Analytica Chimica Acta, accepted, 07/07/2015. This is the accepted version without proofing
corrections. DOI: 10.1016/j.aca.2015.06.011
.
Page 7 of 26
the HSA thermal denaturation experiment. MCR analysis was performed on spectra collected over the entire
temperature range, including the cooling step.
Recovered excitation profiles often appeared to have dual bands, which was an
artefact caused by the TSFS data structure. Basically, RTP and amino acid emission can lie
along the same Δλ line but with a different λ ex . This meant that MCR may not always
recover a pure excitation profile and thus satellite bands were obtained. Nonetheless, it was
clear which band was associated with each component. This effect can be eliminated using
EEM measurements where the real excitation profiles can be extracted. However, TSFS was
preferred over EEM for several reasons. First, with EEM measurements the Rayleigh scatter
is an intrinsic component of the data, whereas in TSFS it is not once Δλ > ~20 nm. Rayleigh
scatter is strongly polarized, thus any scattered light contamination will generate erroneous
anisotropy values [1] [41]. Second, TSFS data can be collected almost twice as fast as the
equivalent EEM data which is an issue considering that four TSFS spectra are required for
each aniso-TSFS plot. Overall, for proteins the advantage of minimizing Rayleigh scatter
artefacts outweighed the excitation profile recovery, however, for unknown analytes, EEM
may be more advantageous. Another caveat with MCR of MDF data must be noted, the
recovered excitation/emission profiles are invariant, which was a consequence of the linearity
requirement for MCR [37]. This meant that the band shifts, which fluorophores experience
during environmental/structural changes [1], were not explicitly extracted by MCR. This
information loss about fluorophore environment was however, offset by increased
quantitative information provided by the MCR scores. In summary, profiles extracted by
MCR represent the best-fit spectra for each population of emitters and the scores give the
relative change in contribution, which incorporates changes due to spectral shifts.
corrections. DOI: 10.1016/j.aca.2015.06.011
.
Page 7 of 26
the HSA thermal denaturation experiment. MCR analysis was performed on spectra collected over the entire
temperature range, including the cooling step.
Recovered excitation profiles often appeared to have dual bands, which was an
artefact caused by the TSFS data structure. Basically, RTP and amino acid emission can lie
along the same Δλ line but with a different λ ex . This meant that MCR may not always
recover a pure excitation profile and thus satellite bands were obtained. Nonetheless, it was
clear which band was associated with each component. This effect can be eliminated using
EEM measurements where the real excitation profiles can be extracted. However, TSFS was
preferred over EEM for several reasons. First, with EEM measurements the Rayleigh scatter
is an intrinsic component of the data, whereas in TSFS it is not once Δλ > ~20 nm. Rayleigh
scatter is strongly polarized, thus any scattered light contamination will generate erroneous
anisotropy values [1] [41]. Second, TSFS data can be collected almost twice as fast as the
equivalent EEM data which is an issue considering that four TSFS spectra are required for
each aniso-TSFS plot. Overall, for proteins the advantage of minimizing Rayleigh scatter
artefacts outweighed the excitation profile recovery, however, for unknown analytes, EEM
may be more advantageous. Another caveat with MCR of MDF data must be noted, the
recovered excitation/emission profiles are invariant, which was a consequence of the linearity
requirement for MCR [37]. This meant that the band shifts, which fluorophores experience
during environmental/structural changes [1], were not explicitly extracted by MCR. This
information loss about fluorophore environment was however, offset by increased
quantitative information provided by the MCR scores. In summary, profiles extracted by
MCR represent the best-fit spectra for each population of emitters and the scores give the
relative change in contribution, which incorporates changes due to spectral shifts.