In Vivo Models for the Study of Fibrosis

Abstract

Significance: Fibrosis and scar formation pose a substantial physiological and psychological burden on patients and a significant public health burden on the economy, estimated to be up to $12 billion a year. Fibrosis research is heavily reliant on in vivo models, but variations in animal models and differences between animal and human fibrosis necessitates careful selection of animal models to study fibrosis. There is also an increased need for improved animal models that recapitulate human pathophysiology.

Recent Advances: Several murine and porcine models, including xenograft, drug-induced fibrosis, and mechanical load-induced fibrosis, for different types of fibrotic disease have been described in the literature. Recent findings have underscored the importance of mechanical forces in the pathophysiology of scarring.

Critical Issues: Differences in skin, properties of subcutaneous tissue, and modes of fibrotic healing in animal models and humans provide challenges toward investigating fibrosis with in vivo models. While porcine models are typically better suited to study cutaneous fibrosis, murine models are preferred because of the ease of handling and availability of transgenic strains.

Future Directions: There is a critical need to develop novel murine models that recapitulate the mechanical cues influencing fibrosis in humans, significantly increasing the translational value of fibrosis research. We advocate a translational pipeline that begins in mouse models with modified biomechanical environments for foundational molecular and cellular research before validation in porcine models that closely mimic the human condition.

Geoffrey C. Gurtner, MD, FACS.

Scope and Significance

Wound healing in adult mammals results in scar formation and fibrosis that can alter the physiology of affected tissues and cause functional impairment.¹ Effective therapeutics to treat and/or prevent skin fibrosis are currently unavailable, largely due to our lack of a comprehensive understanding of the underlying cellular and molecular mechanisms of fibrosis. In this review, we revisit various in vivo models of fibrosis, across different species and disease states, which were developed to better understand the mechanistic basis for scarring and fibrosis following injury.

Translational Relevance

The limited translation of biomedical research investigating fibrosis to date can largely be attributed to variability between animal models and the human condition.^2,3 Among different animals, significant differences have been observed in terms of inflammatory responses and subsequent fibrosis. Increasingly, these differences are attributed to variable biomechanical environments and tissue characteristics.^4–8 In this study, we address the advantages and disadvantages of small and large animal models across a variety of fibrotic disease states and propose a translational pipeline that seeks to leverage the strengths of each.

Clinical Relevance

Fibrosis poses a substantial clinical burden on patients and a significant financial burden on society, with skin fibrosis alone costing up to $12 billion annually.⁹ Fibrosis is most noticeable in the skin, but affects almost all adult tissues. The sequelae of fibrosis range from functional impairments from cutaneous scarring to idiopathic pulmonary fibrosis (IPF) causing restrictive lung disease, to liver fibrosis leading to significant morbidity and mortality. Understanding the mechanisms driving fibrosis to inform the development of therapeutic strategies relies on in vivo models of disease that are as analogous to the human condition, and therefore as clinically relevant as possible.^3,7

Background and Overview

Adult mammalian wound healing is a poorly understood process consisting of three temporally overlapping phases—inflammation, proliferation/tissue formation, and remodeling¹⁰—that results in the formation of a scar.⁷ Scars are largely composed of fibroblasts and type I collagen. This fibrotic “patch” response to injury restores tissue integrity, but fails to recapitulate the form and function of the native tissue.¹ Consequently, fibrotic healing poses a substantial physiological and psychological burden on the individual and a significant public health burden on the economy.⁹

A substantial focus of wound healing research is understanding the processes underlying fibrosis to curtail them in favor of promoting regenerative healing.¹ Fibrosis in response to injury is not limited to the skin, but occurs in almost all adult tissues. For instance, following myocardial infarction, fibrosis prevents ventricular free wall rupture, but as myocardium is replaced by nonfunctional tissue, impaired contractility can lead to heart failure and ventricular arrhythmia.¹¹ IPF, a lung disease of unknown etiology, results in replacement of healthy alveolar tissue with abnormal extracellular matrix (ECM), reducing lung compliance, leading to respiratory failure.¹² Fibrosis is most apparent in skin and when dysregulated, results in formation of hypertrophic scars (HTS) or keloids¹³ after injuries ranging in severity from minor trauma to major burns.^1,10

Both in vitro and in vivo models have improved our understanding of wound healing and fibrosis. Although in vitro studies have the advantage of focusing on specific cell types or targets, these approaches fail to recapitulate the three-dimensional, dynamic environment of the wound, including complex cell–cell interactions.³ Consequently, animal models are critical to elucidate the intricate mechanisms governing wound healing and fibrosis and inform the translation of effective clinical therapies. In this review, we discuss various animal models of fibrosis, highlighting their strengths and weaknesses, with a specific focus on murine and porcine models.

Discussion

Mediators of fibrosis

Our group and others have implicated aberrant signaling in several pathways, including FAK, TGF-β, PI3-Akt, and RhoGTPase, in scar pathophysiology.⁷ Transforming growth factor (TGF)-β has been shown to cause abnormal healing, in part, through mediators such as the Wnt/B-catenin pathway. Many of these pathways also play a role in the fibrosis of solid organs. Lung, cardiac, and renal tissue are responsive to Notch signaling, which drives the expression of TGF-β, alpha smooth muscle actin (α-SMA), and promotes fibrosis.⁷ The aforementioned Wnt/B-catenin pathway has also been implicated in patients with IPF and mouse models of pulmonary fibrosis, as well as cardiac, renal, and hepatic fibrosis.¹³

There is increasing evidence that unique subpopulations of fibroblasts and other cells drive scar formation.¹⁴ During cutaneous wound healing, matrix metalloproteinases remodel the ECM from type III to type I collagen while myofibroblasts mediate wound contraction.³ During dysregulated fibrosis, cells with a myofibroblast phenotype have been shown to excessively deposit ECM and promote contraction.⁴ These cells express α-SMA, collagen I, and fibronectin.⁵ More recently, multiple discrete lineages of fibroblasts have been identified with varying functions.¹⁵ Driskell et al. demonstrated that reticular fibroblasts in the lower dermis were primarily responsible for fibrosis, while papillary fibroblasts in the upper dermis supported epithelization and hair follicle formation (Fig. 1).^14,16 This may partially explain the absence of hair follicles in scar tissue. Our group helped identify a CD26⁺ lineage of fibroblasts, defined by embryonic origin rather than regional location that appears to be solely responsible for ECM deposition in scarring and fibrosis.¹⁴ In a mouse model, these fibroblasts are derived from Engrailed-1-expressing progenitor cells, and their appearance during development coincides with the transition from scarless, fetal healing to adult, fibrotic healing.¹⁴

Figure 1.

Schematic of fibroblast lineages. Fibroblasts are a heterogeneous population of cells with multipotent and differentiated cells. Reticular fibroblasts, located in the lower dermis, are primarily responsible for fibrosis, while papillary fibroblasts in the upper dermis support epithelization and hair follicle formation. Adapted from Driskell et al.¹⁶ Color images are available online.

Our understanding of the cell populations and signaling pathways contributing to fibrosis in various tissues is improving, yet remains incomplete.

Differences between human and animal fibrosis

Understanding the differences in cutaneous wound healing and fibrosis between humans and animals is crucial to conducting fibrosis research in animal models.

We have previously discussed the numerous differences between mouse and human skin,³ some of which we will review here. Although differences in immunology between species may also contribute, this topic is outside the scope of this article. Mouse skin has a subcutaneous muscle layer known as the “panniculus carnosus,” which in humans is largely vestigial, present in the neck as the platysma muscle and the scrotum as the dartos muscle. Contraction of the panniculus carnosus in mice results in rapid constriction of cutaneous wounds, reducing mechanical stress and accelerating healing. In human skin, wound healing largely proceeds without contraction until the final remodeling stages, subsequent to granulation tissue formation. Mouse skin also has a number of architectural differences from human skin. For instance, the hair cycle in mouse skin is 3 weeks on average, but it is more variable in human skin. Apocrine sweat glands, which are present in human skin in the perianal, inguinal, and axillary areas, are not present in mouse skin, and neither are dermal papillae nor rete ridges. For a more extensive list of differences, we refer readers to our prior review.³

In contrast, the skin of red Duroc pigs is highly similar to that of humans.³ Both human and porcine skin have a thick epidermis and comparable dermal–epidermal thickness ratio. Both have dermal papillae and rete ridges and a similar distribution and orientation of blood vessels.¹⁷ While the skin in humans and Duroc pigs display similar characteristics, there are several crucial distinctions. The epidermis of porcine skin has three layers, a stratum germinativum, stratum granulosum, and stratum corneum, unlike human epidermis, which has four layers throughout most of the body and five layers on the palms and soles. Porcine skin has numerous apocrine sweat glands, whereas eccrine sweat glands are limited to areas such as the lips, snout, and carpal organ.¹⁸ From a functional perspective, however, cutaneous wound healing in Duroc pigs closely mimics that of humans. Pigs lack a panniculus carnosus and consequently heal through granulation and reepithelialization rather than contraction, similar to humans. Also, porcine hair follicles cycle independently of neighboring hair follicles and contribute to reepithelialization as hair follicles do in human skin.¹⁸

Mouse models of hypertrophic scarring

HTS are thick, raised dermal scars caused by abnormal wound repair following injury resulting from trauma or surgery.¹⁹ Various animal models have been proposed to study HTS in the literature. One of the earliest strategies used to create HTS animal models is to xenograft human HTS samples onto a skin defect on the dorsum of immunodeficient mice.¹⁹ This strategy enables the creation of sustained HTS-like wounds in mice and has also been used to recreate keloids and burn scars. Genetically modified, humanized mouse skin has also been used as an alternative to human skin for this transplantation-based animal model. However, these immunodeficient mouse models are limited by small sample size, difficulty of maintenance, limited longevity and the bias against the role of immune cells. Chemically induced scarring has also been reported, including treatment of skin wounds with coal tar and subcutaneous injection of bleomycin, although these experiments have produced mixed results.^19,20 The xenograft and chemical induction strategies have also been applied in other animals, such as hamsters, rabbits, and rats.

Increasingly, the role of mechanotransduction in fibrosis is becoming apparent.⁸ It is known that mechanical force contributes both to physiological healing and to pathological fibrosis and scarring.⁷ For instance, human skin wounds with greater tension often form more extensive scars than those that heal with minimal tension.³ Mechanotransduction refers to the conversion of mechanical forces into biological signals that regulate downstream signaling and cellular responses. Loose-skinned animals, such as mice, heal with minimal or no obvious scarring, which we have found in prior work to be, in part, due to the low tension environment.²¹ In 2007, we developed a device to apply human levels of mechanical tension to wounds in mouse models.²¹ We found that such tension results in the formation of HTS similar in histopathology to that of human scars (Fig. 2). Specifically, they were raised, had epidermal thickening and a mast cell infiltrate, were hypervascular and had collagen sheets parallel to the applied force.^5,21 These scars were also hyperplastic, and we discovered that this was due to reduced apoptosis. A follow-up study by our group revealed that T cell-regulated pathways for the recruitment of macrophages and monocyte precursors of fibroblasts were critical to HTS formation.⁴

Figure 2.

Differences in skin elasticity between species inspired the design of a murine HTS model. (A) Human skin demonstrates greater intrinsic tension at rest compared to adult and fetal murine skin as measured by microtensiometry (n = 5 for each group, error bars indicate SD; **p < 0.01) (B) Stress loading curves as measured by microtensiometry demonstrate human skin is relatively stiff compared with murine adult and fetal skin (n = 5 for each group, error bars indicate SD) (C) The biomechanical loading device was engineered from expansion screws and titanium surgical Luhr plates. (D) Loading devices were placed over each of two 2 cm linear incisions on the mouse dorsum. One wound was left unloaded and served as the internal control, while the other was subjected to mechanical loading and served as the experimental wound. Adapted from Aarabi et al.⁴ HTS, hypertrophic scar.

This unique model not only provides a reliable murine model to study HTS but also reveals the crucial role of focal adhesion kinase (FAK) in mechanotransduction.⁵ We have shown that application of mechanical forces leads to activation of FAK in fibroblasts, resulting in increased scar formation (Fig. 3). After cutaneous injury, loading forces activate FAK, promoting collagen production and the secretion of monocyte chemoattractant protein-1 (MCP-1), a chemokine known to play a role in fibrosis, via extracellular-related kinase (ERK).⁵ In these studies, FAK activates numerous downstream components involved in fibrogenic responses such as PI3K/Akt and mitogen-activated protein kinases (MAPK). We have also shown that FAK regulates keratinocyte function.^5,22 FAK-deleted keratinocytes preferentially express integrin-linked kinase (ILK) and suppressed paxillin signaling, which are known components of focal adhesion complexes.²² Similarly, FAK is directly involved in macrophage chemotaxis through integrin signaling and likely plays a role in fibrosis as well.²³ The full extent of FAK's role in fibrosis is still under investigation. Furthermore, we have shown that offloading mechanical stress around scars in large animal models and human subjects attenuates activation of mechanotransduction pathways and results in reduced scarring. Although native murine healing occurs in a reduced skin tension environment, modifying the biomechanical environment permits the use of mouse models to study fibrosis.

Figure 3.

Schematic of FAK mechanotransduction in fibrosis. Schematic of the proposed vicious cycle of hypertrophic scarring driven by mechanical activation of local and systemic fibroproliferative pathways through fibroblast FAK. Adapted from Wong et al.⁵ FAK, focal adhesion kinase. Color images are available online.

Porcine models of hypertrophic scarring

Numerous studies have demonstrated that the pathophysiology of porcine wound healing and HTS formation closely recapitulates that of humans.¹⁷ Duroc pigs have been used as a large animal model of HTS for over 40 years. It can develop thick dermal scars analogous to human HTS (Fig. 4).²⁴ At the cellular level, studies have shown that red Duroc HTS have increased level of fibrotic myofibroblasts, mast cells, and collagen nodules similar to human HTS.²⁵ Moreover, in Duroc pigs, the thickness of the initial inflammatory tissue layer correlates with the thickness of the ultimate scar; another similarity shared with human HTS.²⁴ Many studies have used Duroc pigs to investigate healing of surgical incisions, large excisional wounds, and contact burn wounds.^24,25 In this regard, red Duroc skin serves as an excellent model system to study new experimental therapeutics and medical devices for wound healing and scar management.

Figure 4.

Porcine HTS formation 5 months after creation of excisional wounds of various depths. Acute wounds of various depths (TDS) lead to HTS formation of various thicknesses after 5 months in female red Duroc pigs. Adapted from Zhu et al.²⁴ TDS, total dermatome settings. Color images are available online.

In a previous work by our laboratory, we have also shown that scars in Duroc porcine models form similarly to that of humans, and in particular, that scar formation is closely modulated by endogenous and exogenous mechanical stress on the wound.²⁶ We have also shown that the use of a polymer to offload mechanical stress across a wound during healing resulted in reduced scar formation during excisional wound healing in human phase I participants, and reduced scar formation in both excisional and incisional wound healing in Duroc pigs. The reduction in scar formation was not complicated by delayed healing or wound dehiscence.²⁶

The main challenges associated with porcine models of wound healing are logistical. Pigs are expensive to maintain and challenging to manage due to size.²⁵ Moreover, transgenic pigs are only available for specialized use,²⁵ unlike transgenic mice, which are quite straightforward to obtain. As such, modifying the environment of murine healing to more closely recapitulate that of humans provides a powerful research tool.

Animal models of burn-induced fibrosis

Cutaneous burn injuries, affecting ∼50,000 patients annually, lead to devastating fibrosis and scar formation causing disfigurement and psychological distress.²⁷ Full-thickness burn injuries and deep partial-thickness burns almost always result in HTS. There is an incomplete understanding of the unique mechanisms underlying fibrosis following deep burns, and accordingly, various animal models of burn injury have been developed. Challenges exist with each animal model since resultant scars can significantly vary depending on temperature, time of exposure, and pressure applied with various tools.

Mouse

Although there are substantial differences between rodent skin and human skin, mouse models are the most widely used species in studying burn injury. This is partly due to the availability of a large variety of mouse-specific biochemical reagents and transgenic animals. Several studies report the use of heated metal rods, brass bars, or hot air blowers to create full-thickness burn wounds in mice.²⁸ Partial-thickness burn wounds, however, have been difficult to create using these tools, another limitation of using mice in burn research.^17,24,29,30 Finally, there are studies reporting that in contrast to the hypermetabolic response to burn injury in humans, mice are resilient to the burn-induced metabolic response, even with 30% total body surface area (TBSA) being affected.^28,30 Therefore, although mice are convenient models to study cellular and molecular pathways in burn healing and HTS formation, they may not fully recapitulate the human pathophysiology.

Rat

Although rats and humans share many aspects of organ physiology, rats, such as mice, are small loose-skinned animals and heal wounds by muscle contraction, although the availability of transgenic strains compared to mice is limited. A larger body surface area compared to mice is an advantageous feature, but there are inherent differences to human skin physiology, including expression of enzymes that human skin lack, which are thought to have a significant role in collagen synthesis and wound healing.^28,30 These discrepancies should be considered when using rats for burn wound models.

Domestic pigs

As mentioned previously, pig skin has substantially similar anatomy and physiology to human skin compared to other animals.¹⁷ Recent studies have used heated metal devices with controlled pressure to create full and partial-thickness burn wounds in pigs.²⁴ Following wounding, recapitulation of debridement procedures routinely performed in the clinic is important; hence, standardization of excision methods to remove necrotic tissues may be necessary. For scar formation studies, the red Duroc breed, which develops thick postinjury HTS, has been most extensively studied; although Yorkshire breeds are also being reported as valid burn models.^18,24 Despite their fidelity, pig burn models are unpopular, largely due to the high cost, labor intensity, and ethical issues associated with burning large mammals that require intensive pain management.

Animal models of foreign body reaction

The function and longevity of implanted biomedical devices are significantly limited by the fibrotic reaction they elicit.³¹ The host reaction to implanted biomaterials, termed foreign body reaction (FBR), begins as a wound healing response, but the persistent presence of the biomaterial results in sustained fibrosis and scar tissue formation. The scar tissue isolates the biomaterial from the surrounding microenvironment and leads to implant failure (Fig. 5). A range of small and large animal models have been proposed for investigating FBR to implanted biomaterials, including mice, rats, sheep, canine, porcine, and nonhuman primate models.³² Furthermore, nontraditional models to study FBR such as using zebrafish and multicellular in vitro models have been reported recently, but need to be further developed and validated.^33,34 Mice and rats are the most commonly used animal models to study FBR, but large animal models are preferred for cardiovascular implants such as stents and artificial heart valves as well as some orthopedic implants, including artificial joints or craniofacial implants.³²

Figure 5.

Schematic of the foreign body response. Thrombotic agents and other blood proteins adsorb to the surface of implanted biomaterials to form a provisional matrix. Activated platelets, fibrinogen, and biochemical agents within the matrix direct neutrophils and monocyte-derived macrophages to the implantation site. At the tissue/implant interface, macrophages fuse to form foreign body giant cells. Persistent inflammatory signaling activates collagen-secreting fibroblasts at the biomaterial site, resulting in the formation of a fibrous capsule that can persist for the life of the implant. Adapted from Major et al.³¹ Color images are available online.

The appropriate animal model to study FBR has been the subject of controversy with previous surveys showing that two-thirds of positive animal trials reported in the literature do not translate into human clinical trials.^2,6 This disconnect has been largely attributed to variability in animal models. For example, mice are known to elicit weaker FBR than rats and porcine models.^35–37 Similarly, significant differences have been observed between FBR in the fibrotic reaction in rats and Gottingen minipigs, with the fibrotic reaction in large animal models, which are more similar to humans, being relatively accelerated and more exaggerated. It has been suggested that the differences between small animals such as mice and large animals such as pigs or humans is attributable to the biomechanical environment around the implants. Tight-skinned animals such as pigs and humans are expected to exhibit higher compressive forces on biomedical implants compared with loose-skinned mice.⁶ Regardless, murine models, particularly mice, are preferred because of low cost, ease of handling, and availability of transgenic mice. Therefore, there is a critical need to develop improved murine models to study FBR.

The most common sites of implantation to study FBR are the subcutaneous space (SQ), intraperitoneal cavity (IP), and the brain.^38–41 SQ implantation is commonly used because most implants, such as glucose biosensors, pacemakers, and breast implants are all placed in a subcutaneous pocket in the clinical setting. In the IP model, macrophages and other inflammatory cells are recruited to the implant, but a robust fibrous capsule does not form because the implant is not embedded in tissue, unlike the SQ model. This makes retrieval and study of the implant surface more convenient, but is less reflective of the clinical condition. Biomaterials implanted in the brain are designed to study brain-electrode interfaces.

Research with these animal models has revealed various factors that contribute to fibrosis.^31,32 Delivery of anti-inflammatory and antifibrotic drugs to the implant site have been proposed to reduce fibrosis around biomedical implants.⁴² Increasing evidence points to the fact that biomaterial stiffness, size, shape, and chemical composition are critical to the fibrosis elicited, but more research is needed to define optimal design considerations. Furthermore, there is increasing evidence that mechanical stress, device motion, and overall biomechanical environment around the biomedical implant significantly contributes to FBR.^6,43 In summary, development of improved animal models, which mimic human-like FBR and a targeted focus on the role of mechanotransduction in FBR, may lead to reliable therapies to limit FBR and improve implant longevity.

Other models of fibrotic disease

Fibrotic development underlies a wide range of diseases in various organs, many of which lack effective treatment. Such diseases can cause comorbidities or even death, instilling a need for effective animal models to optimize the study of their etiologies. Thus, several models have been developed over the past few decades to study diseases such as scleroderma, lung fibrosis, and liver fibrosis, which are briefly discussed below.^44–46

Scleroderma, a dermatological disease characterized by abnormal hardening of the dermis via accumulation of ECM proteins, can be experimentally induced using bleomycin injection. Although the bleomycin model presents some of the early pathological features of skin thickening, the model does not replicate the autoimmune nature of scleroderma, and therefore, its clinical relevance is limited.

Among the more widely used animal models is the bleomycin model of pulmonary fibrosis. The exact mechanism by which bleomycin induces a fibrotic response is still unknown, although it is attributed, in part, to injury, inflammation, and reactive oxygen species production.⁴⁵ Bleomycin has been shown to be effective in dogs, sheep, rats, hamsters, and mice.⁴⁵ The development of several IPF-treating drugs has been conducted with the use of bleomycin animal models.⁴⁵ Furthermore, the initial injury following the application of bleomycin has been shown to closely resemble acute lung injuries found in other diseases that may lead to fibrosis, such as those in acute respiratory distress syndrome.⁴⁷ For all of its advantages, the bleomycin model has shortcomings, including its rapid induction of pulmonary fibrosis being inconsistent with the progressive development found in humans.⁴⁸

Several animal models have been created to study hepatic fibrosis as well. One of the more prevalent models uses carbon tetrachloride (CCL₄), which leads to fibrosis in rodents after sustained administration.⁴⁹ Similar to bleomycin, the CCL₄ model varies in effectiveness based on mouse strain.^50,51 Due to its simple and reproducible nature, the model has been used in over various studies worldwide.⁴⁹ These studies have shown increases in hepatic stellate cells as well as expression of TGF-α and amphiregulin associated with CCL₄-induced hepatic fibrosis.⁵²

Summary

Numerous in vivo models of fibrosis exist across various species and for the study of a number of disease states (Table 1). Chemically induced fibrosis models are popular, but are not always reflective of the pathogenesis of the cutaneous diseases one hopes to study. Similarities between pig and human skin make porcine models ideally suited for research on dermal and foreign body related fibrosis. However, these animals are costly and difficult to handle, and their use carries ethical considerations.

Table 1.

In vivo models of fibrosis

Model	Clinical Relevance	References
Mouse mechanical load	Hypertrophic scar (HTS)	4,5
Mouse xenograft	HTS, keloid, burn	19
Dermal bleomycin	HTS	19,20
Mouse contact burn	Full-thickness burn	28
Subcutaneous foreign body	Capsule formation	38
Intraperitoneal foreign body	Inflammatory cell recruitment	39,41
Brain foreign body	Neuroprosthetics	40
Carbon tetrachloride	Hepatic fibrosis	51,52
Lung bleomycin	Idiopathic pulmonary fibrosis, ARDS	45,47
Duroc pig	HTS, full-thickness burn, partial-thickness burn	24–26

ARDS, acute respiratory distress syndrome.

The use of mice is widespread in biomedical research due to the low cost, ease of handling, and the availability of genetically characterized or manipulated strains. As such, mouse models that more closely resemble the human condition, particularly in the context of the biomechanical environment, are an indispensable component of fibrosis research. We advocate a translational pipeline that begins in mouse models with modified biomechanical environments for foundational molecular and cellular research before validation in porcine models that closely mimic the human condition. Familiarity with these models will enable the development of fibrosis research programs with greater potential for clinical impact.

Take-Home Messages.

• Fibrosis research is heavily reliant on in vivo models, but differences between different animal models and between animal and human fibrosis complicate interpretation of findings.

• Three major types of murine HTS models have been described, including the xenograft model, drug-induced fibrosis, and mechanical load-induced fibrosis. Burn-induced fibrosis is initially investigated using murine models, which use heated metal rods, brass bars, or hot air blowers. Porcine models of fibrosis use similar strategies, but better recapitulate human-like fibrosis.

• Fibrosis around biomedical implants is investigated using murine models extensively. Subcutaneous implantation, intraperitoneal model and brain implants are used to study fibrous capsule formation, inflammatory cell recruitment, and neuroprosthetic fibrosis, respectively. Murine models do not recapitulate the mechanical environment around implants that is observed in porcine models and humans, but are still widely used because of the ease of handling and availability of transgenic strains.

• Fibrosis of the lungs, liver, and other organs are investigated using drug-induced fibrosis models, including bleomycin and CCL₄-induced fibrosis.

• The development of murine models, which recapitulate the mechanical cues that influence fibrosis, are the need of the hour in fibrosis research. Combining research using mouse models for foundational molecular and cellular research and subsequent validation in porcine models that closely mimic the human condition is the best strategy for fibrosis research going forward.

Abbreviations and Acronyms

α-SMA

alpha smooth muscle actin

CCL4

carbon tetrachloride

ECM

extracellular matrix

FAK

focal adhesion kinase

FBR

foreign body reaction

HTS

hypertrophic scar

IP

intraperitoneal cavity

IPF

idiopathic pulmonary fibrosis

SQ

subcutaneous space

TGF

transforming growth factor

Author Disclosure and Ghost Writing

There are no conflicts of interests to be disclosed. The article was written by the authors, and ghostwriting services were not used.

About the Authors

Geoffrey C. Gurtner, MD, FACS, is the Johnson and Johnson Professor in the Department of Surgery at Stanford University and his laboratory studies wound repair, fibrosis, and foreign body reaction. Jagannath Padmanabhan, PhD, Sun Hyung Kwon, PhD, and Zeshaan Maan, MD are postdoctoral fellows in Dr. Gurtner's laboratory, studying fibrosis and scar formation. Revanth Kosaraju, BS, and Clark A. Bonham, BS, work as research fellows in the Gurtner's laboratory.