In Vitro Model of Human Cutaneous Hypertrophic Scarring using Macromolecular Crowding

Summary

This protocol describes the use of macromolecular crowding to create an in vitro human hypertrophic scar tissue model that resembles in vivo conditions. When cultivated in a crowded macromolecular environment, human skin fibroblasts exhibit phenotypes, biochemistry, physiology, and functional characteristics resembling scar tissue.

Abstract

It has been shown that in vivo tissues are highly crowded by proteins, nucleic acids, ribonucleoproteins, polysaccharides, etc. The following protocol applies a macromolecular crowding (MMC) technique to mimic this physiological crowding through the addition of neutral polymers (crowders) to cell cultures in vitro. Previous studies using Ficoll or dextran as crowders demonstrate that the expression of collagen I and fibronectin in WI38 and WS-1 cell lines are significantly enhanced using the MMC technique. However, this technique has not been validated in primary hypertrophic scar-derived human skin fibroblasts (hHSFs). As hypertrophic scarring arises from the excessive deposition of collagen, this protocol aims to construct a collagen-rich in vitro hypertrophic scar model by applying the MMC technique with hHSFs. This optimized MMC model has been shown to possess more similarities with in vivo scar tissue compared to traditional 2-dimensional (2-D) cell culture systems. In addition, it is cost-effective, time-efficient, and ethically desirable compared to animal models. Therefore, the optimized model reported here offers an advanced “in vivo-like” model for hypertrophic scar-related studies.

Introduction

Scar tissue represents the endpoint of tissue repair. However, in many individuals, especially those suffering from burns or trauma¹, scarring can be excessive and impose undesirable effects on the morphology and functioning of healed skin. Although the exact mechanisms of pathological (hypertrophic scars and keloids) scar formation are not fully understood, excessive deposition of collagen during wound healing has been demonstrated to be an essential contribution².

It is well-established that transforming growth factor beta 1 (TGF-β1) and alpha smooth muscle actin (αSMA) play key roles in the formation of hypertrophic scars. Evidence suggests that elevated TGF-β1 directly stimulates excessive deposition of collagen via regulating the SMAD signaling pathway³. In addition, αSMA has been found to contribute to hypertrophic scar formation by regulating cell contraction and reepithelialization in the wound healing process⁴. The lack of suitable in vitro and in vivo models is a major impediment towards developing and evaluating interventions and therapies for scar remediation. The aim of this study is to utilize the existing MMC technique to construct an “in vivo-like” hypertrophic scar model that is suitable for evaluating novel and emerging scar-related interventions.

Reproducing living tissue outside of the body has been a goal for years in the scientific community. The development of in vitro techniques in the early twentieth century partly achieved this goal. Current in vitro techniques have slightly evolved from Roux's original demonstration that embryonic cells can survive ex vivo for several days in warm saline⁵. However, in vitro methodologies are mostly limited to single cell types cultivated in 2-D and do not accurately recapitulate tissues in vivo. While useful for examining cell biochemistry, physiology, and genetics, native tissues are 3-D and incorporate multiple cell types. Simple 2-D in vitro systems subject mammalian cells to highly artificial environments in which native tissue-specific architecture is lost⁶. In turn, this affects intracellular and extracellular events, resulting in abnormal cell morphology, physiology, and behaviour⁷.

The interest behind this protocol lies in the development and clinical management of hypertrophic scars and keloids. While it is well-established that dermal fibroblasts are largely responsible for the abundant production of collagens present in scar tissue, cultivating dermal fibroblasts using 2-D in vitro systems fails to reproduce the turnover of collagen observed in vivo⁸. Contemporary in vitro methods still essentially use “warm saline”, an environment completely different from that in living tissues. Tissues in vivo are extremely crowded, with proteins, nucleic acids, ribonucleoproteins, and polysaccharides, occupying 5%–40% of the total volume. As no two molecules can occupy the same space at the same time, there is little free space available and an almost complete absence of free water⁹.

The MMC technique imposes constraints affecting the thermodynamic properties of cytosol and interstitial fluids. Molecular interactions, receptor-ligand signaling complexes, enzymes, and organelles are confined and restricted from interacting freely⁹. Interactions within the pericellular environment (i.e., interstitium) are also constrained. Recent evidence confirms that high concentrations of inert macromolecules in crowded solutions perturb diffusion, physical association, viscosity and hydrodynamic properties¹⁰.

Interestingly, several popular crowding agents (i.e., Ficoll, dextran, polyvinylpyrrolidone [PVP], and sodium 4-styrenesulfonate [PSS]) are not equivalent when applied to different cell types and in different settings. In one previous study, Ficoll was reported to be less cytotoxic for mesenchymal stem cells compared to PVP. These results were interpreted to be the consequence of its neutral charge and relatively small hydrodynamic radius¹¹. In contrast, a second study found that dextran is more effective in stimulating collagen I deposition by human lung fibroblasts compared to Ficoll¹². Data from our own study suggests that Ficoll enhances collagen deposition by hypertrophic scar-derived fibroblasts, whereas PVP is toxic to these cells¹³.

It has been demonstrated that the conversion of procollagen to collagen is faster in a highly crowded in vivo environment¹⁴, while the rate of biological reactions is delayed in a diluted 2-D culture system¹⁵. We have optimized the in vitro protocol here, incorporating MMC to show the cultivation of dermal fibroblasts serving as a more “in vivo-like” model for dermal fibrosis and scar formation. In contrast to the common 2-D culture system, cultivating hHSFs with MMC stimulates the biosynthesis and deposition of collagen significantly¹³. Notably, other characteristics of fibrosis (i.e., increased expression of matrix metalloproteinases [MMPs] and proinflammatory cytokines) are also evident under this optimized MMC protocol¹³. When cultivated using this method, it is shown that dermal cells recapitulate the physiological, biochemical, and functional parameters measured in vivo.

The optimized MMC in vitro protocol has been used to evaluate the expression of collagen and other ECM proteins by dermal fibroblasts isolated from hypertrophic scar dermis and uninvolved adjacent dermis. When cultivated in MMC environments in vitro, it has been observed that hHSFs express certain characteristics (i.e., mRNA, biochemistry, physiology, and phenotype) similar to dermal hypertrophic scar tissue in vivo. The evidence indicates that physical and chemical properties are important considerations when selecting crowders and optimizing MMC conditions for cultivation in vitro.

For proof-of-principle, the MMC protocol is applied here to qualitatively and quantitatively evaluate the ability of Shikonin and its analogues to induce apoptosis. This allows evaluation of the potential applications of these naturally-derived Traditional Chinese Medicine (TCM) compounds for managing dermal scarring¹³. Notwithstanding, the simplicity, cost-effectiveness, and timeliness of this in vitro MMC protocol also satisfies recent regulations to eliminate experimentation in mammals by the EU Directive 2010/63/EU and U.S. Environmental Protection Agency (EPA).

Protocol

1. Cell culture

1. Maintain hHSFs and normal dermal fibroblasts derived from non-pathological tissue (hNSFs) in Dulbecco's modified Eagle's medium (DMEM) containing 10% fetal calf serum (FCS) and 1% v/v penicillin/streptomycin solution (P/S) at 37°C in an incubator with 5% CO₂/95% air.
2. Purchase Ficoll 70, Ficoll 400, and ascorbic acid from the appropriate companies.

2. Construction of MMC hypertrophic scar model

1. Seed the hHSFs or hNSFs (50,000/well) into a 24 well plate containing 1 mL of media in each well.
2. Place in a 37°C incubator at 5% CO₂ and leave overnight.
3. Prepare the MMC media. Based on the total volume required for the experiment, produce 10% FCS/DMEM media by mixing Ficoll 70 (18.75 mg/mL), Ficoll 400 (12.5 mg/mL), and ascorbic acid (100 μM).
4. Place the mixture into a 37°C water bath for 1 h to disperse the crowders into the solution, then sterilize the MMC media using a 0.2 μm filter.
5. Aspirate the spent media and replace with the freshly made MMC media.
6. Incubate the cells for 6 days at 37°C and 5% CO₂, changing the media every 3 days.

3. Expression of the total amount of collagen

1. Prepare Sirius red solution (0.1% w/v). Dissolve 0.2 g of Direct Red 80 powder in 200 mL of distilled deionized water (ddH₂O) with 1 mL of acetic acid.
2. Aspirate the MMC media and add 300 μL of Sirius Red solution into each well. Incubate at 37 °C for 90 min.
3. Gently rinse the Sirius Red solution with tap water and allow the plate to air-dry overnight.
4. Extract the Sirius Red stain by adding 200 μL of sodium hydroxide (0.1 M) into each well. Place the plate onto an orbital shaker for 5–10 min to fully extract the Sirius Red stain.
5. Transfer 100 μL of the extracted Sirius Red stain into a 96 well transparent plate and measure the absorbance at 620 nm using a microplate reader.

4. Expression of collagen I (immunostaining)

1. Wash the wells using 200 μL of phosphate-buffered saline (PBS, pH = 7.35).
2. Fix the cells using methanol (500 μL/well) at 4 °C for 10 min.
3. Block nonspecific interactions with 3% bovine serum albumin for 30 min at room temperature (RT).
4. Aspirate the blocking solution and incubate with 200 μL of anti-collagen I antibody (10 μg/mL) for 90 min at RT.
5. Aspirate the primary antibody and wash 3x with PBS for 5 min each.
6. Incubate with 200 μL of Goat anti-Rabbit-FITC secondary antibody (1:400 dilution) and 4’,6-diamidino-2-phenylindole (DAPI; 1: 2000 dilution) for 30 min at RT. Cover the plate with aluminum foil.
7. Discard both the secondary antibody and DAPI and wash 3x with PBS for 5 min each.
8. Visualize the fluorescence staining directly under a microscope.

5. Western blotting

1. Wash the cells 2x with PBS.
2. Add 40 μL of lysis buffer into each well and scrape the cell layer with a pipette tip. The lysis buffer contains RIPA buffer, protease inhibitor cocktail (PIC), 2 mM sodium vanadate, and 10 mM sodium fluoride.
3. Transfer the protein lysate into microcentrifuge tubes and measure the protein concentration using a protein assay as per the manufacturer's instructions¹⁶.
4. Load 10 μg of protein of each group into the wells of 4%–12% Bis-Tris protein gels. Perform sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) at 200 V for 30 min.
5. Transfer the protein to nitrocellulose membrane by running a western blot at 90 V for 90 min. Avoid formation of air bubbles between the gel and nitrocellulose membrane.
6. Block the membrane with 10 mL of blocking buffer.
7. Incubate with primary antibodies at 4°C overnight. Primary antibodies are: anti-collagen I, anti-collagen III, anti-collagen IV, anti-αSMA, anti-MMP-1, anti-MMP-2, anti-MMP-9, anti-MMP-13, and anti-glyceraldehyde 3-phosphate dehydrogenase (GAPDH).
8. Wash the membrane with 0.1% TBS-Tween 20 (5x for 5 min each). TBS/Tween 20 (1 L) contains the following: 50 mL of 1 M Tris (pH = 7.4), 30 mL of 5 M sodium chloride, 1 mL of Tween 20, and 920 mL of ddH₂O.
9. Incubate with a species appropriate secondary antibody at RT for 1 h.
10. Repeat step 5.8 and visualize fluorescence using an imaging system.

6. RT-PCR

1. Collect the total RNA using the lysis buffer mixed with 2-mercaptoethanol included in the RNA extraction assay kit.
2. Purify the RNA as per the manufacturer's instructions¹⁷.
3. Measure the RNA concentration using a microvolume spectrophotometer.
4. Perform first strand cDNA synthesis using a cDNA synthesis kit as per the manufacturer's instructions.
5. RT-PCR oligonucleotide primers are provided (precoated) in custom 96 well plates. Mix 100 ng of the cDNA samples with 10 μL of SYBR green supermix into the customized plate.
6. Increase the total volume to 20 μL/well using ddH₂O.
7. Run RT-PCR using a thermal cycler following the manufacturer's instructions: 40 cycles of denaturing at 98°C for 15 s and annealing/extension at 60 °C for 60 s. The genes tested in RT-PCR include: COL1A1, COL3A1, ACTA2, SMAD2, SMAD3, SMAD4, SMAD7, IL1A, IL1B, IL6, IL8, MMP1, MMP2, MMP3, TGFB1, and VEGF.

Representative Results

Triplicate samples were performed in each experiment, and each experiment was repeated 3x using cells from three individual patients. Data are expressed as percentages of the control group. One-way ANOVA and Tukey's post-hoc test were applied to analyze statistical differences (*p ˂ 0.05).

MMC using Ficoll at 9% FVO (fractional volume occupancy) enhances the total amount of collagen and collagen I deposition in hHSF¹³. As illustrated in Figure 1A, cell density of hHSFs significantly increased after culturing with Ficoll at 9% and 18% FVO compared to the control and MMC using PVP. Figure 1B,C indicates that Ficoll (at 9% FVO) significantly enhanced the deposition of collagen (including collagen I) compared to other MMC formulations. Quantitative analysis (Figure 1D,E) further demonstrated that Ficoll (at 9% FVO) most effectively improved the deposition of collagen.

hHSFs and hNSFs cultivated in MMC environments were found to regulate the expression of ECM species in addition to collagen¹³. Data reported in Figure 2 indicates that when hHSFs and hNSFs were cultivated with MMC, the expression of collagen IV also increased significantly. Matrix metalloproteinases (MMPs) play an important role during wound healing and scar formation, regulating ECM assembly, and remodelling¹⁸. MMPs also contribute to cell proliferation, cell migration, angiogenesis and apoptosis¹⁹. Notably, an elevated expression of MMPs was found to accumulate in hypertrophic scar tissues compared to native tissues²⁰. It was observed that the expression of MMP-2, -9, and -13 were significantly upregulated in both hNSF and hHSF cultures cultivated in MMC environments.

We also probed for the synthesis of interleukin-6 (IL-6) and vascular endothelial growth factor (VEGF); however, these were undetectable in a western blot. In contrast, RT-PCR analysis (Figure 3) revealed that the expression of IL6 was significantly upregulated, while the expression of VEGF was downregulated in hNSFs and hHSFs cultivated in MMC conditions. An increased expression of IL-6 has been demonstrated to contribute to hypertrophic scar formation²¹. Paradoxically, it is also reported that the formation of hypertrophic scars is associated with an elevated expression of VEGF²². The RT-PCR analysis performed here indicated that the expression of VEGF was greatly attenuated in hNSFs and hHSFs cultivated under MMC conditions.

Finally, the results demonstrated both 1) increased syntheses of collagens, collagen I, collagen IV, MMP-2, MMP-9, and MMP-13 de novo and 2) increased expression of IL6 mRNA in hHSFs and hNSFs. Taken together, these data indicate that cultivation of primary hHSFs and hNSFs in media formulations that include MMC results in retention of the characteristic gene expression, biochemistry, and phenotypes observed in native hypertrophic scar tissue in vivo, leading to a robust “scar-in-a-jar” model.

Figure 1: MMC enhances the total amount of collagen and collagen I deposition in hHSFs. (A) Cell morphology, (B) total collagens, stained with Sirius Red, (C) collagen I expression, demonstrated by immunofluorescence, (D) quantitative analysis of total collagen, and (E) quantitative analysis of collagen I deposition. hHSFs were cultured in media supplemented with Ficoll (9% and 18% FVO), PVP40 (18% FVO), or PVP360 (54% and 72% FVO), for 6 days. Representative images were selected. Image quantitation was performed using ImageJ and is expressed as the average percentage of the control. All experiments were repeated 3x using cells isolated from three unrelated donors. Statistical analysis was performed using one-way ANOVA with Tukey's post-test (*p < 0.05 vs. control group, error bars indicate SEM). Scale bars = (A) 0.5 mm, (B) 2 mm, and (C) 0.5 mm. This figure has been modified from a previous study¹³. Please click here to view a larger version of this figure.

Figure 2: Effects of MMC on cell protein expression. hHSFs and hNSFs were cultured under MMC conditions for 6 days. Whole cell lysates were prepared with RIPA buffer containing protease inhibitor cocktail, sodium vanadate, and sodium fluoride. Protein concentration was measured using the protein assay. Representative images are presented. For quantitative analysis, the intensities of individual protein bands were measured with densitometry, normalized to GAPDH, and converted to percentage of the hNSF in normal medium using ImageJ software. All experiments were performed 3x using cells from three unrelated donors. Statistical analysis was performed using one-way ANOVA with Tukey's post-test (*p < 0.05, error bars indicate SEM). This figure has been modified from a previous study¹³. Please click here to view a larger version of this figure.

Figure 3: Effects of MMC on cell gene expression. hHSFs and hNSFs were cultivated under MMC conditions for 6 days. Total RNA was harvested using an assay kit, and first strand cDNA was synthesized using the cDNA synthesis kit. Target gene expression was normalized to GAPDH and converted to the percentage of the hNSF in normal medium. All experiments were repeated 3x using cells from three unrelated donors. Statistical analysis was performed using one-way ANOVA with Tukey's post-test (*p < 0.05, error bars indicate SEM). This figure has been modified from a previous study¹³. Please click here to view a larger version of this figure.

Discussion

This protocol aims to optimize and authenticate an improved :scar-in-a-jar” in vitro model for human cutaneous scar tissue. Previous studies have reported the application of MMC technique to human lung fibroblasts¹², human bone marrow mesenchymal stem cells²³, and human dermal fibroblasts²³ using dextran¹², Ficoll¹², and PVP²³ as crowders. In the study reported here, the previously published protocol for hypertrophic scar-derived human skin fibroblasts was optimized with Ficoll or PVP as crowders.

The selection and concentration of macromolecular crowders are critical parameters, as they do not yield equivalent results. A previous study has reported that PVP40 (18% FVO) and PVP360 (54% FVO) significantly enhance collagen deposition and cell proliferation in dermal fibroblasts²³ (effects of PVP at FVOs >54% are untested). However, these two conditions do not work consistently for hHSFs using this protocol.

As shown in Figure 1, Ficoll significantly enhances collagen I deposition compared to the control, while PVP has no significant effects. Ficoll at 9% FVO significantly increases total amount of collagen and collagen type I compared to Ficoll at 18% FVO. In addition, it is critical to use cells of a low passage, as primary dermal fibroblasts have a limited lifespan in culture²⁴. It was chosen to use only freshly isolated hHSFs to retain the in vivo phenotype. After prolonged cultivation, primary hHSFs exhibit unusual morphology and atypical functional responses. It is also recommended that the MMC medium be supplemented with ascorbic acid, a key inducer of collagen synthesis in hHSF²⁵. Furthermore, it is suggested by other researchers to use the same antibodies listed in the Table of Materials for hHSF-related studies; however, antibodies need to be validated if applying this protocol to other cell types.

As reported in the representative results, the inclusion of macromolecular crowders was found to stimulate the expression of collagen (i.e., collagen I, collagen IV, MMPs, and IL6) in hHSFs when compared to hHSFs that were cultivated using classical non-MMC conditions. It is argued that the optimized MMC model retains aspects of hHSFs’ in vivo phenotype, recapitulating their characteristic morphology, biochemistry, physiology, and abundant extracellular matrix of cutaneous scar tissue in vivo (in contrast to existing 2-D culture approaches). We are not able to identify any similar in vitro model that is able to recapitulate similar “in vivo-like” properties. When compared to existing animal models, this MMC protocol is quicker as well as more user-friendly, cost-effective, and time-efficient. Cuttle et al. established a porcine hypertrophic scar model using thermal injury, which appears to have similar characteristics to that of human hypertrophic scars²⁶. However, in addition to the costs and time needed to maintain the animals, the experiment requires more than 3 months to complete²⁶. This optimized MMC model requires about 1 week of preparation before ready for use.

The protocol offers an advanced in vitro model for the examination of novel anti-scarring therapies. The MMC model has been used to evaluate Shikonin, a molecule previously reported to inhibit de novo formation of scars, for remediation of mature hypertrophic scars^27,28. Similar evaluation of novel compounds and interventions using classical approaches to drug discovery and proof-of-concept would require considerable resources, funds, and time. This study required minimal finances and can be completed within several months. The protocol is flexible and readily adaptable for applications in the development and assessment of novel hypertrophic scar treatments prior to animal studies.

In addition, this protocol can be further modified to develop more “in vivo-like” properties. For example, the presence of overabundant TGF-β1 is a consistent finding in hypertrophic scar tissues, mediating scar formation by stimulating collagen synthesis and deposition²⁸. TGF-β1 could be readily incorporated into the MMC protocol and further improve recapitulation of in vivo pathology. We have not yet explored the model’s full potential, which may be useful for other collagen- and ECM-related pathologies (i.e., scleroderma, pulmonary fibrosis, endomyocardial fibrosis, etc). Moreover, it would be interesting to observe the effects of MMC on cells over a longer culture period, such as 2 or 3 weeks. It is also worthwhile to further evaluate the effects of MMC on cell and ECM hierarchical architecture and alignment, particularly the orientation of collagen, as these characterizations are essential for in vivo scar tissue formation.

One of the major limitations of this protocol is the restriction of cell types. Hypertrophic scar formation involves the participation of various cell populations, and the interactions between different cell types plays an essential role in scar formation. For example, keratinocytes also play an important role in cutaneous wound healing and scar formation²⁹. Incorporating additional cell populations into this model will greatly improve its significance in future research.

Materials

Name	Company	Catalog Number	Comments
0.2 μm filter	Sartorius	16534
2-Mercaptoethanol	Sigma-Aldrich	M6250
4’,6-diamidino-2-phenylindole (DAPI)	Thermo Fisher Scientific	P36962
Alexa Fluor 680	Thermo Fisher Scientific	A-21076
Alexa Fluor 800	Thermo Fisher Scientific	A-11371
alpha smooth muscle actin (αSMA) primary antibody	Abcam	ab5694
Applied Biosystems 7500 Fast Real-Time PCR System (thermal cycler )	Thermo Fisher Scientific	4351106
Ascorbic acid	Wako	#013-12061
Bovine serum albumin	Sigma-Aldrich	#A2153
Bradford protein assay	Bio-Rad	500-0006
Collagen I primary antibody (for immunostaining)	Abcam	6308
Collagen I primary antibody (for western blot)	Abcam	ab21286
Collagen III primary antibody	Abcam	ab7778
Collagen IV primary antibody	Abcam	ab6586
Direct Red 80	Sigma-Aldrich	2610108
Dulbecco's Modified Eagle's Medium (DMEM)	Life Technologies	11996-065
Fetal calf serum (FCS)	Life Technologies	6000-044
Ficoll 400	GE HealthCare	#17-0300-10
Ficoll 70	GE HealthCare	#17-0310-10
GAPDH primary antibody	Sigma-Aldrich	G8795
Goat Anti-Rabbit secondary antibody	Abcam	ab97050
Human hypertrophic scar/normal fibroblasts (hHSF/hNSF)	Cell Research Corporation	106, 107, 108
iScript cDNA Synthesis Kit	Bio-Rad	#1708890
MMP-1 primary antibody	Abcam	ab38929
MMP-13 primary antibody	Abcam	ab39012
MMP-2 primary antibody	Abcam	ab37150
MMP-9 primary antibody	Abcam	ab38898
NanoDrop Microvolume Spectrophotometers	Thermo Fisher Scientific	N/A
Nitrocellulose membrane	Bio-Rad	10484060
NuPAGE 4-12% Bis-Tris Protein Gels	Thermo Fisher Scientific	NP0321BOX
Odyssey blocking buffer	LI-COR Biosciences	927–40000
Odyssey Fc Imaging System	LI-COR Biosciences	N/A
Olympus IX-81 HCS microscope (for immunostaining)	Olympus	N/A
Penicillin/streptomycin solution (P/S)	Life Technologies	15140-122
PrimePCR Assays	Bio-Rad		Customized primers pre-coated in 96-well plates based on requirement
Protease inhibitor cocktail (PIC)	Sigma-Aldrich	11697498001
PVP 360	Sigma	#PVP360
PVP 40	Sigma	#PVP40
RIPA buffer	Merck	R0278
RNeasy Plus Mini Kit	QIAGEN	#74134
Sodium vanadate	Sigma-Aldrich	450022
Sodium vanadate	Sigma-Aldrich	450243
SpectraMax M5 Multi-Mode microplate reader	Molecular Devices	N/A
SsoAdvanced universal SYBR green supermix	Bio-Rad	#172-5270
Tween 20	Sigma-Aldrich	P9416