely separate the heme signals.



Figure 1 The CellSpex 0.3 metre spectrograph with the CellSpex CCD camera on the right. Compare this with an Ocean Optics miniature spectrograph (foreground).

Hemes of the Electron Transport Chain



There are 7 hemes in the mammalian mitochondrial ETC: heme a and heme a3 in cytochrome oxidase, cytochrome c, heme c1 which is part of the high potential chain in the bc1 complex, hemes bL and bH which are part of the low potential chain in the bc1 complex and heme b560 in complex II.

The high absorption originates from the heme porphyrin ring and the spectrum changes when the iron is reduced (Fe2+) or oxidized (Fe3+). Typically, the reduced heme has a strong peak in the α-band, and the oxidized heme has a weaker broader spectrum. We use the reduced minus oxidized difference spectrum to fit the change in attenuation giving us concentration changes from baseline (Figure 1). Heme b560 in complex II has a very low midpoint potential and remains fully oxidized in the absence of an extrinsic strong reductant and so does not contribute to the change in attenuation spectrum under physiological conditions.


Figure 2 Heme reduced minus oxidized difference spectra of the mitochondrial transport chain.

Full Spectra Fitting



We have found that heme a of the mammalian cytochrome oxidase has two forms: the canonical form with a peak at 605nm and a second form with a peak at 602nm, and both forms are included in the fitting model. The model is a simple linear combination of model spectra (LCMS) using the heme spectra, e.g.



Where ΔA(λ) is the modelled change in attenuation spectrum, λ is the wavelength, ρ is the differential pathlength, ΔCc is the change in concentration of the cth heme and εc(λ) is the extinction spectrum of the cth heme.

The pathlength has been measured which allows concentrations to be calculated in units of nanomolar (nM).

Figure 3 shows the model fitted to data from RAW 264.7 cells. The residuals are the difference between the data and the model, and give an indication of the quality of the fit. The flat residuals of figure 3 show that all the components present in the data have been accounted for in the model and that the fit is of high quality.



Figure 3 Attenuation spectra under anoxic conditions from RAW 264.7 cells, the fit to the data and the residuals (difference between the data and the model).
Figure 4 shows the components of the fit. Each component is the spectrum of the heme scaled by the concentration change and the fit is the sum of these components.

The shoulder at 565nm originates from the b-hemes of the
bc1 complex. Although the b-hemes are not spectrally resolved, this is the only combination of signal that can account for the shoulder and hence the fitting is able to accurately separate the two signals.

Likewise for the two forms of heme a. The large width of the peaks and small separation prevents their spectral resolution, and the centre of the heme a peak in the data is observed to shift with differing contributions of these two components. However, the signals can be separated because there is only one combination of the two signals that can account for the data.

Figure 4 Individual heme components of the fit from Figure 3.

NAD(P)H



The Iberius measures NAD(P)H fluorescence simultaneously with the heme measurements using a second spectrograph/CCD system. We use 365nm excitation and full spectral fitting to achieve high signal to noise from the weak fluorescence.

the fluorescence signal originates from mitochondrial and cytosolic pools of NADH and NADPH and each component is spectrally identical. The mitochondrial NADH signal is separated from the other components by using anoxia and an uncoupler to fully reduce and fully oxidize, respectively, only the mitochondrial NADH using the assumption that the other components do not change.

This allows the oxidation state of mitochondrial NADH to be back calculated and the redox potential to be calculated.

Figure 5 NAD(P)H fluorescence from RAW 264.7 cells.
CellSpex Ltd © 2024