Critical role of Lin28‐TNFR2 signalling in cardiac stem cell activation and differentiation

Abstract Tumour necrotic factor receptor‐2 (TNFR2) has been to be cardiac‐protective and is expressed in cardiac progenitor cells. Our goal is to define the mechanism for TNFR2‐mediated cardiac stem cell activation and differentiation. By employing a protocol of in vitro cardiac stem cell (CSC) differentiation from human inducible pluripotent stem cell (hiPSC), we show that expression of TNFR2 precedes expression of CSC markers followed by expression of mature cardiomyocyte proteins. Activation of TNFR2 by a specific agonist promotes whereas inhibition of TNFR2 by neutralizing antibody diminishes hiPSC‐based CSC differentiation. Interestingly, pluripotent cell factor RNA‐binding protein Lin28 enhances TNFR2 protein expression in early CSC activation by directly binding to a conserved Lin28‐motif within the 3'UTR of Tnfr2 mRNA. Furthermore, inhibition of Lin28 blunts TNFR2 expression and TNFR2‐dependent CSC activation and differentiation. Our study demonstrates a critical role of Lin28‐TNFR2 axis in CSC activation and survival, providing a novel strategy to enhance stem cell‐based therapy for the ischaemic heart diseases.

that stem cell-based therapies could improve cardiac function, attenuated matrix remodelling, decrease infarct size and improve haemodynamic parameters in animal models and even in clinical trials.
These two clinical trials have been reported [3][4][5][6] . However, many hurdles have to be overcome before this strategy becomes practical.
These hurdles include generating sufficient number of cardiac stem cell (CSC) and mature cardiomyocytes (CMs), and incorporating the cells efficiently and seamlessly into the host myocardium to ensure their synchronous contraction via electromechanical junctions.
Therefore, a better understanding the regulation of stem cell-derived differentiation of CSC/CMs is needed.
Based on currently available data and work from embryonic stem cells with in vivo lineage-tracing results, a working model of heart cell lineage diversification has been recently proposed 7  into FLK1 + ISL (islet-1) + multipotent cardiovascular progenitor cells 8,9 which can generate the three major types of cardiac cells: CMs, smooth muscle cells and endothelial cells 10 . CM commitment occurs with the induction of transcription factors such as NKX2.5 (NK2 transcription factor related, locus5) and GATA4 (GATA-binding protein 4), which control its initial differentiation and further maturation 11 . A heart lineage map has been derived from relatively specific molecular markers, HCN4 (hyperpolarization-activated cyclic nucleotide-gated channel 4) for the first heart field which committed to cardiomyogenic cell lineage, ISL1 for second heart field which represent a multiple progenitors differentiating into various cell lineage in the heart, WT1 (wilms tumour 1) and TBX18 (T-box family member 18) for the proepicardium, and WNT and PAX3 (paired box gene 3) for the neural crest 9,10,[12][13][14][15][16][17][18] . Maturation of these CM precursor cells is characterized by the expression of cardiac contractile proteins such as myosin heavy chain (MHC) and cardiac troponin T (cTnT).
Tumour necrotic factor-α (TNF) is a major mediator of inflammation and inflammatory diseases, and it has also been implicated in several cardiovascular diseases 19 . TNF elicits a broad spectrum of biological effects including proliferation, differentiation and apoptosis 20,21 . These differences in TNF-induced responses are mostly due to the differential signalling via its two distinct receptors: type I 55 kDa TNF receptor (TNFR1) and type II 75 kDa TNF receptor (TNFR2) 22 . TNFR1 is expressed ubiquitously, whereas TNFR2 expression is tightly regulated and found predominantly in CMs, vascular endothelial cells and haematopoietic cells 23 . Our in vitro and in vivo studies reveal that TNFR2 via Akt mediates cell survival and tissue repair 24,25 . Our previous data have shown that in human ischaemic heart disease (IHD)TNFR2 and phospho-histone H3 (pH3 S10 ) dramatically increased. TNFR2 + pH3 S10+ CSCs are increased and co-expressed pluripotent stem cell protein Lin28 in IHD, and these cells were CD45-negative and VEGFR2-negative. In vitro experiment showed hypoxia and/or TNF induce up-regulation of TNFR2 and TNFR2 + pH3 S10+ CSCs 26 . These results suggest that both Lin28 and TNFR2 signalling may trigger CSC activation and differentiation. However, the functional connections between Lin28 and TNFR2 are not clear.
In the present study, we attempt to define the mechanism for TNFR2-mediated CSC activation and differentiation. By employing a protocol of in vitro CSC differentiation from human inducible pluripotent stem cell (hiPSC), we show that TNFR2 is up-regulated by pluripotent factor Lin28. Moreover, we demonstrate a critical role of Lin28-TNFR2 axis in CSC activation and differentiation.

| Cardiomyocyte differentiation
To produce human CMs from pluripotent stem cells, hiPSCs were differentiated into hiPSC-CMs with a chemically defined CM differentiation protocol 27 . Briefly, hiPSCs were first treated with a small molecule inhibitor of GSK3β signalling, CHIR99021 (STEMCELL Technologies Inc., Vancouver, Canada), to activate the Wnt signalling pathway.   Expression data were normalized to the level of human GAPDH transcripts. The primers (NKX2.5, GATA4, SOX2, Nanog, OCT4, TNFR2, TNFR1, 18sRNA) for quantitative RT-PCR are listed in Table S1.

| Immunofluorescence-staining analysis
Cells or frozen tissue slides were fixed with 4% paraformaldehyde, permeabilized with 0.2% Triton X-100 in PBS, blocked with a solution of protein blocker for an hour and incubated with primary antibodies at 4°C overnight. Antibodies used are listed in Table S2. Secondary antibodies conjugated with Alexa Fluor 488 or 594 (Invitrogen, Carlsbad, USA) were then added, and the incubation was performed at room temperature for an hour in the dark. Nuclei were stained with 4'6-diamidino-2-phenylindole (DAPI) (Vector Laboratories, Burlingame, CA, USA).

| Statistical analysis
All figures are representative of at least three experiments unless otherwise noted. All graphs report mean ± SEM values of biological replicates. Comparisons between two groups were performed by unpaired, two-tailed t test, between more than two groups by one-way ANOVA followed by Bonferroni's post-hoc or by twoway ANOVA using Prism 6.0 software (GraphPad). P values were two-tailed and values <0.05 were considered to indicate statistical significance. P < 0.05, P < 0.01 and P < 0.001 are designated in all figures with *, **, ***, respectively.

| Data availability
All other data supporting the presented findings are available from the corresponding author upon request.

| Differentiation of hESCs and iPS cells into CSC and CMs
In vitro differentiation from hESC or hiPSC has provided a useful approach to define the gene function in cell specification. A matrix sandwich protocol with the GSK3 inhibitor and Wnt inhibitor (GiWi protocol) has produced high yield preparations of CSC from hESC or hiPSC 27 . We employed the differentiation protocol from hiPSC Functional maturity of the differentiated CMs was evaluated by electrophysiology, which were determined through single cell dissection from random areas and followed by action potential and calcium influx recordings in the whole cell patchclamp configuration.
A typical Ca 2+ (but not K + or Na + ) action potential was observed in hiPS-derived CMs (Figure 2A-D). These data suggest that differentiated CMs not only express correct cellular markers but also exhibit functional properties of mature CMs.

| TNFR2 expression precedes the expression of CSC markers in an in vitro differentiation system
We examined gene expression of TNFR2 during differentiation and found that TNFR2 was highly up-regulated upon differentiation but F I G U R E 2 Functional maturity of differentiated CMs evaluated by electrophysiology. hiPSC-based cardiac differentiation was performed and hiPSC-derived CMs after day 30 differentiation were subjected to electrophysiology through single cell dissection from random areas and followed by action potential and calcium influx recordings in the whole cell patchclamp configuration. Representative traces of membrane potentials recorded from beating cells before, during and after the application of blockers of Na + channel Tetrodotoxin (TTX, 1 μmol/L, A); Ca 2+ channel (Co 2+ , 100 μmol/L, B); and K + channel (Ba 2+ , 20 μmol/L, C) peaked at day 3 followed by a decline thereafter. In contrast, TNFR1 was ubiquitously expressed in all stages ( Figure 3A). We evaluated expression of TNFR2 proteins and CSC markers by immunostaining. Ki67 on day 7 followed by a decline on day 12 of differentiation ( Figure 3B and C). Taken together, the early kinetics of TNFR2 expression suggests that TNFR2 may play a role in CSC differentiation, proliferation and maturation.

| Inhibition of TNFR2 attenuates whereas TNFR2-specific agonist enhances cardiac cell activation/differentiation
We then tested our hypothesis that TNFR2 plays a critical role in  with a site-specific mutation (D143F) preferentially binds to TNFR2 and activates TNFR2-specific signalling such as Akt ( Figure S3). In contrast, TNFR2 neutralization antibody has been shown to block TNFR2-dependent signalling 28,29 . We observed that the presence of αTNFR2 or R2-TNF in the differentiation media had no effect on gene expression of stem cell markers (such as OCT4). However, αTNFR2 drastically reduced, whereas R2-TNF significantly increased, gene expression of CSC marker GATA4 and CM marker cTnT ( Figure 4A and B). Accordingly, αTNFR2 attenuated while R2-TNF augmented CSC differentiation and maturation as measured for GATA4 and cTnT immunostaining ( Figure 4C-F).
To gain insight into the potential molecular mechanisms through which TNFR2 mediates CSC proliferation, differentiation and maturation, we examined the TNFR2 downstream signalling in CSC. We have previously reported that TNFR2 in vascular endothelial cells activates Akt and STAT3, leading to endothelial cell proliferation and migration 21,40,43 . These reports prompted us to determine if TNFR2 signalling induces Akt and STAT3 activation during CSC activation/ differentiation. We detected both Akt and STAT3 were highly activated at early phase of CSC differentiation, coinciding with the kinetics of TNFR2 expression. Importantly, the presence of anti-TNFR2 antibody (αTNFR2) blocked phosphorylation of Akt and STATA3 (4g), consistent with its effect on CSC activation/differentiation. These data suggest that TNFR2-mediated Akt and STAT3 signalling is required for CSC proliferation, differentiation and maturation.

| TNFR2 is up-regulated by Lin28 at an early phase of CSC activation/differentiation
Distinct from TNFR1,TNFR2 expression is restricted in certain cell types 30 . Expression of TNFR2 at an early stage of differentiation prior to CSC generation promoted us to examine if stem cell/pluripotent factors could regulate TNFR2 expression. Lin28 is an RNA-binding protein that regulates microRNA generation and stability. It also regulates protein translation by binding to the 3'-untranslated region (3'UTR) on mRNAs 31 . It has been reported that three conserved sequences 'GGGCAGA', 'GAT' and 'GGAG' on mRNA 3'-UTR are within the consensus recognition motif for Lin28 32 . Sequence analyses indicated that the Tnfr2 mRNA 3'-UTR contains such a motif ( Figure 5A).

The 3'UTR of Tnfr2 was inserted into a luciferase reporter plas-
mid (Luc-Tnfr2-3'UTR) followed by mutations at one or all three of the Lin28-binding sequences (Tnfr2-3'UTR-ΔM1, ΔM2, ΔM3 and ΔM123) ( Figure 5B). To determine if Lin28 enhances TNFR2 translation via the Tnfr2 3'UTR, an effect of Lin28 co-expression on the Luc-Tnfr2-3'UTR reporter gene activity was analysed. Co-expression of Lin28 increased activity of the Luc-Tnfr2-3'UTR reporter gene in H9C2 cardiomyoblast cells. However, a deletion at any one of three conserved sites diminished the effect of Lin28 on the reporter gene ( Figure 5C). We further assessed the ability of Lin28 binds to the Tnfr2 3'-UTR during CSC differentiation by an RNA-binding protein immunoprecipitation (RIP) assay. Consistent with the kinetics of Lin28 and TNFR2 expression, the binding of Lin28 to the 3'-UTR of Tnfr2 mRNA was not detectable in hiPSC at day 0, but was strongly detected in cells at day 3 of differentiation when Lin28 + TNFR2 + cells peaked followed by a decline in day 7 when TNFR2 + GATA4 + cells peaked ( Figure 5D). Taken together, these results indicate that Lin28 up-regulates TNFR2 expression at an early phase of CSC differentiation by transiently binding to the Tnfr2 3'-UTR.

| Inhibition Lin28 attenuates TNFR2 expression and cardiac cell activation/differentiation
We examined gene expression of Lin28 and TNFR2 during differentiation and found that Lin28, like TNFR2 was highly upregulated upon differentiation but peaked at day 3 followed by a decline thereafter ( Figure 6A). TNFR2 + cells could co-express Lin28 and Lin28 + TNFR2 + cells peaked on day 3, prior to appearance of TNFR2 + GATA4 + and TNFR2 + NKX2.5 + cells during differentiation ( Figure 6B and C).
We then determined the role of Lin28-mediated TNFR2 expression in hiPSC-derived CSC differentiation. To this end, we examined effects of Lin28 inhibition on CSC differentiation. hiPSC-based CSC differentiation was performed in the absence or presence of a Lin28 inhibitor Lin28 1632. Inhibition of Lin28 significantly reduced the number of total TNFR2 + cells and proliferating TNFR2 + cells as well as differentiated GATA4 + and cTnT + cells as measured by immunostaining ( Figure 6D and E).

| D ISCUSS I ON
TNFR2 has been implicated to have cardiac-protective functions.
Ablation of the TNFR2 gene exacerbates heart failure and reduces survival, whereas ablation of TNFR1 blunts TNF-induced heart failure and improves survival in TNF-transgenic mice 33,34 . We have reported that TNFR1 and TNFR2 are differentially expressed in human ischaemic myocardium and proposed a cardioprotective role of TNFR2 in ischaemic heart 29 . Subsequently, we have shown that TNFR2 + cells with phospho-histone H3 S10 (pH3 S10 ) are detected in human ischaemic heart and co-express pluripotent stem cell protein Lin28 26 . However, it is unknown if and how TNFR2 signalling F I G U R E 4 TNFR2 inhibition attenuates whereas TNFR2-specific agonist enhances cardiac cell differentiation. hiPSC-based cardiac differentiation was performed in the presence of isotype IgG or TNFR2 neutralization antibody (αTNFR2; 100 ng/ml) (A, C and D), or in the presence of Saline or R2-TNF (100 ng/ml) (B, E and F). A and B, Relative expression of various markers during differentiation was determined by qRT-PCR. Experiments were repeated three times. C to F, Representative immunostaining images of GATA4 + cTnT + cells are shown (C,E) and quantifications of GATA4 + cTnT + cells are presented (D,F). G, hiPSC-based cardiac differentiation was performed in the presence of isotype IgG or TNFR2 neutralization antibody (αTNFR2; 100 ng/ml). hiPSC and D3 CSC lysates were subjected to Western blotting. Data are from three independent experiments. Scale bar: 50 μm. *P<0.05; ***P<0.001  TNFR2-3′UTR-ΔM1   TNFR2-3′UTR-ΔM2   TNFR2-3′UTR-ΔM3   TNFR2-3′UTR-ΔM1M2M3   TNFR2-3′UTR-ΔM1M2   TNFR2-3′UTR-ΔM2M3   TNFR2-3′UTR- is required for CSC differentiation and how TNFR2 is regulated and activated during CSC differentiation. In this report, we have taken an in vitro approach of differentiation from hiPSC to CSC and we have found that TNFR2 expression is induced at an early phase of CSC differentiation. Specifically, Lin28 up-regulates TNFR2 protein expression by directly binding to a conserved Lin28-motif within the 3'UTR of Tnfr2 mRNA. Further kinetics analyses indicate that Lin28-TNFR2 expression not only precedes the expression of CSC markers and mature CM proteins, but also is required for CSC generation. This is supported by the result that inhibition of Lin28 orTNFR2 diminishes, whereas TNFR2 activation by a specific ago- One important mechanistic finding in our study is that TNFR2 is up-regulated in cardiogenic cells. It is known that TNFR2 expression is restricted to specific cell types such as endothelial cells and CMs , and can be induced under various pathological conditions, primarily at a transcriptional level. TNFR2 promoter contains several consensus elements for transcriptional factors SP1, AP1 and NF-κB; all of these factors could be activated by inflammatory cytokines. Therefore, TNFR2 expression has been shown be regulated by cytokines, including interleukin-1β and TNF itself 21,30,36 .
Since TNFR2 is co-expressed with MESP1, Lin28 as well as cardiogenic factors GATA4 and NKX2.5, we have reasoned that cardiogenic cells exhibit unique ability to turn on TNFR2 expression.
Indeed, TNFR2 is up-regulated in the in vitro hiPSC differentiation system. We further demonstrate that the pluripotent factor Lin28, an RNA-binding protein, could directly bind to a consensus Lin28-motif within the 3'UTR of Tnfr2 mRNA to up-regulate TNFR2 protein expression. Lin28 is best known to regulate generation of miRNA let-7, but also acts in let-7-independent fashion by either promoting or suppressing protein translations 37,38 . Our data suggest that Lin28 promotes the TNFR2 translation by binding to its 3'UTR. Interestingly enough, it has been shown that Lin28 transcription can be strongly induced by inflammation-activated NF-κB and Lin28 in turn further enhance the NF-κB-dependent inflammatory responses, forming a positive feedback loop 32,39 . It is plausible that inflammation activates Lin28 to induce TNFR2 expression in ischaemic heart. It needs to be determined how Lin28 is up-regulated in the in vitro hiPSC differentiation system in the absence of inflammatory cytokines. Our data show that blockade TNFR2 reduces whereas R2-TNF sustains Lin28 expression in the in vitro system, suggesting TNFR2 by activating NF-κB could form feedback loop with Lin28. Of note, TNFR2-specific activation promotes cell survival without enhancing inflammation as we have previously demonstrated in TNFR2-transgenic mice 40 . Therefore, R2-TNF together with hESC/hiPSC-derived CSCs would provide an effective treatment for ischaemic heart disease.
A remaining question is the molecular mechanisms through which TNFR2 mediates CSC proliferation, differentiation and maturation.
Recent studies suggest that both Akt and STAT3 are critical for CSC proliferation and differentiation from ESCs 41,42 . We observe that both Akt and STAT3 are highly activated at early phase of CSC differentiation. Consistent with the effects of TNFR2 neutralization antibody on CSC activation/differentiation, anti-TNFR2 antibody blocks activation of Akt and STATA3 during CSC differentiation. Previously we have identified Bmx, a non-receptor tyrosine kinase implicated in cell migration, as the first TNFR2-specific tyrosine kinase. TNFR1, via an adaptor molecule ASK1-interacting protein-1 (AIP1), activates ASK1-JNK-dependent cell apoptosis. In contrast, TNFR2 via Bmx promotes cell activation, migration, growth or proliferation in vascular endothelial cells 21,40,43 . Furthermore, we show that Bmx binds to the C-terminal 16 aa sequence of TNFR2 to mediate TNFR2-induced Akt and STATA3dependent cell migration and angiogenesis 21,40,44 . Importantly, both TNFR2 and Bmx have been implicated to have cardiac-protective F I G U R E 7 A model for the role of Lin28-TNFR2 signalling in cardiac stem cell activation and differentiation. Lin28 induces TNFR2 expression in iPSCs. Proliferative TNFR2 + cells in turn become CSCs which subsequently become cTnT + mature cardiomyocytes. TNFR2 may mediate Akt and STAT3 signalling to induce CSC activation and differentiation. TNFR2 inhibition attenuates whereas TNFR2-specific agonist enhances cardiac cell activation/differentiation. CSC: cardiac stem cells; CM: cardiomyocytes; cTnT: cardiac troponin T; αR2: TNFR2 neutralization antibody; R2-TNF: TNFR2-specific agonist functions 29,43,[45][46][47][48] . It needs further investigations to determine if Bmx mediates TNFR2-dependent Akt/STATA3 activation during CSC activation/differentiation.

Collectively, we have defined the important function of TNFR2
in CSCs activation and differentiation. Therefore, specific activation of TNFR2 signalling may be a novel strategy for the treatment of ischaemic diseases in humans.

AUTH O R CO NTR I B UTI O N S
The following people designed, performed research and analysed data: QX, BY, LL,BQ, CQ, XBG, HJZ and WM; HJZ and WM wrote the paper.

CO N FLI C T O F I NTE R E S T
The authors confirm that there are no conflicts of interest.