Human bioengineered artery equivalents may represent a first
step toward the future replacement of the animal models
used for studying the initial phases of atherosclerosis.
According to the US National Center for Health Statistics, cardiovascular diseases — including coronary heart disease and cerebrovascular disease — are leading causes of death in the western world. Sixty eight percent of all cardiovascular diseases are related to atherosclerosis, resulting in a substantial morbidity and mortality, as well as a major socioeconomic burden.1 As a result of current demographic changes, namely, the significant increase in the elderly population, the burden of atherosclerosis is on the rise.2 It is therefore not surprising that research on atherosclerosis is becoming more and more important, and substantial financial resources are being transferred into this particular field of biomedical research. Alongside this development, the use of animal models in atherosclerosis research has also exponentially increased over the last few decades.
According to a MEDLINE search conducted in May 2014, 29,795 publications can be found when entering the search terms “atherosclerosis AND animal”. Assuming that only half would represent animal in vivo studies, and assuming that only 15 animals were used per study, then this would render a load of 223,463 animals used in atherosclerosis trials published in the MEDLINE database alone. Interestingly, the number of publications was stable until the beginning of the 1990s, at approximately 200 entries per year. However, starting at around 1990, a dramatic increase in publications is observed (see Figure 1), demonstrating the increase of research in this area of cardiovascular disease.
An overview of atherosclerosis
Atherosclerosis represents a chronic inflammatory disorder that is the underlying cause of most cardiovascular diseases.1–3 As a first event in the pathogenesis of atherosclerosis, LDL (a cholesterol-rich lipoprotein) accumulates in the intimal layer of the healthy vessel wall.4,5 As a next step, the activated endothelial layer promotes the adhesion and transmigration of monocytic/macrophagic blood cells, which themselves take up the accumulated LDL particles via a receptor-mediated process. This process is the basis for the formation of the initial ‘fattystreak lesion’ in the vascular wall.4,6 This lesion then serves as the basis for a progressive immunological reaction, leading to the formation of more and more complex, highly organised vascular lesions, which finally — after a number of years — result in a ‘vulnerable plaque lesion’.4,7 If these vulnerable plaque lesions rupture at the location of the ‘fibrous cap’, a thrombotic (potentially lethal) response is evoked. This may lead to the occlusion of vessels, such as in the case of myocardial infarction.8 So far, these two processes, i.e. lipoprotein accumulation and the transmigration of macrophages, have received the greatest attention with regard to achieving optimal mimicry of human atherogenesis in vitro.
Current animal models
According to the recent review by Getz et al.3, several different animal models are used for testing therapeutic agents and studying the etio-pathogenesis of atherosclerosis. Getz et al. state that no model is ideal, as each has its own advantages and limitations with respect to manipulation of the atherogenic process and modelling human lipoprotein profiles. Due to the relative ease of genetic manipulation and the short time-frame for the development of atherosclerosis, murine models are currently the mostextensively used animals in atherosclerosis research. However, murine atherosclerosis development significantly differs from human atherosclerosis.9 Therefore, a range of further larger animal models have been widely used, including rabbits, rats, guinea-pigs, hamsters, birds, dogs, swine and nonhuman primates (NHPs).3,9,10 In particular, the development of porcine11 and NHP models12–14 has been intensified in recent years. In pigs, especially the diabetic hypercholesteraemic model,6 the Rapacz-familial hypercholesteraemia model,11 the PCSK9-gain-of-function model and different mini-pig models for investigating metabolic syndrome11,14 were seen as important steps toward more-representative animal research.11 Predominantly, the more-human analogous genetic expression patterns, as well as the development of high-risk atherosclerotic lesions, were seen as major advantages of these transgenic porcine models.11 Also, NHPs have been used for atherosclerotic research12,15 — mainly rhesus and cynomolgous monkeys,12 African green monkeys,12,13 and baboons15 — as they provide unique opportunities to evaluate effects of long-term diets and/or pharmaceutical agents. Particularly, the aspect of dyslipidaemia profiles similar to those of humans seems outstanding in these models, as summarised by Chen et al.12: NHPs “… closely resemble humans in lipid metabolism and disease physiology compared to lower species.” Also, a large cross-species study revealed that only NHPs and certain dog species (e.g. obese beagle models) provide a close match to dyslipidaemic humans.12
This development toward higher (more-clinically relevant) animal species is associated with several problems. Besides the profound ethical considerations surrounding large scale animal research in higher species (dogs, pigs, NHPs), the financial hurdles associated with these experiments also limit further and broad-scale scientific progress. This is even more important when considering the results of large cross-species studies, which reveal that even these higher species show significant differences, and in particular, when it comes to research on lipid metabolism. 14
New in vitro models
Given the ethical and scientific obstacles associated with the animal models, the development of more representative in vitro models based on human cells seems to be of outstanding importance, and ultimately represents the only solution to this situation. In addition, such in vitro models could not only overcome the ethical discussions resulting from large scale animal experiments, they could also potentially allow for more broad-scale screening platforms, thus making atherosclerosis research more effective.
Therefore, given the lack of models that fully represent the human atherothrombotic pathology, several in vitro studies have been initiated that focus on the development of a human cell-based model for studying atherosclerosis ex vivo. These in vitro models hold the potential to significantly reduce and replace animal in vivo experiments and — at the same time — to increase the predictive value of the experiments with regard to human pathologies. Most of the initial experimental in vitro studies used two dimensional (2-D) cell culture systems (of endothelial cells or macrophages) to study basic mechanisms of atherosclerosis.16 In spite of the importance of these pioneering investigations, these 2-D models lacked both a) the three-dimensional (3-D) structure and b) the pulsatile blood-flow environment present in native human arteries. Therefore, these studies on 2-D culture dishes were (and are) associated with significant limitations, as they do not account for the complexity of the native artery environment, including all of the cell–cell and cell–matrix interactions that are present in vivo. Particularly, the signalling between the luminal endothelial cell component and the sub-endothelial vascular interstitial cell (media) component is missing. Given that vascular smooth muscle cell activation, proliferation, and migration have been identified as key processes of atherosclerosis development, the single cell nature or the 2-D aspect of these models represent major shortcomings.17 In order to address these limitations, several groups developed 2-D co-culture models of human and animal endothelial cells and vascular interstitial cells (smooth muscle cells).18–20 In these models, the two different cell types were separated by a preformed collagen matrix, which is intended to mimic the presence of the basal lamina, resembling for the first time a 3-D-like cell arrangement. However, these attempts were also limited by the non-physiological attachment of cultured cells to plastic dishes (or synthetic trans-well membranes) on the 2-D scale.16 In addition, in these types of co-culture systems, the culture times were too short for the development of cell–extracellular matrix interactions that are typical of the in vivo situation.16 Seeking further improvement, Dorweiler et al.21 were able to create one of the first long-term co-culture models through the use of fibrin gels. Besides an improvement in the overall culture duration, they also reported the accumulation of LDL and immune cells in the (bioartificial) sub-endothelial matrix, representing a significant improvement compared to previous systems. 21,22 However, in spite of the proof of lipoprotein accumulation, this non-dynamic co-culture system lacks the full 3-D cylindrical architecture of a native vessel. Even more importantly, it also lacks the haemodynamic pressure/flow/shear-stress environment present in the native situation, which plays a fundamental role in atherosclerosis development in vivo.
As part of therapeutic bioengineering approaches, several groups have successfully demonstrated the manufacture of tissue-engineered artery equivalents (with a 3-D cylindrical native-like architecture) under pulsatile, native-like flow (‘bioreactor’) conditions. These bioengineered vascular grafts showed a native analogous microstructure and approximation of the biomechanical behaviour of native tissues.23 These autologous tissue-engineered vascular grafts were originally designed for the surgical repair of congenital cardiac malformations. They were successfully implanted into large animal models, and exhibited adequate functionality in vivo for periods of up to 240 weeks.24,25 In an initial attempt to overcome the above-mentioned limitations, Robert et al.16 combined the technologies of atherosclerosis research with vascular bioengineering techniques. After creating a native-like human vascular cell-based layered bioengineered artery equivalent by using a pulsatile flow bioreactor system (also containing a basement membrane), LD- and HD-lipoproteins were integrated into the flow loop. The accumulation of cholesterol-rich lipoproteins represents a first, essential step in atherosclerosis development.5 In the engineered artery model, endothelial as well as sub-endothelial lipoprotein recovery of LDL, as well as HDL, could be observed over time. As a next step of human atherosclerosis development, lipoprotein-mediated endothelial activation leads to monocyte infiltration and transmigration.6 In the bioengineered artery model, monocytes injected into the flow loop were also found to adhere to the (activated) endothelium and to transmigrate into the vascular interstitium, which was comparable to the events that occur in the native counterpart.16 In spite of the several improvements on the previous 2-D and 3-D co-culture in vitro models, the proposed model only focused on short time-points. The formation of ‘foam’ cells, an integral part of atherosclerosis development, has not yet been demonstrated.
Summary and conclusions
In biomedical research, there has been an increasing interest over the last few years in the creation of systems for modelling human disease in vitro. The major difficulties encountered in these attempts stems from the complexity of native structures and processes, which are difficult, or sometimes impossible, to model in vitro. The resulting simplification of in vitro model systems holds the risk of limited, or even misleading, scientific output, with questionable representation of the native human situation. Atherosclerosis is an excellent example of this development, as it displays a highly complex multi-organmediated pathology. To make matters worse, atherosclerosis is not only a complex disease, but it is also very species-specific — indicating that animal research also has limited representative value. This scientific ‘dilemma’, combined with the astronomic health as well as socioeconomic impact of this disease, has recently stimulated the development of alternative in vitro models. What would be needed is a fully dynamically-perfused human native-analogous artery model, in order to study pathogenic processes in atherothrombosis development — processes which might harbour potential for therapeutic intervention. The latest advances in the field of biomedical engineering have led to the development of bioengineered artery models that overcome at least some of the limitations of previous models. However, these latest model systems suffer from extensive simplification, including: a) the use of cell culture media containing xenogenic and immunogenic elements (instead of blood); b) the use of low pressure conditions; c) the lack of defined shear-stress; d) extensive ex vivo cell expansion; and e) the lack of neurogenic as well as lymphogenic vascular elements. However, most importantly, the ‘vulnerable plaque lesion’ — which forms after years (or even decades) in humans — would be the most interesting structure to be studied, as this is the pre-stage to any potentially lethal, ruptured vascular plaque. Considering the short cultivation periods in the currently available bioengineered artery models (i.e. of only a few weeks), so far, the in vitro modelling of human atherogenesis has been limited to the study of the very initial phases of the disease.
Therefore, the in vitro mimicry of all phases of human atherosclerosis and vulnerable plaque development still represents ‘fiction’ rather than a scientific fact. However, the recent developments, as well as the rapid progress in bioengineering and biotechnology, create significant hope that the realisation of in vitro modelling of human atherosclerosis is within our grasp.
1 National Heart, Lung & Blood Institute (2012). NHLBI Fact Book, Fiscal Year 2012 — By Section, 201pp. Bethesda, MD, USA: National Institutes of Health, US Department of Health & Human Services. Available at: http://www.nhlbi.nih.gov/about/factbook/toc.htm (Accessed 25.05.14).
2 Edo, M.D. & Andrés, V. (2005). Aging, telomeres, and atherosclerosis. Cardiovascular Research 66, 213–221.
3 Getz, G.S. & Reardon, C.A. (2012). Animal models of atherosclerosis. Arteriosclerosis, Thrombosis, & Vascular Biology 32, 1104–1115.
4 Robert, J., Weber, B., Frese, L., Emmert, M.Y., Schmidt, D., von Eckardstein, A., Rohrer, L. & Hoerstrup, S.P. (2013). A three-dimensional engineered artery model for in vitro atherosclerosis
research. PLoS One 14, e79821.
5 Williams, K.J. & Tabas, I. (1998). The response-toretention hypothesis of atherogenesis reinforced. Current Opinion in Lipidology 9, 471–474.
6 Glass, C.K. & Witztum, J.L. (2001). Atherosclerosis: The road ahead. Cell 104, 503–516.
7 Rosenson, R.S., Brewer, H.B., Jr, Davidson, W.S., Fayad, Z.A., Fuster, V., Goldstein, J., Hellerstein, M., Jiang, X.C., Phillips, M.C., Rader, D.J., Remaley, A.T., Rothblat, G.H., Tall, A.R. & Yvan-Charvet, L. (2012). Cholesterol efflux and atheroprotection: Advancing the concept of reverse cholesterol transport. Circulation 125, 1905–1919.
8 Brokopp, C.E., Schoenauer, R., Richards, P., Bauer, S., Lohmann, C., Emmert, M.Y., Weber, B., Winnik, S., Aikawa, E., Graves, K., Genoni, M., Vogt, P., Lüscher, T.F., Renner, C., Hoerstrup, S.P. & Matter, C.M. (2011). Fibroblast activation protein is induced by inflammation and degrades type I collagen in thin-cap fibro – atheromata. European Heart Journal 32, 2713–2722.
9 Whitman, S.C. (2004). A practical approach to using mice in atherosclerosis research. Clinical Biochemistry Reviews 25, 81–93.
10 Xiangdong, L., Yuanwu, L., Hua, Z., Liming, R., Qiuyan, L. & Ning, L. (2011). Animal models for the atherosclerosis research: A review. Protein Cell 2, 189–201.
11 Hamamdzic, D. & Wilensky, R.L. (2013). Porcine models of accelerated coronary atherosclerosis: Role of diabetes mellitus and hypercholesterolemia. Journal of Diabetes Research 2013, 761415. [doi:10.1155/2013/761415]
12 Chen, Z., Strack, A.M., Stefanni, A.C., Chen, Y., Wu, W., Pan, Y., Urosevic-Price, O., Wang, L., McLaughlin, T., Geoghagen, N., Lassman, M.E., Roddy, T.P., Wong, K.K., Hubbard, B.K. & Flattery, A.M. (2011). Validation of human ApoB and ApoAI immunoturbidity assays for non-human primate dyslipidemia and atherosclerosis research. Journal of Cardiovascular Translational Research 4, 373–383.
13 Rayner, K.J., Esau, C.C., Hussain, F.N., McDaniel, A.L., Marshall, S.M., van Gils, J.M., Ray, T.D., Sheedy, F.J., Goedeke, L., Liu, X., Khatsenko, O.G., Kaimal, V., Lees, C.J., Fernandez-Hernando, C., Fisher, E.A., Temel, R.E. & Moore, K.J. (2011). Inhibition of miR-33a/b in nonhuman primates raises plasma HDL and lowers VLDL triglycerides. Nature, London 478, 404–407.
14 Yin, W., Carballo-Jane, E., McLaren, D.G., Mendoza, V.H., Gagen, K., Geoghagen, N.S., cNamara, L.A., Gorski, J.N., Eiermann, G.J., Petrov, A., Wolff, M., Tong, X., Wilsie, L.C., Akiyama, T.E., Chen, J., Thankappan, A., Xue, J., Ping, X., Andrews, G., Wickham, L.A., Gai, C.L., Trinh, T., Kulick, A.A., Donnelly, M.J., Voronin, G.O., Rosa, R., Cumiskey, A.M., Bekkari, K., Mitnaul, L.J., Puig, O., Chen, F., Raubertas, R., Wong, P.H., Hansen, B.C., Koblan, K.S., Roddy, T.P., Hubbard, B.K. & Strack, A.M. (2012). Plasma lipid profiling across species for the identification of optimal animal models of human dyslipidemia. Journal of Lipid Research 53, 51–65.
15 Wang, X.L., Wang, J., Shi, Q., Carey, K.D. & VandeBerg, J.L. (2004). Arterial wall-determined risk factors to vascular diseases: A nonhuman primate model. Cell Biochemistry Biophysics 40, 371–388.
16 Robert, J., Weber, B., Frese, L., Emmert, M.Y., Schmidt, D., von Eckardstein, A., Rohrer, L. & Hoerstrup, S.P.(2013). A three-dimensional engineered artery model for in vitro atherosclerosis research. PLoS One 8, e79821.
17 Gomez, D. & Owens, G.K. (2012). Smooth muscle cell phenotypic switching in atherosclerosis. Cardio -vascular Research 95, 156–164.
18 Navab, M., Hough, G.P., Stevenson, L.W., Drinkwater, D.C., Laks, H. & Fogelman, A.M. (1988). Monocyte migration into the subendothelial space of a coculture of adult human aortic endothelial and smooth muscle cells. The Journal of Clinical Investigation 82, 1853–1863.
19 Takaku, M., Wada, Y., Jinnouchi, K., Takeya, M., Takahashi, K., Usuda, H., Naito, M., Kurihara, H., Yazaki, Y., Kumazawa, Y., Okimoto, Y., Umetani, M., Noguchi, N., Niki, E., Hamakubo, T. & Kodama, T. (1999). An in vitro coculture model of transmigrant monocytes and foam cell formation. Arteriosclerosis, Thrombosis, & Vascular Biology 19, 2330–2339.
20 Wada, Y., Sugiyama, A., Kohro, T., Kobayashi, M., Takeya, M., Naito, M. & Kodama, T. (2000). In vitro model of atherosclerosis using coculture of arterial wall cells and macrophage. Yonsei Medical Journal 41, 740–755.
21 Dorweiler, B., Torzewski, M., Dahm, M., Ochsenhirt, V., Lehr, H.A., Lackner, K.J. & Vahl, C.F. 2006). A novel in vitro model for the study of plaque development in atherosclerosis. Thrombosis & Haemostasis 95, 182–189.
22 Dorweiler, B., Torzewski, M., Dahm, M., Kirkpatrick, C.J., Lackner, K.J. & Vahl, C.F. (2008). Subendothelial infiltration of neutrophil granulocytes and liberation of matrix-destabilizing enzymes in an experimental model of human neo-intima. Thrombosis & Haemostasis 99, 373–381.
23 Cummings, I., George, S., Kelm, J., Schmidt, D., Emmert, M.Y., Weber, B., Zünd, G., Hoerstrup, S.P. (2012). Tissue-engineered vascular graft remodeling in a growing lamb model: Expression of matrix metalloproteinases. European Journal of Cardiothoracic Surgery 41, 167–172.
24 Hoerstrup, S.P., Cummings, I., Lachat, M., Schoen, F.J., Jenni, R., Leschka, S., Neuenschwander, S., Schmidt, D., Mol, A., Günter, C., Gössi, M., Genoni, M. & Zund, G. (2006). Functional growth in tissueengineered living, vascular grafts: Follow-up at 100 weeks in a large animal model. Circulation 114, Suppl. 1, I159–I166.
25 Kelm, J.M., Emmert, M.Y., Zürcher, A., Schmidt, D., Begus Nahrmann, Y., Rudolph, K.L., Weber, B., Brokopp, C.E., Frauenfelder, T., Leschka, S., Odermatt, B., Jenni, R., Falk, V., Zünd, G. & Hoerstrup, S.P. (2012). Functionality, growth and accelerated aging of tissue engineered living autologous vascular grafts. Biomaterials 33, 8277–8285.