REVIEWS

Circulating microRNAs as novel biomarkers

for diabetes mellitus
Claudiane Guay and Romano Regazzi
Abstract | Diabetes mellitus is characterized by insulin secretion from pancreatic cells that is insufficient
tomaintain blood glucose homeostasis. Autoimmune destruction of cells results in type1 diabetes mellitus,
whereas conditions that reduce insulin sensitivity and negatively affect cell activities result in type2 diabetes
mellitus. Without proper management, patients with diabetes mellitus develop serious complications that
reduce their quality of life and life expectancy. Biomarkers for early detection of the disease and identification
of individuals at risk of developing complications would greatly improve the care of these patients. Small noncoding RNAs called microRNAs (miRNAs) control gene expression and participate in many physiopathological
processes. Hundreds of miRNAs are actively or passively released in the circulation and can be used to
evaluate health status and disease progression. Both type1 diabetes mellitus and type2 diabetes mellitus
are associated with distinct modifications in the profile of miRNAs in the blood, which are sometimes
detectable several years before the disease manifests. Moreover, circulating levels of certain miRNAs seem
to be predictive of long-term complications. Technical and scientific obstacles still exist that need to be
overcome, but circulating miRNAs might soon become part of the diagnostic arsenal to identify individuals
atrisk of developing diabetes mellitus and its devastating complications.
Guay, C. & Regazzi, R. Nat. Rev. Endocrinol. advance online publication 30 April 2013; doi:10.1038/nrendo.2013.86

Introduction
Diabetes mellitus affects >350million people worldwide
and makes a considerable contribution to morbidity and
mortality globally.1 In industrialized countries, diabetes
mellitus is the leading cause of blindness, renal failure
and nontraumatic lower limb amputations. Patients
with diabetes mellitus also have an increased risk of
developing cardiovascular disorders and having a stroke,
which means that this disease is a heavy socioeconomic
burden.2 Unfortunately, the prevalence of diabetes mel
litus is increasing at a dramatic pace both in children and
in adults as a result of changes in lifestyle (reduced physi
cal activity and increased overnutrition and obesity),
but also as a consequence of population ageing. Indeed,
according to estimates from the International Diabetes
Federation, 552million people are expected to have
diabetes mellitus in 2030.1
This Review focuses on type1 diabetes mellitus
(T1DM) and type2 diabetes mellitus (T2DM), the two
principal forms of the disease. T1DM is an autoimmune
disorder in which pancreatic cells are attacked and
eliminated by the immune system. During the immune
response, leucocytes infiltrating the pancreatic islets
secrete proinflammatory cytokines that recruit cyto
toxic Tlymphocytes and contribute to cell dysfunction
and death.3 The immune reaction leads to progressive
destruction of cells, which results in severe or com
plete insulin deficiency. T1DM generally develops during
Competing interests
The authors declare no competing interests.

childhood or in young adults and accounts for 58% of

all cases of diabetes mellitus.4 The majority of the other
cases are attributable to T2DM. The pathogenesis of
this disease is closely linked to genetic, environmental
and/or lifestyle factors, such as hypercaloric nutrition,
lack of exercise and obesity. T2DM occurs when target
tissues lose insulin sensitivity, including the liver, skeletal
muscles and adipose tissues. This state of insulin resist
ance can usually be compensated for by expansion of
the functional cell mass and by an increase in insulin
secretion. However, in individuals who are genetically
predisposed to develop T2DM, the cells are unable to
sustain the increased demand for insulin, which leads
tochronic hyperglycaemia and the onset of T2DM.5
Despite intensive research, the causes of T1DM and
T2DM remain incompletely understood and a definitive
cure is still not available. The efficacy of the current treat
ments to delay progression of diabetes mellitus would
be drastically improved if they could be implemented
during the initial phases of the disease and targeted at
individuals with the highest probability of benefitting
from the therapeutic intervention. This goal can only
be achieved by identifying new biomarkers for predict
ing and/or monitoring the progression of T1DM and
T2DM and their long-term complications. The aim of
this Review is to summarize the weaknesses of the blood
parameters and biomarkers that are currently used to
detect people at risk of developing T1DM and T2DM and
to discuss the potential use of circulating microRNAs
(miRNAs) as a novel class of biomarkers.

NATURE REVIEWS | ENDOCRINOLOGY

University of Lausanne,
Department of
Fundamental
Neurosciences, Rue
duBugnon 9, 1005
Lausanne, Switzerland
(C. Guay, R. Regazzi).
Correspondence to:
R. Regazzi
romano.regazzi@unil.ch

ADVANCE ONLINE PUBLICATION | 1

2013 Macmillan Publishers Limited. All rights reserved

REVIEWS
Key points
New biomarkers are needed to improve the identification of individuals at risk
of developing diabetes mellitus and its associated complications, monitor
disease progression and assess the efficacy of therapeutic interventions
Circulating microRNAs (miRNAs) are attractive biomarker candidates as they
can be easily collected, are stable under different storage conditions and can
be measured using assays that are specific, sensitive and reproducible
Pioneering studies have identified characteristic changes in blood levels of
miRNAs in samples from a range of cohorts of patients with diabetes mellitus
However, definitive miRNA signatures for type1 diabetes mellitus, type2 diabetes
mellitus or their associated complications remain to be defined and agreed upon
Although measuring circulating miRNAs is a promising approach in individuals
at risk of developing diabetes mellitus, several key issues still need to
be addressed, including the determination of the most appropriate blood
sampling protocols

Box 1 | Blood parameters and biomarkers for diabetes

T1DM and T2DM
Blood parameters
Glycaemia
HbA1c
T1DM
Autoantigens
Glutamate decarboxylase (GADA)
Insulin (IAA)
Islet cells (ICA)
Tyrosine phosphatases (IA2 and IA2)
Zinc transporter8 (ZnT8)
T2DM
Traditional biomarkers
Cholesterol*
Creatinine*
Free fatty acids (FAA)*
-hydroxybutyrate (-HB)
Lipoproteins (HDL)*
Triacylglycerol (Tg)*
Novel biomarkers
Adipokines*
C-reactive protein (CRP)*
Ferritin*
Incretins (e.g. glucagon-like peptide1)
Linoleoylglycero-phosphocholine (L-GPC)*
*These biomarkers are not specific to T2DM but are also
predictive of other metabolic disorders. Abbreviations: T1DM,
type1 diabetes mellitus; T2DM, type2 diabetes mellitus.

Classic diabetes mellitus biomarkers

According to the WHO, a diagnosis of diabetes
mellitus should be based on measurements of blood
levels of glucose in the fasted state and following an
oralglucose tolerance test.6 A diabetic state is defined
by glucoselevels >7.0mmol/l (126mg/dl) in the fasted
state and >11.1mmol/l (200mg/dl) after an oral glucose
tolerance test.6 Other serum parameters such as levels of
HbA1c or residual Cpeptide can also be helpful in the
diagnosis of diabetes mellitus.

Biomarkers for T1DM

T1DM is generally diagnosed when >8090% of the
pancreatic cells have been destroyed by the immune
system.7 The progression of the disease is slow (itcan take
months to years for the patient to become symptomatic),

which provides a potentially long period of time inwhich

to identify and treat individuals at risk. In the past
2years, progress has been made to preserve the function
of residual cells at the onset of T1DM using immuno
suppressive medications.8,9 However, the efficacy of these
treatments is currently limited, although considerable
improvements could probably be made if therapies could
be initiated at earlier stages of the disease when many
cells are still present.
Autoantibodies against islet antigens are often used
as biomarkers for T1DM, as their presence in the blood
is characteristic of the disease. Several autoantibodies
have been described, but those directed against islet
cells, insulin, tyrosine phosphatase IA2 and IA2,
glutamate decarboxylase and zinc transporter8 are the
most reliable for identifying individuals at risk of devel
oping T1DM (Box1).7,10 However, the use of islet auto
antibodies as biomarkers has some important limitations.
Firstly, autoantibodies appear fairly late in the course
of T1DM, so cannot be used to initiate treatment early
in the disease course. Secondly, although most indivi
duals at the onset of T1DM are positive for at least some
autoantibodies, many autoantibody-positive individuals
will never develop the disease. Thirdly, autoantibodies are
not suitable for monitoring therapeutic outcomes because
they are not immediately removed from the circulation
if the autoimmune reaction is stopped. 10 Additional
biomarkers for T1DM are therefore needed to comple
ment the information obtained from the presence of
autoantibodies and other risk factors (such as age, family
history, susceptibility genes and environmental triggers).

Biomarkers for T2DM

Individuals at risk of developing T2DM are currently
identified by a combination of easily accessible serum
parameters (including levels of glucose, triacylglycerol,
cholesterol, lipoproteins and HbA1c), physical charac
teristics (BMI, waist-to-hip ratio, blood pressure and
sex) and lifestyle factors (food consumption, physical
inactivity and smoking). By combining all these classic
biomarkers and risk factors, the probability of predict
ing the development of the disease ranges from 0.85 to
0.90 in a period of 510years before the onset of T2DM.11
Other molecules have emerged as potentially useful bio
markers (Box1), including incretins such as glucagon-like
peptide1, cytokines, adipokines, ferritin and Creactive
protein.12 Individually, none of these so-called novel
biomarkers can predict the manifestation of T2DM effi
ciently, but in combination they can achieve predictive
values similar to those possible with classic biomarkers.13
All these serum parameters can predict the develop
ment of T2DM a few years in advance of disease mani
festation in individuals already displaying metabolic
alterations. However, the biomarkers are not specific for
diabetes mellitus and cannot be used to assess disease
susceptibility in the general population. Genotype analy
sis could potentially complement the use of these bio
markers for the identification of individuals susceptible
to developing T2DM later in life. However, the predic
tive values of genotypic traits has yet to exceed 0.60. 12

2 | ADVANCE ONLINE PUBLICATION

www.nature.com/nrendo
2013 Macmillan Publishers Limited. All rights reserved

REVIEWS
Thus, early and lifestyle-independent predictive factors
are currently required to enable physicians to recognize
individuals at risk of developing T2DM.

miRNAs
miRNAs are small non-coding RNA molecules of 2123
nucleotides that function as translational repressors by par
tially pairing to the 3' untranslated region of target mRNAs
(Box2). These regulators of gene expression were first dis
covered in Caenorhabditis elegans,14,15 and then in verte
brates and plants.16 According to the latest estimates, the
human genome encodes >1,600 miRNA precursors, gener
ating up to 2,237 mature miRNAs,17 each of which has the
potential to control hundreds of targets. Now, miRNAs are
universally recognized as major regulators of gene expres
sion and key controllers of several biologic and pathologic
processes.18 They are produced from stem-loop precursor
RNAs that are generated from independent transcriptional
units or from introns of genes that encode proteins (Box2).
These primary transcripts (pri-miRNAs) are initially pro
cessed to produce short RNA molecules (pre-miRNAs)
and are then exported to the cytosol where they are
further cleaved to generate the mature forms of miRNAs
(Figure1). The mature miRNAs can either be included in
the RNA-induced silencing complex to guide translational
repression of target mRNAs or bereleasedbythe cells.
Ifthey are to be released, the miRNAs become attached to
proteins or lipoproteins or are loaded inside vesicles that
are released into the extracellular space during plasma
membrane blebbing or after fusion of multivesicular bodies
with the plasma membrane (Figure1).

Role of miRNAs in diabetes pathogenesis

Pancreatic cells and the tissues targeted by insulin
express a well-defined set of miRNAs. Most of the
miRNAs are not cell-specific, but are widely distributed
throughout the tissues of the human body. A notable
exception is miR375, a miRNA highly enriched in
pancreatic islets that regulates the expression of genes
involved in hormone secretion and in cell mass expan
sion in response to insulin resistance.19,20 The miRNA
expression profile of cells and tissues targeted by insulin
is altered in patients with T1DM and T2DM, which prob
ably contributes to the impaired function of these tissues
under disease states.2123 Indeed, the islets of prediabetic
nonobese diabetic (NOD) mice, a model of T1DM,
contain increased levels of several miRNAs, including
miR21, miR34a, miR29 and miR146a, which have
deleterious effects on cell function.24,25 The expres
sion of most of these miRNAs, as well as that of many
other miRNAs, is also altered in the islets of ob/ob and
db/db mice, which are models of obesity and T2DM.26,27
Interestingly, the expression of miR29 and miR34a is
also increased in tissues targeted by insulin in these mouse
models, which possibly contributes to insulin resist
ance.28,29 Other miRNAs that are dysregulated in tissues
targeted by insulin in ob/ob mice, dietary mouse models of
obesity and diabetic Goto-Kakizaki rats include miR143,
miR802 and two closely related miRNAs, miR103 and
miR107.2831 Strong experimental evidence indicates

Box 2 | MicroRNA biogenesis

MicroRNAs (miRNAs) are generated from intergenic genomic sequences or
fromintronic regions of protein-coding genes. Precursors of miRNAs generatedfrom
intronic regions of protein-coding genes are therefore cotranscribed with their
hosting gene.91 Most mammalian miRNAs are transcribed by RNA polymeraseII
as long precursor molecules that contain a characteristic stem-loop structure.92
These primary transcripts are cleaved by the ribonucleaseIII enzyme Drosha to
produce hairpin-structured pre-miRNAs of ~70 nucleotides. These molecules are
then transported via an Exportin5-dependent process into the cytoplasm, where
they are further cleaved by the endoribonuclease Dicer to generate imperfect
duplexes of ~22 nucleotides consisting of a guide strand (miRNA) and apassenger
strand (miRNA*). The guide miRNA strand is incorporated together with Argonaute
proteins into the RNA-induced silencing complex (RISC). Thepassenger miRNA*
strand is usually degraded but can sometimes also be loaded into the RISC
complex and be functional. The majority of the miRNAs are generated by this
canonical pathway, but alternative pathways have now been described.93,94
Mature miRNAs exert their action by guiding the RISC complex to complementary
sequences within the 3' untranslated region of target mRNAs, leading to
translational repression and/or transcript degradation.16 Target recognition is
mainly determined by miRNA complementarity with the so-called seed sequences
(that is, residues 28) of target mRNAs.16,95 Interestingly, a single miRNA can
potentially bind and regulate the expression of hundreds of targets, whereas
asingle 3' untranslated region can be targeted by numerous different miRNAs.95

Drosha

2
Exportin-5

pri-miRNA

pre-miRNA

Dicer

Nucleus

Multivesicular bodies
RISC

4
miRNA

mRNA target

Argonaute-2
HDL
Protein complex

Lipoproteins

Microvesicles

Exosomes

Figure 1 | Biogenesis and release of miRNAs. Pre-miRNAs are generated in the

nucleus by the ribonucleaseIII enzyme Drosha after cleavage of pri-miRNAs (1).
The pre-miRNAs are then transported in the cytoplasm through a process involving
Exportin5 and the GTP-binding protein Ran (2) and further cleaved by Dicer to yield
2123 nucleotide duplexes (3). One strand of the miRNA duplex can either
associate to the RISC complex and guide translational repression of target mRNAs
(4) or be released by the cells. In the latter case, the mature miRNA binds to
RNAbinding proteins such as Argonaute2 (5) or to lipoproteins (6). Alternatively,
the miRNAs can be loaded in microvesicles formed by plasma membrane blebbing
(7) or in exosomes that are released in the extracellular space upon exocytic
fusion of multivesicular bodies with the plasma membrane (8). Abbreviations:
miRNA, microRNA; pre-miRNA, miRNA precursor; pri-miRNA, primary miRNA
transcript; RISC, RNA-induced silencing complex.

a contribution of these miRNAs to the development of

insulin resistance in these obesity models.2931
Changes in the miRNA profile that are related to diabetes
mellitus have also been reported in human tissues. More

NATURE REVIEWS | ENDOCRINOLOGY

ADVANCE ONLINE PUBLICATION | 3

2013 Macmillan Publishers Limited. All rights reserved

REVIEWS

RNA-binding protein
Lipoprotein
Exosome

Blood
vessel

1
HDL

Receiving cells

Fusion
Receptors
Endocytosis

Figure 2 | Blood and other body fluids contain active miRNAs. miRNAs can
bereleased or shed by cells and are found in a stable form in body fluids.
ThemiRNAs present in the blood are associated with protein complexes such
asArgonaute2 or with HDL particles, or are transported inside membrane-bound
vesicles such as exosomes (1). Evidence suggests that circulating miRNAs can be
taken up in active form through different mechanisms, including receptor-mediated
capture, endocytosis or fusion of exosomes with the plasma membrane of
receiving cells (2). Transfer of miRNAs between distantly located cells constitutes
a potentially new communication mode. Abbreviation: miRNA, microRNA.
Box 3 | Exosomes
Exosomes are microvesicles of 30120nm that were first observed to be
released by reticulocytes in the 1980s.9698 Since then, exosomes have been
detected in the culture media of most cell types and are abundant in several
body fluids, such as blood, urine, saliva and breast milk.43 The concentration
of exosomes in the blood has been estimated to be ~3million per l.99 These
vesicles display a characteristic size, form and lipid bilayer structure that can be
visualized by electron microscopy. Exosome markers, such as the tetraspanins
CD63 and CD81 and the Escort proteins Alix and Tsg101, are also routinely used
to distinguish exosomes from other vesicles secreted by cells. Detailed analysis
of the exosome content revealed that they can carry proteins and nucleic acids
(such as mRNAs, microRNAs and long non-coding RNAs).39,100 The interest in
exosomes has increased dramatically in the past 5years, following the discovery
that their cargo can be transferred in active form to recipient cells,39,101,102 which
led to the exciting concept of exosomes as messengers of a new cell-to-cell
communication mode. However, the precise mechanisms leading to exosome
release and/or uptake remain to be elucidated. Exosomes are produced inside
cells via the multivesicular endosomal pathway,97,98 but very little is known about
the selective packaging of nucleic acids and proteins composing the exosome
cargo. An improved understanding of these mechanisms might reveal new biologic
roles for exosomes in the healthy state, but also in pathophysiological conditions.

than 60 differentially expressed miRNAs were detected in

human skeletal muscle biopsy samples from patients with
T2DM, including miR143, which is upregulated, and two
muscle-specific miRNAs, miR206 and miR133a, which
are downregulated.32 Interestingly, the levels of about 15%
of these miRNAs were already modified in individuals
with impaired glucose tolerance, which suggests that the

miRNAs are involved in the early phases of the disease

process. The expression of some of these miRNAs is con
trolled by insulin, but this regulatory mechanism seems to
be impaired in patients with diabetes mellitus.33
As well as the changes in tissues targeted by insulin that
are described above, diabetes mellitus results in consider
able modifications in miRNA expression in blood vessels,
heart, retina and kidneys. This finding indicates that
these non-coding RNAs are involved in the development
oflong-term complications of diabetes mellitus.22,34,35

A functional role for circulating miRNAs?

In addition to regulating gene expression inside the
cells that produce them, several miRNAs are found in
blood and other body fluids in association with proteins,
microvesicles and/or lipoprotein complexes (Figure2).3638
The function of circulating miRNAs remains to be estab
lished, but invitro studies indicate that miRNAs trans
ported by exosomes (Box3) or HDL can be transferred
in an active form to recipient cells.38,39 This observation
raises the intriguing possibility of an involvement of
miRNAs in a novel cell-to-cell communication mode.
Circulating miRNAs are very stable and resistant to treat
ment with ribonucleases, freezing/thawing cycles and
other drastic experimental conditions.40,41 Consequently,
serum or plasma samples can be stored at 20C or
80C for up to several months without notable degra
dation of miRNAs,42 which suggests that these small RNA
molecules are sufficiently robust to serve as biomarkers.
Circulating miRNAs have several other advantages as
potential biomarkers: they are found not only in blood
but also in other easily accessible biologic fluids (such as
urine, saliva, amniotic fluid and breast milk),43 they can be
detected by highly sensitive and specific quantitative realtime PCR, and most of them are evolutionarily conserved,
which facilitates the translation of results obtained from
invivo animal studies to human health care. Moreover,
profiles of the miRNAs in serum of healthy donors are
fairly homogeneous and constant over 24h and miRNAs
can be measured in both serum and plasma.41,44,45

miRNAs as diabetes mellitus biomarkers

The idea of using miRNAs found in blood as biomarkers
is fairly new and was first proposed for detecting various
forms of cancer,41,46,47 autoimmune diseases48 and sepsis.49
Studies have now also analysed the miRNA profile in
serum, plasma or blood cells in an attempt to develop
new approaches to predict the development and pro
gression of diabetes mellitus (Table1). Zampetaki and
colleagues50 were the first to identify a characteristic
expression profile of miRNAs in blood that was related
to T2DM. In their prospective study, they analysed blood
samples from >800 individuals randomly selected from
the Bruneck population (Bolzano Province, Italy) and
identified a subset of five miRNAs (miR15a, miR28-3p,
miR29b, miR126 and miR223) that displayed a charac
teristic deregulation in 80 participants with either pre
diabetes or T2DM. Importantly, the levels of these
miRNAs were already modified 510years before
theonset of the disease, providing initial evidence for

4 | ADVANCE ONLINE PUBLICATION

www.nature.com/nrendo
2013 Macmillan Publishers Limited. All rights reserved

REVIEWS
Table 1 | Blood miRNA changes associated with diabetes mellitus
Study

Study design

Source

Method of analysis

miRNAs identified

Observations

Zampetaki
etal.
(2010)50

800 individuals from

the Bruneck cohort

Plasma

Microarray profiling
(confirmed by qPCR)

miR15a, miR28-3p,
miR29b, miR126 and
miR223

First study to suggest

a blood miRNA signature
for T2DM

Kong etal.
(2011)51

56 individuals with
health conditions
related to diabetes
mellitus

Serum

qPCR on specific
miRNAs

miR9, miR29a, miR30d,

miR34a, miR124a,
miR146a and miR375

Deregulated in T2DM

Karolina
etal.
(2012)52

265 individuals with

health conditions
related to the
metabolic syndrome

Blood

Microarray profiling
(confirmed by qPCR)

miR150, miR192,
miR27a, miR320a
andmiR375

Correlation between
raised levels of fasting
glucose and altered levels
of miR27a and miR320a

Nielsen
etal.
(2012)53

Danish and Hvidoere

cohorts of newly
diagnosed children

Serum

Microarray profiling
(confirmed by qPCR)

miR24, miR25,
miR26a,miR27a,
miR27b, miR29a,
miR-30a-5p, miR148a,
miR152, miR181a,
miR200a and miR210

miR25 negatively
correlated with residual
cell function

Sebastiani
etal.
(2012)54

20 patients newly
diagnosed with
T1DM

Serum

Microarray profiling
(confirmed by qPCR)

miR9, miR31, miR34a,

miR146a, miR155,
miR181a and miR199a

Deregulated in T1DM

Salas-Perez
etal.
(2012)56

20 patients with
T1DM

PBMCs

qPCR on specific
miRNAs

miR21a and miR93

Underexpressed in T1DM

Sebastiani
etal.
(2011)57

19 patients with
T1DM

Lymphocytes

qPCR on specific
miRNAs

miR326

Correlation with islet

autoimmune attack

T2DM

T1DM

Abbreviations: T1DM, type1 diabetes mellitus; T2DM, type2 diabetes mellitus; miRNA, microRNA; PBMC, peripheral blood mononuclear cell; qPCR, quantitative
real-time PCR.

the usefulness of circulating miRNAs as early predictors

of T2DM and its vascular complications.
The miRNA content of serum from patients with pre
diabetes and/or who are newly diagnosed with T2DM
has also been analysed by other groups. Kong and coworkers51 detected an increase in the expression of seven
diabetes-related miRNAs (miR9, miR29a, miR30d,
miR34a, miR124a, miR146a and miR375) in patients
with T2DM compared with patients who had prediabetes
or were susceptible to T2DM. However, no differences
were observed between individuals with normal glucose
tolerance and those with prediabetes, which indicates
that the level of these miRNAs in serum is not suitable
for predicting susceptibility to T2DM. In a study pub
lished in 2012, Karolina etal.52 measured the miRNAs
presentinthe blood and exosomes of 265 patients with
different health conditions associated with the metabolic
syndrome. They detected an upregulation of miR27a,
miR150, miR192, miR320a and miR375 in patients
with T2DM and observed a strong correlation between
raised fasting levels of glucose and the increase in levels
ofmiR27a and miR320a. These pioneering studies
demonstrate the potential of miRNAs as biomarkers for
T2DM. However, the heterogeneity of the results obtained
underscoresthe need for large prospective studies to
identify reliable miRNA signatures for diagnosing T2DM.
A similar approach was used to identify new biomarkers
to predict destruction or regeneration of residual cells

in T1DM. Nielsen and colleagues compared two cohorts

(Danish and Hvidoere) of patients newly diagnosed with
T1DM with an age-matched control group. 53 Global
miRNA sequencing, followed by quantitative real-time
PCR verification and regression analysis to adjust for age,
sex and multiple testing, highlighted a group of miRNAs
(miR24, miR25, miR26a, miR27a, miR27b, miR29a,
miR30a-5p, miR148a, miR152, miR181a, miR200a
and miR210) that were differentially expressed in cohorts
of patients with T1DM compared with control groups.
Several of these miRNAs modulate the expression of genes
involved in apoptosis and/or important cell regulatory
networks.53 Moreover, levels of miR25 were found to cor
relate with residual cell function (determined by levels
ofCpeptide) and adequate glycaemic control (measured
by levels of HbA1c) 3months after disease onset in the
Danish cohort. However, this correlation was not observed
in the Hvidoere cohort, in which glycaemic control was
evaluated 12months after the diagnosis of T1DM, possibly
because of the loss of residual cells.
In a study presented at the 2012 meeting of the European
Association for the Study of Diabetes, Sebastiani and coworkers54 compared the profile of miRNAs in the blood
of 20 patients newly diagnosed with T1DM with that of
healthy control individuals. Of 206 miRNAs detected in
the serum of both groups, 64 were found to be differently
expressed in the patients with T1DM. Interestingly, some
of these miRNAs regulate the functions of immune cells

NATURE REVIEWS | ENDOCRINOLOGY

ADVANCE ONLINE PUBLICATION | 5

2013 Macmillan Publishers Limited. All rights reserved

REVIEWS
(miR31, miR146a, miR155, miR181a and miR199a)
or of cells (miR9 and miR34a). A miRNA abundantly
expressed in the islets of Langerhans, miR375, has been
proposed as a suitable biomarker to detect cell death and
to predict the development of T1DM in animal models.55
Indeed, massive cell loss elicited by administration of
streptozocin caused a dramatic rise in circulating levels
ofthis miRNA.55 Moreover, plasma levels of miR375 were
considerably increased in NOD mice 2weeks before the
onset of T1DM.55 The changes in levels of miR375 con
sequent to cell death were short-lived (<1week). Thus,
these promising findings obtained in mice will need to
be verified in humans, as the decline in cell mass takes
much longer in humans than in NOD mice.
Instead of analysing plasma samples, other studies have
focused their attention on blood cells and measured the
expression of specific miRNAs that are thought to have
important roles in the immune reaction that leads to T1DM.
Salas-Prez etal.56 observed diminished expression of
miR21a and miR93 in peripheral blood mononuclear cells
(PBMC) of patients with T1DM compared with healthy
control individuals. The reduction of miR21a (but not
of miR93) expression could be reproduced by incubating
PBMC from control individuals in a medium containing
25mmol/l of glucose, suggesting that the reduced levels of
miR21a might be the consequence of chronic hypergly
caemia. Finally, Sebastiani and colleagues analysed miRNA
expression in blood lymphocytes from patients with T1DM
and observed a rise in levels of miR326 that correlated with
the timing of the islet autoimmune attack.57 The primary
goal of the latter two studies was to identify miRNAs that
could be involved in the development of T1DM. However,
as a large proportion of miRNAs in serum are released by
blood cells, it is possible that the miRNA changes in PBMC
and/or lymphocytes observed in these two studies might
yield detectable differences in plasma levels that would
enable the autoimmune reaction to be monitored.

Prediction of diabetes mellitus complications

T1DM and T2DM are both associated with long-term
microvascular and macrovascular complications that
can have a devastating effect on quality of life and life
expectancy. The discovery of biomarkers that could
be used to identify individuals at risk of experiencing
serious complications such as retinopathy, nephropathy
or cardiovascular disorders would enable clinicians to
tailor therapeutic approaches and minimize the expected
effects of the disease. However, reliable biomarkers for
these long-term complications are still lacking.
Cardiovascular complications are a major concern in
this group of patients as they account for up to 80% of
premature mortality in patients with diabetes mellitus.58
Prevention, or even a delay, of these complications would
be a huge advancement in the treatment of diabetes
mellitus. As discussed previously, Zampetaki and col
leagues50 identified a unique plasma miRNA signature in
patients with T2DM. Among the miRNA-related char
acteristic changes observed, reduced levels of miR126
showed the strongest association with T2DM, and cor
related with the occurrence of subclinical and overt artery

diseases. Interestingly, another study also reported down

regulation of miR126 in blood samples obtained from
patients with coronary artery disease.59 This miRNA is
highly enriched in endothelial cells, where it has important
roles in cell homeostasis and vascular integrity in differ
ent animal models.60,61 Moreover, the levels of miR126
released in apoptotic bodies are reduced by chronic expo
sure of endothelial cells to raised blood levels of glucose,50
which makes this miRNA an ideal candidate biomarker
for monitoring diabetic vascular complications. Another
miRNA in endothelial cells that deserves further attention
is miR503. The levels of this miRNA are upregulated in
muscle biopsy samples and in peripheral blood-derived
plasma of patients with diabetes mellitus who have limb
ischaemia.62 Interestingly, local inhibition of miR503
in a mouse model of limb ischaemia improved vascular
healing and blood flow recovery. 62 Other circulating
miRNAs have also been suggested as diagnostic markers
for various cardiovascular diseases,63,64 but their use in
predicting or monitoring cardiovascular complications
in patients with diabetes mellitus has yet to be investigated.
Kidney disease affects 30% and 50% of patients with
T1DM and T2DM, respectively.6567 Microalbuminuria has
been proposed as a biomarker that could be used to predict
the occurrence of this important complication; however,
studies published in the past decade have revealed that
anoteworthy proportion of patients with diabetes mellitus
undergo renal failure before microalbuminuria is detect
able, or even before it occurs.6870 Circulating miRNAs
represent a viable alternative for monitoring renal failure
in patients with diabetes mellitus. They are not elimi
nated by haemodialysis71 and have already been tested
in different renal diseases with promising results in both
animal models and human patients.7274 Indeed, a corre
lation was observed between levels of certain circulating
miRNAs, such as miR-16, miR-21, miR-210 and miR-638,
and glomerular filtration rate, a well-known parameter of
the progression of kidney disease.72 Large-scale prospec
tive studies focusing on patients with diabetes mellitus
undergoing renal failure will be necessary to identify
aspecific miRNA profile in plasma or urine that can be
used to predict the appearance of this complication. Urine
represents an ideal source of miRNAs as it can be easily
collected noninvasively and in large amounts. Moreover,
urinary exosomes originate from various cell types that
span the entire urinary track and would be ideally suited
for use in monitoring the progression of renal diseases.7578
Indeed, profiles of miRNAs in urine have been reported
to differ across the stages of diabetic nephropathy,79 sug
gesting that they could be used as tools to follow the pro
gressive alteration of the renal processes in patients with
diabetes mellitus.
To our knowledge, no reports have been published to
date about the use of circulating miRNAs to predict the
occurrence of diabetic retinopathy. However, an involve
ment of some specific miRNAs, such as miR29b and
miR200b, in the development of this complication has
been demonstrated.35 These findings might open the door
to future investigations aimed at assessing the potential
use of miRNAs as predictors of this debilitating condition.

6 | ADVANCE ONLINE PUBLICATION

www.nature.com/nrendo
2013 Macmillan Publishers Limited. All rights reserved

REVIEWS
Gestational diabetes mellitus
Circulating miRNAs could in principle also be used
to identify women at increased risk of developing
gestational diabetes mellitus. Most screening protocols
are currently based on a glucose challenge test performed
around weeks2428 of gestation. Therefore, interven
tions such as diet, exercise or medication are sometimes
started as late as 32weeks of gestation. In a multistage
retrospective study, the miRNAs in serum of women at
1619weeks of gestation were screened. Three miRNAs
(miR29a, miR122 and miR132) were identified that
were deregulated in women developing gestational dia
betes mellitus before changes in blood levels of glucose
became detectable.80 Placental-specific miRNAs can also
be detected in maternal serum.44,81 Thus, it is possible
that gestational diabetes mellitus might lead to changes
in the levels of miRNAs in blood that differ from those
of T1DM or T2DM.

Future directions
The findings described in this article confirm that
miRNAs are attractive novel biomarkers for diabetes
mellitus. Indeed, changes in the levels of a subset of these
small RNA molecules in body fluids promise to provide
new clues for early identification of individuals at risk of
developing diabetes mellitus and the complications asso
ciated with this disorder, for following disease progression
and for assessing the efficacy of therapeutic interventions.
However, major scientific and technical advances need to
be made before these goals can be achieved.
On the basis of the current data, circulating miRNAs
should be able to be used to substitute or complement
other routine measurements in the future. Thus, their
efficacy in predicting the occurrence of diabetes mellitus
or its complications needs to be systematically compared
with biomarkers that are already available. In particu
lar, it will be essential to scrutinize whether miRNAs
are indeed able to provide earlier and/or more precise
detection of individuals at risk of developing the disease
or its long-term complications than current biomarkers.
Thesituation will hopefully evolve in the future; however,
there is currently no obvious advantage to replacing
other traditional biomarkers with measurements of levels
of circulating miRNAs.
Different papers have reported changes in levels of
miRNAs that are associated with diabetes mellitus or its
complications but, for various reasons, we are still very
far from a consensus about the most relevant miRNAs
to be measured. This lack of agreement can in part
be attributed to the heterogeneity of the approaches
selected by investigators. Some of the published studies
carried out systematic profiling of a large number of
miRNAs,50,52,53 whereas others focused on a small group
of selected miRNAs that were supposedly most likely
to be affected in patients with diabetes mellitus.51,5557
These approaches yielded a large number of potentially
interesting candidate biomarkers, but most of them still
awaitconfirmation by independent researchers and/or
in larger cohorts. Standardized protocols for sample
preparation, RNA extraction and miRNA analysis are

urgently needed to facilitate comparisons between differ

ent studies and to reach a consensus about the miRNAs
most suitable to function as early biomarkers of diabetes
mellitus. In fact, the level of certain miRNAs, such as
miR15b and miR16, in plasma is greatly influenced by
the degree of haemolysis and by cellular contamination,
making them less suitable as clinical biomarkers.82
Diabetes mellitus is a complex disorder that involves
major metabolic alterations and important adaptations
in the activity of several organs. miRNAs are being pro
posed as potential biomarkers for an increasing number
of diseases. Therefore, it will be essential to determine
whether the detected miRNA changes are exclusively
indicative of a prediabetic or diabetic condition or if they
are also observed in other physiological or pathological
situations,45 such as in response to modifications in the
nutritional state, inflammation, autoimmunity or cancer.
At present, the origin of miRNAs in the blood is largely
unknown and their levels do not directly mirror changes
in pancreatic cells or tissues targeted by insulin that
occur in diabetes mellitus. miRNAs can be actively or
passively released by a variety of cells and can be carried
by membrane-bound vesicles, protein complexes or lipo
protein particles (Figures1 and 2). So far, the majority of
studies have measured levels of miRNAs directly in the
plasma (or serum). This approach is obviously conveni
ent and straightforward. However, modifications in the
level of miRNAs originating from specific groups of cells
that are not in direct contact with the blood or are not
very abundant are unlikely to have a notable effect on
the pool of miRNAs in the plasma and will probably go
undetected. Although technically demanding, protocols
enabling a specific assessment of the miRNAs carried by
exosomes, protein complexes or lipoproteins will proba
bly provide more detailed information about the physio
logical or pathological status of the organs of interest
than measuring serum levels of miRNAs. Furthermore,
it might be possible to affinity-purify membrane-bound
vesicles originating from a specific group of cells, taking
advantage of the presence of characteristic proteins on
the vesicle surface.12 This approach will be particularly
appropriate for estimating the residual cell mass in
patients newly diagnosed with T1DM. Indeed, except
inthe case of a sudden loss of a large number of cells,
the miRNAs released by insulin-secreting cells represent
only a negligible fraction of all non-coding RNAs circu
lating in the blood. This problem might be partially over
come by searching for miRNAs that are highly expressed
in cells, such as miR375.19,55 However, the expression
of this particular miRNA is modified in a variety of
tumorous cells and several authors have suggested using
circulating levels of miR375 for the diagnosis of differ
ent types of cancer.8386 Several insulin-secreting cell lines
release notable amounts of exosomes8789 and we have
initial evidence indicating that this is also the case for
rodent and human islet cells.90 Protocols permitting an
enrichment in exosomes derived from islet cells would
probably improve the detection of specific changes in
insulin-secreting cells and provide a better evaluation
ofthe functional cell mass.

NATURE REVIEWS | ENDOCRINOLOGY

ADVANCE ONLINE PUBLICATION | 7

2013 Macmillan Publishers Limited. All rights reserved

REVIEWS
Conclusions
miRNAs are emerging as important regulators of gene
expression and central players in many physiological and
pathological processes. Their presence in extracellular
fluids in astonishingly stable forms has led to the idea
of using them as biomarkers for a variety of diseases.
These properties have also attracted the attention of
diabetologists who have initiated the search for miRNAs
that enable early detection of T1DM and T2DM and
their associated complications. A number of scientific
and methodological issues need to be addressed before
circulating miRNAs can attain the status of biomarkers of
diabetes mellitus, but the available data suggest that they
will soon serve as valuable new blood parameters that will
help physicians to refine their therapeutic interventions.
1.

2.

3.

4.

5.

6.

7.

8.

9.

10.
11.

12.

13.

14.

15.

Whiting, D.R., Guariguata, L., Weil, C. & Shaw, J.

IDF diabetes atlas: global estimates of the
prevalence of diabetes for 2011 and 2030.
Diabetes Res. Clin. Pract. 94, 311321 (2011).
Li, R., Zhang, P., Barker, L.E., Chowdhury, F.M.
&Zhang, X. Cost-effectiveness of interventions
to prevent and control diabetes mellitus:
asystematic review. Diabetes Care 33,
18721894 (2010).
Pirot, P., Cardozo, A.K. & Eizirik, D.L. Mediators
and mechanisms of pancreatic beta-cell death
intype1 diabetes. Arq. Bras. Endocrinol.
Metabol. 52, 156165 (2008).
American Diabetes Association. Diagnosis and
classification of diabetes mellitus. Diabetes Care
28 (Suppl. 1), S37S42 (2005).
Prentki, M. & Nolan, C.J. Islet cell failure in
type2 diabetes. J. Clin. Invest. 116, 18021812
(2006).
Stumvoll, M., Goldstein, B.J. & vanHaeften,T.W.
Type2 diabetes: principles of pathogenesis and
therapy. Lancet 365, 13331346 (2005).
Winter, W.E., Harris, N. & Schatz, D. Type1
diabetes islet autoantibody markers. Diabetes
Technol. Ther. 4, 817839 (2002).
Orban, T. etal. Co-stimulation modulation
withabatacept in patients with recent-onset
type1 diabetes: a randomised, double-blind,
placebo-controlled trial. Lancet 378, 412419
(2011).
Sherry, N. etal. Teplizumab for treatment of
type1 diabetes (Protg study): 1year results
from a randomised, placebo-controlled trial.
Lancet 378, 487497 (2011).
Purohit, S. & She, J.X. Biomarkers for type1
diabetes. Int. J. Clin. Exp. Med. 1, 98116 (2008).
Schulze, M.B. etal. Use of multiple metabolic
and genetic markers to improve the prediction
oftype2 diabetes: the EPIC-Potsdam Study.
Diabetes Care 32, 21162119 (2009).
Muller, G. Microvesicles/exosomes as potential
novel biomarkers of metabolic diseases. Diabetes
Metab. Syndr. Obes. 5, 247282 (2012).
Kolberg, J.A. etal. Development of a type2
diabetes risk model from a panel of serum
biomarkers from the Inter99 cohort. Diabetes
Care 32, 12071212 (2009).
Lee, R.C., Feinbaum, R.L. & Ambros, V.
TheC.elegans heterochronic gene lin4 encodes
small RNAs with antisense complementarity
tolin14. Cell 75, 843854 (1993).
Wightman, B., Ha, I. & Ruvkun, G.
Posttranscriptional regulation of the
heterochronic gene lin14 by lin4 mediates
temporal pattern formation in C. elegans.
Cell75,855862 (1993).

Review criteria
We systematically searched PubMed using different
combinations of the following words: diabetes, type1
diabetes, type2 diabetes, gestational diabetes,
diabetes complications, diabetic retinopathy,
diabetic nephropathy, cardiovascular complications,
biomarkers, microRNAs, circulating miRNAs
and exosomes. We also scrutinized the appropriate
references cited in the selected papers. Two abstracts
from the 2012 European Association for the Study of
Diabetes meeting held in Berlin were also included. Owing
to space limitations, not all original articles on the subject
could be listed in the present review. Whenever possible,
the most recent and complete reviews on the topic were
chosen. All selected articles are in English.

16. Bartel, D.P. MicroRNAs: genomics, biogenesis,

mechanism, and function. Cell 116, 281297
(2004).
17. miRBase. The microRNA database [online],
www.mirbase.org (2013).
18. Flynt, A.S. & Lai, E.C. Biological principles of
microRNA-mediated regulation: shared themes
amid diversity. Nat. Rev. Genet. 9, 831842
(2008).
19. Poy, M.N. etal. A pancreatic islet-specific
microRNA regulates insulin secretion.
Nature432, 226230 (2004).
20. Poy, M.N. etal. miR375 maintains normal
pancreatic - and -cell mass. Proc. Natl Acad.
Sci. USA 106, 58135818 (2009).
21. Kumar, M., Nath, S., Prasad, H.K., Sharma, G.D.
& Li, Y. MicroRNAs: a new ray of hope for diabetes
mellitus. Protein Cell 3, 726738 (2012).
22. Shantikumar, S., Caporali, A. & Emanueli, C.
Role of microRNAs in diabetes and its
cardiovascular complications. Cardiovasc.
Res.93, 583593 (2012).
23. Guay, C., Roggli, E., Nesca, V., Jacovetti, C.
&Regazzi, R. Diabetes mellitus, a microRNArelated disease? Transl. Res. 157, 253264
(2011).
24. Roggli, E. etal. Involvement of microRNAs in
thecytotoxic effects exerted by proinflammatory
cytokines on pancreatic -cells. Diabetes 59,
978986 (2010).
25. Roggli, E. etal. Changes in microRNA expression
contribute to pancreatic -cell dysfunction in
prediabetic NOD mice. Diabetes 61, 17421751
(2012).
26. Lovis, P. etal. Alterations in microRNA
expression contribute to fatty acid-induced
pancreatic -cell dysfunction. Diabetes 57,
27282736 (2008).
27. Zhao, E. etal. Obesity and genetics regulate
microRNAs in islets, liver, and adipose of diabetic
mice. Mamm. Genome 20, 476485 (2009).
28. Herrera, B.M. etal. Global microRNA
expressionprofiles in insulin target tissues
inaspontaneous rat model of type2 diabetes.
Diabetologia 53, 10991109 (2010).
29. Trajkovski, M. etal. MicroRNAs 103 and 107
regulate insulin sensitivity. Nature 474,
649653 (2011).
30. Jordan, S.D. etal. Obesity-induced
overexpression of miRNA143 inhibits insulinstimulated AKT activation and impairs glucose
metabolism. Nat. Cell Biol. 13, 434446 (2011).
31. Kornfeld, J.W. etal. Obesity-induced
overexpression of miR802 impairs glucose
metabolism through silencing of Hnf1b. Nature
494, 111115 (2013).

8 | ADVANCE ONLINE PUBLICATION

32. Gallagher, I.J. etal. Integration of microRNA

changes invivo identifies novel molecular
features of muscle insulin resistance in type2
diabetes. Genome Med. 2, 9 (2010).
33. Granjon, A. etal. The microRNA signature
inresponse to insulin reveals its implication
inthe transcriptional action of insulin in human
skeletal muscle and the role of a sterol
regulatory element-binding protein1c/myocyte
enhancer factor2C pathway. Diabetes 58,
25552564 (2009).
34. Kantharidis, P., Wang, B., Carew, R.M.
&Lan,H.Y. Diabetes complications: themicroRNA
perspective. Diabetes 60, 18321837 (2011).
35. Natarajan, R., Putta, S. & Kato, M. MicroRNAs
and diabetic complications. J. Cardiovasc. Transl.
Res. 5, 413422 (2012).
36. Arroyo, J.D. etal. Argonaute2 complexes carry a
population of circulating microRNAs independent
of vesicles in human plasma. Proc. Natl Acad.
Sci. USA 108, 50035008 (2011).
37. Gibbings, D.J., Ciaudo, C., Erhardt, M.
&Voinnet,O. Multivesicular bodies associate
with components of miRNA effector complexes
and modulate miRNA activity. Nat. Cell Biol. 11,
11431149 (2009).
38. Vickers, K.C., Palmisano, B.T., Shoucri, B.M.,
Shamburek, R.D. & Remaley, A.T. MicroRNAs
are transported in plasma and delivered to
recipient cells by high-density lipoproteins.
Nat.Cell Biol. 13, 423433 (2011).
39. Valadi, H. etal. Exosome-mediated transfer of
mRNAs and microRNAs is a novel mechanism
ofgenetic exchange between cells. Nat. Cell Biol.
9, 654659 (2007).
40. Kroh, E.M., Parkin, R.K., Mitchell, P.S.
&Tewari,M. Analysis of circulating microRNA
biomarkers in plasma and serum using
quantitative reverse transcription-PCR (qRTPCR).
Methods 50, 298301 (2010).
41. Mitchell, P.S. etal. Circulating microRNAs
asstable blood-based markers for cancer
detection. Proc. Natl Acad. Sci. USA 105,
1051310518 (2008).
42. Mraz, M., Malinova, K., Mayer, J. &Pospisilova,S.
MicroRNA isolation and stability in stored RNA
samples. Biochem. Biophys. Res. Commun. 390,
14 (2009).
43. Weber, J.A. etal. The microRNA spectrum
in12body fluids. Clin. Chem. 56, 17331741
(2010).
44. Gilad, S. etal. Serum microRNAs are promising
novel biomarkers. PLoS ONE 3, e3148 (2008).
45. Keller, A. etal. Toward the blood-borne miRNome
of human diseases. Nat. Methods 8, 841843
(2011).

www.nature.com/nrendo
2013 Macmillan Publishers Limited. All rights reserved

REVIEWS
46. Chen, X. etal. Characterization of microRNAs
inserum: a novel class of biomarkers for
diagnosis of cancer and other diseases.
CellRes. 18, 9971006 (2008).
47. Lawrie, C.H. etal. Detection of elevated levels
of tumour-associated microRNAs in serum of
patients with diffuse large Bcell lymphoma.
Br.J.Haematol. 141, 672675 (2008).
48. Alevizos, I. & Illei, G.G. MicroRNAs as
biomarkers in rheumatic diseases. Nat. Rev.
Rheumatol. 6, 391398 (2010).
49. Wang, J.F. etal. Serum miR146a and miR223
as potential new biomarkers for sepsis.
Biochem. Biophys. Res. Commun. 394, 184188
(2010).
50. Zampetaki, A. etal. Plasma microRNA profiling
reveals loss of endothelial miR126 and other
microRNAs in type2 diabetes. Circ. Res. 107,
810817 (2010).
51. Kong, L. etal. Significance of serum microRNAs
in pre-diabetes and newly diagnosed type2
diabetes: a clinical study. Acta Diabetol. 48,
6169 (2011).
52. Karolina, D.S. etal. Circulating miRNA profiles
inpatients with metabolic syndrome. J. Clin.
Endocrinol. Metab. 97, E2271E2276 (2012).
53. Nielsen, L.B. etal. Circulating levels of
microRNA from children with newly diagnosed
type1 diabetes and healthy controls: evidence
that miR25 associates to residual beta-cell
function and glycaemic control during disease
progression. Exp. Diabetes Res. 2012, 896362
(2012).
54. Sebastiani, G. etal. MicroRNA expression
fingerprint in serum of type1 diabetic patients.
Diabetologia 55, S48 (2012).
55. Erener, S., Mojibian, M., Fox, J.K., Denroche,H.C.
& Kieffer, T.J. Circulating miR375 as a biomarker
of -cell death and diabetes in mice. Endocrinology
154, 603608 (2013).
56. Salas-Perez, F. etal. MicroRNAs miR21a and
miR93 are down regulated in peripheral blood
mononuclear cells (PBMCs) from patients with
type1 diabetes. Immunobiology 218, 733737
(2013).
57. Sebastiani, G. etal. Increased expression of
microRNA miR326 in type1 diabetic patients
with ongoing islet autoimmunity. Diabetes Metab.
Res. Rev. 27, 862866 (2011).
58. Winer, N. & Sowers, J.R. Epidemiology of
diabetes. J. Clin. Pharmacol. 44, 397405
(2004).
59. Fichtlscherer, S. etal. Circulating microRNAs in
patients with coronary artery disease. Circ. Res.
107, 677684 (2010).
60. Fish, J.E. etal. miR126 regulates angiogenic
signaling and vascular integrity. Dev. Cell 15,
272284 (2008).
61. Wang, S. etal. The endothelial-specific microRNA
miR126 governs vascular integrity and
angiogenesis. Dev. Cell 15, 261271 (2008).
62. Caporali, A. etal. Deregulation of microRNA503
contributes to diabetes mellitus-induced
impairment of endothelial function and
reparative angiogenesis after limb ischemia.
Circulation 123, 282291 (2011).
63. Creemers, E.E., Tijsen, A.J. & Pinto, Y.M.
Circulating microRNAs: novel biomarkers and
extracellular communicators in cardiovascular
disease? Circ. Res. 110, 483495 (2012).
64. van Empel, V.P., De Windt, L.J.
&daCostaMartins, P.A. Circulating miRNAs:
reflecting oraffecting cardiovascular disease?
Curr. Hypertens. Rep. 14, 498509 (2012).
65. Pambianco, G. etal. The 30-year natural history
of type1 diabetes complications: the Pittsburgh
Epidemiology of Diabetes Complications Study
experience. Diabetes 55, 14631469 (2006).

66. Parving, H.H., Lewis, J.B., Ravid, M.,

Remuzzi,G. & Hunsicker, L.G. Prevalence and
risk factors for microalbuminuria in a referred
cohort of typeII diabetic patients: a global
perspective. Kidney Int. 69, 20572063 (2006).
67. Thomas, M.C., Groop, P.H. & Tryggvason, K.
Towards understanding the inherited
susceptibility for nephropathy in diabetes. Curr.
Opin. Nephrol. Hypertens. 21, 195202 (2012).
68. Macisaac, R.J. & Jerums, G. Diabetic kidney
disease with and without albuminuria. Curr. Opin.
Nephrol. Hypertens. 20, 246257 (2011).
69. Perkins, B.A. etal. Microalbuminuria and the
risk for early progressive renal function decline
in type1 diabetes. J. Am. Soc. Nephrol. 18,
13531361 (2007).
70. Perkins, B.A. etal. Regression of
microalbuminuria in type1 diabetes. N. Engl.
J.Med. 348, 22852293 (2003).
71. Martino, F. etal. Circulating microRNAs are
noteliminated by hemodialysis. PLoS ONE
7,e38269 (2012).
72. Neal, C.S. etal. Circulating microRNA
expression is reduced in chronic kidney
disease. Nephrol. Dial. Transplant 26,
37943802 (2011).
73. Wang, G. etal. Elevated levels of miR146a
andmiR155 in kidney biopsy and urine from
patients with IgA nephropathy. Dis. Markers 30,
171179 (2011).
74. Wang, N. etal. Urinary microRNA10a and
microRNA30d serve as novel, sensitive and
specific biomarkers for kidney injury. PLoS ONE
7, e51140 (2012).
75. Alvarez, M.L. & Distefano, J.K. The role of noncoding RNAs in diabetic nephropathy: potential
applications as biomarkers for disease
development and progression. Diabetes Res.
Clin. Pract. 99, 111 (2013).
76. Miranda, K.C. etal. Nucleic acids within
urinaryexosomes/microvesicles are potential
biomarkers for renal disease. Kidney Int. 78,
191199 (2010).
77. van Balkom, B.W., Pisitkun, T., Verhaar, M.C.
&Knepper, M.A. Exosomes and the kidney:
prospects for diagnosis and therapy of renal
diseases. Kidney Int. 80, 11381145 (2011).
78. Beltrami, C., Clayton, A., Phillips, A.O.,
Fraser,D.J. & Bowen, T. Analysis of urinary
microRNAs in chronic kidney disease. Biochem.
Soc. Trans. 40, 875879 (2012).
79. Argyropoulos, C. etal. Urinary microRNA profiling
in the nephropathy of type1 diabetes. PLoS ONE
8, e54662 (2013).
80. Zhao, C. etal. Early second-trimester serum
miRNA profiling predicts gestational diabetes
mellitus. PLoS ONE 6, e23925 (2011).
81. Chim, S.S. etal. Detection and characterization
of placental microRNAs in maternal plasma.
Clin.Chem. 54, 482490 (2008).
82. McDonald, J.S., Milosevic, D., Reddi, H.V.,
Grebe, S.K. & Algeciras-Schimnich, A.
Analysisof circulating microRNA: preanalytical
and analytical challenges. Clin. Chem. 57,
833840 (2011).
83. Bryant, R.J. etal. Changes in circulating
microRNA levels associated with prostate
cancer. Br. J. Cancer 106, 768774 (2012).
84. Komatsu, S. etal. Circulating microRNAs in
plasma of patients with oesophageal squamous
cell carcinoma. Br. J. Cancer 105, 104111
(2011).
85. Li, L.M. etal. Serum microRNA profiles serve
asnovel biomarkers for HBV infection and
diagnosis of HBV-positive hepatocarcinoma.
Cancer Res. 70, 97989807 (2010).
86. Madhavan, D. etal. Circulating miRNAs as
surrogate markers for circulating tumor cells and

NATURE REVIEWS | ENDOCRINOLOGY

prognostic markers in metastatic breast cancer.

Clin. Cancer Res. 18, 59725982 (2012).
87. Lee, H.S., Jeong, J. & Lee, K.J. Characterization
of vesicles secreted from insulinoma NIT1 cells.
J. Proteome Res. 8, 28512862 (2009).
88. Palmisano, G. etal. Characterization of membraneshed microvesicles from cytokine-stimulated
-cells using proteomics strategies. Mol. Cell
Proteomics 11, 230243 (2012).
89. Sheng, H. etal. Insulinoma-released exosomes
or microparticles are immunostimulatory and
can activate autoreactive Tcells spontaneously
developed in nonobese diabetic mice.
J.Immunol. 187, 15911600 (2011).
90. Guay, C., Menoud, V., Gattesco, S. & Regazzi, R.
MicroRNA transfer as a new celltocell
communication mode between pancreatic
cells. Diabetologia 55, S95 (2012).
91. Rodriguez, A., Griffiths-Jones, S., Ashurst, J.L.
&Bradley, A. Identification of mammalian
microRNA host genes and transcription units.
Genome Res. 14, 19021910 (2004).
92. Lee, Y. etal. MicroRNA genes are transcribed
byRNA polymeraseII. EMBO J. 23, 40514060
(2004).
93. Cifuentes, D. etal. A novel miRNA processing
pathway independent of Dicer requires
Argonaute2 catalytic activity. Science 328,
16941698 (2010).
94. Ruby, J.G., Jan, C.H. & Bartel, D.P. Intronic
microRNA precursors that bypass Drosha
processing. Nature 448, 8386 (2007).
95. Doench, J.G. & Sharp, P.A. Specificity of
microRNA target selection in translational
repression. Genes Dev. 18, 504511 (2004).
96. Trams, E.G., Lauter, C.J., Salem, N.Jr &Heine,U.
Exfoliation of membrane ecto-enzymes in the
form of micro-vesicles. Biochim. Biophys. Acta
645, 6370 (1981).
97. Pan, B.T., Teng, K., Wu, C., Adam, M.
&Johnstone, R.M. Electron microscopic
evidence for externalization of the transferrin
receptor in vesicular form in sheep reticulocytes.
J. Cell Biol. 101, 942948 (1985).
98. Harding, C., Heuser, J. & Stahl, P. Receptormediated endocytosis of transferrin and
recycling of the transferrin receptor in rat
reticulocytes. J. Cell Biol. 97, 329339 (1983).
99. Vlassov, A.V., Magdaleno, S., Setterquist, R.
&Conrad, R. Exosomes: current knowledge
oftheir composition, biological functions, and
diagnostic and therapeutic potentials. Biochim.
Biophys. Acta 1820, 940948 (2012).
100. Johnstone, R.M., Adam, M., Hammond, J.R.,
Orr, L. & Turbide, C. Vesicle formation during
reticulocyte maturation. Association of plasma
membrane activities with released vesicles
(exosomes). J. Biol. Chem. 262, 94129420
(1987).
101. Kosaka, N. etal. Secretory mechanisms and
intercellular transfer of microRNAs in living cells.
J. Biol. Chem. 285, 1744217452 (2010).
102. Zhang, Y. etal. Secreted monocytic miR150
enhances targeted endothelial cell migration.
Mol. Cell 39, 133144 (2010).
Acknowledgements
The authors are supported by grants from the Swiss
National Science Foundation, from the European
Foundation for the Study of Diabetes and from
theSocit Francophone du Diabte (SFD)-Servier.
C.Guay is supported by a fellowship from Fonds
delaRecherche en Sant du Qubec, the SFD
andthe Canadian Diabetes Association.
Author contributions
Both authors contributed equally to all aspects
ofthisarticle.

ADVANCE ONLINE PUBLICATION | 9

2013 Macmillan Publishers Limited. All rights reserved