Heart Failure Immunology

This theme is led primarily by Professor Dominique de Kleijn.

It is accepted that proinflammatory cytokines serve as the downstream “effectors” of the innate immune system by facilitating tissue repair within the heart.1,2 Like in HFREF3,4mouse models, HFPEF is accompanied in an increased influx of macrophages in the heart together with an increase in proinflammatory factors. In both mouse5and human1,2cytokines like IL-6 and TNFalpha are increased during HF identifying the immunological system as an important player in HF. Angiotensin inhibition, a cornerstone in the treatment of HF, also reduces activation of the innate immune system 6,7 and subsequent cytokine production. Discovery of the Toll-like receptors (TLRs) as the “upstream” molecular innate immune component identified these TLRs as potential targets for HF.

Toll-like receptors

TLRs are transmembrane receptors that recognize ‘pathogen-associated molecular patterns’ (PAMPs) of different microorganisms. Recognition of PAMPs enables host defence and mediates the release of IL-6 and TNFalpha. Danger-associated molecular patterns (DAMPs) 8-10  are also recognized by TLRs and are thought to mediate repair mechanisms of injured tissue 11  as in HF.

Toll-like receptors and the heart

We showed  TLRs are involved in the response of the heart after myocardial infarction (MI) and remodeling leading to HF 3,4,12,13. After MI, these receptors are probably activated by endogenous ligands 4 and not by bacteria.

For TLR2, we showed that TLR2 on the leucocytes and not the heart, determines infarct size after ischemia/reperfusion (I/R) injury and remodeling13 . For HF, independent of infarct size, survival was promoted and incidence of HF was reduced in the TLR2 KO and determined by the leucocytic TLR2 deficiency 14.

Leucocytic TLR2 as a target to reduce infarct size and adverse remodeling

As TLR2 on the leucocytes is the receptor to block, we successfully used the TLR2 blocking antibody OPN301 in the mouse 13. Using a pig model, we showed the therapeutic efficacy of  clinical grade humanized anti-TLR2 antibody, OPN-305 15, showing that OPN305 is a promising intervention for HF. The use of blocking TLR2 antibodies to prevent HFREF, has not been explored and a role of TLR2 in HFPEF is unknown. Different DAMPs act through different TLRS, suggesting a range of TLRs are potentially involved in HF.

Extra-cellular plasma vesicles

The immune system uses cyto- & chemokines as well as inflammatory cells to communicate. A third, relatively unknown, way of immunological communication is the use of vesicles as a way of distributing antigen by inflammatory cells 16 .

Extracellular vesicles, including microvesicles, microparticles and exosomes 17 are abundant in  plasma. These vesicles are secreted by all cells especially after stimulation with PAMPs and are important in a variety of processes including coagulation 18, antigen presentation16 and tissue damage 19. Plasma extracellular vesicles can be easily isolated from frozen plasma or serum. They contain protein, miRNA and RNA depending on the source and stimulus involved.

Until now, research has focused on number and vesicle type 20-22 but not their content. Vesicle number and type have been associated with HF. Furthermore, miRNAs mostly originating from plasma vesicles have been associated with HF 23.

We showed that vesicle content is gender-dependent and associated with ACS diagnosis as well as prediction of secondary CV events.

Contents of vesicle sub-populations, have not been studied in any cardiovascular disease but will reflect the biological processes of the originating cells and for this can be used to monitor the development of HF that can be used in the clinic. Furthermore, we showed in pigs that mesenchymal stem cell (MSC) conditioned medium (CM) reduces incidence of HF after MI and that among the microvesicle sub-populations exosomes are the active component 35.

For more info on vesicles see http://www.socrates-singapore.org

Research Plan and Strategy

A  Involvement of TLRs in HF

1) Establish the nucleotide-detecting TLRs as a therapeutic target for HF.

2) Establish TLR2 as a therapeutic target for HFPEF and validate for HFREF.

3) Establish TLR and Tissue Factor as a combined therapeutic target in HF.

4) Explore interaction of TLR with Angiotensin and validate combined therapy in HF.


Experiment 1  The nucleotide-detecting TLR

Self-nucleic acids and mitochrondial DNA that escapes degradation can activate TLR9 and contribute to chronic inflammation during HF.24,25 TLR3 can be stimulated by plasma mRNA and TLR7 is involved in the promotion of cardiac fibrosis. 26,27 This strongly suggests that endogenous nucleotides can activate “nucleotide detecting” TLRs (TLR3, 7 and 9) after MI leading to HF. Oligodeoxynucleotide (ODN) inhibitors (IRS 166) of TLR7 are available and have been used in vivo.28,29 The involvement of TLR3 &7 in HFPEF and HFREF has not been shown. Mouse and pig models for HFREF and HFPEF are available as well as TLR3 and 7 null mice and ODN inhibitors for application in mouse & pig.


Experiment  2   TLR2 as target for HF

OPN301 will be used in a mouse HFREF model to determine optimal time and dose for intervention followed by the pig HFREF model (OPN305). For HFPEF, the TLR2 null mouse will be used to establish involvement followed by antibody interventions.


Experiment 3  TLRs and Coagulation

From sepsis, a strong bilateral interaction is established between the innate immune system and coagulation resulting in microvascular failure.30,31 Binding of Tissue factor (TF) leads to activation of macrophages and tissue infiltration of neutrophils.31 Clinically, coagulation and inflammation appear to play important roles after MI during reperfusion, no reflow and angiogenesis that are important features during the development of HF.32-34 Understanding of the TLR-TF balance during HF might lead to target combination of both TLR and TF. Furthermore, this can identify coronary microvasculature failure as an underlying cause of HF. Mouse delta TF & TLR2 null mice are available as TLR2 (OPN301/5) and TF (Ixolaris) blockers are available and can be applied in mouse and pig model.


Experiment 4  TLRs and Angiotensin

Inhibition of Angiotensin reduces activation of TLR4 signalling.6,7  It is, however, unknown what role Angiotensin (AT) has in the prominent actions of TLRs 2 and 4 during HF. It is possible this is mediated via AT Type I receptor (ATR) and/or TLR2 and 4.

ATR and TLR2 & 4 null mice are available as well are ATR and TLR2 & 4 blockers.


B  Plasma extracellular vesicle content as biomarker source for HF

5) Identify and validate plasma vesicle content to monitor TLR intervention In HF

6) Characterization of plasma extracellular subpopulations in pig HF models.

7) Identification & validation of vesicle proteins in existing human HF cohorts.


Experiment  5    Plasma vesicle content during TLR interventions

Isolation of 3 vesicle sub-populations (Annexin V (AXV), Cholera Toxin B (CTB) and CD14-associated ) will be applied to the mouse and pig TLR interventions (see A). Via Q-proteomics, miRNA array and subsequent validation, vesicle content will be studied for changes in pathways and identification of efficacy markers.


Experiment 6    Characterization of plasma and saliva vesicles in pigs during HF

For HFPEF, HFREF and sham and for both male and female pigs, plasma will be collected daily in individual pigs as well as HF parameters with MRI at Baseline, 7 days, 14 days and 28 days. From the plasma, the 3 (AXV, CTB, CD14) plasma vesicle subpopulations will be isolated and used for iTraq quantitative proteomics with subsequent MRM validation in the individual pigs and miRNA levels. A selection will be made based using IPA for validation in human HF cohorts.


Experiment 7     Identification and validation of vesicle proteins in human HF cohorts

Availability of our expanding, well-annotated human biobank cohorts (“4B”, SHOP, SLAS & ASIAN-HF) opens 2 possibilities. One is the validation of the vesicle markers identified in pig models (see Expt 6 above ). A second  is undertaking discovery analyses on pooled plasma samples followed by subsequent validation in individual patient samples. The latter will be done comparing within gender HFREF vs HFPEF, HF with MI vs HF without MI as well as between ethnicities.


Experiment 8    Validation of MSC-exosomes in pig HF models

Clinical grade MSC exosomes will be tested on pig HFREF and HFPEF models in dose-response studies of efficacy in augmentation of cardiac performance.


Details of Existing Projects

Animal HF models in Singapore

With the recent opening of MD2 at NUS, mouse & pig cardiovascular models are achievable in Singapore. This will be enhanced by this grant for several reasons.

First is the collaboration with Experimental Cardiology (Invented and developed the Coronary Bypass “Octopus stabilizer” in pigs at the end of the 90’s) at UMC Utrecht, the Netherlands. The Utrecht group has more than 20 years of experience and one of the largest pig research facilities in Europe. Mouse intervention studies started in 2000 and led to numerous high impact publications (3,4,11-15). On a 100% appointment at CVRI and Surgery (NUS/NUHS) in Singapore and a 0% professorship appointment at UMC Utrecht and Inter-university Cardiology Institute the Netherlands (ICIN), Prof. Dominique de Kleijn (PI of this theme) will bring this experience to Singapore. For this, he is using his start-up grant and a Royal Dutch Academy grant (ICIN, 800,000 Euro) for training of local staff in Utrecht and in Singapore by Utrecht staff in mouse and pig models of myocardial infarction and HF. Assoc. Prof Roger Foo will bring in the HFPEF mouse model.

Second are major investments made in MD2 including the purchase of a mouse Echocardiography (Vevo 2100), dedicated pig MRI and CT to study heart pump function. This makes MD2, a unique facility for Asia and probably the world to perform cardiovascular research on pigs.

The strong combination of long term background experience with a uniquely well-equipped facility is rare. We will have mouse and pig HF models in Singapore operational in the first quarter of 2013, experience that can be used by NUS, Singapore and beyond.

The existing collaboration between Singapore and the Netherlands has already resulted, in addition to publications, in a new clinical application for MSC exosomes tested within this grant in a multicenter Phase I & II trial (Theme 5).


Translational research with Industry and Clinic


In addition to the earlier mentioned MSC exosomes, Toll-like receptors are considered as elegant targets as they are high in the immunological hierarchy and for MI and HF must be targeted on circulating leucocytes. Collaborations have been established with world-leading experts including Prof Luc O’Neill and Prof Nicolas Frangogiannis and also with the TLR company (www.opsona.com) Opsona Ltd. Our aim (as with the TLR2 antibodies) is to develop the HF application towards the clinic together with Opsona. Other TLRs are chosen, as a pharmaceutical intervention has been created enabling translation from KO-mouse to mouse intervention and pig intervention to clinical trial. Experience to do this exists within this CG with the MSC exosomes and the TLR2 antibody as well as to perform clinical trials (Assoc. Prof Huay Cheem Tan, Prof Mark Richards, Assoc. Prof Carolyn Lam & Dr Mark Chan).


With IMB, NTU, UMC Utrecht, ICIN and Cavadis BV (www.cavadis.com), we were the first to consider the content of plasma extracellular vesicles in the context of diagnosis of MI and prediction of secondary cardiovascular events This will now be intensified, by studying vesicle sub-populations.

The finding of gender-dependent diagnostic and prognostic information within the (sub) vesicle content opens a new research area that will be applied beyond cardiovascular application. With IMB, we are perfecting the isolation of vesicle sub-populations to semi-automated isolation and analysis and are searching for other subpopulations.

With NTU, we optimized the iTraq quantitative proteomics on a minimal amount (250 ul) of plasma. In addition, post-translational modifications (ie phosphorylation, acetylation and oxidation) can be determined in the vesicles with quantification in individual samples (125 ul plasma) no longer dependent on antibody detection with the validation of Mass Spectrometry based Multiple Reaction Monitoring (MRM). Extension of studies to other biological sources, including saliva and urine, is possible.

Vesicle isolation and proteomics technology combined with the availability of well-annotated biobanks in Singapore and abroad will make this a unique approach with broad application that already is receiving a lot of attention from the diagnostics industry (Cavadis BV) and pharma (companion diagnostics).

Existing plasma markers re-assessed for their (possibly enriched) presence in vesicles will become patentable again.


1) Gordon JW etal Circ Res. 2011;108:1122-1132

2) Mann DL Circ Res. 2011;108:1133-1145

3) Timmers L, et al, Circ Res. 2008 Feb1;102(2):257-64.

4) Arslan F, et al, Circ Res. 2011 Mar 4;108(5):582-92.

5) Nagai et al., Hypertension 2011, 57, 208-215

6) Takahashi Y, et al Clin Sci (Lond). 2010 Jul 23;119(9):395-405.

7) Dasu MR, Riosvelasco AC, Jialal I. Atherosclerosis. 2009Jan;202(1):76-83.

8) Matzinger, P. Science 296, 301-305 (2002).

9) Seong, S. Y. & Matzinger, P. Nat. Rev. Immunol. 4, 469-478 (2004).

10) O'Neill LA. Immunol Rev. 2008;226:10-8.

11) Arslan F, de Kleijn DP, Pasterkamp G. Nat Rev Cardiol. 2011;8(5):292-300.

12) Timmers L, et al, Circ Res. 2009 Mar 13;104(5):699-706.

13) Arslan F, et al, Circulation. 2010 Jan 5;121(1):80-90.

14) Arslan F, Thesis UMC Utrecht, the Netherlands 2011

15) Arslan F, et al. Circ Cardiovasc Interv. 2012 Apr;5(2):279-87

16) Théry C, et al Nat Rev Immunol. 2009 Aug;9(8):581-93.

17) Tushuisen et al Arterioscler Thromb Vasc Biol 2011;31;4-9.

18) Del Conde I, et al, Blood. 2005 Sep 1;106(5):1604-11.

19) Lai RC, et al Stem Cell Res. 2010;4:214-222.

20) Gyorgy B, et al, Cell Mol Life Sci. 2011 Aug;68(16):2667-88.

21) Nozaki T, et al, Eur J Heart Fail. 2010 Nov;12(11):1223-8.

22) Garcia S, et al, J Heart Lung Transplant. 2005 Dec;24(12):2184-9.

23) Creemers EE, et al, Circ Res. 2012 Feb3;110(3):483-95.

24) Guiducci C, et al., Nature. 2010 Jun 17;465(7300):937-41.

25) Oka T, et al., Nature. 2012 May 10;485(7397):251-5.

26) Nakada E, et al., Placenta. 2011 Jul;32(7):500-5.

27) Alvarez D, et al., J Biol Chem. 2011 Sep 2;286(35):30444-54.

28) Barrat FJ, et al., J Exp Med. 2005 Oct 17;202(8):1131-9.

29) Pawar RD, et al., J Am Soc Nephrol. 2007Jun;18(6):1721-31.

30) Levi M, van der Poll T, Buller HR. Circulation. 2004 Jun 8;109(22):2698-704.

31) van der Poll T, de Boer JD, Levi M. Curr Opin Infect Dis. 2011;24(3):273-8.

32) Pawlinski R, Mackman N J Thromb Haemost. 2009;7(2):288-9.

33) Merlini PA, et al Am J Cardiol. 2004 Apr 1;93(7):822-5

34) Pawlinski R, et al Thromb Haemost. 2004 Sep;92(3):444-50.

35) Timmers L, et al Stemcell Res. 2011, 6, 206-214.