phosphate, pH 7.4, 50 mM NaCl, 0.5 mM TCEP, with
0.5 mg/ml BSA, 0.002% (v/v) Tween-20, and RNAse inhibitor (RNAsin, Promega) at 40–100 u/ml.
3. Oligonucleotides should be de-protected by following the
manufacturer’s instructions, lyophilized, and resolubilized in
the appropriate buffer. Final oligonucleotide concentrations
are then calculated from absorption spectroscopy by using the
Beer-Lambert law (A ¼ c ε l, where A is the absorption, c the
concentration in mol·L
À1 , ε the extinction coefficient in
L·mol
À1 ·cm
À1 , and l the pathlength in cm).
4. The lowest concentration of immobilized oligonucleotide that
gives enough signal in the protein association step should be
selected as overloading the biosensor may lead to overcrowding
and steric hindrance. It is generally the case that slow loading
for a long time is preferable to fast loading in a short time.
5. It can sometimes happen that not all sensors give the same
response in this step. In this case, the measured association or
dissociation amplitudes can be normalized for the different
loading levels. With some oligonucleotides, the response during the loading phase can be very small, making it difficult to
ensure equal loading of all the biosensors. In some cases, we
have found that using a different salt concentration in step (b)
increases the size of the response so that equal loading can be
confirmed. The sensors then need to be returned to the experimental buffer in step (c).
6. The baseline signal after loading in step (b) should be stable,
that is there should be no leaching of the bound RNA. This is
almost always the case with streptavidin (SA) biosensors but
may not be with other sensor types. If leaching does occur then
it is generally the case that reducing the concentration in the
loading step reduces the extent of the leaching.
7. In the ideal case, the length of the association phase should be
long enough to allow all response curves to approach close to
equilibrium but this is not always possible, particularly in studies of high affinity interactions which require the use of low
concentrations and therefore slow binding kinetics. When
repeating the measurement with a different set of protein concentrations in order to cover the appropriate concentration
range, the loading level reached in step (b) must be the same
in each measurement as the instrument response, but not the
kinetics of the response, is directly proportional to the loading
level. If this is not the case, the instrument response (signal
amplitude of the association or dissociation phase) can be
normalized for the loading level (see Note 5).
8. The manufacturers recommend that the duration of the dissociation phase should be long enough to give at least 5% dissociation. Although this may be reasonable in some cases, it is
BLI: Protein-RNA Interactions
365
0.5 mg/ml BSA, 0.002% (v/v) Tween-20, and RNAse inhibitor (RNAsin, Promega) at 40–100 u/ml.
3. Oligonucleotides should be de-protected by following the
manufacturer’s instructions, lyophilized, and resolubilized in
the appropriate buffer. Final oligonucleotide concentrations
are then calculated from absorption spectroscopy by using the
Beer-Lambert law (A ¼ c ε l, where A is the absorption, c the
concentration in mol·L
À1 , ε the extinction coefficient in
L·mol
À1 ·cm
À1 , and l the pathlength in cm).
4. The lowest concentration of immobilized oligonucleotide that
gives enough signal in the protein association step should be
selected as overloading the biosensor may lead to overcrowding
and steric hindrance. It is generally the case that slow loading
for a long time is preferable to fast loading in a short time.
5. It can sometimes happen that not all sensors give the same
response in this step. In this case, the measured association or
dissociation amplitudes can be normalized for the different
loading levels. With some oligonucleotides, the response during the loading phase can be very small, making it difficult to
ensure equal loading of all the biosensors. In some cases, we
have found that using a different salt concentration in step (b)
increases the size of the response so that equal loading can be
confirmed. The sensors then need to be returned to the experimental buffer in step (c).
6. The baseline signal after loading in step (b) should be stable,
that is there should be no leaching of the bound RNA. This is
almost always the case with streptavidin (SA) biosensors but
may not be with other sensor types. If leaching does occur then
it is generally the case that reducing the concentration in the
loading step reduces the extent of the leaching.
7. In the ideal case, the length of the association phase should be
long enough to allow all response curves to approach close to
equilibrium but this is not always possible, particularly in studies of high affinity interactions which require the use of low
concentrations and therefore slow binding kinetics. When
repeating the measurement with a different set of protein concentrations in order to cover the appropriate concentration
range, the loading level reached in step (b) must be the same
in each measurement as the instrument response, but not the
kinetics of the response, is directly proportional to the loading
level. If this is not the case, the instrument response (signal
amplitude of the association or dissociation phase) can be
normalized for the loading level (see Note 5).
8. The manufacturers recommend that the duration of the dissociation phase should be long enough to give at least 5% dissociation. Although this may be reasonable in some cases, it is
BLI: Protein-RNA Interactions
365