Review
Decoding the Cardiac Message
The 2011 Thomas W. Smith Memorial Lecture
Gerald W. Dorn II
Abstract: This review reflects and expands on the contents of my presentation at the Thomas W. Smith Memorial
Downloaded from http://circres.ahajournals.org/ by guest on June 17, 2017
Lecture at American Heart Association Scientific Sessions, 2011. “Decoding the cardiac message” refers to
accumulating results from ongoing microRNA research that is altering longstanding concepts of the mechanisms for,
and consequences of, messenger RNA (mRNA) regulation in the heart. First, I provide a brief historical perspective
of the field of molecular genetics, touching on seminal research that paved the way for modern molecular
cardiovascular research and helped establish the foundation for current concepts of mRNA regulation in the heart.
I follow with some interesting details about the specific research that led to the discovery and appreciation of
microRNAs as highly conserved pivotal regulators of RNA expression and translation. Finally, I provide a personal
viewpoint as to how agnostic genome-wide techniques for measuring microRNAs, their mRNA targets, and their
protein products can be applied in an integrated multisystems approach to uncover direct and indirect effects of
microRNAs. Experimental designs integrating next-generation sequencing and global proteomics have the potential to
address unanswered questions regarding microRNA–mRNA interactions in cardiac disease, how disease alters mRNA
targeting by specific microRNAs, and how mutational and polymorphic nucleotide variation in microRNAs can affect
end-organ function and stress response. (Circ Res. 2012;110:755-763.)
Key Words: RNA silencing 䡲 transcriptional regulation 䡲 genetic mutation 䡲 microRNA
W
e are members of a biomedical research community
whose world-view has been completely altered over
the past 150 years, first by the discovery of the principles
of genetics, then by the development of the techniques of
molecular biology, and most recently by the application of
computerized/roboticized “omics” approaches to broad biological questions. The scientific topic of this session, microRNAs in
the cardiovascular system, is an example of an area of
research that did not exist even a decade ago. Yet the ongoing
explosion of information on microRNAs has provided insight
into integration and regulation of biological systems. We
must therefore react by adjusting our concepts and experimental approaches.
Because “a dwarf on a giant’s shoulders sees farther of the
two,”1 I offer here some perspective on the scientific road that
has brought molecular genetics to its current state. I then
propose a conceptual paradigm for investigating microRNA–
mRNA–protein regulation that extends the typical reductionist experimental approach in favor of global analyses that
lend themselves to integrative interpretation of biological
pathways in terms of functional modules. Finally, I discuss
future developments that have the potential to integrate basic
molecular, proteomic, and physiological research with clinical genetics of microRNAs.
The Genesis of Molecular Biology
The underlying principle of molecular biology can be summarized as follows: genes are vertically transmitted packets
of biological information that determine phenotype, and DNA
is the stuff of which genes are made. The experimental basis
for this paradigm, now accepted as natural law, was developed by 2 19th-century Europeans who were contemporaries,
and yet apparently unaware of each other’s work. Fr. Gregor
Mendel is familiar to every middle school student. During the
period immediately before and during the American Civil
War, his meticulous observation and quantitative documentation of patterns for vertical phenotype transmission of
specific traits in Pisum sativum (peas) identified dominant
and recessive genetic characteristics and enabled him to
derive the “laws of inheritance.”2 Mendel’s original work was
misunderstood and largely ignored until it was independently
reproduced and acknowledged over 30 years later by Hugo de
Vries and Carl Correns.3,4
Contemporaneous with Mendel’s botanical studies, the
young physician Friedrich Miescher was embarking on what
would now be considered a postdoctoral research fellowship
studying the chemistry of blood cells. Although his project in
Felix Hoppe-Seyler’s program was to have focused on
Original received December 12, 2011; revision received January 5, 2012; accepted January 10, 2012. In December 2011, the average time from
submission to first decision for all original research papers submitted to Circulation Research was 14.29 days.
From the Center for Pharmacogenomics, Department of Internal Medicine, Washington University School of Medicine, St. Louis, Missouri.
Adapted from a presentation delivered at the American Heart Association Scientific Sessions, Orlando, Florida, November 12–16, 2011.
Correspondence to Gerald W. Dorn II, MD, Philip and Sima K. Needleman Professor, Washington University Center for Pharmacogenomics, 660 S.
Euclid Ave., Campus Box 8220, St. Louis, MO 63110. E-mail [email protected]
© 2012 American Heart Association, Inc.
Circulation Research is available at http://circres.ahajournals.org
DOI: 10.1161/CIRCRESAHA.111.256768
755
756
Circulation Research
March 2, 2012
Non-standard Abbreviations and Acronyms
CLIP
PABP
RISC
RISCome
RISC-Seq
RNA-Seq
UTR
cross-linking immunoprecipitation
poly(A) binding protein
RNA-induced silencing complex, ie, a microRNA bound to
Argonaute 2 protein
the population of mRNAs in a RISC, ie, those mRNAs
targeted by present microRNAs
identification of microRNA-targeted mRNAs by nextgeneration sequencing of mRNAs from Argonaute 2
immunoprecipitates
quantitative transcriptional profiling by next-generation
sequencing of mRNA
untranslated region
Downloaded from http://circres.ahajournals.org/ by guest on June 17, 2017
lymphocytes, they were difficult to obtain. By contrast,
leukocytes were abundant in the purulent suppurations of
bandages obtained from the local surgical clinic. From
leukocyte nuclei, Miescher isolated an acid-insoluble, sodium
hydroxide- and disodium phosphate-soluble nonproteinaceous substance he designated “nuclein,”5 later termed nucleic acid, or DNA. Publication of Miescher’s results in his
mentor’s journal was delayed for 2 years while Hoppe-Seyler
reproduced them, extended them to yeast, and published his
own data.6
The work of Mendel and Miescher and their successors
established the foundations of molecular biology by defining
genetic principles and discovering DNA. However, the revelation that genes are made of DNA required 70 additional
years, until World War II. As Professor Emeritus at Rockefeller University, 66-year-old Oswald Avery, MD, and his
colleagues reported that DNA was the substance that induced
virulent transformation of nonpathological pneumococcus.7
The identity of DNA as the “transforming principle” of genes
was confirmed by tracing bacteriophage 32P-DNA in studies
performed by Alfred Hershey and Martha Chase.8 Together,
the Avery-MacLeod-McCarty and Hershey-Chase experiments established the molecular basis for heredity and helped
to focus attention on the mechanism by which DNA encodes
information, and on defining its molecular structure. Accord-
ingly, X-ray crystallographic studies performed by Rosalind
Franklin and Maurice Wilkins9,10 helped James Watson and
Francis Crick divine the double helical nature of DNA,11
which in turn permitted the physicist George Gamow to
propose triplet nucleotide codons for amino acids.12
From DNA to RNA
Understanding DNA was the key that unlocked the door of
molecular genetics. A timeline of Nobel Prizes given for
DNA research since World War II reveals a burst of DNA
research that characterized the second half of the 20th century
(Figure 1, top). Discovery of the enzymes responsible for
DNA replication and RNA transcription, elucidation of the
mechanism of bacterial transduction, and development of the
technique for saturation mutagenesis were acknowledged in
the late 1950s, followed by recognition of Watson and
Crick’s elucidation of the 3-dimensional structure for DNA in
1962. Over the next 15 years the attention of the Nobel
committee was elsewhere, but in the 15 years that followed
(ie, 1978 –1993), 4 Nobel prizes were awarded for DNA
research, 3 related to development of DNA manipulation or
diagnostic techniques (restriction endonucleases in 1978,
rDNA expression and DNA sequencing in 1980, and polymerase chain reaction and site-directed mutagenesis in 1993).
Just 5 years ago the Nobel Prize in Physiology or Medicine
was awarded for targeted gene manipulation in mice. Thus,
by the time I graduated from Medical School over 30 years
ago, we had a detailed understanding of the fundamental
mechanisms of DNA replication and transcription. An appreciation of the impact of DNA mutations provided a mechanistic basis for “Mendelian” human disease, including the
first-recognized example, alkaptonuria, the heritable features
of which had originally been described in 1902.13
Our understanding of the roles for genetic mutation and
regulated gene transcription in heart disease is built on these
fundamental observations of the middle and late 20th century.
Using a combination of genetic techniques and functional
genomics, Christine and Jon Seidman first described in
1990 a molecular basis for the most common Mendelian
cardiac disease, familial hypertrophic cardiomyopathy.14,15
This seminal pair of publications inaugurated an era of
molecular genetics in cardiology that has, to date, uncovered
Figure 1. Timeline of Nobel prizes for
DNA- and RNA-based research.
The year, classification of Nobel prize
(P o M⫽Physiology or Medicine;
Chem⫽Chemistry), and surnames of
awardees are listed for DNA-related (top)
and RNA-related (bottom) research. In
italics immediately below each is a brief
description of the prize-winning research.
Note that the 1959 Nobel awardee Arthur
Kornberg, who isolated and characterized
DNA polymerase I, is the father of Roger
Kornberg, who received a Nobel prize in
2006 for elucidating the role of RNA polymerase II and other factors in transcription. Detailed information on all the Nobel
laureates and their work is available at
www.nobelprize.org/nobel_prizes/.
Dorn
Decoding the Cardiac Message
757
Downloaded from http://circres.ahajournals.org/ by guest on June 17, 2017
over 1000 mutations of more than 30 different genes causing
“monogenic” or Mendelian heritable hypertrophic and dilated
cardiomyopathies. (For a fascinating personal perspective of
the Seidmans’ body of work and the overall field, the
interested reader is referred to Seidman & Seidman.16)
Likewise, the pathways by which molecular regulation of
muscle gene expression determines cardiac development, and
the contribution of dysregulated cardiac gene expression to
heart disease, were initially uncovered by Eric Olson and his
colleagues (reviewed in Olson,17 Potthoff & Olson,18 Hill &
Olson,19 and Haberland et al20). The general paradigm established by these and other researchers engendered detailed
investigations of genetic reprogramming in heart disease,
provided a foundation for molecular diagnostics in human
and experimental heart failure, and suggested novel therapeutic avenues aimed at normalizing the pathological effects of
disease-regulated cardiac genes.21–23
Observations that there were distinctive gene expression
patterns within disease was one of many factors prompting
mechanistic investigations that moved beyond DNA to examine its messenger RNA gene products. The evolutionary
advantage of DNA as the archive for biological design lies in
its chemical stability, ability to undergo repair, and high
fidelity of replication permitting faithful vertical transmission
of encoded phenotypes. In other words, DNA is comparatively static and immutable. For these reasons, results from
the Human Genome Project have provided not only a blueprint of our species’ molecular architecture in comparison
with that of other species,24,25 but also revealed to what extent
past Homo sapiens genetic admixture with other ancient
hominids impacts modern humans.26 Metaphorically, however, when purchasing a car, one does not just compare parts
lists and schematic diagrams of various models; one takes the
car for a test drive. By the same logic, in moving beyond the
genetic blueprint it was necessary to determine how mRNA
and protein gene products are regulated and interact to direct
biological processes.
In comparison with DNA, mRNA is evanescent and
susceptible to enzymatic and physical degradation. Thus, the
major insights into RNA biology followed and built on those
of DNA. Returning to the Nobel prize timeline (Figure 1,
bottom), the first RNA Nobel prize was awarded in 1968 for
deciphering the mechanisms of translation, followed by a
20-year hiatus that was brought to an end by recognition of
the discoveries that RNAs can have catalytic activity similar
to proteins (1989) and how RNA splicing produces protein
diversity (1993). By 2006, RNA science had clearly arrived
when the Nobel prizes for both Chemistry and Physiology or
Medicine were awarded for RNA biology, the former for
deciphering the mechanisms of transcription and the latter for
the discovery of modulated gene expression by doublestranded RNAs, ie, RNA interference. This brings us to the
modern era and the overall topic of this scientific session,
microRNAs.
curiosity limited to nematodes: it was widely accepted that
an antisense RNA complementary to a given mRNA could
suppress that mRNA, but it was not appreciated how
endogenous (or exogenous) double-stranded small RNAs
could function as modulators of biological function. This
apparent contradiction was addressed by Andrew Fire and
Craig Mello’s subsequent discovery that RNA interference
by double-stranded RNA was more efficient than that by
single-stranded antisense RNA28 (for which they received
the 2006 Nobel Prize in Physiology or Medicine). This
observation revealed the presence of previously unknown
mechanisms for processing double-stranded small RNAs
and presenting the active single strand to its mRNA target.
These revelations were rapidly followed by identification
and characterization of the second microRNA, let-7, by
Gary Ruvkun,29,30 setting the stage for 3 concurrent 2001
reports in Science that demonstrated unexpected abundance and evolutionary conservation of microRNAs, and
suggested their broad impact on biological function.31–33
These 3 papers inaugurated the burst of microRNA research that continues to the present; they also received the
AAAS Newcomb Cleveland Prize for the most significant
papers published in Science during that year.
Conceptually, what do microRNAs do? To quote a past
student from my laboratory, “MicroRNAs make mRNAs not
make protein.” Although there may be some exceptions, the
general function of microRNAs is to prevent mRNA translation into protein. Two possible mechanisms have been
identified, mRNA destabilization and translational suppression (although, again, this is an area of research that continues
to evolve). As schematically depicted in Figure 2A, a fully
processed single-stranded microRNA is incorporated into an
Argonaute2-containing macromolecular complex, the RNAinduced silencing complex or RISC,34 to which target
mRNAs are recruited via complementary Watson–Crick
binding to the incorporated microRNA. mRNA destabilization takes place through nucleolytic cleavage and degradation of the complementary mRNA. Accordingly,
steady state expression levels of mRNAs targeted by
microRNAs may be decreased due to degradation. Alternatively, mRNAs incorporated into RISCs may undergo
translational suppression via deadenylation, ie, removal of
the polyA⫹ tail, which may or may not require participation in the RISC of poly(A) binding protein (PABP) and
associated deadenylases35,36 in association with GW182
family proteins37 (Figure 2B). It is likely that translational
suppression and mRNA degradation both occur, or that
they take place sequentially.
My goal here is not to provide an encyclopedic review of
cardiac microRNAs encompassing the explosion of data from
the past few years. Others have contributed more to the field, and
any such compendium would soon be made obsolete by ongoing
progress and developments. Rather, I would like to offer a perspective on investigative and analytic approaches that may be useful for
interrogating cardiac microRNAs in health and disease.
RNA Interference and MicroRNAs
An Unbiased and Genome-Wide Approach to
MicroRNA Target Analysis
The first microRNA, lin-4, was described in Caenorhabditis elegans in 1993 by Victor Ambros and colleagues.27
At the time, lin-4 was largely considered to be a biological
Association of an mRNA with a microRNA–RISC complex
is required for gene silencing. This association is reversible
758
Circulation Research
March 2, 2012
Downloaded from http://circres.ahajournals.org/ by guest on June 17, 2017
Figure 2. Schematic representations of microRNA–mRNA
interactions in the RNA-induced silencing complex (RISC).
A, mRNA destabilization: a red microRNA is incorporated into a
green Argonaute 2 complex, the RISC, and recruits its target
mRNA (black) through sequence complementarity to binding
sites in the 3⬘ region; nucleolytic cleavage degrades the mRNA.
B, mRNA translational suppression: here, additional RISC cofactors poly(A) binding protein (PABP, blue) and GW182 proteins
(red), deadenylate the target mRNA and suppress ribosomal
translation. C, (left) Equation adapted from steady state receptor–ligand interactions showing the relationship of microRNA
(miR) and target mRNA concentrations and microRNA–mRNA
binding energy on RNA duplex formation at the RISC. (right)
Nucleotide sequence of miR-499 (top) and 5 bioinformatically
predicted binding sites of its validated target, Sox6 (bottom).
Note that the computer algorithm selects binding sites based on
perfect seed-sequence (underlined) complementarity. Nucleotides in red are mismatches between miR-499 and target site.
There are no available data as to which of these sites (or others
not bioinformatically predicted) are responsible for miR-499mediated Sox6 suppression.
and therefore governed by the same general principles as
noncovalent receptor–ligand binding38: the microRNA–RISC
is analogous to an unoccupied receptor, the mRNA is like its
ligand, and the microRNA–mRNA duplex is like the active
receptor–ligand complex. Thus, the steady state probability
that a given microRNA–mRNA pair will form in the RISC is
determined by the concentration of microRNA, the concentration of its mRNA target, and the binding affinity of the
microRNA–mRNA duplex (Figure 2C). For RNA duplexes,
binding affinity is more properly described by binding energy, which is a function of Watson–Crick complementarity
and can be calculated using relatively straightforward methods.39 The formula for calculating the steady state interaction
between a microRNA and a given mRNA target is simple, but
each of the formulaic variables exhibits tremendous biological variability. The concentration (expression level) of ⬇200
cardiac microRNAs varies widely in human and experimental
heart disease.21,40,41 Likewise, the concentration (expression
level) of cardiac mRNAs varies independently due to transcriptional regulation,21,42,43 and dependently due to destabilization by cardiac-expressed microRNAs. Finally, the binding energy for a given microRNA–mRNA complex will vary
with the primary sequence complementarity and can be
influenced by secondary mRNA structures that affect binding
site accessibility to RISC-associated microRNAs. Figure 2C
(right panel) provides an example of sequence variability
within 5 bioinformatically predicted miR-499 binding sites in
one of its thoroughly validated cardiac mRNA targets,
Sox6.44
One way to account for confounding biological variability is to quantify each of the variables simultaneously and
in the proper clinical or experimental context. Multiple
microRNAs and mRNAs are simultaneously regulated in
cardiac disease (vide supra). Furthermore, a given microRNA has the potential to independently target multiple
mRNAs, and a given mRNA is likely to be targeted by
more than 1 microRNA.45 Thus, agnostic whole-genome
approaches are essential to deconvoluting and understanding the pathophysiology of microRNA–mRNA interactions
in heart disease. Toward this end we have used nextgeneration sequencing of cardiac mRNA (RNA-Seq46) to
digitally measure absolute mRNA expression levels (ie,
the transcriptome) in normal and diseased hearts, and have
adapted a similar approach to quantifying the mRNAs
within cardiac RISC complexes (RISC-Seq47). The latter
utilizes Argonaute 2 immunoprecipitation combined with
microextraction and next-generation sequencing of associated mRNAs to identify RISC-targeted mRNAs (the
RISCome), ie, those mRNAs that are enriched in the RISC
in comparison with the transcriptome due to the actions of
endogenous microRNAs. It is therefore also essential to
quantify all the endogenously expressed microRNAs, for
which next-generation sequencing protocols are available.48,49 By comparing the results of microRNA sequencing, RNA-Seq, and RISC-Seq performed under different
sets of pathophysiological conditions (Figure 3A), it is
possible to obtain an overview of how different diseases
alter the cardiac microRNA–mRNA interactome. To date,
this type of comprehensive and unbiased genome-wide
approach to cataloguing disease-induced miR–mRNA interactions has not been performed for the heart.
The combination of RISC-Seq and RNA-Seq will identify
microRNA targets, but does not indicate which of the
expressed microRNAs is directly targeting which of the
RISC-associated mRNAs. Available bioinformatics platforms
can help refine the universe of potential microRNA–mRNA
pairs, but their results can be inconsistent and are heavily
biased toward near-perfect seed-sequence (microRNA nucleotides 2– 8) complementarity.50 We therefore adapted the
procedure for “RISC programming” with a microRNA of
interest51 to our in vivo analyses of cardiac microRNA–
mRNA interactions. 47 Conceptually, we compare the
RISComes of unprogrammed tissue (in which RISC-Seq will
identify direct mRNA targets of all endogenous microRNAs)
to those from tissue expressing higher levels of the microRNA of interest (in which the mRNA targets for that
microRNA will be further enriched, in comparison with the
unprogrammed RISCome). In vivo cardiac RISC programming can be achieved by conventional or inducible cardiomyocyte-specific transgenic overexpression, producing cellautonomous data. We take care to express the programming
microRNA(s) at levels that are observed in either healthy or
diseased tissue, to prevent recruitment of nonphysiological
mRNA targets to the RISC. Our analyses also evaluate
Dorn
Decoding the Cardiac Message
759
Downloaded from http://circres.ahajournals.org/ by guest on June 17, 2017
Figure 3. Schematic depiction of in vivo wholegenome approaches for measuring microRNA–
mRNA interactions. A, Next-generation sequencing techniques are used to digitally quantify
microRNA (red), mRNA (blue), and RISCassociated mRNA (ie, microRNA–mRNA duplexes;
green). The result is a catalog of microRNA–mRNA
expression and interaction that can be compared
between pathophysiological conditions. B, Schematic depiction of cardiac RISC programming by
transgenesis. Transgenic expression of a
microRNA of interest (red) will specifically increase
its mRNA targets in the RISCome, which can be
identified by comparison to nontransgenic (or
other microRNA programmed) RISComes.
RISC-enrichment of mRNAs as a function of mRNA levels in
the respective transcriptomes, thereby accounting for secondary effects of the programming microRNA (or any induced
phenotype) on transcriptionally regulated mRNA expression.
For the most part, comparing programmed (transgenic) with
unprogrammed (nontransgenic) RISComes identifies direct
mRNA targets of the programming microRNA (Figure 3B),
although mRNAs indirectly enriched in the RISCome
through the effects of secondarily upregulated microRNAs
will also be detected. When we used this approach to compare
cardiac mRNA targets of 2 muscle-specific micro-RNAs that
are from different families, but expressed in the heart at
similar levels (miR-133a and miR-499), RISC-Seq identified
mRNA targets for each that were largely nonoverlapping and
were consistent with the previously reported effects of these
microRNAs on cardiac development and myofibrillar gene
expression, respectively.47
The approach of assaying Argonaute 2–associated mRNAs
(typically by microarray) to define mRNA targets has been used
in several technical variations, including HITS-CLIP52 and
PAR-CLIP.53 Thus, this general approach is increasingly accepted as the gold standard for biologically defining mRNA
targets of microRNAs. The decision of whether or not to use
some form of chemical RNA cross-linking (CLIP: cross-Linking
immunoprecipitation) depends on what type of information is
desired from the specific study. CLIP detects direct microRNA–mRNA binding and provides spatial information on
microRNA binding and flanking sites at the mRNAs at a cost of
biasing the results toward microRNA–mRNA pairs exhibiting the
strongest RNA-binding protein interactions; CLIP may also be less
sensitive to dynamic regulation of microRNAs than techniques that
do not use cross-linking.54
There are some important caveats for microRNA target
identification by RISC-Seq: first, to account for interindividual variability (as after a physiological manipulation such
as aortic banding or myocardial infarction), one should
analyze results of RISC-Seq in comparison to RNA-seq from
the same sample. In this manner, RISC enrichment is assessed per individual tissue sample, and not grouped by
experimental treatment. Second, levels of statistical significance should be appropriate for the intent of the study. In our
initial RISC-Seq validation studies, we set particularly rigorous statistical filters for RISC enrichment (2-fold change,
P⬍0.0001, false discovery rate ⬍0.01) to avoid false-positive
values.47 In so doing we knowingly failed to identify many
weak mRNA targets; this will not always be the preferred
approach. Finally, assessment of RISC enrichment (ie, level
of RISC-associated mRNA/level of that mRNA in the transcriptome) might theoretically overlook the strongest mRNA
targets of highly expressed microRNAs, which may already
be highly represented in the nonprogrammed RISC. Thus,
RISC-programming data need to be interpreted in the context
of parallel miR-Seq, RNA-Seq, and RISC-Seq data from
nonprogrammed tissues.
The Multiple Degrees of Separation Between a
MicroRNA and Its Induced Phenotype
The above discussion suggests how cardiac-directed forced
expression of a microRNA can be useful for identifying its
760
Circulation Research
March 2, 2012
Downloaded from http://circres.ahajournals.org/ by guest on June 17, 2017
Figure 4. Differing conceptual approaches linking a
microRNA with its induced phenotype. A, The classic linear
approach as applied to microRNAs (miR). The miR of interest
suppresses a particular mRNA that encodes a specific protein,
silencing of which is responsible for the observed phenotype.
Operating under this concept, gain- and loss-of-function manipulations have predictable and discreet effects. B, Alternate conceptual paradigm incorporating parallelism and redundancy at
the level of miRs and their many mRNA targets. Direct effects
on multiple proteins are shown with potential feedback regulation of microRNA expression, mRNA target expression, and secondary and tertiary posttranslational modifications, all of which
may integrate to induce the observed phenotype. This complexity produces unpredictable affects that can be detected using
global analytic approaches.
mRNA targets. By extension, this approach can also help
define the impact of differential mRNA targeting after
genetic reprogramming, as in cardiac hypertrophy or heart
failure. The conventional experimental transgenic approach overexpresses a factor, defines the resulting phenotype, and then fills in the intervening biological pathway
to establish a plausible mechanism. When a microRNA is
overexpressed, the underlying assumption is often that it
will suppress a specific messenger RNA that encodes a
particular protein, and that suppression of the protein
produces the phenotype. This linear view, and its relatively
modest variations (ie, some proteins regulate gene expression, etc., providing for feed-forward and feed-back regulation) has conceptual appeal because its simplicity and
structure lend it to interrogation by molecular perturbation
(Figure 4A). The cardiac literature is replete with manuscripts that follow this general experimental and analytic
pattern, and much has been learned over the course of
these investigations. Using miR-133a as a well-studied
example, 6 relatively early manuscripts each reported
single or small groups of mRNA targets implicated as the
causal factor for various observed phenotypes. Validated
miR-133a targets include Whsc2,55 Ccnd2,56 Ctgf,57
Casp9,58 HCN4,59 and Hand2.60 Indeed, these mRNAs
were also revealed as miR-133a targets in heart by RISCSeq performed in miR-133a cardiac transgenic mice.47 The
point is that conclusions based on an underlying assump-
tion that one miR primarily affects one mRNA whose
protein product is largely responsible for an observed
phenotype are likely to be significant oversimplifications
of the true situation.
The following considerations suggest a more complex
interpretation of microRNA actions and support a global/
genome-wide approach to their delineation: first, many microRNAs are members of families having several other
members with similar sequences; these different family members will have overlapping mRNA targets. For this reason,
genetic ablation of one member of a microRNA family will
not necessarily have a major effect when other family
members remain. miR-133, for which there are 3 family
members, provides an example of functional compensation
with single gene ablation that was revealed with multigene
ablation.56 Second, as noted above, a single microRNA
typically targets dozens or more different mRNAs, and in
many instances the effects on individual mRNAs and proteins
are modest.61 Thus, the aggregate phenotype induced by a
given microRNA results from the cumulative effects on all of
its targets. Because multiple mRNA targets of a given microRNA or microRNA family tend to cluster within specific
functional pathways (such as cell growth, differentiation, motility, or programmed death), this action of microRNAs can be
considered as modular with respect to biological processes,
rather than specific to one or a few mRNAs. Finally, the direct
targets of microRNAs may have multiple degrees of separation from the end-organ phenotype, being a culmination of
first-, second-, and third-generation effects. Figure 4B schematically depicts a microRNA signaling pathway that takes
into account multiple microRNAs acting simultaneously on
their mRNA targets, multiple mRNA targets being suppressed by a given microRNA, and reflecting likely avenues
of regulatory feedback that can produce indirect secondary
and tertiary effects. The best characterized cardiac example of
indirect microRNA effects relates to myomiR (miRs-208a,
-208b, and -499) – directed cardiac myosin isoform switching
in cardiac hypertrophy.44,62 These effects are so striking that
antimiR 208a therapy has been shown to prevent hypertensive cardiac remodeling,63 and yet myosin heavy chain
mRNAs are not direct targets of any myomiR. Thus, the
major myomiR action on the heart, regulation of myosin
isoforms, is indirect.
An unbiased approach to protein regulation by microRNAs
would seem to complement the agnostic techniques of
microRNA target identification and transcriptional profiling
described above. Indeed, global proteomics have been used to
connect molecular mechanism with microRNA-conferred
phenotype, with some unexpected results. An important
finding of the first in vitro global proteomics analyses of
noncardiac cells was that the vast majority of proteins
regulated by microRNAs are not direct micrRNA targets.61
The likely explanations for the disconnect between
microRNA-targeted mRNA and protein include secondary
and tertiary effects noted above, and the inherent bias of
conventional proteomics toward the most highly expressed
genes, which may not be the direct microRNA targets.
Nevertheless, proteomics will be a useful addition to assays
of bioinformatically predicted candidate mRNA targets and
Dorn
steady-state quantification of their specific protein products
when determining the mechanistic basis for microRNAinduced phenotypes.
The Pathological Potential for Genetic
MicroRNA Variation
Downloaded from http://circres.ahajournals.org/ by guest on June 17, 2017
The unique ability of microRNAs to impact multiple effectors
within a biological pathway optimally positions them as
therapeutic targets. MicroRNA-based therapeutics offer an
avenue to deal with the inherent plasticity of biological
systems, in which critical responses are characteristically
mediated through multiple parallel and redundant pathways.
Biological redundancy provides an obvious evolutionary
advantage: when the primary biological pathway is under
attack by a pathological organism or genetic mutation, the
essential response can be retained through induction of
alternate mechanisms. For the same reasons, the reaction of
an organism to experimental manipulation of a single molecular effector does not represent the normal role of that
effector in the response. The stimulus–response is more like
that of a water balloon than a linear pathway. When one side
is pushed in, redundancy in the rest of the structure/pathway
compensates and maintains overall systematic integrity. The
ability of a single microRNA to coordinately regulate multiple effectors within a given biological pathway avoids this
compensation by functionally overlapping factors. Extending
the water balloon metaphor, by simultaneously targeting
multiple effectors in a given pathway, one can consider
microRNAs to adjust the overall volume of water in the
balloon.
A largely overlooked aspect of microRNA biology is the
potential impact of naturally occurring nucleotide sequence variation on microRNA function, ie, microRNA
mutations. It is striking that the nucleotide sequence of
mature microRNAs tends to be very highly conserved
across species, whereas sequence variability within putative microRNA binding sites in mRNA 3⬘ UTRs is
common.64,65 This makes intuitive sense: nucleotide sequence is the primary determinant of microRNA function
because it determines the binding energy of a microRNA–
mRNA duplex, and therefore the efficiency of mRNA
targeting to RISC and resulting mRNA suppression. With
rare exceptions,66 and even though mRNA binding sites far
outnumber microRNAs, sequence variation within a microRNA binding site will most likely have modest impact
because it will affect the binding of only 1 family of
microRNAs to only 1 binding site in only 1 target mRNA.
By contrast, sequence variation in a mature microRNA has
the potential to alter the mRNA targeting profile of that
microRNA for any mRNA with a binding site that either
loses or gains sequence complementarity as a consequence
of the microRNA mutation. Consistent with this paradigm,
a human seed sequence mutation in miR-96 produces
heritable hearing loss.67 Recently, we uncovered a nonseed
sequence mutation of human miR-499 that alters its mRNA
targeting profile and modified the functional and proteomic phenotype induced by miR-499 overexpression in
mouse hearts.68 A u to c mutation at nucleotide position 17
altered the pattern of cardiac RISC mRNA enrichment for
Decoding the Cardiac Message
761
a subset of mRNA targets whose binding sites had imperfect seed sequence complementarity and tended to have
appropriately complementary nucleotides corresponding to
miR-499 position 17. These findings demonstrate for the
first time that a nonseed sequence mutation can alter
microRNA function and the consequent organ/organism
phenotype. These results support genetic screening for,
and functional analysis of, microRNA sequence variants in
human disease.
Final Thoughts
The ongoing explosion of microRNA research reflects a
propitious convergence of biology and technology in an
area in which nucleotide sequence is the key determinant
of function. Next-generation sequencing is being used to
globally profile microRNA and mRNA expression, to
examine microRNA–mRNA interactions in different
pathophysiological contexts, and to discover and evaluate
the function of microRNA and binding site mutations.
Sophisticated proteomics are being applied to assays of
protein expression and posttranslational modification.
Computerized algorithms can be used for in silico modeling of microRNA–mRNA interactions and to understand
the structural determinants of biological function. Pathway
analysis of micoRNA–mRNA–protein interactions can be
applied to results of genome-wide profiling studies to
provide unbiased insight into integrated biological function of specific microRNAs. Future efforts will therefore
likely have greatest success when they take an agnostic
approach, when they integrate multiple techniques, and
when they bridge scientific disciplines.
Sources of Funding
Supported by National Institutes of Health R01 HL59888,
HL087871, HL1072726, and HL108943.
Disclosures
None.
References
1. Bartlett J. [comp]. Familiar Quotations. 10th ed., rev. and enlarged by
Nathan Haskell Dole. Boston: Little, Brown; 1919. Bartleby.com; 2000.
2. Mendel G. Versuche über Pflanzenhybriden. Verhandlungen des naturforschenden Vereins Brünn. 1866;4:3– 47.
3. De Vries H. Sur la loi de disjonction des hybrides. CR Acad Sci Paris.
1900;130:845– 848.
4. Correns CG. Mendel’s Regel über das Verhalten der Nachkommenschaft
der Rassenbastarde. Ber Deut Botan Gessel. 1900;18:158 –168.
5. Miescher F. Ueber die chemische Zusammensetzung der Eiterzellen.
Medizinisch–chemische Untersuchungen. 1871;4:441– 460.
6. Hoppe-Seyler F. Ueber die chemische Zusammensetzung des Eiters.
Medizinisch–chemische Untersuchungen. 1871;4:486 –501.
7. Avery OT, Macleod CM, McCarty M. Studies on the chemical nature of
the substance inducing transformation of pneumococcal types: induction
of transformation by a desoxyribonucleic acid fraction isolated from
pneumococcus type III. J Exp Med. 1944;79:137–158.
8. Hershey AD, Chase M. Independent functions of viral protein and nucleic
acid in growth of bacteriophage. J Gen Physiol. 1952;36:39 –56.
9. Franklin RE, Gosling RG. Molecular configuration in sodium thymonucleate. Nature. 1953;171:740 –741.
10. Wilkins MH, Stokes AR, Wilson HR. Molecular structure of deoxypentose nucleic acids. Nature. 1953;171:738 –740.
11. Watson JD, Crick FH. Molecular structure of nucleic acids; a structure for
deoxyribose nucleic acid. Nature. 1953;171:737–738.
12. Hayes B. The invention of the genetic code. Am Sci. 1998;86:8 –14.
762
Circulation Research
March 2, 2012
Downloaded from http://circres.ahajournals.org/ by guest on June 17, 2017
13. Garrod AE. The incidence of alkaptonuria: a study in chemical individuality. Lancet. 1902;ii:1616 –1620.
14. Tanigawa G, Jarcho JA, Kass S, Solomon SD, Vosberg HP, Seidman JG,
Seidman CE. A molecular basis for familial hypertrophic cardiomyopathy: an alpha/beta cardiac myosin heavy chain hybrid gene. Cell. 1990;
62:991–998.
15. Geisterfer-Lowrance AA, Kass S, Tanigawa G, Vosberg HP, McKenna
W, Seidman CE, Seidman JG. A molecular basis for familial hypertrophic
cardiomyopathy: a beta cardiac myosin heavy chain gene missense
mutation. Cell. 1990;62:999 –1006.
16. Seidman CE, Seidman JG. Identifying sarcomere gene mutations in
hypertrophic cardiomyopathy: a personal history. Circ Res. 2011;108:
743–750.
17. Olson EN. Gene regulatory networks in the evolution and development of
the heart. Science. 2006;313:1922–1927.
18. Potthoff MJ, Olson EN. MEF2: a central regulator of diverse developmental programs. Development. 2007;134:4131– 4140.
19. Hill JA, Olson EN. Cardiac plasticity. N Engl J Med. 2008;358:
1370 –1380.
20. Haberland M, Montgomery RL, Olson EN. The many roles of histone
deacetylases in development and physiology: implications for disease and
therapy. Nat Rev Genet. 2009;10:32– 42.
21. Matkovich SJ, Van Booven DJ, Youker KA, Torre-Amione G, Diwan A,
Eschenbacher WH, Dorn LE, Watson MA, Margulies KB, Dorn GW 2nd.
Reciprocal regulation of myocardial microRNAs and messenger RNA in
human cardiomyopathy and reversal of the microRNA signature by biomechanical support. Circulation. 2009;119:1263–1271.
22. Heidecker B, Kittleson MM, Kasper EK, Wittstein IS, Champion HC,
Russell SD, Hruban RH, Rodriguez ER, Baughman KL, Hare JM. Transcriptomic biomarkers for the accurate diagnosis of myocarditis. Circulation.
2011;123:1174–1184.
23. Frey N, Katus HA, Olson EN, Hill JA. Hypertrophy of the heart: a new
therapeutic target? Circulation. 2004;109:1580 –1589.
24. Lander ES, Linton LM, Birren B, et al. Initial sequencing and analysis of
the human genome. Nature. 2001;409:860 –921.
25. Venter JC, Adams MD, Myers EW, et al. The sequence of the human
genome. Science. 2001;291:1304 –1351.
26. Green RE, Krause J, Briggs AW, et al. A draft sequence of the Neandertal
genome. Science. 2010;328:710 –722.
27. Lee RC, Feinbaum RL, Ambros V. The C. elegans heterochronic gene
lin-4 encodes small RNAs with antisense complementarity to lin-14. Cell.
1993;75:843– 854.
28. Fire A, Xu S, Montgomery MK, Kostas SA, Driver SE, Mello CC. Potent
and specific genetic interference by double-stranded RNA in Caenorhabditis elegans. Nature. 1998;391:806 – 811.
29. Reinhart BJ, Slack FJ, Basson M, Pasquinelli AE, Bettinger JC, Rougvie
AE, Horvitz HR, Ruvkun G. The 21-nucleotide let-7 RNA regulates
developmental timing in Caenorhabditis elegans. Nature. 2000;403:
901–906.
30. Pasquinelli AE, Reinhart BJ, Slack F, et al. Conservation of the sequence
and temporal expression of let-7 heterochronic regulatory RNA. Nature.
2000;408:86 – 89.
31. Lee RC, Ambros V. An extensive class of small RNAs in Caenorhabditis
elegans. Science. 2001;294:862– 864.
32. Lau NC, Lim LP, Weinstein EG, Bartel DP. An abundant class of tiny
RNAs with probable regulatory roles in Caenorhabditis elegans. Science.
2001;294:858 – 862.
33. Lagos-Quintana M, Rauhut R, Lendeckel W, Tuschl T. Identification of
novel genes coding for small expressed RNAs. Science. 2001;294:
853– 858.
34. Rivas FV, Tolia NH, Song JJ, Aragon JP, Liu J, Hannon GJ, Joshua-Tor
L. Purified Argonaute2 and an siRNA form recombinant human RISC.
Nat Struct Mol Biol. 2005;12:340 –349.
35. Fabian MR, Mathonnet G, Sundermeier T, et al. Mammalian miRNA
RISC recruits CAF1 and PABP to affect PABP-dependent deadenylation.
Mol Cell. 2009;35:868 – 880.
36. Fukaya T, Tomari Y. PABP is not essential for microRNA-mediated
translational repression and deadenylation in vitro. EMBO J. 2011;30:
4998 –5009.
37. Liu J, Rivas FV, Wohlschlegel J, Yates JR III, Parker R, Hannon GJ. A
role for the P-body component GW182 in microRNA function. Nat Cell
Biol. 2005;7:1261–1266.
38. Limbird LE. Cell Surface Receptors: A Short Course on Theory and
Methods. 3rd ed. New York, NY: Springer; 2004.
39. Rehmsmeier M, Steffen P, Hochsmann M, Giegerich R. Fast and
effective prediction of microRNA/target duplexes. RNA. 2004;10:
1507–1517.
40. van Rooij E, Sutherland LB, Liu N, Williams AH, McAnally J, Gerard
RD, Richardson JA, Olson EN. A signature pattern of stress-responsive
microRNAs that can evoke cardiac hypertrophy and heart failure. Proc
Natl Acad Sci U S A. 2006;103:18255–18260.
41. Thum T, Galuppo P, Wolf C, Fiedler J, Kneitz S, vanLaake LW,
Doevendans PA, Mummery CL, Borlak J, Haverich A, Gross C,
Engelhardt S, Ertl G, Bauersachs J. MicroRNAs in the human heart: a
clue to fetal gene reprogramming in heart failure. Circulation. 2007;116:
258 –267.
42. Margulies KB, Matiwala S, Cornejo C, Olsen H, Craven WA, Bednarik
D. Mixed messages: transcription patterns in failing and recovering
human myocardium. Circ Res. 2005;96:592–599.
43. Barth AS, Kumordzie A, Frangakis C, Margulies KB, Cappola TP,
Tomaselli GF. Reciprocal transcriptional regulation of metabolic and
signaling pathways correlates with disease severity in heart failure. Circ:
Cardiovasc Genet. 2011;4:475– 483.
44. van Rooij E, Quiat D, Johnson BA, Sutherland LB, Qi X, Richardson JA,
Kelm RJ Jr, Olson EN. A family of microRNAs encoded by myosin genes
governs myosin expression and muscle performance. Dev Cell. 2009;17:
662– 673.
45. Lewis BP, Burge CB, Bartel DP. Conserved seed pairing, often flanked
by adenosines, indicates that thousands of human genes are microRNA
targets. Cell. 2005;120:15–20.
46. Matkovich SJ, Zhang Y, Van Booven DJ, Dorn GW II. Deep mRNA
sequencing for in vivo functional analysis of cardiac transcriptional
regulators: application to Galphaq. Circ Res. 2010;106:1459 –1467.
47. Matkovich SJ, Van Booven DJ, Eschenbacher WH, Dorn GW 2nd. RISC
RNA sequencing for context-specific identification of in vivo microRNA
targets. Circ Res. 2011;108:18 –26.
48. Morin RD, O’Connor MD, Griffith M, Kuchenbauer F, Delaney A,
Prabhu AL, Zhao Y, McDonald H, Zeng T, Hirst M, Eaves CJ, Marra
MA. Application of massively parallel sequencing to microRNA profiling
and discovery in human embryonic stem cells. Genome Res. 2008;18:
610 – 621.
49. Buermans HP, Ariyurek Y, van Ommen G, den Dunnen JT, t Hoen PA.
New methods for next generation sequencing based microRNA
expression profiling. BMC Genomics. 2010;11:716.
50. Dorn GW 2nd. MicroRNAs: redefining mechanisms in cardiac disease.
J Cardiovasc Pharmacol. 2010;56:589 –595.
51. Karginov FV, Conaco C, Xuan Z, Schmidt BH, Parker JS, Mandel G,
Hannon GJ. A biochemical approach to identifying microRNA targets.
Proc Natl Acad Sci U S A. 2007;104:19291–19296.
52. Chi SW, Zang JB, Mele A, Darnell RB. Argonaute HITS-CLIP decodes
microRNA–mRNA interaction maps. Nature. 2009;460:479 – 486.
53. Hafner M, Landthaler M, Burger L, et al. Transcriptome-wide identification of RNA-binding protein and microRNA target sites by PAR-CLIP.
Cell. 2010;141:129 –141.
54. Wen J, Parker BJ, Jacobsen A, Krogh A. MicroRNA transfection and
AGO-bound CLIP-seq data sets reveal distinct determinants of miRNA
action. RNA. 2011;17:820 – 834.
55. Care A, Catalucci D, Felicetti F, et al. MicroRNA-133 controls cardiac
hypertrophy. Nat Med. 2007;13:613– 618.
56. Liu N, Bezprozvannaya S, Williams AH, Qi X, Richardson JA,
Bassel-Duby R, Olson EN. MicroRNA-133a regulates cardiomyocyte
proliferation and suppresses smooth muscle gene expression in the heart.
Genes Dev. 2008;22:3242–3254.
57. Duisters RF, Tijsen AJ, Schroen B, Leenders JJ, Lentink V, van dMI,
Herias V, van Leeuwen RE, Schellings MW, Barenbrug P, Maessen
JG, Heymans S, Pinto YM, Creemers EE. MiR-133 and miR-30
regulate connective tissue growth factor: implications for a role of
microRNAs in myocardial matrix remodeling. Circ Res. 2009;104:
170 –178.
58. Xu C, Lu Y, Pan Z, Chu W, Luo X, Lin H, Xiao J, Shan H, Wang Z, Yang
B. The muscle-specific microRNAs miR-1 and miR-133 produce
opposing effects on apoptosis by targeting HSP60, HSP70 and caspase-9
in cardiomyocytes. J Cell Sci. 2007;120:3045–3052.
59. Luo X, Lin H, Pan Z, Xiao J, Zhang Y, Lu Y, Yang B, Wang Z.
Down-regulation of miR-1/miR-133 contributes to re-expression of
Dorn
60.
61.
62.
63.
64.
pacemaker channel genes HCN2 and HCN4 in hypertrophic heart. J Biol
Chem. 2008;283:20045–20052.
Vo NK, Dalton RP, Liu N, Olson EN, Goodman RH. Affinity purification
of microRNA-133a with the cardiac transcription factor, Hand2. Proc
Natl Acad Sci U S A. 2010;107:19231–19236.
Baek D, Villen J, Shin C, Camargo FD, Gygi SP, Bartel DP. The impact
of microRNAs on protein output. Nature. 2008;455:64 –71.
van Rooij E, Sutherland LB, Qi X, Richardson JA, Hill J, Olson EN.
Control of stress-dependent cardiac growth and gene expression by a
microRNA. Science. 2007;316:575–579.
Montgomery RL, Hullinger TG, Semus HM, Dickinson BA, Seto AG,
Lynch JM, Stack C, Latimer PA, Olson EN, van Rooij E. Therapeutic
inhibition of miR-208a improves cardiac function and survival during
heart failure. Circulation. 2011;124:1537–1547.
Chen K, Rajewsky N. Natural selection on human microRNA binding
sites inferred from SNP data. Nat Genet. 2006;38:1452–1456.
Decoding the Cardiac Message
763
65. Saunders MA, Liang H, Li WH. Human polymorphism at microRNAs
and microRNA target sites. Proc Natl Acad Sci U S A. 2007;104:
3300 –3305.
66. Brest P, Lapaquette P, Souidi M, Lebrigand K, Cesaro A, VouretCraviari V, Mari B, Barbry P, Mosnier JF, Hebuterne X, Harel-Bellan
A, Mograbi B, Darfeuille-Michaud A, Hofman P. A synonymous
variant in IRGM alters a binding site for miR-196 and causes deregulation of IRGM-dependent xenophagy in Crohn’s disease. Nat Genet.
2011;43:242–245.
67. Mencia A, Modamio-Hoybjor S, Redshaw N, Morin M, Mayo-Merino F,
Olavarrieta L, Aguirre LA, del Castillo I, Steel KP, Dalmay T, Moreno F,
Moreno-Pelayo MA. Mutations in the seed region of human miR-96 are
responsible for nonsyndromic progressive hearing loss. Nat Genet. 2009;
41:609 – 613.
68. Dorn GW 2nd, Matkovich SJ, Eschenbacher WH, Zhang Y, Hu Y, Dorn
LE. A human 3⬘ miR-499 mutation alters cardiac mRNA targeting and
function. Circ Res. In press.
Downloaded from http://circres.ahajournals.org/ by guest on June 17, 2017
Decoding the Cardiac Message: The 2011 Thomas W. Smith Memorial Lecture
Gerald W. Dorn II
Downloaded from http://circres.ahajournals.org/ by guest on June 17, 2017
Circ Res. 2012;110:755-763
doi: 10.1161/CIRCRESAHA.111.256768
Circulation Research is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231
Copyright © 2012 American Heart Association, Inc. All rights reserved.
Print ISSN: 0009-7330. Online ISSN: 1524-4571
The online version of this article, along with updated information and services, is located on the
World Wide Web at:
http://circres.ahajournals.org/content/110/5/755
Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally published
in Circulation Research can be obtained via RightsLink, a service of the Copyright Clearance Center, not the
Editorial Office. Once the online version of the published article for which permission is being requested is
located, click Request Permissions in the middle column of the Web page under Services. Further information
about this process is available in the Permissions and Rights Question and Answer document.
Reprints: Information about reprints can be found online at:
http://www.lww.com/reprints
Subscriptions: Information about subscribing to Circulation Research is online at:
http://circres.ahajournals.org//subscriptions/