Albeit very powerful and usually straightforward, HDX-MS has
also some disadvantages, e.g., rather limited spatial resolution and
need for sample dilution as the experiment often starts with
ten-fold dilution into a deuterated buffer. However, due to the
coupling to sensitive mass spectrometric detection, micro- and
submicromolar concentrations are easily managed. Although
HDX in combination with mass spectrometry was introduced
more than two decades ago and there are hundreds of research
articles describing its uses to study protein folding/unfolding,
protein–protein and protein–ligand interactions [4, 5], the number
of reports utilizing HDX for protein–nucleic acid interaction is
rather limited. This is most likely due to the nature of nucleic
acids that complicate HDX-MS analyses. During the quenching
step (lowering the pH) in an HDX-MS workflow, the DNA backbone becomes protonated, which often results in its poor solubility
and precipitation [6]. The protein interacting with DNA then tends
to precipitate as well. The presence of DNA can also have adverse
effects on chromatographic separation. Such behavior, however,
depends on the DNA composition and length. Therefore, reports
dealing with larger DNA stretches always utilize technical tricks to
deal with the troublesome DNA while those using very short
oligonucleotides may sometimes be spared these adverse effects.
The first ever attempt to use HDX for protein–nucleic acid
interaction study enabled the determination of a protein–DNA
complex dissociation constant by a combination of HDX and
matrix-assisted laser desorption/ionization mass spectrometry
[7]. However, it took five more years until Sperry et al. in 2008
truly successfully probed structural features of protein–DNA interactions of two biological systems (a protein binding to a telomeric
oligonucleotide and the protease thrombin interacting with an
aptamer). These studies were successfully accomplished thanks to
the incorporation of strong anion exchange trap column prior to
the liquid chromatography-mass spectrometric (LC-MS) analysis
to selectively trap and remove the oligonucleotides [8, 9]. Besides
an ion exchanger, the use of protamine sulfate has also been
reported as beneficial in the HDX-MS analysis of oligonucleotidecontaining samples. Poliakov and coworkers concluded that the
protein–DNA co-precipitation upon the lowering of pH is primarily of electrostatic origin. Thus, inclusion of small, highly basic, and
hence, positively charged protamine improved protein recovery
[10]. Alternatively, in another study, Roberts et al. counteracted
the negative effect of DNA on the formation of peptides, solubility,
and chromatographic separation in HDX by tuning the concentration of a positively charged denaturing agent (guanidine hydrochloride) in their quench buffer [11]. Use of another denaturant
has recently been exploited by Graham et al. to explain structural
and mechanistic aspects of DNA unwinding when guanidine
194
Ruzena Filandrova et al.
also some disadvantages, e.g., rather limited spatial resolution and
need for sample dilution as the experiment often starts with
ten-fold dilution into a deuterated buffer. However, due to the
coupling to sensitive mass spectrometric detection, micro- and
submicromolar concentrations are easily managed. Although
HDX in combination with mass spectrometry was introduced
more than two decades ago and there are hundreds of research
articles describing its uses to study protein folding/unfolding,
protein–protein and protein–ligand interactions [4, 5], the number
of reports utilizing HDX for protein–nucleic acid interaction is
rather limited. This is most likely due to the nature of nucleic
acids that complicate HDX-MS analyses. During the quenching
step (lowering the pH) in an HDX-MS workflow, the DNA backbone becomes protonated, which often results in its poor solubility
and precipitation [6]. The protein interacting with DNA then tends
to precipitate as well. The presence of DNA can also have adverse
effects on chromatographic separation. Such behavior, however,
depends on the DNA composition and length. Therefore, reports
dealing with larger DNA stretches always utilize technical tricks to
deal with the troublesome DNA while those using very short
oligonucleotides may sometimes be spared these adverse effects.
The first ever attempt to use HDX for protein–nucleic acid
interaction study enabled the determination of a protein–DNA
complex dissociation constant by a combination of HDX and
matrix-assisted laser desorption/ionization mass spectrometry
[7]. However, it took five more years until Sperry et al. in 2008
truly successfully probed structural features of protein–DNA interactions of two biological systems (a protein binding to a telomeric
oligonucleotide and the protease thrombin interacting with an
aptamer). These studies were successfully accomplished thanks to
the incorporation of strong anion exchange trap column prior to
the liquid chromatography-mass spectrometric (LC-MS) analysis
to selectively trap and remove the oligonucleotides [8, 9]. Besides
an ion exchanger, the use of protamine sulfate has also been
reported as beneficial in the HDX-MS analysis of oligonucleotidecontaining samples. Poliakov and coworkers concluded that the
protein–DNA co-precipitation upon the lowering of pH is primarily of electrostatic origin. Thus, inclusion of small, highly basic, and
hence, positively charged protamine improved protein recovery
[10]. Alternatively, in another study, Roberts et al. counteracted
the negative effect of DNA on the formation of peptides, solubility,
and chromatographic separation in HDX by tuning the concentration of a positively charged denaturing agent (guanidine hydrochloride) in their quench buffer [11]. Use of another denaturant
has recently been exploited by Graham et al. to explain structural
and mechanistic aspects of DNA unwinding when guanidine
194
Ruzena Filandrova et al.