Valvular Interstitial Cells: Physiology, Isolation, and Culture

Abstract

Valvular interstitial cells (VICs) are the primary cellular component of the heart valve. Their function is to maintain the structure of the valve leaflets as they endure some three billion beats in the course of a human lifespan. Valvular pathology is becoming ever more prevalent in our ageing world, and there has never been a greater need for understanding of the pathological processes that underpin these diseases. Despite this, our knowledge of VIC pathology is limited. The scientific enquiry of valve disease necessitates stable populations of VICs in the laboratory. Such populations are commonly isolated from porcine and human tissue. This is achieved by digesting valve tissue from healthy or diseased sources. Understanding of the many VIC phenotypes, and the biochemical cues that govern the transition between phenotypes is essential for experimental integrity. Here we present an overview of VIC physiology, and a tried-and-true method for their isolation and culture. We make mention of several biochemical cues that the researcher may use in their culture media to ensure high quality and stable VIC populations.

1. Introduction

The four valves of the mammalian heart are delicate, complex structures that ensure unidirectional flow of blood through the heart’s four chambers. Two atrioventricular valves, the tricuspid and mitral, sit between the atria and ventricles. During systole, these valves are closed to prevent the backflow of blood into the atria. The semilunar valves—so named for their half-moon shaped leaflets—sit between the ventricles are their outflows: the pulmonary trunk and the aorta. During systole, these valves are open, allowing blood to exit the heart. During diastole, they slam shut, preventing the blood from flowing backwards. The valves themselves comprise two main structural features: the valve leaflets (also called ‘cusps’) and the fibrocartilaginous ring onto which they are mounted. The leaflets are beautifully thin sheets of tissue organised into three distinct layers each with a definite function. The upstream layer, the ‘ventricularis’ of the semilunar valves and the ‘atrialis’ of the atrioventricular valves is rich in radially oriented elastin, allowing the leaflet to hinge open and closed. The middle layer, the ‘spongiosa’, is rich in glycosaminoglycans, forming the bulk of the matrix and structurally linking the superficial layers. Finally, the downstream surface, the ‘fibrosa’, is rich in circumferentially oriented collagen that confers strength against the mechanical load of the blood pressure when the valve is closed [1, 2].

Two key cell types are responsible for maintaining the structure and function of the heart valve leaflet. Valvular endothelial cells (VECs) form a confluent monolayer on the surface of the valve. These cells are phenotypically distinct from endothelial cells elsewhere in the vasculature.

The second key cell type, and the object of this chapter, is the valvular interstitial cell (VIC). These are effectively fibroblasts that populate the extracellular matrix of the valve leaflets.

Functionally, VICs are responsible for maintaining the matrix of heart valve over the course of growth and development. This is no mean feat—the heart valves must endure some three billion beats over their lifetime.

2. Valvular interstitial phenotypes and their physiology

There are five recognised phenotypes of valvular interstitial cell: embryonic, quiescent, progenitor, activated, and osteoblastic.

2.1 Embryonic VICs

Embryonic VICs are responsible for valvulogenesis—the formation of the valves from the endocardial cushion during fetal development. The precise detail of how this occurs in not well understood, but several key features of these embryonic VICs are known. Firstly, they are highly proliferative, a feature which distinguishes them from other VIC phenotypes [3]. Second, is that these cells are endothelial in origin, having undergone the endothelial-to-mesenchymal transition coordinated by a complex sequence of cell signallers. Notch and TGF family ligands (such as BMP-2 and VEGF) induce TGF-β and Wnt/β-catenin signalling which supresses endothelial-type gene expression in the endothelial cells, and instead encourages the cell to become embryonic VICs, a mesenchymal cell type [4, 5, 6]. Once this transition is complete, the maturing valve and its embryonic VICs ‘striate’ the matrix into the three aforementioned layers. Precisely how this occurs is unknown, but recent evidence suggests that the differentiation is aided by the haemodynamic forces of the fetal circulation [7].

2.2 Progenitor VICs

The VIC population is not a static one. Rather, a subpopulation of VICs present in the valve tissue in adulthood express surface markers of progenitor cells. Studies on healthy porcine valves show that approximately 5% of VICs express ABCG2, NG2, or SSEA-4 [8]. The origin of these progenitor cells in not entirely clear. Evidence suggests that some of these progenitor cells migrate to the valve from the blood stream—studies in mice have shown that bone marrow progenitor cells marked with green fluorescent protein appear in the valve, and assume the role of qVICs by secreting collagen 1 [9]. Other researchers suggest that another source of pVICs are VICs already residing within the valve, and are ‘activated’ in response to injury [3].

2.3 Quiescent VICs and activated VICs

Quiescent VICs (qVICs) maintain the matrix throughout adult life [3]. They are neither very metabolically nor mitotically active. Instead, their role in maintaining the valve matrix is by sensing mechanical forces, and responding by differentiating into activated VICs (aVICs). The activated VIC is a myofibroblastic phenotype—characterised by the synthesis of ECM proteins and α-smooth muscle actin (a contractile protein) [10]. The differentiation from qVIC to aVIC is the key feature of mechano-regulated growth of this tissue from fetus to adult. As the heart grows, the pressures across the valve leaflet increase, introducing undue stress to the ECM. qVICs resident in the ECM can sense this stress, and differentiate into aVICs that then secrete matrix proteins in order to strengthen the matrix, until such a time when the matrix is able to support the increased pressure. At this time the ‘stress’ stimulus is lost and the aVIC then dedifferentiates back into a qVIC. It should be no surprise that the fetal heart valve contains a high proportion of aVICs, and that the healthy adult valve contains nearly no aVICs [11].

This mechano-regulation of heart valve growth and development relies on the ability of the qVIC’s mechano-sensing ability. This is achieved through three surface protein classes: integrins, cadherins, and mechano-sensitive ion channels [12]. Integrins bind the VIC to the matrix, forming a continuation of the cytoskeleton intracellularly with the ECM extracellularly. Cadherins are transmembrane proteins that bind other cells—in the case of VICs, cadherin 2 is upregulated in qVICs, and cadherin 11 in aVICs [12]. Mechano-sensitive ion channels, particularly calcium channels, also enable mechano-sensation by altering ion concentration within the cell in response to stress-induced deformation [13]. Collectively, these three mechano-sensing modalities induce a transcriptional change within a qVIC in response to stress, which causes aVIC differentiation, upregulation of ECM component synthesis, and the production of α-SMA [10].

2.4 Osteoblastic VICs

Osteoblastic VICs (obVICs) are, as the name suggests, bone-producing cells that arise from VICs in the context of heart valve calcification. Up to 13% of the cellular content of end-stage diseased valve leaflets are positive for osteoblast markers RUNX2 and ALP, and histology of diseased leaflets shows evidence of lamellar bone [14]. Exactly where the obVICs come from is not certain. Some authors suggest that obVICs arise exclusively from qVICs in the context of age-related valve disease [10], while others suggest that obVICs come from a restricted range of susceptible VICs or perhaps pVICs [14].

3. Valvular interstitial cells in health and disease

3.1 From qVIC to aVIC and back again: dynamic reversibility of VIC activation

The key feature of VIC physiology that underpins normal growth and development, and the pathogenesis of heart valve disease, is the reversibility of the transition between qVIC and aVIC [15]. This reversibility (depicted in Figure 1), or the failure thereof, is perhaps best illustrated by two examples: growth and development by fetal heart valves, and the development of calcific aortic valve disease. Each involves substantial activation of VICs and the subsequent production of matrix.

Figure 1.

Shows the ‘dynamic reversibility’ of VICs as they transition between their quiescent and activated states, as well as the key protein expression markers associated with each phenotype.

The fetal heart valve is an immature structure—not yet required to bear the forces of systolic blood pressure while the fetal circulation, with its shunts, remains active. Once a baby is born, however, those shunts close, and the semilunar valves must immediately assume their role in preventing blood from flowing backwards into the ventricles from their respective outflows. The transition from immature fetal valve to fully competent valve requires several histological changes. First, the immature, loosely organised ECM network of the early valve must become more structured and load-bearing. Secondly, the resident VICs must proliferate in order to adequately populate the valve matrix, and will need to change their phenotype to meet the increased mechanical needs of the growing valve [11].

The immature valve matrix has a lower proportion of elastin than the adult valve, and a much greater proportion of glycosaminoglycan [11]. Interestingly, the proportion of collagen—the primary ECM component of the valve by weight—appears to be fairly stable from the first trimester through to adulthood. What does change markedly, however, is the degree of ‘organisation’, assessed by various authors using methods such as picrosirius red staining viewed under polarised light [11], or by quantifying the degree of hydroxyproline cross-linking [16]. In the first instance, collagen fibre thickness and alignment is significantly increased from fetus to child, and in the second, hydroxyproline crosslinking is significantly elevated. The net effect is substantial increase in the strength of the valve matrix.

How do the resident VICs achieve this? The answer lies in the presence of aVICs. The cell population of the fetal valve is up to 30% activated VIC, as determined by α-SMA staining [11]. These cells contribute to the changes in valve ECM makeup by simultaneous upregulation of matrix proteins (collagen, for example) and matrix metalloproteases (MMPs) [17]. The combination of upregulated ECM structural proteins and the enzymes that break them down is ‘remodelling’—i.e., the overall proportion of collagen remains static, but the maturation into strong, organised bundles proceeds.

In the fetal valve, the laydown of collagen in the correct orientation is driven by the mechano-sensing ability of the resident VICs. But as we age, there are situations where the dynamic reversibility of qVIC and aVIC phenotypes is upset, and the resulting products of VIC activation lead to disease. The prescient example is calcific aortic valve disease (CAVD)—the most common valvular pathology of the developed world, with prevalence rates of up to 15% in those over 75 years old [18]. In the early phase of this disease, aVIC activation and the accumulation of structural ECM components mimics the changes seen in fetal valves, only in this circumstance, the interruption of the ‘dynamic reversibility’ of aVICs is lost, and the valves become stiff and non-compliant [19, 20]. Valves affected by CAVD are characterised by the overproduction of collagen [21] and the dysregulation of MMP expression [22].

The ‘dynamic reversibility’ of VIC phenotypes and the mechanical properties of the valve matrix are a true ‘chicken and egg’ problem. VIC activation invariably leads to stiffening and matrix remodelling (either physiological in the case of the fetal valve, or pathological in the case of CAVD), which, in turn, results in a stiffer matrix, which then serves a signal for qVICs to activate [23].

3.2 VIC calcification: murky origins

The origin if the calcification in heart valve disease is not perfectly understood. What is known is that VICs can calcify via two distinct mechanisms: calcific nodule formation and obVIC differentiation (depicted in Figure 2) [10]. The relative contributions of these two mechanisms to heart valve disease, or if the two mechanisms are independent of one another is not known.

Figure 2.

Shows the two phenotypes of VIC calcification: the dystrophic calcific nodules that occur after aVIC apoptosis, and the ossification of obVICs. What is unclear is if the obVIC phenotype arises directly from qVICs or aVICs.

Current evidence suggests that calcific nodules arise from aVICs, and are in fact calcified apoptotic bodies left after the cell dies. This type of calcification is termed ‘dystrophic’ to distinguish it from the matrix calcification caused by obVICs [24]. Where the obVICs arise from, is not clear. Some have suggested that obVICs arise from the qVIC population independently, and in response to changes in matrix mechanics or the accumulation of calcium [25], while others suggest that obVICs arise from aVICs in a manner that depends ‘activating’ signalling [26].

Support for the idea that the obVIC and aVIC calcification are distinct pathways is seen in the apparent changes in α-SMA expression, which is markedly reduced in presence of osteogenic medium and markedly increased in the presence of TGF-β. Conversely, qVICs cultured in osteogenic medium will demonstrate upregulation in key osteoblast markers: ALP and OCN [25]. To further confuse the true origin of obVICs, the same pattern of gene expression occurs when valvular endothelial cells are cultured in osteogenic media [27].

3.3 VIC signalling molecules

The various signalling molecules that induced change in VICs has been extensively reviewed elsewhere [10]. There are, however, three molecules that commonly feature in culture medium preparations, and the physiology that underpins their use is included here.

3.3.1 TGF-β

Perhaps the most well characterised stimulus for VIC activation is transforming growth factor β (TGF-β) [28, 29, 30, 31, 32, 33, 34]. TGF-β, or more specifically the isoform TGF-β1 is one member of a much larger ‘TGF-β superfamily’, which includes 30 species such as bone morphogenic proteins (BMPs), growth and differentiation factors, and various other peptides key to growth and development [35]. In adulthood, TGF-β1 is involved in a myriad of cell functions, from cell fate determination, cell migration, response to injury, and ECM regulation [34, 36, 37]. In the context of VICs, the key function of TGF-β is its ability to regulate activation of VICs and maintain homeostasis of the valve ECM by induction of the aVIC phenotype. As well as structural proteins, ECM contains large amounts of ‘latent’ signalling molecules—signallers bound up with latency-associated peptides. These peptide complexes are hidden from cells, and so do not enact their functions until they are cleaved from the matrix during proteolysis following injury [36, 38]. TGF-β, as a promoter of ECM production, serves as a ‘switch’ in heart valves: when the valve is unable to cope with a tension force, microscopic injury releases TGF-β from the matrix, which in turn signals to cells to secrete more ECM, until such time that the matrix can withhold the tension and the TGF-β switch is turned off. It is through delicate control of TGF-β stimulated ECM secretion that a valve grows and remodels throughout the changing haemodynamic environment from fetal development to adulthood [6, 39].

TGF-β acts on cells via several biochemical pathways. The most well characterised pathway with respect to VIC physiology is the ‘canonical TGF-β’ pathway, or Smad pathway [34, 38, 40]. TGF-β forms a homodimer in the extracellular space and binds two type 2 TGF-β receptors (Tgfbr2), which then recruit two type 1 TGF-β receptors (Tgfbr1) forming an activated receptor-ligand complex [39]. The activated Tgfbr1 is then able to phosphorylate intracellular messenger proteins Smad2 and Smad3, which form a complex with Smad4 and move to the nucleus as transcription factors to activate the associated genes [39, 41]. It is this process that is responsible for the activation of qVICs to myofibroblasts and for the development of dystrophic nodules [3].

Non-Smad pathways initiated by TGF-β are termed ‘non canonical’. These include the ERK/MAPK pathway [41], which has been shown to regulate key aspects of aVIC expression in a manner independent of the canonical TGF-β pathway [42].

3.3.2 FGFs

The fibroblast growth factor (FGF) family comprises 23 distinct proteins, 18 of which are ligands to various FGF receptors, effecting a wide range of cellular functions: from migration to proliferation to cell survival [43, 44]. Of interest to the VIC researcher is FGF-2, also called basic FGF (bFGF), so named for its discovery in the alkaline fraction of pituitary extracts [45]. FGF-2 has been implicated in numerous biological processes, from cell migration and fate determination to pathological signalling in cancer metastasis [46]. Most interesting is that while this molecule is present and active throughout the body, a distinct physiological role has not been identified. Homozygous FGF-2 knockout mice survive perfectly well and give birth to healthy young.

FGF-2’s signalling pathway first involves heparin binding [46]. Two heparin-FGF2 complexes then bind two FGF receptors (FGFR) to form a stable dimer, which then activates intracellular tyrosine kinase domains and effects changes in gene expression through a number of signalling cascades, among them MAPK and AKT pathways [43].

FGF-2 has been shown to be particularly important in regulating the quiescent VIC phenotype [34, 47, 48, 49]. The mechanism of FGF-2’s effect appears to be multifaceted: both repressing the canonical TGF-β pathway by blocking Smad nuclear translocation [48], and by activating of the Akt/mTOR pathway [34, 49]. The downstream effects of this activation have been shown to oppose the effect of TGF-β. FGF-2 appears to promote the qVIC phenotype: VICs treated with FGF-2 have a spindle-shaped morphology, and express more vimentin and less α-SMA [17].

3.3.3 Nitric oxide

The function of nitric oxide (NO) in the vasculature is understood to be a local control mechanism for maintaining blood pressure and perfusion. First, NO is produced in endothelial cells by endothelial nitric oxide synthase (eNOS) in response to stretch or biochemical stimuli [50]. Then, NO is able to rapidly diffuse into the surrounding tissue by virtue of its small size, where it binds guanylyl cyclase (a cytoplasmic receptor) [51] and leads to a myriad of downstream effects including relaxation of the vascular smooth muscle [50]. With respect to heart valves, NO production by VECs and its subsequent action of VICs is an important paracrine regulator of valve cell function [3, 10]. More importantly, dysregulation of NO signalling is implicated in the development of heart valve disease [52, 53, 54, 55]. In the initial stages of CAVD, eNOS is upregulated on the aortic-side endothelium [56], where it contributes to pathological neovascularisation of the valve tissue [57]. The early eNOS response is a protective mechanism to temper the proliferation of VICs, but as overexpression of eNOS continues, resistance to NO develops—a phenomenon termed ‘NOS decoupling’ [58].

With respect to VIC activation, exogenous NO inhibits the formation of calcific nodules in porcine VIC cultures under TGF-β stimulation [59]. Culturing VICs in media supplemented with DETA-NONOate—a biologically inert molecule that spontaneously releases two NO molecules in a predictable manner [60]—limits the formation of new nodules.

This effect is not purely induced by exogenous NO, rather it can be replicated by co-culturing VICs with NO-producing VECs [61]. Under these conditions, VECs can inhibit the formation of aVIC-related calcific nodules in a manner that is reversed by addition of a NO scavenger. This suggests that of the multitude of paracrine signals that travel between the endothelium and the interstitium, NO is paramount in maintaining the valve in its quiescent state.

4. Sources of valvular interstitial cells

There are a number of VIC sources for a researcher’s consideration. No one option is without issue. Perhaps the easiest source of VICs is from animals. Pigs are by far the most common source of VICs for research [61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81], as they are very easily obtainable from meat processing facilities. As a biproduct of industry, obtaining and using porcine VICs requires little or no ethical oversight [82]. Another notable benefit is the anatomical similarity between the pig and human heart. This is of particular importance as the mechanical and structural forces regulate VIC physiology and gene expression [23, 68], and so choosing a structurally similar heart approximates a physiologically similar VIC. In a similar vein, pigs housed for research can be induced into diseased states for investigation of a particular pathology. For example, pigs fed on high-cholesterol diets experience atherosclerotic lesions, though no porcine model of calcific valve disease has been trialled to date [52, 83]. It should be noted that VICs derived from meat industry pigs can be considered healthy qVICs by definition—their valves are not experiencing rapid growth, nor has age-related disease begun to influence their phenotype. This fact is either a boon or an issue for the cell culturist. Those interested in studying the development of valve disease might be eager to use a healthy population of qVICs, while those searching for therapeutic options for persistent aVIC or obVIC presence will likely have to choose another source. In our experience, porcine VICs are easily obtained and fairly forgiving when it comes to cell culture. Cultures derived from digested porcine valves are adequately proliferative and less sensitive to aVIC formation when cultured on hard plastic.

Sheep may represent a promising animal alternative for obtaining VICs liable to calcify. Sheep exhibit an elevated calcium metabolism compared to humans, and this has led to their use as an experimental model of heart valve calcification [84]. Several authors have made use ovine VICs for experimentation [85, 86].

Human VICs are the best option for studying heart valve disease, as none of the animal models approximate our valvular pathologies perfectly. The most ideal source is a young, healthy adult human heart valve. Cells from this source, in theory, are almost entirely qVICs. However, there are a number of constraints that make sourcing healthy human tissue difficult.

First is availability—unlike other cell types, VICs are not sourced from a healthy tissue that is routinely removed from people undergoing surgery (as it the case in, for example, adipose tissue removed during breast cancer surgery), and suitable heart valve tissue from younger donors is contained within a heart that is likely more suitable for heart transplantation. To complicate matters further, in the instance where a young donor heart is not suitable for transplant, the heart valve may be harvested as a ‘homograft’ (a heart valve transplant).

The second constraint is time. While VICs are relatively robust in the absence of nutrient supply [10], good laboratory practice dictates that cells should be isolated and incubated as quickly as possible. Very rarely are healthy heart valves available within temporal or spatial proximity to a laboratory. Despite these constraints, various researchers have made use of healthy human VICs, typically from donor hearts not suitable for either transplantation or homograft use [47, 87].

A much more commonplace source of human VICs is those isolated from patients undergoing valve replacement surgery. Valve replacement surgeries are very common—our experience at Auckland City Hospital saw an average of 5 valves available per week. The obvious problem with this source is the inherent disease state of the cells within the valve. Though our protocol above recommends removal of the more calcified portion of the valve, the pathology is not localised at the calcification, and altered stiffness characteristics across the whole leaflet ensures that all cells present will have some degree of pathological programming [53].

The diseased state of these cells may actually be useful for researchers concerned with the study of aortic calcification—the most common valvular pathology in the developed world. Many examples of diseased valves used for this purpose are available in the literature [47, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99, 100, 101, 102].

The key consideration to address with diseased human VICs is ethics. Patients must consent to their tissue being harvested and used in research, and protection of the tissue and any data generated must be ensured by the researcher. In addition, many cultures have unique requirements for the disposal or return of human tissue [103].

5. Isolation protocol

This isolation protocol was developed by our laboratory for the isolation of human VICs from patients undergoing valve replacement surgery [104]. An abbreviated flow chart is included in Figure 3.

Figure 3.

Shows and abbreviated flow chart of steps 7–18.

5.1 Consumables

• Collagenase II (Sigma, #C6885)

• Sterile phosphate buffered saline

• Sterile DMEM (Sigma, #D5649) supplemented with sodium bicarbonate per manufacturer’s instructions

• Fetal bovine serum (FBS)

• 50 mL containers

• 15 mL centrifuge tubes

• 50 mL centrifuge tube

• 90 mm petri dishes

• Sterile guards

• #10 Scalpel blades and handle

• Sterile forceps

• 70 μm cell strainer (Bio-Strategy, #BDAA352350)

• 0.2 μm sterile filter (Bio-Strategy, #SART16534-K)

• 30 mL syringe

• T25 cell culture flasks (Sigma, #C6356)

• 25 mL serological pipettes

• Disposable ‘squeezy’ pipettes

5.2 Hardware

• Tissue culture hood

• Incubator, 37°C at 5% CO₂

• Orbital shaker-incubator, 37°C

• Sterilisation services (i.e., autoclaving service, sterile filtration capabilities)

• Balance

• Centrifuge

• Electric pipette filler

• Microscope with haemocytometer slide

• Tube rack

5.3 Method

Prior to tissue collection:

1. In the tissue culture hood, prepare at least 100 mL of DMEM supplemented with 10% FBS, and warm to 37°C

2. Prepare a > 20 mL aliquot of sterile PBS for tissue collection, and warm to 37°C

3. Turn on the shaker-incubator to equilibrate to 37°C

4. Weigh out 30kU collagenase II. Note that the mass of the collagenase is lot-number specific - refer to the specification sheet for the number of units per milligramme. Add the collagenase II to a 50 mL container containing 30 mL PBS, gently agitate to dissolve

5. In the tissue culture hood, aspirate the collagenase II into the 30 mL syringe. Attach the 0.2 μL filter to the syringe and gently push the collagenase through the filter into a sterile container. Collagenase II loses its activity within hours, and should be prepared as close to the time of digest as possible. Divide the collagenase into 2 portions of 15 mL each in 50 mL containers. Label them ‘Digest 1’ and ‘Digest 2’

6. Once sealed, remove ‘Digest 2’ from the hood and weigh the container, recording the mass to the nearest milligramme. This will allow weighing of the usable valve tissue in a sterile manner

   In the operating room:

7. Make sure to bring the warmed aliquot of PBS for tissue collection. When the surgeon removes the valves, ensure the scrub nurse or assisting surgeon offers the container in a sterile manner, and that the tissue is added to the container and sealed shut all while in the sterile field. Return to the laboratory as quickly as possible

   In the laboratory:

8. Using sterile forceps, add the leaflets to the ‘Digest 1’ container

9. Add to the shaker-incubator for 10 min at 37°C with gentle agitation

10. Once complete, arrange the leaflets on a 90 mm petri dish, and scrape down the outside using a scalpel blade. If the calcifications make this too difficult, then consider removing the calcified sections before scraping. Rinse the valves using a squeezy pipette and sterile PBS

11. If hVECs are desired, collect the rinsing media and pool this with suspension in ‘Digest 1’, centrifuge, and resuspend the cell pellet in 1 mL of DMEM +10% FBS. Note that the expected pellet is small, and the suspension should be plated in a 6-well plate or smaller

12. If not already done so, remove calcified sections of the valve. Finely mince the tissue using the scalpel, and transfer to the tissue to ‘Digest 2’

13. Weigh ‘Digest 2’ and calculate the difference from recorded mass in step 7. If the mass is less than 500 mg, then there is a low likelihood that the culture will generate enough cells to succeed. We recommend proceeding only if the tissue mass exceeds 500 mg

14. Add ‘Digest 2’ to the shaker-incubator at 37°C for 2 h with gentle agitation

15. In a tube rack, assemble the 50 mL tube with the 70 μm cell strainer. Pour the digestate through the strainer, and agitate the contents using a squeezy pipette

16. Centrifuge at 1200 RPM for 2 min

17. Discard the supernatant, and, without disturbing the pellet, rinse the tube with 15 mL of DMEM +10% FBS to remove any residual collagenase. Spin briefly to re-pellet any cells that may have been disturbed

18. Discard the supernatant. Resuspend the cell pellet in 2 mL of DMEM +10% FBS. Take and 30 μL sample and pipette directly onto the haemocytometer. Do not perform cell count immediately - instead, get the cells into the incubator as quickly as possible: immediately add the suspension to T25 culture flask and make up to final volume with 8 mL of DMEM +10% FBS. Add to the incubator at 37°C

19. Perform a cell count, and record. Note that digests that yield <100,000 cells are unlikely to be successful

20. After the digest:

21. Monitor the culture daily for confluency and signs of myofibroblastic activation. hVIC cultures generally reach 80% confluency in 8 days. Monitor for signs of myofibroblast differentiation, including pleiomorphism and non-fusiform morphologies

22. Replace entire medium every four days

23. When using hVICs in experiments, make sure to seed at a density of 10,000 cells/cm², and anticipate that growth will plateau within 5 days

6. Quality control considerations

Regardless of cell source, the goal for VIC culture is a population of phenotypically stable cells (i.e. few aVICs) with as little endothelial contamination as possible.

To assess phenotypic stability, we encourage researchers to plate out a sample of their VIC culture and assess it for features of myofibroblast activation in aVICs. Typically, we >5000 cells from our digestate and plated these in a 24-well plate, and left to grow for up to 6 days. We then assessed for the presence of aVICs first by visual inspection on phase microscopy, and then by α-SMA staining. aVICs differ from pVICs in their morphology. On phase microscopy, qVICs are slender, spindle-shaped cells with a central body surrounded by two or three thin processes. Their area rarely exceeds 5000μm², and their aspect ratio (the ratio between their major and minor axes) should be greater than 3 [17]. aVICs on the other hand are much larger, with areas often exceeding 10,000μm². Their shape is much more circular, with a large cell body and loss of the thin processes. Similar cell area measurements were found by Porras and colleagues, who used FIB media to prevent (or reverse) aVIC formation in porcine cultures [72].

We suggest staining for α-SMA and vimentin to assess the degree of aVIC presence (an example of which can be seen in Figure 4). aVICs have dense bands of α-SMA running across the cell, terminating in adhesion complexes that link the intracellular force generated by this protein to the ECM [105]. True qVICs will have very little α-SMA staining, if at all.

Figure 4.

Shows two cultures of VICs: one rich in qVICs on the left, and another rich in aVICs. Cells are stained for vimentin (green) and α-SMA (red). Cells were cultured for four days on plastic. Scale bars represent 100 μm.

Many researchers will quantify aVIC numbers by using α-SMA positivity, either by fluorescent intensity measurement or thresholding numbers of positive cells [47, 48]. In our experience, there is little utility in attempting to quantify the degree of VIC activation using α-SMA staining or phase microscopy, as all VICs will show some α-SMA expression and variation in morphology. More intensive methods for quantifying α-SMA include western blotting and real time PCR [47, 106]. These are comparative methods (i.e., they require two groups), and so are of little use in the ‘quality control’ stage. Experimentally, we have found that expression of ACTA2 (α-SMA’s gene) and a number of other aVIC-associated genes correlate with the presence of ‘aVIC-looking’ cells on phase microscopy and α-SMA staining. We suggest that researchers subjectively assess the appearance of their VIC cultures on day 6 post-isolation.

The second main feature of VIC quality control is endothelial contamination (Figure 5). Despite the initial digest and scraping we suggest in the protocol, it is impossible to remove all VECs. Endothelial contamination can run as high as 10% [10]. In our experience, VIC cultures generated by the means outlined above will produce cultures with less than 1% von Willebrand factor (vWF) positivity. To assess VEC contamination, we suggest first saving and plating the digestate from digest 1 in a 6-well plate or smaller (see step 11) for use as a positive control. In the same plate, we suggest reserving a small volume of the VIC suspension from digest 2, and culturing both for up to 6 days. We then suggest staining the two wells for an endothelial marker. The choice of marker is not especially important: VECs express vWF, CD31 (PECAM) and CD34, and are negative for α-SMA [74, 107].

Figure 5.

Shows a quality control procedure performed on a VIC culture (left). The yellow arrow indicates VEC contamination. This was performed alongside a VEC culture as a positive control. This culture was grown from ‘Digest 1’ aspirate. Both cultures were grown for four days on plastic. Scale bars represent 100 μm.

7. Culture conditions

The majority of the literature has VICs cultured on flat plastic in Dublecco’s modified Eagle medium (DMEM) supplemented with 10% fetal bovine serum (FBS). We have had good results with this formulation. There are, however, a number of other culture media that should be considered. Alternatives to DMEM include M199 medium [29, 108] and DMEM-F12 [73, 109].

Good laboratory practice stipulates that isolated cells should be cultured in such a way that preserves their in vivo phenotype as faithfully as possible. And so, it befalls the VIC researcher to consider the inherent aVIC promoting properties of chosen media. The supplementation of media with 10% FBS is standard practice for the culture of any fibroblast [110]. However, careful thought should be put into the inherent content of peptides that stimulate VIC activation. TGF-β, for example, is present in FBS in concentrations ranging from 10 to 20 ng/mL. TGF-β is a strong promoter of VIC activation, and several authors have mitigated its effect by using a 2% FBS medium supplemented with insulin as a mitotic agent. This so-called ‘FIB’ media, first used by Latif and colleagues [47, 72], is also supplemented with 10 ng/mL FGF-2—an agent shown to reduce VIC activation in favour of the quiescent phenotype. We have demonstrated that VICs cultured in FIB media show significantly downregulated expression of aVIC-associated genes (α-SMA in particular), and a morphology more typical of qVICs (greater aspect ratio, and reduced total cell area) compared to VICs cultured in the standard DMEM preparation [17]. The inherent ability of FIB medium to suppress aVIC formation can be potentiated further by culturing VICs on collagen-coated plates [72].

Induction of the obVIC phenotype requires osteogenic medium. This invariably contains ascorbic acid, dexamethasone, and β-glycerophosphate [28, 109, 111], supplemented in some instances with TGF-β or BMP-2 [89].

Seeding densities vary widely in the literature, and will differ depending on the surface on which the VICs are grown. On hard plastic or glass, we have found VICs to be rather sensitive to contact inhibition and myofibroblastic activation. We suggest that VICs are seeded at a density of 10,000 cells/cm² for experiments running for less than five days. At this density, cell growth reaches a plateau on day 4 [104].

8. Conclusion

VICs are the specialised fibroblasts that populate the human valve and maintain its structure and function throughout our lives. The increasing incidence of valvular pathology means that there is no better time to gain understanding about the physiological and pathological processes that affect these cells. Despite this, our understanding of VIC health and disease is in its infancy. Culture of VICs is an important part of the scientific exploration of heart valve disease. Several key methodological considerations need to be addressed for the jobbing scientist:

• VICs come in a variety of phenotypes. Some are implicated in embryogenesis, while other in disease

• Biochemical signals govern the phenotypic transition of VICs between these phenotypes

• A normal, physiological dynamism exists between the qVIC and the aVIC. This is the key process that underpins normal valve growth and development

• Upsetting this dynamic transition is implicated in the development of valvular disease. Chief among them is calcific aortic valve disease

• Phenotypically stable VIC cultures can come from pigs and humans

• Diseased VICs are the most readily attainable human cell source, and with careful consideration of culture medium preparations, stable cultures of qVICs can be readily and easily obtained

• qVICs are liable to become aVICs in culture. The researcher is therefore tasked with ensuring the culture conditions avoid errant VIC activation

Acknowledgments

This work was supported by the Green Lane Research and Educational Fund and the Health Research Council of New Zealand.