"Thermodynamics of RNA melting, one base pair at a time".
Evgenia N. Nikolova and Hashim M. Al-Hashimi
Chemical Biology Doctoral Program, Department of Chemistry and Biophysics, University of Michigan, Ann Arbor, Michigan 48109, USA
Reprint requests to: Hashim M. Al-Hashimi, Chemical Biology Doctoral
Program, Department of Chemistry and Biophysics, 930 N. University Avenue,
University of Michigan, Ann Arbor, MI 48109, USA;
e-mail: hashimi@umich.edu;
fax: (734) 647-4865.
The melting of base pairs is a ubiquitous feature of RNA structural
transitions, which are widely used to sense and respond to cellular stimuli.
A recent study employing solution nuclear magnetic resonance (NMR)
imino proton exchange spectroscopy provides a rare base-pair-specific view
of duplex melting in the Salmonella FourU RNA thermosensor, which regulates
gene expression in response to changes in temperature at the translational
level by undergoing a melting transition. The authors observe “microscopic”
enthalpy–entropy
compensation—often seen “macroscopically” across a series of related molecular
species—across base pairs within the same RNA. This yields variations in
base-pair stabilities that are an order of magnitude smaller than corresponding
variations in enthalpy and entropy. A surprising yet convincing link is
established between the slopes of enthalpy–entropy correlations and RNA
melting points determined by circular dichroism (CD), which argues
that unfolding occurs when base-pair stabilities are equalized. A single
AG-to-CG mutation, which enhances the macroscopic hairpin thermostability
and folding cooperativity and renders the RNA thermometer inactive in
vivo, spreads its effect microscopically throughout all base
pairs in the RNA, including ones far removed from the site of mutation.
The authors suggest that an extended network of hydration underlies this
long-range communication. This study suggests that the deconstruction of
macroscopic RNA unfolding in terms of microscopic unfolding events will
require careful consideration of water interactions.
The melting and forming of RNA duplex structure are reoccurring structural
transitions that are used in a variety of functional contexts. For example,
transient short hairpins can direct RNA cotranscriptional folding by masking
elements prone to kinetic traps and releasing them via melting transitions
once downstream elements involved in long-range native contacts are transcribed
(Nagel and Pleij 2002). Duplex melting, in general
catalyzed by protein chaperones (Herschlag 1995)
or induced by competing trans-acting small RNAs (Kozak
2005), is often a requirement for rescuing RNA from kinetic traps (Treiber
and Williamson 2001) and allowing transformations to occur between
functionally different RNA structures (Al-Hashimi
and Walter 2008). The melting of duplex mRNA prior to ribosome entry
is believed to provide a general mechanism for periodically stalling translation
and promoting proper cotranslational protein folding and other processes
(Kozak 2005; Watts et al. 2009).
While melting transitions have been studied for decades using a variety
of biophysical techniques, most studies have relied on bulk spectroscopic
measurements such as UV absorbance, circular dichroism (CD), and
calorimetry (Jaeger et al. 1993), which typically
provide a macroscopic view of melting that fails to resolve potential differences
in the melting behavior of individual base pairs. By measuring the thermodynamic
stability of individual base pairs in an RNA thermosensor and an inactive
mutant, Schwalbe and co-workers (Rinnenthal et al.
2010) provide a rare site-specific view of duplex melting and revive the
power of a long-standing nuclear magnetic resonance (NMR) technique
to study RNA unfolding at atomic resolution.
The target of their studies is the temperature labile Salmonella
FourU RNA thermometer, which regulates gene expression in response to changes
in temperature by sequestering the Shine–Dalgarno ribosomal-binding sequence
within a hairpin at low temperatures and liberating it via a melting transition
at high temperatures. The AU-rich hairpin structure also contains several
noncanonical base pairs that are believed to confer its reduced thermostability.
Also examined is a single-point mutant (A8C) that exhibits a repressed
phenotype in vivo, in which an AG mismatch within the stem–loop structure
is replaced with a CG Watson–Crick (WC) base pair.
The NMR experiments take advantage of the fact that relatively acidic
imino protons of base-paired guanines and uridines/thymines (pKa
~ 9.2–9.6) can exchange with basic solvents when they melt to adopt an
“open” conformation but not when they are in a “closed,”
intrahelical conformation. By measuring the exchange rate (kex)
of single imino protons with solvent, it is possible to probe the opening/closing
of individual base pairs, and deduce lifetimes of the closed and open states
and the kinetics of their interconversion. This NMR technique is not new
in its basic principle. Crothers (Crothers et al.
1974), Patel (Patel and Hilbers 1975), and their respective
coworkers used one-dimensional (1D) NMR imino proton line broadening
back in the early 1970's to estimate thermodynamic parameters associated
with helical melting and end-fraying in tRNA and short DNA duplexes. Subsequent
developments in NMR pulse sequences and their application under variable
base catalyst concentration and temperature allowed for a direct quantitative
measurement of proton exchange rates and the elucidation of a generalized
two-step exchange mechanism in nucleic acids by Leroy and Guéron
(Leroy et al. 1985, 1988; Gueron
et al. 1987) involving a single base-pair opening mode. In this mechanism,
transient and independent base-pair opening is followed by direct, or water-mediated,
imino proton transfer to a proton acceptor (Fig. 1A).
The acceptor can be either an external base catalyst (e.g., NH3 or phosphate)
or an intrinsic catalyst—nitrogen from the opposite nucleobase (N1 of A
or N3 of C) in the oligonucleotide—that has unfavorably low pKa
(3.7 or 4.2) but high local concentration (Gueron and
Leroy 1995). This framework allowed Leroy and Guéron (Leroy
et al. 1988) to accurately quantify site-specific base-pair lifetimes
and activation enthalpies in DNA and subsequently in RNA by Varani and
coworkers (Leroy et al. 1988; Varani
et al. 1989). Current NMR methodology allows measurement of intrinsic
exchange times ranging between ~1 msec and 2 sec using relaxation techniques,
and longer than ~ 2 min using real-time H-D exchange kinetics, while the
window in between is currently inaccessible to NMR (Gueron
and Leroy 1995).
FIG. 1, RNA imino proton exchange by NMR
FIG. 1. RNA imino proton exchange by NMR.
(A) Schematic of a two-step direct (or water-mediated) imino proton exchange reaction for a GC base pair with a base catalyst, which is followed by reversible base-pair formation (data not shown). Secondary structure and base-pair thermodynamic parameters for the wild-type (wt) Salmonella FourU RNA thermometer (B) and the A8C mutant RNA (C) showing color-coded free energy (DGdiss) values relative to the open state (left) and a plot of free energy (DGdiss), enthalpy (DHdiss), and entropy (T DSdiss) at 20°C versus nucleobase examined (right).
Temperature dependence of DGdiss values derived from DHdiss and DSdiss using the Gibbs-Helmholtz equation for the wt RNA (D) and A8C mutant RNA (E). Corresponding Tm values obtained from CD are indicated (inset) to show coincidence with the intersection point of linear curves ([B,C,D,E]; adapted from the original article [Rinnenthal et al. 2010] with permission from the authors and Oxford University Press © 2010).
Since DHdiss and DSdiss
are assumed to be independent of temperature, the observed behavior implies
that variations in DHdiss across
base pairs are linearly compensated by changes in DSdiss,
such that DGdiss converges to a common
value at a compensatory temperature Tc, which is equal
to Tm. While an excellent linear DHdiss(DSdiss)
correlation is observed for both the wild-type (wt) and mutant sequences,
such correlations should in general be treated with caution, particularly
when a narrow range of DGdiss is
observed relative to the spread of DHdiss,
as is the case here, from which DGdiss
= DHdiss - T DSdiss
~const. naturally implies a linear relationship and/or when |DGdiss
< |DHdiss| with DSdiss
calculated by subtraction of the independently measured free energy and
enthalpy (Lumry and Rajender 1970; Sharp
2001). There, large correlated errors in DHdiss
and DSdiss could produce an artificial
linear relationship between the two with high correlation coefficients.
However, for this study, the correlations appear to be real because the
data satisfy a statistical test for error compensation effects (Sharp
2001); the slope of the correlation for the wt and mutant RNAs are
different and in both cases coincident with Tm, and the
DGdiss(T)
van't Hoff plots for multiple residues seemingly intersect at a common
point. As would be expected for an intrinsic correlation, the slope represents
a temperature at which variations in entropy and enthalpy are balanced,
and this appears to coincide with Tm for cooperative
duplex melting. Thus, the melting thermodynamics of various base pairs
in the RNA seem to be enslaved along a predefined DHdiss(DSdiss)
line.
What might be the common thread that links the thermodynamics of
different base pairs? Enthalpy–entropy compensation has been observed
in a variety of biological systems and is often linked to interactions
with water. For example, early studies on small molecules and proteins
in aqueous solution have shown that the slopes of DHdiss(DSdiss)
cluster consistently around a common temperature of ~280 K for widely different
processes, and thus could reflect a common solvent property (Lumry
and Rajender 1970). Solvent interactions have also been invoked
to explain enthalpy–entropy compensation during protein unfolding (Liu
et al. 2000) and base-opening in imino proton exchange studies of a
DNA dodecamer (Chen and Russu 2004). Typically, the
effect is observed macroscopically across related systems, for example,
in comparing binding energies of related ligands with target receptors
(Breslauer et al. 1987; Lis and
Sharon 1998) or the unfolding energies of different proteins (Liu
et al. 2000), and not microscopically within a given
system as observed here for the melting of individual base pairs
within the same RNA. Indeed, the observation of enthalpy–entropy
compensation within different base pairs supports the Tinoco–Uhlenbeck
approximation of decomposing nucleic acid helices into independent base-pair
units with nearest-neighbor effects (Tinoco et al.
1971). This long-standing link between solvent and enthalpy–entropy compensation,
together with the large DHdiss and
DSdiss
values derived, which rule out an intrinsic RNA effect, lead Rinnenthal
et al. (2010) to invoke interactions with water as the driving
force for the observed enthalpy–entropy compensation. Rinnenthal
et al. (2010) suggested that base-pair opening likely results
in the release of highly ordered water molecules from the major groove,
giving
rise to a loss of solvent enthalpy that is compensated for by a gain in
solvent entropy, and that these opposing contributions are larger
for GC versus AU base pairs. They speculated that the AG mismatch is
responsible
for destabilizing its neighbors and, in turn, the overall helix
through solvent-mediated communication with remote sites. They concluded
that melting of the water shell should
precede RNA melting,
at which point favorable RNA–solvent interactions on duplex stability
should be neutralized.
There is ample precedence for such a role of solvent in RNA structural
stability. Molecular dynamics (MD) simulations by Auffinger
and Westhof (2002) reveal a “pre-melting” phase due to water
shell disruption before global RNA duplex melting. MD simulations
also suggest that DNA base-opening induces water shell perturbations
(Giudice et al. 2003). Biophysical studies using
neutron scattering provide evidence that hydrated RNA undergoes a dynamic
glass transition that is driven by solvent interactions, much
like
hydrated proteins (Caliskan et al. 2006). Analysis
of RNA crystal structures and MD simulations reveals highly specific
AG and GU hydration pockets with long-lived waters and cooperative
contacts with neighboring water/RNA groups (Auffinger
and Westhof 1997; Reblova et al. 2003).
Thus, it is possible that local differences in conformation and
respective hydration patterns between wt and mutant RNA that are
somehow transduced down the helix to remote sites affect the relative DHdiss
and DSdiss values for individual base pairs as well as increase
the overall thermostability of the mutant hairpin. Diminished stacking
interactions in the mismatched duplex are used to explain the dramatically
lower cooperativity of the wt RNA melting transition. Single mismatched
GA base pairs are known to destabilize nucleic acid duplexes (Aboul-ela
et al. 1985) and reduce coaxial stacking relative to CG base pairs
when placed at helical interfaces (Schroeder et al.
1996). Diminished stacking can arise from structural perturbations and
conformational exchange between different base-pairing geometries reported
in structural studies of GA mismatches in DNA and RNA (see Allawi
and SantaLucia 1998). In fact, sensitivity to next-nearest neighbor
context has been observed for RNA GA mismatch structures, arguing for potential
long-range effects dictated by adjacent elements (Morse
and Draper 1995). The lower cooperativity of the wt is consistent
with single-molecule force-pulling experiments on DNA hairpins,
which show that inclusion of various mismatches (TT, GA, GT, etc.) reduces
the cooperativity of the unfolding transition due to appreciably populated
intermediate, partially unfolded, states (Woodside
et al. 2006). Thus, the thermodynamic stability of the RNA structure
seems to be intimately linked to solvent interactions, which are
very difficult to visualize using techniques such as X-ray crystallography
and NMR spectroscopy.
As indicated by previous research, for some nucleobases, the dependence
of kex(T) with temperature deviates from the behavior
expected based on the assumed two-state model and transition state theory.
This could reflect temperature-induced changes in the RNA structure (and
therefore that DH and DS
are not independent of temperature as assumed) and/or deviations in the
assumption that base-pair opening is uncorrelated at higher temperatures
near Tm. Such effects can be interrogated in future studies
by acquiring additional data to allow the fitting of more complicated statistical
models. In addition, the precise nature of the “open” state probed
by NMR exchange measurements, and whether it is the same state probed by
bulk optical melting measurements, is yet to be established and indeed
there are studies that suggest they could be different (Benight
et al. 1988).
Through their study, Rinnenthal et al.
(2010) have resurrected and modernized a long-standing NMR technique in
obtaining site-specific information regarding the thermodynamic stability
of RNA structure. Imino proton exchange is limited to specific base-pair
opening transitions that occur with time constants in the range of approximately
1 msec to 2 sec and, therefore, provides primarily a cross-sectional view
of RNA stability. However, other advances in solution (Furtig
et al. 2007; Zhang et al. 2007; Johnson
and Hoogstraten 2008; Hansen et al. 2009; Bailor
et al. 2010) and solid state (Olsen et al.
2010) NMR and other biophysical techniques (Al-Hashimi
and Walter 2008) are making it possible to quantitatively describe
microscopic structural transitions along the entire RNA structure taking
place over timescales spanning picoseconds to several minutes. While it
appears that the dissection of macroscopic RNA stability in terms of constituent
microscopic contributions is within reach, the present study calls for
new experimental tools for characterizing dynamic interactions between
water and RNA. Perhaps there is yet another well-known NMR technique
that we can draw on to achieve this goal.
NetworkEditors' Perspectives: "NMR reveals solvent
effects on RNA helical structures".
This fascinating study by Evgenia Nikolova and Hashim Al-Hashimi
details the use of NMR techniques in studying the important solvent
effects on helical regions of RNA molecules within intact cells. Helical
structures are "closed" to activity, while extended RNA molecules are "open"
to activity (Draper DE, "RNA Folding: Thermodynamic and Molecular Descriptions
of the Roles of Ions", Biophysical
J. 95: (11), 5489-5496 (December 15, 2008). Helices are stabilized
as hairpin RNA molecules, and are initiated by purine and pyrimidine
bases pairing within the helix interior in a largely hydrophobic
environment.
"Models of
successive levels of resolution during individual gene transcription".
Additional References:
Reprint requests to: Hashim M. Al-Hashimi, Chemical Biology Doctoral Program, Department of Chemistry and Biophysics, 930 N. University Avenue, University of Michigan, Ann Arbor, MI 48109, USA; e-mail: hashimi@umich.edu; fax: (734) 647-4865.
1. Taft RJ, Simons C, Nahkuri S, Oey H, Korbie
DJ, Mercer TR, Holst J, Ritchie W, Wong JJ-L,
Rasko JEJ, Rokhsar DS, Degnan BM and Mattick JS,
"Nuclear-localized
tiny RNAs are associated with transcription initiation and splice sites
in metazoans".
2. Mestdagh P, Fredlund E, Pattyn F, Rihani A, Van Maerken
T, Vermeulen J, Kumps C, Menten B, De Preter K, Schramm A, Schulte J, Noguera
R, Schleiermacher G, Janoueix-Lerosey I, Laureys G, Powel R, Nittner D,
Marine J-C , Ringnér M, Speleman F, and Vandesompele
J,
"An integrative
genomics screen uncovers ncRNA T-UCR functions in neuroblastoma tumours".
3. Dunoyer P, Schott G, Himber C, Meyer D, Takeda A, Carrington JC,
and Voinne O,
"Small
RNA Duplexes Function as Mobile Silencing Signals Between Plant Cells".
4. Kosaka N, Iguchi H, Yoshioka Y, Takeshita F, Matsuki Y, and Ochiya
T,
"Secretory Mechanisms
and Intercellular Transfer of MicroRNAs in Living Cells?"
5. DePaul AJ, Thompson EJ, Patel SS, Haldeman K, and Sorin EJ,
"Equilibrium
conformational dynamics in an RNA tetraloop from massively parallel molecular
dynamics".
6. Gao JS, Zhang Y, Li M, Tucker LD, Machan JT, Quesenberry P, Rigoutsos
I , and Ramratnam B.
"Atypical transcription
of microRNA gene fragments".
7. Gupta RA, Shah N, Wang KC, Kim J, Horlings HM, Wong DJ, Tsai M-C,
Hung T, Argani P, Rinn JL, Wang Y, Brzoska P, Kong B, Li R, West RB, van
de Vijver MJ, Sukumar S, and Chang HY,
"Long non-coding
RNA HOTAIR reprograms chromatin state to promote cancer metastasis".
8. Carninci P,
"RNA Dust:
Where are the Genes?"
9. Inui M, Martello G, and Piccolo S,
"MicroRNA
control of signal transduction".
10. Frenster JH, and Hovsepian JA,
"Analysis
of Intra-Nuclear Entropy Changes during EMT Activation".
11. Deng N-J, and Cieplak P,
"Free Energy
Profile of RNA Hairpins: A Molecular Dynamics Simulation Study".
12. Zhang H, Li Y, and Lai M,
"The microRNA network
and tumor metastasis".
13. Sheehy JP, Davis AR, and Znosko BM,
"Thermodynamic
characterization of naturally occurring RNA tetraloops".
14. Schudoma C, May P, Nikiforova V, and Walther D,
"Sequence–structure
relationships in RNA loops: establishing the basis for loop homology modeling".
15. Li L, Feng T, Lian Y, Zhang G, Garen A, and Song X.
"Role of human noncoding
RNAs in the control of tumorigenesis".
16. Frenster JH, and Hovsepian JA .
"Functional Embryomas
as a Result of Embryonic Gene Re-expression".
17. Grinchuk OV, Jenjaroenpun P, Orlov YL, Zhou J, and Kuznetsov
VA,
"Integrative
analysis of the human cis-antisense gene pairs, miRNAs and their transcription
regulation patterns".
18. Chen TS, Lai RC , Lee MM, Choo ABH, Lee CN, and Lim
SK,
"Mesenchymal stem
cell secretes microparticles enriched in pre-microRNAs".
19. Mohamed JS, Gaughwin PM, Lim B, Robson P, and Lipovich
L,
"Conserved long
noncoding RNAs transcriptionally regulated by Oct4 and Nanog modulate pluripotency
in mouse embryonic stem cells".
20. Nicodemi M, and Prisco A,
"Thermodynamic
Pathways to Genome Spatial Organization in the Cell Nucleus".
1. Each cell retains all of its embryonic genes for a lifetime.
2. Controls for embryonic genes are often absent in adults.
3. Uncontrolled embryonic genes can replicate wildly.
4. Replicating genes participate in intra-cellular competition.
5. The basis for gene competition is selective transcription.
6. MicroRNAs can reprogram embryomic transcription.
7. Gene reprogramming can produce normal phenotypes.
8. Normal phenotypes can by-pass chromosomal lesions.
9. MicroRNA therapy may need to be permanent.
10. Transplantation of microRNAs could be preferred.
1. Pathways within cell genomes involve a flow of information.
2. Information can flow by direct contact or by third parties.
3. Direct contact within whole genomes is difficult to regulate.
4. DNA-DNA direct contects are influenced by agents.
5. Nuclear agents include hydrophilic ionic and hydrophobic conforming ligands.
6. Third parties within genomes involve RNAs and proteins.
7. RNAs and proteins are easy to regulate or reverse.
8. Information can be shared, lost, or transformed.
9. System information can be hidden during system isolation.
10. Local information can be permanently lost during system entropy.
Links to Current
Research in Euchromatin:
Links to
Euchromatin Activator RNA Reviews:
Links to
Euchromatin Activator RNA Research:
Links to Ultrastructural
Probes of DNase I-Sensitive Sites:
Links to
RNA as a Therapeutic Agent:
Links to Hodgkin Lymphoma
Immuno-Pathology:
Links to Activated
T-Lymphocyte Immunotherapy:
Links to Medical
Systems Biology:
Links to Selective
Gene Transcription:
Links to RNA-Induced
Epigenetics:
Links to RNA-Induced
Embryogenesis:
Links to RNA and
Biological Causality:
Links to Reprogramming
and Neoplasia:
A Brief History of Activator RNA:
"Ultrastructural
Probes of Active DNA Sites, and the RNA Activators of DNA".
(PowerPoint Presentation).
Top of Page - Euchromatin
Network - Euchromatin
Research - Research
in Quantitative Radiology
For Further Information and Feedback:
Jeannette A. Hovsepian, M.D.
E-mail: frensasc@ix.netcom.com
Phone: +1 650 367 6483