3. Nucleic acids have high absorbance per mass unit in comparison to protein, so they will easily saturate the 280 nm absorbance signal even at submicromolar concentrations (see Table 3
for observed UV/RI ratio of proteins in comparison to nucleic
acids). A good starting point is to monitor absorbance simultaneously at 280 nm and 295 nm, but for higher concentrations, or longer nucleic acids, it may be necessary to move the
detection even further away from 280 nm to avoid saturation
(for FIR–DNA complexes, 310 nm was used).
4. The noise level of the RI signal on a Wyatt Optilab REX
instrument should be around Æ2 Â 10
À9 RIU. The RI signal
is extremely sensitive to temperature, so the Optilab REX is
designed to thermostat the internal measurement cell precisely
to Æ0.005
C, but experience has shown that the location of
the RI unit and the SEC column in the laboratory should also
be considered; for example, placing the unit or the SEC column under air conditioning outflow will generate fluctuations
in signals that follow the periodicity of the room temperature.
The column should be placed in a temperature control chamber (if available) or isolated by wrapping it in several layers of
bubble wrap or other forms of thermal insulation.
5. There are many SEC columns suitable for SEC/MALS systems. The Supredex and Superose HR and GL 10/300 series
columns are routinely used by our laboratory; the Superdex
Increase columns provide greater resolution. Other commonly
used columns are the TosoHaas TSK-GEL series SEC columns.
The silica-based TSK-GEL columns provide excellent resolution but are limited in chemical stability to the pH range
extending from 2 to 8. Although they provide lower resolution
than silica-based columns, the polymer-based TSK-GEL columns are a very good choice for samples with high
M w s. Laboratory-packed columns can also be used when a
specific media is required to minimize the interactions with
the SEC matrix. The critical factor in a column’s compatibility
with the SEC/MALS analysis is the quality of the LS baseline
after a new column has been conditioned. Every new column
sheds particles and the noise in the LS baseline is initially high,
but after 24–72 h of continuous pumping, the baseline noise
should decrease significantly. A major challenge for SEC column selection is analysis of large assemblies with M w above
1 Â 10
6 Da (1 MDa), for which field-flow fractionation
(reviewed in Ref. 19) would be a preferred method of fractionation; FFF/MALS have been used to study protein–nucleic
acid complexes [20].
6. For protein samples in aqueous buffers, the 0.1 or 0.22 μm
Durapore
®
, low-protein-binding membrane (Millipore, Bedford, MA) has proven to be an excellent choice; centrifugal
392
Ewa Folta-Stogniew
Précédent

- 389/484

Suivant