2.2.3. Models of atherosclerotic plaque complications
In advanced stages of lesion development, the atherosclerotic plaque is
more vulnerable, unstable and prone to rupture. Indeed, plaque rupture
frequently leads to luminal thrombosis which is the direct cause of
acute ischaemia responsible for cardiovascular deaths and this
phenomenon is accelerated in the diabetic setting[66, 67]. However,
plaque rupture is rarely detected in mouse models of atherosclerosis, a
limitation with respect to replication of human pathology to
recapitulate the human disease. Nonetheless, several experimental models
have proven advantageous in analysing post-rupture thrombotic events. In
general, to induce unstable plaques, an invasive approach such as the
placement of a perivascular carotid cuff or ligation of the common
carotid artery under western diet-fed conditions in the
ApoE-/- mouse can achieve an environment of altered
haemodynamic flow to replicate the advanced stages of atherosclerotic
plaque development with evidence of an occlusive thrombus[68].
Another method commonly used to develop the unstable plaque phenotype is
through the elevation of Angiotensin II, either endogenously via
surgical clipping of the renal vasculature, or via Angiotensin II
infusion through the placement of osmotic mini pumps. High angiotensin
II levels in the western-diet fed ApoE−/− mice exhibit
accelerated lesions with characteristic features of a thinner fibrous
cap, larger lipid necrotic core, and increased macrophage content, all
of which contribute to plaque destabilisation and vulnerability[69,
70]. Another interesting model that displays features of advanced
unstable atherosclerotic plaque is a model harbouring a heterozygous
mutation in the fibrillin-1 gene (C1039+/-) which
leads to the fragmentation of elastic fibres in the vessel wall,
increased arterial wall stiffness and highly sporadic plaque rupture
when crossed to the ApoE-/-[71, 72]. When fed a
western diet for 35 weeks,
ApoE-/-Fbn1C1039G+/- mice exhibit a
3-fold increase in necrotic core, which was associated with increased
T-cell infiltration, augmented neovascularization and intraplaque
haemorrhage[72]. Furthermore, spontaneous plaque rupture was
observed in the ascending aorta and brachiocephalic arteries in more
than 50% of ApoE-/-Fbn1C1039G+/-mice[72]. In line with human complications, these
ApoE-/-Fbn1C1039G+/- mice showed
coronary plaques and myocardial infarction as well as sudden
death[72]. Thus, this reproducible model of advanced atherosclerosis
and plaque rupture with added complications of myocardial infarction is
clinically relevant.
While these models are useful in studying the pathways involved in
plaque rupture, they are not entirely ideal as they are generated by an
artificial environment often driven by genetic manipulation or altered
biomechanical stress. Therefore, there is a need to develop preclinical
models that represent the spontaneous plaque rupture that occurs in
humans. A more innovative approach has been developed by Chen et al who
showcase a unique mouse model of atherosclerotic plaque instability that
is proving to be useful for mechanistic insights and advantageous for
therapeutic testing[73]. With the hypothesis that low shear stress
and high tensile stress contribute to plaque instability in patients,
computational haemodynamic modelling was used to develop a unique
tandem stenosis mouse model of plaque instability/rupture. The tandem
stenosis model involves applying two sutures to the right common carotid
artery in the ApoE-/- mouse fed a high-fat western
diet, resulting in an unstable plaque reflecting the characteristics
seen in humans. These features include ruptured fibrous caps,
intraplaque haemorrhage, large necrotic cores, plaque inflammation,
intravascular occlusive thrombus formation and near-infrared
autofluorescence. Importantly, this model of plaque instability and
rupture is suitable for testing the responsiveness of pharmacological
interventions including plaque-stabilising drugs and other inflammatory
mediators, however, its role in the diabetic setting is yet to be
determined.
Experimental analysis of diabetes-associated atherosclerosis
In most preclinical animal models, atherosclerotic lesion size is the
first quantitative measure performed to determine if genetic loss or
overexpression as well as pharmacological intervention has an impact on
diabetes-associated atherosclerosis. Generally, lesions are observed in
1) aortic root/sinus and ascending aorta, 2) aortic tree which spans the
arch, thoracic and abdominal regions and 3) the brachiocephalic
(innominate) artery. Quantification of lesion area in the aortic sinus
and ascending aorta is described in detail by Paigen et al[74] and
Daughtery and Whitman[75]. This involves sectioning the whole frozen
aortic root spanning approximately 300µm and the ascending aorta
(400µm). Acquaintance with the architecture of the sinus region and the
three valve leaflets is essential for the accurate sectioning of this
region of tissue requiring considerable skill. Staining of selected
sections 30µm apart is then performed using Oil Red O, a fat soluble dye
for the staining of triglycerides and lipids in atherosclerotic plaque.
The advantage of obtaining serial sections through the sinus region is
that cellular composition of the plaque (collagen, macrophages, smooth
muscle cells and necrotic core) can be determined by histological (H&E,
picosirius red, masson’s trichrome) and immunohistochemical staining
(CD68 antibody to detect macrophages). Additionally, inflammation and
oxidative stress markers can be quantified by performing
immunohistochemistry and immunofluorescence using specific antibodies
[31, 76-79].
A relatively quick and simple method to quantify atherosclerotic lesions
on the intimal surface of the aortic tree is by an en facemethod, whereby the formalin-fixed aorta is stained with Sudan IV, a
lipid soluble dye, after which the aorta is cut open, pinned flat and
imaged. This mode of analysis is normally one dimensional with lesion
area being the main assessment, however, the aorta can then be embedded
in paraffin for immunohistochemical analysis to determine inflammatory
and oxidative stress markers. Furthermore, using a separate cohort of
mice, the aorta can be snap frozen for RNA and protein extraction to
determine atherogenic proteins and pathways by qRT-PCR and Western
Blotting respectively[20, 76, 80, 81]. Similarly, the innominate
artery which is connected to the aortic arch can be dissected out and
sectioned to assess lesion area using histological stains. Lesions in
the innominate artery develop the complexities of late stage plaque,
characterized by an atrophic media, perivascular inflammation, and
thinning of the fibrous cap which in some instances can lead to plaque
rupture and intraplaque haemorrhage.
A major limitation of these methods is that the analysis does
not provide volumetric three-dimensional assessment of the plaque, nor
the degree of vascular occlusion. Often tissue integrity is compromised
during preparation and processing for histology. To overcome these
issues, over the past decade, there have been improvements in imaging
technologies to provide a more in-depth view of plaque complexity. For
example, an ex vivo scanning method, known as microscopic
computed tomography (microCT) has been adapted for the visualisation of
atherosclerotic plaque[82]. Excised hearts and vessels from diabetic
ApoE-/- mice, can be stained en bloc with metal
solutions and scanned using microCT. Then, an image analysis software is
able to recreate a virtual histological section to view multi-focal
images in both 2-dimensional and 3-dimensional formats, allowing
volumetric assessments of the atherosclerotic lesion[82]. The tissue
is intact and can further be processed for traditional histology and
immunohistochemistry[82]. In addition, ex vivohigh-resolution three dimensional MRI is able to sensitively and
volumetrically quantitate atherosclerotic lesions in the mouse aortic
root and innominate artery with specific insight into microanatomy of
the lipid-rich necrotic core and fibrous cap components which can be
verified by histology, allowing for tracking of lesion progression over
time[83]. Using fluorescence emission computed technology, Htun et
al were able to detect autofluorescence in the near-infrared range that
correlated with plaque instability and vulnerability, establishing this
technique as a vital imaging tool for the detection of high risk plaques
in patients[84].
The small size of mouse aortae can limit the availability of
tissue for analysis, often requiring separate cohorts of mice to be run
for extensive histological analysis as well as gene expression and
protein quantification, making these studies time, labour and cost
intensive. To circumvent this, a new emerging technology, known as Laser
Capture Microdissection (LCM) is gaining significant interest. This
technique can be performed on frozen or paraffin-embedded sectioned
tissue. LCM enables subpopulations of tissue cells to be collected under
the microscope using a low-power infrared laser, which harvests cells of
interest directly by first removing unwanted cells to give
histologically pure enriched cell populations[85, 86]. Furthermore,
this technology preserves the integrity of the tissue and does not
damage RNA, DNA and protein, therefore, enabling a number of downstream
molecular analyses to be performed. The precision with which tissues are
dissected enables the extraction of lesion-specific areas allowing
researchers to explore cell populations and differentially expressed
genes that make up the different plaque regions. Indeed, LCM of the
atherosclerotic lesion showed a significant enrichment of foam-cell
specific RNA transcripts[87]. In addition, this cell-specific
approach enables analysis of genetic and pharmacological manipulations
of atherosclerotic lesions.
Over the last decade, research has focused on examining cell-specific
populations and markers that are responsible for disease progression. As
such, flow cytometry has proven useful to sort, identify and quantify
cell subpopulations within the atherosclerotic aorta. In particular,
flow cytometry has been used to track leukocyte infiltration,
composition and subtypes in aortae, which is significantly heightened in
disease settings[88]. A major advantage of flow-cytometry is the
ability to simultaneously stain for surface and intracellular markers
using a range of antibodies to identify different cell types and their
functional stage[88, 89]. This enables both quantitative and
qualitative results from one sample. Using this technique, Nagareddy et
al elegantly demonstrated significant increases in pro-inflammatory
circulating monocytes (Ly6-Chi subset) and neutrophils
but not lymphocytes, which were significantly lessened after glucose
lowering, suggesting a role for leukocytosis in promoting
diabetes-associated atherosclerosis[90]. Recently, Brennan et al
cultured diseased carotid arteries ex vivo from diabetic and
non-diabetic patients who underwent carotid endarterectomies and
analysed tissue cytokine responses in response to anti-inflammatory
treatment with a pro-resolving mediator known as LipoxinA4[91].
Additionally, other groups have utilised this method to perform flow
cytometry of immune cell subsets[92]. This reproducible technique
allows the investigation of important aspects of atherogenesis,
particularly the function of immune cell and cell-cell interactions
within the tissue from a human disease perspective[91, 92].
In line with flow cytometry data, more recently, single-cell RNA
sequencing (scRNAseq) technologies allow the evaluation of transcriptome
information from a single cell to reveal cell population differences at
the RNA level. Moreover, a key feature is the ability to detect
heterogeneity among individual cells to define the cellular landscape
via cell maps[93]. Since its inception a few years ago, several
research groups have employed scRNAseq technology to uncover the
transcriptome-based cellular landscape of human and murine
atherosclerotic plaque, with a particular focus on unrecognized immune
cell populations and their role in atherogenesis[94-97]. Indeed,
current work is underway to profile the transcriptional landscape in
diabetic atherosclerotic aortae using scRNAseq.
Lastly, as atherosclerosis is a dynamic process involving the
interaction between different cell types, it is pertinent to explore the
use of real-time functional assays. As an initial step in atherogenesis,
vascular inflammation triggers the cell adhesion cascade, contributing
to the recruitment and infiltration of leukocytes to the endothelium.
Intravital microscopy is now widely used to visualise and record
leukocyte rolling and adhesion using whole blood perfusion in
real-time[98]. As this method is performed on vessels ex
vivo , it does not require invasive surgery in animals. Additionally,
this technique has been validated in diabetes-induced vasculopathies
enabling localised inhibitor treatment to evaluate the role of
therapeutics in modulating leukocyte-endothelial interactions[20, 98,
99]. Vessel myography ex vivo is another important technique to
assess endothelial viability and function, which are critical early
markers for the development of diabetes-associated atherosclerosis. This
technique allows researchers to evaluate various parameters including
the assessment of vascular tone, endothelium-dependent and independent
function and nitric oxide bioavailability, which can be determined in
the presence or absence of pharmacological interventions/genetic
manipulations[20, 100]. A summary of experimental and research
outputs to evaluate diabetes-associated atherosclerosis is shown in
Table 1.
Animal models of diabetes associated microvascular diseases
In diabetic patients, it is apparent that microvascular and
macrovascular complications often develop simultaneously. Microvascular
complications are characterised by functional and structural organ
damage arising from changes affecting capillaries and arterioles. Three
major manifestations of diabetic microvascular disease are nephropathy,
retinopathy and neuropathy. While animal models that describe diabetic
microvascular diseases have been extensively reviewed[101-103], we
aim to highlight the most promising murine models that most closely
align with three debilitating human microvascular pathologies, diabetic
nephropathy, retinopathy and neuropathy.
Diabetic nephropathy