Insulin resistance and insulin sensitizing agents

Insulin resistance and insulin sensitizing agentsContents lists available at ScienceDirect
Metabolism Clinical and Experimental
j ourna l homepage: www.metabo l i smjourna l .com
Articles from the An insulin centennial: Past, Present, and Future Special Issue, Edited by Alexander Kokkinos and Eleuterio Ferrannini
Insulin resistance and insulin sensitizing agents
Lucia Mastrototaro a,b, Michael Roden a,b,c, a Institute for Clinical Diabetology, German Diabetes Center, Leibniz Center for Diabetes Research at Heinrich-Heine University Düsseldorf, Düsseldorf, Germany b German Center for Diabetes Research, Partner Düsseldorf, München-Neuherberg, Germany c Department of Endocrinology and Diabetology, Medical Faculty and University Hospital, Heinrich-Heine University Düsseldorf, Düsseldorf, Germany
Corresponding author at: Department of Endocrino Faculty and University Hospital, Heinrich-Heine Unive Diabetes Center at Heinrich-Heine University, Auf Düsseldorf, Germany.
E-mail address: [email protected] (M. Roden).
https://doi.org/10.1016/j.metabol.2021.154892 0026-0495/© 2021 The Authors. Published by Elsevier Inc
a b s t r a c t
a r t i c l e i n f o
Article history: Received 23 June 2021 Received in revised form 8 September 2021 Accepted 20 September 2021
Insulin resistance is a common feature of obesity and type 2 diabetes, but novel approaches of diabetes subtyping (clustering) revealed variable degrees of insulin resistance in peoplewith diabetes. Specifically, the severe insulin resistant diabetes (SIRD) subtype not only exhibits metabolic abnormalities, but also bears a higher risk for car- diovascular, renal and hepatic comorbidities. In humans, insulin resistance comprises dysfunctional adipose tis- sue, lipotoxic insulin signaling followed by glucotoxicity, oxidative stress and low-grade inflammation. Recent studies show that aside from metabolites (free fatty acids, amino acids) and signaling proteins (myokines, adipokines, hepatokines) also exosomes with their cargo (proteins, mRNA and microRNA) contribute to altered crosstalk between skeletal muscle, liver and adipose tissue during the development of insulin resistance. Reduc- tion of fat mass mainly, but not exclusively, explains the success of lifestyle modification and bariatric surgery to improve insulin sensitivity. Moreover, some older antihyperglycemic drugs (metformin, thiazolidinediones), but also novel therapeutic concepts (new peroxisome proliferator-activated receptor agonists, incretinmimetics, so- dium glucose cotransporter inhibitors, modulators of energy metabolism) can directly or indirectly reduce insu- lin resistance. This review summarizes molecular mechanisms underlying insulin resistance including the roles of exosomes and microRNAs, as well as strategies for the management of insulin resistance in humans. © 2021 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND license (http://
creativecommons.org/licenses/by-nc-nd/4.0/).
1. Introduction
Insulin resistance (IR) describes an impairment of insulin sensitivity, reflected by a shift of the insulin concentration-effect curve towards higher insulin concentrations [1]. Of note, common chronic IR, as ob- served in obesity and type 2 diabetes mellitus (T2DM), mostly also in- cludes an impairment of insulin responsiveness, reflected by a reduction of the maximal insulin effect. In humans, IR is generally de- fined by the inability of insulin-target tissues to adequately dispose blood glucose, suppress endogenous glucose production (EGP) and li- polysis as well as to stimulate glycogen synthesis at elevated plasma in- sulin concentrations [1]. In contrast to common chronic IR, conditions of increasedmetabolic demand, such as fasting, dehydration, stress and in- fection, can induce a state of reversible IR mediated by stress hormones and pro-inflammatory cytokines, which inhibit insulin action in target tissues and promote energy mobilization [2]. Finally, mutations in the insulin signaling cascade can cause genetic forms of IR, such as leprechaunism, Rabson–Mendenhall syndrome and type A insulin
logy and Diabetology, Medical rsity, Düsseldorf, c/o German dem Hennekamp 65, 40225
. This is an open access article under
resistance syndrome [3]. These specific genetic causes and further epi- genetic mechanisms possibly underlying IR [4] are beyond the scope of this review.
Common IR mainly results from an imbalance of energy intake and expenditure [5], although genetic predisposition is also involved [6]. The still rising prevalence of obesity and T2DM [7] requires the assess- ment of IR not only for refined phenotyping, but even more to enable stratified prevention and treatment. The hyperinsulinemic-euglycemic clamp test (HEC) is the gold standardmethod for quantifying in vivo in- sulin sensitivity, which is calculated as glucose disposal during steady state and expressed as M-value or rate of glucose disappearance (Rd) [1]. Under these conditions, skeletalmuscle (SkM) takes up themajority of glucose so that whole-body or peripheral insulin sensitivity by M- value and Rdmainly reflect SkM insulin sensitivity [1]. In addition, min- imal model analysis of the frequently sampled intravenous or oral glucose tolerance tests allow to assess insulin sensitivity (e.g. Si, Matsuda and oral glucose insulin sensitivity indices) during dynami- cally changing glycemia and insulinemia. For larger studies, surrogate indices of insulin sensitivity have been developed from fasting insulin, C-peptide and glucose levels, such as the quantitative insulin sensitivity check index and the homeostatic model assessment IR (HOMA-IR) [1], which although assessed under differentmetabolic conditions correlate reasonably well with clamp-derived measures [8].
the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
L. Mastrototaro and M. Roden Metabolism Clinical and Experimental 125 (2021) 154892
Under typical clamp conditions, which simulate postprandial hyperinsulinemia, the IR observed in people with T2DM and their insulin-resistant offspring results mainly from reduced non-oxidative glucose disposal due to lower SkMglycogen synthesis, as a consequence of lower glucose uptake [5]. The lower glucose transport is caused by ab- normal insulin signaling via the insulin receptor kinase (IRK) and insu- lin receptor substrate 1 (IRS1) [5]. Of note, a recent study showed that the pattern of IRS1 phosphorylation is not uniformly altered in all T2DM individuals, suggesting that other post-translational modifica- tions likely contribute to the attenuation of insulin signaling [9]. In white adipose tissue (WAT), IRmay be a very early event during the de- velopment of T2DM, triggered by invasion of macrophages with subse- quent local release of pro-inflammatory cytokines, which inhibit IRK activity [10]. Dysfunctional insulin resistant WAT exhibits excessive li- polysis with release of free fatty acids (FFA), which enter other tissues such as the liver and form lipid toxic intermediates, resulting in inhibi- tion of IRK and hepatic IR with subsequently elevated gluconeogenesis, reduced glycogen synthesis and higher EGP [5]. Higher cellular glucose availability also favors synthesis of fatty acids (FA) via de novo lipogen- esis (DNL), which augments intrahepatic lipid (IHL) content and corre- lates negativelywith insulin sensitivity [5,11]. In addition tometabolites and cytokines also gut microbiota can contribute to IR. Gut dysbiosis is associated with the disruption of tight-junctions and the translocation of bacterial lipopolysaccharides (LPS) into the systemic circulation, resulting in the development of inflammation and IR [12]. Conversely, gut microbiota-derived short-chain FA could prevent obesity and T2DM by reducing inflammation and increasing lipid storage capacity by WAT [13].
2. Mechanisms underlying the development of insulin resistance
Common IR associates with defective insulin signaling mediated by several mechanisms in humans, including accumulation of specific lipid mediators, abnormal features of mitochondrial function as well as increases in stress-activated protein c-Jun-N-terminal-kinase (JNK) and inflammatory pathways (Fig. 1) [5].
2.1. Lipid-induced insulin resistance
Diacylglycerols (DAG) and ceramides have been largely studied as mediators of lipid-induced IR in liver and SkM [13]. During lipid over- supply from high-fat high-calorie nutrition or excessive adipose lipoly- sis, FFA accumulate ectopically exceeding the rate of intracellular FA oxidation (FAO) and storage [5].Moreover, reducedmitochondrial den- sity and function in insulin-resistant people impede FAO and exacerbate lipid-induced IR [5].
Chronic elevation of certain DAG species impairs insulin signaling via activation of conventional (α, βI, βII, γ) and novel protein kinase C (nPKC) isoforms (δ, ε, ν, θ). DAG localized in the plasma membrane (PM) activate nPKCs, whereas sequestration of DAG and PKCs in lipid droplets may protect from IR [5]. Accumulation of sn-1,2-DAG in the PM activates PKCε and inhibitory IRK Thr1160-phosphorylation in the liver [14] as well as in WAT at least in mice [15], resulting in IR. Besides PKCε, PKCδ can be also increased in livers of obese humans and may in- duce hepatic IR via decreased Tyr612-phosphorylation of IRS1 and Ser473-phosphorylation of protein kinase B (Akt) [16]. In SkM of obese, T2DM or lipid-infused individuals, DAG activates PKCθ, which in turns causes IR via inhibitory Ser1101-phosphorylation of IRS1 [5,17]. Addi- tionally, lipid infusion-induced elevation of DAG content in SkM of healthy men can activate PKCβII, PKCδ and nuclear factor (NF)-κB in- flammatory pathway [18].
The role of intramyocellular ceramides in IR is controversial, indeed not all studies confirmed a relationship between SkM IR and ceramides in human obesity and T2DM [17,19]. Nevertheless, hepatic total ceramides and certain dihydroceramides (e.g. C16:0) are increased in insulin-resistant humans with non-alcoholic steatohepatitis (NASH)
2
and correlate positively with hepatic oxidative stress and inflammation [20]. Interestingly, the ratio of ceramide/dihydroceramide 16:0 posi- tively associateswithmarkers of inflammation in visceral adipose tissue (VAT), whereas it relates negatively with IHL, mitochondrial capacity and lipid peroxidation in liver [21]. These data suggest tissue-specific roles for these lipid species without supporting an involvement of cer- amide C16:0 in obesity-induced IR in humans, as shown in previous studies [22].
2.2. Abnormal features of mitochondrial function
Impaired muscle mitochondrial functionality, which includes mito- chondrial dynamics, turnover and plasticity, and lower content are fre- quent features of the elderly and people with IR or overt T2DM [23]. Downregulation of peroxisome proliferator-activated receptor (PPAR)γ coactivator 1-α (PGC-1α) suggests altered mitochondrial bio- genesis in T2DM, whereas reduced expression of lipoprotein lipase (LPL) and PPARδ is responsible for decreased mitochondrial content in SkM of nondiabetic insulin-resistant offspring [24]. Mitochondrial fu- sion is also decreased in SkM of obese individuals and the transcript levels of Mfn2 correlate positively with markers of insulin sensitivity [23]. Decreasedmitochondrial content and transcript levels of oxidative phosphorylation genes have been demonstrated in WAT of obese and T2DM persons and might contribute to the development of systemic IR [23]. In contrast, hepatic mitochondrial respiration can be upregu- lated in obesity with and without steatosis, but decreases in NASH, de- spite cellular substrate excess and higher mitochondrial content [25], which might be a consequence of accumulation of dysfunctional mito- chondria rather than increased biogenesis [26].
2.3. Endoplasmic reticulum (ER) stress
Although lipids likely play the primary role in mediating IR and in- flammation [27], not all insulin-resistant individuals display elevated FFA concentrations. Evidence from human studies point to enhanced ER stress and JNK activation in liver and in WAT of obese individuals and to a positive correlation with BMI and percent fat [28,29]. Indeed, ER stress is reduced after surgically-induced body weight (BW) loss, underlining that ER stress affects insulin action in obesity [29]. Con- versely, the use of chemical chaperones, which enhance protein folding and protectmice against ER stress, improves hepatic andmuscle insulin sensitivity, without affectingWAT insulin sensitivity and the expression of ER stress markers in obese humans [30].
2.4. Low-grade inflammation
Certain metabolites (e.g. cholesterol), gut-derived LPS and bacteria activate the resident hepatic macrophages which release pro- inflammatory cytokines, leading to hepatic inflammation and systemic IR [31]. A mouse study reveals that the liver-specific transgenic expres- sion of NF-κB in absence of steatosis and adiposity inhibits insulin- stimulated suppression of EGP and reduces glucose uptake and glyco- gen synthesis in SkM, suggesting a crosstalk between liver and muscles mediated by hepatic cytokines [32].
Emerging evidences reveal that inflammation occurs also directly in SkM in obesity [33], probably as a consequence of excessive lipid accu- mulation. Mechanistically, lipid oversupply activates NF-κB in myotubes, resulting in reduced mitochondrial respiratory capacity and increased reactive oxygen species (ROS) production, mitochondrial fragmentation and mitophagy [34].
Obesity-related IR is accompanied by macrophage infiltration and inflammatory cytokines production in WAT, which activate JNK and NF-κB causing local and systemic IR [35]. However, WAT IR can exist without concomitant macrophage infiltration and inflammation, as shown in healthy humans after oral lipid ingestion [36], suggesting that IR likely precedes systemic low-grade inflammation.
Fi g.
cr os st al k un
de rl yi ng
H ig h en
hi ch
w it h el ev
at io n of
G LP
-1 an
d SC
ic ci rc ul at io n,
an d ex
st or ag
le ad
s to
tr it io n or
ba ri at ri c su
rg er y is th e m os te
ffe ct iv e st ra te gy
to im
pr ov
e W
m et fo rm
lF X -R
ia af fe ct in g gu
t m ic ro bi ot a. M et fo rm
in an
G LP
-1 ,w
hi ch
m ed
d BW
ip on
d fa vo
uc ed
FF A re le as e. M ac ro ph
ag es
tr ig ge
ce an
d su
bs eq
u en
pr o- in fla
to ki ne
an d ex
cr os st al k.
Se ve
ra lp
ha rm
PA R α /δ
re du
d in su lin
re si st an
s, su
ch as
d an
ta go
re as e pl as m a co
nc en
id ac cu
Th e PP
e an
ci rc ul at in g FF A le ve
ls .O
er ,t he
y fo rm
to xi c in te rm
ed ia te s (s n- 1, 2- D A G an
d Ce
r) ,w
hi ch
in su
e, re su
bs eq
rt ed
to FA
co nt ri bu
in de
cr ea
of G PD
ob es it y- re la te d in su
lin re si st an
ce ,h
H ep
at ic
in su
ic al
ag en
in no
nh um
yo ce llu
la r sn
-1 ,2 -D
sp ho
in su
gl uc
de cr ea
se d Sk
se s in su
en ha
up ta ke
m it oc ho
bi og
en es is .F ur th er m or e, N M N an
d M et
T4 tr an
br an
e by
en ha
ph os ph
d A M PK
re sp
se B;
te d pr ot ei n ki na
se ;B
CA A =
br an
ch ed
-c ha
in am
= bo
er =
RM P =
ho nd
ri al
op ho
re ;D
A G =
di ac yl gl yc er ol s; D N L =
de no
vo lip
og en
lu m ;F
O =
FF A =
fr ee
F 21
LP -1 R =
PD 2 =
sp ha
te de
IH L =
1 =
su bs tr at e 1;
K H K =
ke to he
xo ki na
id e m on
ra to r- ac ti va
te d
TP 1B
ph os ph
(S )E V =
tr ac el lu la r ve
si cl es ;S
Sk M
1, 2- di ac yl gl yc er ol s;
TN Fα
α ;T
.
L. Mastrototaro and M. Roden Metabolism Clinical and Experimental 125 (2021) 154892
3
2.5. Amino acids
Several plasma metabolites, including phospholipids, ketoacids and amino acids (AA) contribute to the development of IR and are increased in T2DM [13]. Elevation of plasma AA impairs insulin-stimulated glu- cose disposal in SkM, likely via overactivation of the mammalian target of rapamycin (mTOR)/S6 kinase pathway and phosphorylation of IRS1 [37]. Branched-chain (BC) AA are important for protein and glucoseme- tabolism, and their plasma concentration positively correlate with IR andpredict the development of T2DM innormoglycemic adults [38]. Al- thoughBCAA are precursor of hepatic gluconeogenesis, a short-term ex- posure to increased plasma BCAA does not affect suppression of EGP [39]. In SkM, BCAA impair glucose disposal during hyperinsulinemia and increase maximal ATP synthesis without affecting mitochondrial DNA abundance [40,41]. Conversely, a short-term dietary reduction of BCAA decreases postprandial insulin secretion, improves WAT metabo- lism and increases the Bacteroidetes in the gut microbiota of T2DM in- dividuals [42].
3. Other mediators of interorgan crosstalk
In addition to metabolites, cytokines (adipokines, myokines, hepatokines), microRNAs or exosomes (Exo), which carry different bio- logically active cargo, can contribute to the development of IR [13,43–46].
3.1. Cytokines
WAT secretes several adipokines, which participate to the interor- gan crosstalk and are associatedwith IR (reviewed in [44]). Adiponectin and omentin are anti-inflammatory adipokines, whose levels are in- versely associated with IR [44]. The main pro-inflammatory cytokines, tumor necrosis factor (TNF)α and interleukin (IL)-6, are elevated in obesity and T2DM [47]. The infusion of IL-6 increases insulin- mediated glucose uptake in healthy but not in T2DM humans [48], in- stead TNFα administration in healthy humans induces SkM IR by acti- vating JNK [49]. Conversely, IL-10 negatively correlates with prevalence of T2DM, BMI and body fat, and positively correlateswith in- sulin sensitivity in humans [50].
SkM can release myokines, which either exert their effects within SkM or in other tissues and may protect from detrimental effects of adipokines [51]. SkM-derived IL-6 during exercise enhances EGP and SkM lipolysis, but it also triggers the production of anti-inflammatory cytokines and hinders the release of TNFα resulting in reduced systemic inflammation and risk of IR [51]. IL-13, IL-15 and irisin, recently de- scribed as myokines, are negatively associated with IR, obesity and T2DM, and increased after exercise, as reviewed in [52,53].
Proteomic studies revealed that the liver also secretes cytokines (hepatokines), which can contribute to IR and inflammation (e.g. fetuin A, hepassocin, selenoprotein P) or improve IR (e.g. adropin, sex hormone-binding globulin) [54]. Fibroblast growth factor (FGF)21 has complex effects, as it positively associates with T2DM and obesity, but administration of its analogue reduces BW and increases adiponectin in T2DM [54].
3.2. MicroRNA
MicroRNAs are small-non coding RNA molecules that regulate glu- cose homeostasis through the modulation of several components of in- sulin signaling pathway (Fig. 2) and might be used as biomarkers for monitoring the development and progression of metabolic disease in humans, as described in this section (Table 1) [46,55]. MicroRNA-34a, key component of adipocyte-secreted Exo, inhibits a shift of macro- phage to M2 anti-inflammatory phenotype and its expression in VAT of overweight/obese humans correlates positively with parameters of IR and systemic inflammation [56]. The levels of several extracellular
4
microRNAs differ betweenobese and lean individuals and are correlated with metabolic parameters like circulating FFA, HbA1c and HOMA-IR [57]. MicroRNA-122 is associated with IR, inflammation and adiposity in overweight adults and regulates insulin signaling pathways, as re- vealed by functional analysis [58]. Moreover, the exosomal microRNA profile correlates with BMI, transaminases and uric acid in children with non-alcoholic fatty liver disease (NAFLD) [59] and differs between healthy and BMI-matched T2DM humans, who display higher levels of microRNA-20b-5p, which modulates the Akt signaling when transfected in SkM cells [60]. People with T2DM display also upregula- tion of microRNA-144 and downregulation of its predicted target IRS1, which supports the relevance of microRNA-144 for T2DM development [61]. The levels of circulating microRNAs are affected by surgically- induced BW loss, exercise and glucose lowering treatment [62–66], sug- gesting a new potential use of microRNA as predictive biomarkers for monitoring therapy response and moving towards precision medicine for IR prevention.
3.3. Exosomes
There is growing evidence that extracellular vesicles (EV), in partic- ular Exo derived from endosomes (Table 1, Fig. 2), are also involved in the pathophysiology of obesity and IR.WAT in vivo secrete Exo contain- ing adipocyte-molecules, whose expression and abundance are regu- lated by obesity [67]. The metabolic effects resulting from EV released by WAT and WAT-macrophages of obese mice have been investigated by injecting them in leanmice, where they cause IR, glucose intolerance and inflammation [68,69]. Conversely, anti-inflammatory M2 macro- phages secrete Exo, which improve glucose tolerance and insulin sensi- tivity in obesemice [70]. The deletion of AMPKα1 inWAT enhances the release of Exo, which exacerbate the lipid deposition and inflammation when injected in hepatocytes, suggesting a role of Exo…