computational analysis and prediction techniques. Computational
methods can be used to predict RNA-binding sites on RBPs solved
in the apoform (unbound) state, to dock RNA molecules into
unbound RBPs, or to study the molecular dynamics of RNA–
protein complexes. Such modeling exercises can help to predict
the minimal RNA-binding protein fragment or domain to include
in crystallography construct design. Several methods are available
to predict RNA-binding sites on the structures of unbound RBPs
(see Note 4) including:
l
IDEEPE predicts RNA–protein-binding sites from RNA
sequences by combining global and local convolutional neural
networks (CNNs) and has been shown to identify experimentally verified binding motifs [11].
l
NPDOCK (Nucleic acid–Protein Dock) is a web server for
modeling
RNA–protein
and
DNA–protein
complex
structures [12].
Furthermore, fragment-based approaches are being developed
that are capable of docking highly flexible single-stranded RNA
bound to a protein based on the protein structure, RNA sequence,
and conserved contacts [13]. These predictions can be followed up
experimentally to confirm that RNA binds to the RBP or RBD (see
Subheading 3.8).
3.3 Preparing
RNase-Free Solutions
Even minute amounts of RNase contamination can ruin crystallization experiments with RNA by degrading RNA in complex with
protein leading to sample heterogeneity. Therefore, it is advisable
to use disposable RNase-free certified plasticware throughout protein production to minimize the likelihood of RNase contamination (see Note 1). Buffers, RNA, and protein preps can be
conveniently checked for RNase activity using the Ambion RNaseAlert
® lab test kit, which uses a fluorescent cleavable RNA oligonucleotide (excitation/emission maxima 490/520 nm) to detect
RNase contamination. The test involves pipetting 5 μL of 10Â
RNaseAlert
® lab test buffer into a tube containing the lyophilized
RNA oligonucleotide substrate [14]. Subsequently up to 45 μL of
the solution to be tested can be added to the tube and incubated for
30–60 min at 37
C. Solutions with RNase contamination will
cleave the substrate producing a fluorescent signal that is proportional to RNase activity. Most solutions can be tested apart from
solutions that interfere with the fluorophore’s excitation or emission, e.g., gel loading buffers, or inhibit RNases, e.g., high ionic
strength solutions or solutions outside the pH range 4–9. Furthermore, crystallographic protein samples can be incubated with RNA
oligonucleotides at room temperature for several days to ensure
that there is no RNase activity and to confirm that the integrity of
Studying RNA–Protein Complexes Using X-Ray Crystallography
427
methods can be used to predict RNA-binding sites on RBPs solved
in the apoform (unbound) state, to dock RNA molecules into
unbound RBPs, or to study the molecular dynamics of RNA–
protein complexes. Such modeling exercises can help to predict
the minimal RNA-binding protein fragment or domain to include
in crystallography construct design. Several methods are available
to predict RNA-binding sites on the structures of unbound RBPs
(see Note 4) including:
l
IDEEPE predicts RNA–protein-binding sites from RNA
sequences by combining global and local convolutional neural
networks (CNNs) and has been shown to identify experimentally verified binding motifs [11].
l
NPDOCK (Nucleic acid–Protein Dock) is a web server for
modeling
RNA–protein
and
DNA–protein
complex
structures [12].
Furthermore, fragment-based approaches are being developed
that are capable of docking highly flexible single-stranded RNA
bound to a protein based on the protein structure, RNA sequence,
and conserved contacts [13]. These predictions can be followed up
experimentally to confirm that RNA binds to the RBP or RBD (see
Subheading 3.8).
3.3 Preparing
RNase-Free Solutions
Even minute amounts of RNase contamination can ruin crystallization experiments with RNA by degrading RNA in complex with
protein leading to sample heterogeneity. Therefore, it is advisable
to use disposable RNase-free certified plasticware throughout protein production to minimize the likelihood of RNase contamination (see Note 1). Buffers, RNA, and protein preps can be
conveniently checked for RNase activity using the Ambion RNaseAlert
® lab test kit, which uses a fluorescent cleavable RNA oligonucleotide (excitation/emission maxima 490/520 nm) to detect
RNase contamination. The test involves pipetting 5 μL of 10Â
RNaseAlert
® lab test buffer into a tube containing the lyophilized
RNA oligonucleotide substrate [14]. Subsequently up to 45 μL of
the solution to be tested can be added to the tube and incubated for
30–60 min at 37
C. Solutions with RNase contamination will
cleave the substrate producing a fluorescent signal that is proportional to RNase activity. Most solutions can be tested apart from
solutions that interfere with the fluorophore’s excitation or emission, e.g., gel loading buffers, or inhibit RNases, e.g., high ionic
strength solutions or solutions outside the pH range 4–9. Furthermore, crystallographic protein samples can be incubated with RNA
oligonucleotides at room temperature for several days to ensure
that there is no RNase activity and to confirm that the integrity of
Studying RNA–Protein Complexes Using X-Ray Crystallography
427