The Laryngoscope C 2014 The American Laryngological, V
Rhinological and Otological Society, Inc.
Establishing Principles of Macromolecular Crowding for In Vitro Fibrosis Research of the Vocal Fold Lamina Propria Matthias Graupp, MD; Hans-J€ urgen Gruber, PhD; Gregor Weiss, PhD; Karl Kiesler, MD; Sophie Bachna-Rotter, MD; Gerhard Friedrich, MD; Markus Gugatschka, MD Objectives/Hypothesis: Vocal fold fibrosis represents a major disease burden. Screening of antifibrotic compounds could be facilitated by an in vitro fibrogenesis system. Limitations of existing models might be overcome by implication of the excluded volume effect. Study Design: In-vitro study. Methods: Vocal fold fibroblasts obtained from rats’ lamina propria were cultured in four different settings: in standard medium, under “crowded” conditions by adding inert macromolecules, under external administration of transforming growth factor (TGF)ß-1, and under a combination of both. After 5 days, supernatant and cell layer were collected and analyzed by enzyme-linked immunosorbent assay. Immunofluorescence was additionally performed. Results: Collagen-alpha1(I) deposition increased significantly under crowded conditions and after administration of TGFb21. Amounts of collagen in the cell layer were significantly higher under crowding conditions with TGFb21 compared to administration of TGFb21 alone. Conclusion: Crowding enhanced collagen deposition, resulting in more favorable conditions for studying fibrogenesis. This can be the first step toward developing a robust in vitro model for testing antifibrotic compounds. Key Words: Vocal fold scar, in vitro fibrogenesis, fibrosis, macromolecular crowding. Level of Evidence: NA Laryngoscope, 125:E203–E209, 2015
INTRODUCTION Treatment of fibrotic and scarred vocal folds (VF) is still an unresolved chapter in laryngology. To date, it has not been possible to restore scarred VFs to preinjury status due to the highly complex micro-architecture of the VF, and in particular of the trilayered VF lamina propria. Standard therapy is comprised of conservative speech therapy and surgical medialization procedures that relieve symptoms of dysphonia but cannot restore to preinjury status.1 During the last decade, significant progress occurred in exploring the cellular and biochemical mechanisms of the fibrotic cascade. Basic knowledge of cellular mechanisms of VF injury and scarring has emerged, giving deeper insights and new understandings to the complex interplay between interstitial proteins (e.g., fibronectin, decorin, fibromodulin),
From the ENT University Hospital Graz, Department of Phoniatrics (MATTHIAS G., S.B-R., K.K., G.F., MARKUS G.), Clinical Institute of Medical and Chemical Laboratory Diagnostics (H-J.G.), and the Institute of Cell Biology, Histology and Embryology (G.W.), Medical University Graz, Graz, Austria Editor’s Note: This Manuscript was accepted for publication December 1, 2014. This study received public funding from the City of Graz. The authors have no other funding, financial relationships, or conflicts of interest to disclose. Send correspondence to Markus Gugatschka, Assistant Professor MD, DMSci, Department of Phoniatrics, ENT University Hospital Graz, Medical University Graz, Auenbruggerplatz 26, A-8036 Graz, Austria. E-mail: [email protected]
glycosaminoglycans (e.g., hyaluronic acid [HA]); various extracellular matrix fibers such as collagen, procollagen, elastin) and inflammatory cytokine signaling.2 Despite this increase in knowledge, substantial obstacles to successful development of regenerative therapies remained, one of them being the difficulty of trials in humans. Although in vitro and in vivo studies in animal models of several antifibrotic compounds, such as fibroblast growth factor3 and hepatocyte growth factor (HGF),4 have shown very promising results, no such therapies have been approved for treatment in humans. Likewise, the approval of stem cell therapy is hindered by the risk of malignant transformation. The development of an in vitro fibrogenesis model that mimics VF fibrosis/fibrogenesis and antifibrosis is highly desirable. Such a model would be an invaluable tool for several reasons: It would allow rapid screening of antifibrotic compounds before going into larger expensive clinical trials. A technically well-characterized fibrogenesis model would allow targeting of the fibrotic cascade at different stages that are invisible under routine culture conditions. In addition, in vitro models using VF fibroblasts (VFF) could be a step closer to potentially eliminate animal studies. However, the development of a fibrogenesis model bears several biological and technical obstacles. Perhaps the biggest challenge is to guarantee a sufficient in vitro deposition of collagen and its incorporation into a pericellular matrix. Fibrogenic cells in monolayer culture do not lay down significant amounts of collagen in an
DOI: 10.1002/lary.25103
Laryngoscope 125: June 2015
Graupp et al.: MMC for Vocal Fold Fibrosis Research
E203
appropriate time for testing antifibrotic compounds. This is also due to the fact that cells under culture conditions find themselves in an aqueous environment that is hardly representative of its in vivo microenvironment from where they originally derive. Furthermore, collagen is a central target in fibrosis research, yet there are additional features characteristic for fibrosis such as asmooth-muscle-actin (a-SMA).5 The aforementioned transition of fibroblasts to fibrotic myofibroblasts during fibrogenesis and a survey of production and deposition of noncollagenous extracellular matrix (ECM)-proteins must be considered as well. Many of these obstacles can be overcome by implicating the principles of the excluded volume effect. This biophysical effect is created by additional macromolecules that occupy a given volume, thereby confining other molecules to the remaining space. The biochemical and biophysical implications of this effect have consequences on reaction kinetics and molecular assembly. Macromolecules drive reaction partners into closer collaboration, resulting in improved proteinfolding and protein–protein interactions.6 The confounding factor of cell culture in standard aqueous media is its characteristic difference from the natural, “crowded” state. That means that in living systems biochemical processes proceed in an environment containing a high concentrations of macromolecules (50– 400 mg/ml). The experimental approach of macromolecular crowding (MMC) has been successfully employed in areas other than laryngology. Lareu et al. showed that collagen matrix deposition is dramatically enhanced in vitro when crowded with macromolecules. Both collagen deposition and collagen cross-linking were found to be increased significantly.7 The aim of our study group was to develop an in vitro fibrogenesis model by applying the principles of MMC to laryngeal fibrosis research. In a first step, we proved the transferability of these principles to VFF and compared the production of different ECM components and collagen deposition under conventional cell culture conditions and under crowding conditions.
MATERIALS AND METHODS VFF were obtained from one male Sprague-Dawley rat, aged 5 months. The rat was humanely euthanized by intracardiac injection of ketamine 100 mg/ml after sedation with isoflurane (5%) in an induction chamber with an oxygen flow of 1.5 l/ min. After that, the larynges were excised. All study procedures were conducted according to the Austrian guidelines for animal experiments and were approved by the Austrian Ministry of Science. Under stereo-magnification, epithelium and underlying lamina propria of both VFs were carefully excised, as described earlier.8 Specimens were minced, and fibroblasts were extracted and cultured following established tissue-explant protocols.9 Growth medium consisted of Dulbecco’s modified Eagle’s medium (DMEM) (Life Technologies, Gaithersburg, MD) enriched with 10% fetal calf serum (FCS) and 1% penicillin/ streptomycin (P/S). After confluence of 90%, VFF were passaged using 0.5 % trypsin. Cells at passage 3 were used for the experiment and seeded at a concentration of 5 3 104 cells into 24 well plates (Nalgen Nunc International, NY) in 10% FCS, 5% CO2 at 37 C. After 24 hours, medium was changed to serum-free
Laryngoscope 125: June 2015
E204
medium for starvation. Another 24 hours later, wells were allocated to the different groups by changing medium again (each group consisted of three replicates). The control group was set on DMEM containing 0.5% FCS, 1% P/S, and 100 mM of Lascorbic acid 2-phosphate (as the magnesium salt, a more stable form of ascorbate10). To achieve myofibroblast differentiation, 5 ng/mL21 transforming growth factor beta-1 (TGFb21) (Sigma-Aldrich St. Louis, MO) was added, as described earlier.11 Another group was cultured under “crowded” conditions by adding macromolecules, a mixture of 37.5 mg/mL21 70 kDa Ficoll (Fc) (Sigma-Aldrich St. Louis, MO) with 25 mg/mL21 400 kDa Fc5; and for the last group, TGFb21 and macromolecules were added simultaneously. A specific formulation of Ficoll (Sigma-Aldrich) was used because it proved to show the best results in terms of collagen deposition time and crosslinking.12 After 5 days, supernatants were sampled and immediately stored at 280 C. The cell layer was detached and lysed using cOmplete Lysis M (Roche Applied Sciences, Basel, SUI, Indianapolis, IN) and stored at 280 C. Assessments of collagen-alpha1(I), collagen-III, HA, and fibronectin levels from supernatants and cell layers were carried out separately using enzyme-linked immunosorbent assays (ELISA) kit (USCN Life Science Inc., Houston, TX, USA), PRC for collagen-I-a, rat-fibronectin ELISA (from ALPCO, Salem, NH, USA), Quantikine ELISA for hyaluronan (R&D Systems, Minneapolis, MN, USA), and rat collagen type III ELISA (Qayee-Bio, Shanghai, PRC). Alpha-smooth-muscleactin (a-SMA) is a commonly used marker for myofibroblast formation and an essential part of the cytoskeleton. As such, ELISA was performed only from the cell layer (a-SMA ELISA by Cusabio, Wuhan, PRC). All assays were performed in duplicates (technical duplicates).
Immunofluorescence Fibroblasts were grown on glass chamber slides (Lab-Tek II, Nalgene Nunc International, Naperville, IL, USA) and washed with phosphate buffered saline (PBS) before being harvested. Slides were air-dried overnight and stored at 220 C. Cells were immuno-labeled using the UltraVision LP Detection System (Thermo Scientific, Fremont, CA) according to the manufacturer’s instructions. The following antibodies were diluted in antibody diluent (Dako, Glostrup, Denmark) and applied for 30 minutes at room temperature: Collagen-1 (rabbit IgG, 8.5 mg/ml, Proteintech, Chicago, IL); ACTA-2 (goat IgG, 20 mg/ml, LifeSpan BioSciences, Seattle, WA). Cells were washed three times with PBS followed by incubation with the secondary antibody (Alexa Fluor 555 Donkey Anti-goat IgG, 10 mg/ml, Alexa Fluor 488 Donkey Anti-rabbit IgG, 10 mg/ml; both Life Technologies, Carlsbad, CA) for 30 minutes. Slides were again washed with PBS, mounted with ProLong Gold antifade reagent (Life Technologies, Gaithersburg, MD, USA) and observed with a Leica DM600B fluorescent microscope (Leica, Wetzlar, Germany) connected to an Olympus DP72 digital camera (Olympus, Tokyo, Japan).
Statistical Analysis Differences of the mean were analyzed by paired t tests after proof of normal distribution using Predictive Analytics Software (PASW) statistics 18.0 (SPSS Inc., Chicago, IL, USA). 0.05 was chosen as level for statistical significance. Normal distribution was given in all parameters, as confirmed by the Kolmogorov-Smirnov test. Biological replicates were treated as independent variables, whereas the technical duplicates were averaged.
Graupp et al.: MMC for Vocal Fold Fibrosis Research
TABLE I. Results for Different Parameters in Different Conditions. Collagen Ia (ng/ml) Uncrowded
Supernatant Control TGF P valueA Cell layer
Fibronectin (mg/ml) Crowded
P valueB
Uncrowded
Crowded
P ValueB
379.5 6 23.8
3,152.8 6 94.6
0.05
9.2 6 0.9
13.1 6 1.8
0.124
3,382.8 6 122.5
3,855.5 6 226.9
0.099
15.8 6 1.6
28.3 6 3.2
0.05
0.05
0.199
0.05
0.05
Control
336.1 6 55.0
1,251.8 6 142.9
0.05
0.5 6 0.04
1.3 6 0.1
0.05
TGF P valuea
709.8 6 136.4 0.05
1,303.0 6 136.2 0.839
0.05
0.7 6 0.03 0.05
1.7 6 0.04 0.05
0.05
Hyaluronic Acid (ng/ml)
a-SMA (ng/ml) P Valueb
Uncrowded
Crowded
P Valueb
Uncrowded
Crowded
Control
429.8 6 25.5
1,053.3 6 31.9
0.05
–
–
–
TGF P valuea
1,064.8 6 31.9 0.05
1,295.8 6 58.7 0.05
0.05
– –
– –
–
Control TGF
180.0 6 20.5 323.0 6 45.0
642.5 6 73.1 712.0 6 67.2
0.05 0.05
Below range 7.7 6 0.6
9.33 6 3.8 10.0 6 1.7
– 0.118
P valuea
0.05
0.124
–
0.774
Supernatant
Cell layer
P valuea 5 comparing control and TGF; P valueb 5 comparing uncrowded and crowded; TGF 5 transforming growth factor.
RESULTS Collagen-Alpha1(I) Collagen-alpha1(I) deposition was significantly enhanced under crowded conditions compared to standard culture medium in supernatant (379.5 6 23.8 ng/ml vs. 3152.8 6 94.6 ng/ml; P 0.05) as well as in the cell layer (336.1 6 55.0 ng/ml vs. 1251.8 6 142.9 ng/ml; P 0.05). See Table I for all results. The administration of TGF-b alone significantly increased the deposition of collagen-I-a in the supernatant (379.5 6 23.8 ng/ml vs. 3382.8 6 122.5 ng/ml; P 0.05) and in the cell layer (336.1 6 55.0 ng/ml vs. 709.8 6 136.4 ng/ml; P 0.05) compared to the control. The combination of crowding and TGF-b also significantly increased collagen-alpha1(I) deposition in the supernatant and in the cell layer compared to the uncrowded control (379.5 6 23.8 ng/ml vs. 3855.5 6 226.9 ng/ml; P 0.05; resp. 336.1 6 55.0 ng/ml vs. 1303.0 6 136.2 ng/ml; P 0.05) (see Fig. 1). Moreover, the combination of crowding and TGF-b administration resulted in a significant increase of collagen-alpha1(I) levels in the cell layer compared to TGF-b administration alone (1303.0 6 136.2 ng/ml vs. 709.8 6 136.4 ng/ml; P 0.05) but not when compared to crowding alone (1303.0 6 136.2 ng/ml vs. 1251.8 6 142.9 ng/ml; P 5 0.839) (see Table I). Results of immunocytochemistry correlated to results achieved from ELISA trials. Deposition of collagen-alpha1(I) was more dense Laryngoscope 125: June 2015
under crowded conditions compared to the uncrowded control (see Fig. 2, right column).
Hyaluronic Acid (HA) Crowding led to significantly increased levels of HA in supernatants (429.8 6 25.5 ng/ml vs. 1053.3 6 31.9 ng/ ml; P 0.05) as well as in the cell layer (180.0 6 20.5 ng/ ml vs. 642.5 6 73.1 ng/ml; P 0.05) when compared to the uncrowded control. Administration of TGF-b also enhanced the amount of HA significantly compared to the control under uncrowded conditions in the supernatant (429.8 6 25.5 ng/ml vs. 1064.8 6 31.9 ng/ml; P 0.05) and in the cell layer (180.0 6 20.5 ng/ml vs. 323.0 6 45.0 ng/ml; P 0.05). In addition, the combination of crowding and TGF-b increased HA in supernatant (429.8 6 25.5 ng/ml vs. 1295.8 6 34.6 ng/ml; P 0.05) and the cell layer (180.0 6 20.5 ng/ml vs. 712.0 6 67.2 ng/ml; P 0.05) (see Fig. 3).
Fibronectin (Fn) Fn was significantly increased in the cell layer in all conditions compared to the uncrowded control (uncrowded control 0.5 6 0.04 mg/ml vs. crowding 1.3 6 0.1 mg/ml; P 0.05 vs. uncrowded TGF-ß 0.7 6 0.03 mg/ml; P 0.05 vs. crowding TGF-ß 1.7 6 0.04 mg/ml; P 0.05). There was no significant difference between uncrowded control and crowding in the supernatant Graupp et al.: MMC for Vocal Fold Fibrosis Research
E205
Fig. 1. Deposition of Collagen I-a in different conditions. *P . 05 versus control in supernatant respectively cell layer.
(9.2 6 0.9 mg/ml vs. 13.1 6 1.8 mg/ml; P 5 0.124); whereas administration of TGF-ß significantly increased Fn levels under crowded and uncrowded conditions (9.2 6 0.9 mg/ml vs. 15.8 6 1.6 mg/ml; P 0.05; resp. 9.2 6 0.9 mg/ml vs. 28.3 6 3.2 mg/ml; P 0.05) (see Fig. 4).
crowding and TGF-b (10.0 6 1.7 ng/ml). There was no significant difference between the groups (see Fig. 5).
DISCUSSION
As described above, alpha-smooth muscle actin (aSMA) was determined only from the cell layer after lysing because it is part of the cytoskeleton. In uncrowded control samples, a-SMA was below range, indicating their nonscar condition and expression patterns. Therefore, this group had to be excluded from statistical analysis. In the other groups, a-SMA was detectable, with the lowest value after administration of TGF-b (7.7 6 0.6 ng/ml). Crowding alone resulted in slightly elevated values (9.33 6 3.8 ng/ml), as well as the combination of
Similar to other fields in modern medicine, big hopes were also put in tissue engineering techniques and regenerative medicine in the field of laryngology with a strong focus on VF injury and VF scars. Based on deeper insights in the molecular mechanisms between local and chemo-attracted cells and components of ECM, novel therapeutic attempts have been created such as (stem) cell therapy, implantation of matrices and scaffolds, or growth factor therapy.13 Hirano and coworkers published numerous articles investigating the effects of HGF on normal and scarred VFF in vivo and in vitro.14–16 Despite the clear antifibrotic effect, HGF it is not approved for treatment in humans thus far due to safety reasons.
Fig. 2. Deposition of hyaluronic acid in different conditions. *P .05 versus control in supernatant respectively cell layer.
Fig. 3. Deposition of fibronectin in different conditions. *P . 05 versus control in supernatant respectively cell layer.
Alpha-Smooth Muscle Actin (a-SMA)
Laryngoscope 125: June 2015
E206
Graupp et al.: MMC for Vocal Fold Fibrosis Research
Fig. 4. Deposition of a-SMA in different conditions.
Significant burden arise from the fact that the laryngeal structures are only poorly accessible. As a consequence, most examinations are restricted to simple endoscopy for which often only indirect signs of pathophysiological processes can be detected. Furthermore,
monitoring disease or therapeutic interventions at a histological level is highly problematic; biopsies are virtually impossible under normal conditions because this inevitably leads to deterioration of the delicate structures and thus again to fibrosis. Therefore, the VFF as the central cell type of fibrogenesis/inflammation is practically unapproachable and thus poorly characterized and understood. Of note, there is only one publication in the field that successfully investigated the scar fibroblast (aka. myofibroblast) differentiation in an in vitro model.11 Recent publications by our group showed that age is, among others, an important factor when dealing with VFF. The cultured scar VFF of younger animals produced significantly higher levels of HA, even 3 months after injury, compared to the unaffected contralateral side and to the cultured scar VFF of older animals.17 We observed the same phenomenon under external stimulation for which the cultured sVFF of younger animals responded stronger to HGFstimulation.8 An in vitro test system that mimics laryngeal fibrogenesis might overcome some of the aforementioned
Fig. 5. Immunocytochemistry for a-SMA (left column: a, d, e) and collagen I (right column: b, d, f). Each row represents different conditions: uncrowded control (a, b), crowded control (c, d), and crowded TGF-b (e, f). Scale bar: 100 mm. [Color figure can be viewed in the online issue, which is available at www.laryngoscope.com.]
Laryngoscope 125: June 2015
Graupp et al.: MMC for Vocal Fold Fibrosis Research
E207
problems. Raghunath et al. published the scar-in-a-jar approach based on the principles of MMC.18 They demonstrated that they could resolve several problems with in vitro fibrogenesis by adding inert macromolecules into the culture medium. Within a very short time, this dramatically increased the collagen deposition in the pericellular matrix, with a higher degree of cross-linking compared to earlier models.18 As mentioned earlier, the fibrogenic cells in aqueous monolayer culture do not lay down sufficient amounts of collagen in a useful time for testing antifibrotic compounds. Fibrillar collagen-I is synthesized and exported into the extracellular space as water-soluble procollagen, where it undergoes the removal of the C- and N-terminal propeptides. Furthermore, the deposition of a collagen matrix depends on the conversion of newly synthesized procollagen to collagen in the extracellular space or shortly before its release into the same.6 The supramolecular assembly of collagen into fibrils via the removal of the C-terminal propeptide from the procollagen I molecule is driven by the C-proteinase/bone morphogenetic protein and is enzymatically rate-limited and occurs at a very slow rate.5 This results in an accumulation of unprocessed water-soluble procollagen in the cell culture medium, where it is discarded with every medium change. Therefore, although analytic methods have mostly settled for the measurement of procollagen secreted into culture medium, these conditions barely mimic a fibrotic situation. As another consequence of this slow enzymatic conversion, the effects of C-proteinase inhibitor as an antifibrotic agent are undetectable under routine culture conditions.6 The macromolecules present in the cell occupy a significant part of the total volume of the medium (about 40%), which reduces the accessible volume in the cell. In multicellular organisms, crowding is not confined to cellular interiors but also occurs in the extracellular matrix of tissues.5,7,18 In 2009, Chen et al. presented a MMCmodel for which they tested novel antifibrotic compounds at the epigenetic, posttranscriptional/translational/secretional level using human lung fibroblasts.5 Their setting allowed for the evaluation of several antifibrotic compounds by optical and biochemical quantitation methods. MMC enabled scaffold-free approaches that are coherent enough to be manipulated by the presence of sufficient ECM to stabilize the structures. Ascorbate is another crucial cosubstrate of the enzymes responsible for the posttranslational hydroxylation of prolyl and lysyl residues for rendering the collagen triple helix thermostable and for regular extracellular crosslinking.18 The addition of ascorbic acid, in the form of a magnesium salt, resulted in enhanced production and deposition of collagen on the cell layer. However, this aspect was not reviewed in VF scar models thus far. Our experiments with MMC in cultured rats VFF confirmed the previous results in terms of col-1-a deposition. In supernatant as well as in the cell layer, MMC conditions lead to a four- to nine-fold increase of col-1-a levels compared to the standard medium (control). The same effect was observed after the administration of TGF-b under crowded and uncrowded cell culture settings (see Fig. 1). In the cell layer, we again observed the strong Laryngoscope 125: June 2015
E208
effects of both TGF-b and crowding as reflected by elevated collagen-alpha1(I) levels. The combination of TGF-ß with crowding resulted in significantly higher levels of collagen compared to TGF-ß alone in the cell layer, but not in the supernatant. This finding is important and can be interpreted by the fact that crowding strongly supports the incorporation of collagen into the ECM, which is represented by the cell layer.12 This is desirable for an in vitro fibrogenesis model because considerable amounts of (water-soluble) newly produced collagen in supernatants get lost for analysis with every medium change under conventional culture conditions. Whereas the growth factors seem to influence both compartments (supernatant and cell layer) to a similar extent, crowding works more effectively in the cell layer. These results were confirmed by immunocytochemistry, which showed a considerably more dense deposition of collagen under crowded conditions compared to the uncrowded control. Similar effects were observed when checking for HA levels in supernatants, where the levels of HA were elevated in all settings. In the cell layer, TGF-b and crowding increased HA production. The combination of TGF-b and crowding significantly increased the levels of HA compared to the uncrowded control. These findings suggest a possible role of the crowding for formation of several ECM components in vitro (e.g., scaffold-free ECM production). Fibronectin is a glycoprotein that functions as an adhesion molecule for cell-to-cell and cell-to-ECM interaction. Yamashita et al. demonstrated the increased abundance of fibronectin in a postinjury mouse model.19 Interestingly, the Fn levels were not elevated in supernatants compared between the crowded and uncrowded conditions. The administration of TGF-ß, and hence the conversion of VFF into myofibroblasts, led to increased Fn levels under both uncrowded and crowded conditions. a-SMA, as an important marker for myofibroblast differentiation, was increased under the administration of TGF-b, as well as under crowding conditions compared to the uncrowded controls. Applying the principles of MMC to VF fibrogenesis might open new perspectives for the understanding and treatment of chronic VF scars. However, considerable limitations exist. VFF behave differently under static conditions in terms of gene/protein expression.20 In addition, we cannot yet integrate other cell types such as mononuclear, which play decisive roles in vivo, into our in vitro model. Likewise, the influence and effects of the overlying epithelium after injury or disruption must be considered when further developing this MMC approach. Every in vitro model is just an approximation and simulation of in vivo reactions. In the special case of the larynx, which is difficult to access and very sensitive to trauma, the presented in vitro fibrogenesis model can substantially improve the characterization of the fibrotic cascade and the development and testing of antifibrotic compounds. The next steps must include a nearer morphological characterization of the newly built collagen structures, the testing of established and novel antifibrotic agents, and a transition to human cells. Graupp et al.: MMC for Vocal Fold Fibrosis Research
CONCLUSION Our study demonstrated that MMC is a suitable approach to enhance collagen-alpha1(I) deposition in cultured rats’ VFF. Collagen-alpha1(I) deposition increased significantly under crowded conditions in supernatant (up to 9-fold) as well as in the cell layer (up to 4-fold). Levels of numerous other important ECM components (HA, fibronectin, collagen-III) were also found to be increased. This experiment was the first step to establish MMC as a valuable in vitro fibrogenesis tool for VF scar research. This model has the potential for fast and effective in vitro collagen deposition and may provide a valid alternative to animal experiments.
Acknowledgments The authors would like to thank Prof. Michael Raghunath MD, PhD for fruitful discussion and training in macromolecular crowding; Richard Ackbar, PhD, for proofreading and discussing the manuscript; and Anita Leitner, MTA, and Sabine Pailer, MTA, for laboratory analyses.
BIBLIOGRAPHY 1. Friedrich G, Dikkers FG, Arens C, et al. Vocal fold scars: current concepts and future directions. Consensus report of the phonosurgery committee of the European laryngological society. Eur Arch Otorhinolaryngol 2013; 270:2491–507. doi: 10.1007/s00405-013-2498-9. 2. Allen J. Cause of vocal fold scar. Curr Opin Otolaryngol Head Neck Surg 2010;18:475–480. 3. Hirano S, Kishimoto Y, Suehiro A, Kanemaru S, Ito J. Regeneration of aged vocal fold: first human case treated with fibroblast growth factor. Laryngoscope 2008;118:2254–2259. 4. Hirano S, Bless DM, Nagai H, et al. Growth factor therapy for vocal fold scarring in a canine model. Ann Otol Rhinol Laryngol 2004;113:777– 785. 5. Chen CZ, Peng YX, Wang ZB, et al. The Scar-in-a-Jar: studying potential antifibrotic compounds from the epigenetic to extracellular level in a single well. Br J Pharmacol 2009;158:1196–1209.
Laryngoscope 125: June 2015
6. Lareu RR, Arsianti I, Subramhanya HK, Yanxian P, Raghunath M. In vitro enhancement of collagen matrix formation and crosslinking for applications in tissue engineering: a preliminary study. Tissue Eng 2007;13:385–391. 7. Lareu RR, Subramhanya KH, Peng Y, et al. Collagen matrix deposition is dramatically enhanced in vitro when crowded with charged macromolecules: the biological relevance of the excluded volume effect. FEBS Lett 2007;581:2709–2714. 8. Graupp M, Kiesler K, Friedrich G, et al. Vocal fold fibroblast response to growth factor treatment is age dependent: results from an in vitro study. J Voice 2014;28:420–423. doi: 10.1016/j.jvoice.2013.11.005. 9. Thibeault SL, Li W, Bartley S. A method for identification of vocal fold lamina propria fibroblasts in culture. Otolaryngol Head Neck Surg 2008; 139:816–822. 10. Hata R, Senoo H. L-ascorbic acid 2-phosphate stimulates collagen accumulation, cell proliferation, and formation of a three-dimensional tissuelike substance by skin fibroblasts. J Cell Physiol 1989;138:8–16. 11. Vyas B, Ishikawa K, Duflo S, Chen X, Thibeault SL. Inhibitory effects of hepatocyte growth factor and interleukin-6 on transforming growth factor-beta1 mediated vocal fold fibroblast-myofibroblast differentiation. Ann Otol Rhinol Laryngol 2010;119:350–357. 12. Chen C, Loe F, Blocki A, Peng Y, Raghunath M. Applying macromolecular crowding to enhance extracellular matrix deposition and its remodeling in vitro for tissue engineering and cell-based therapies. Adv Drug Deliv Rev 2011;63:277–290. 13. Gugatschka M, Ohno S, Saxena A, Hirano S. Regenerative Medicine of the Larynx. Where are we Today? A Review. J Voice 2012;26:670.e7–e13. doi: 10.1016/j.jvoice.2012.03.009. 14. Kishimoto Y, Hirano S, Kitani Y, et al. Chronic vocal fold scar restoration with hepatocyte growth factor hydrogel. Laryngoscope 2010;120:108– 113. 15. Ohno T, French LC, Hirano S, Ossoff RH, Rousseau B. Effect of hepatocyte growth factor on gene expression of extracellular matrix during wound healing of the injured rat vocal fold. Ann Otol Rhinol Laryngol 2008; 117:696–702. 16. Hirano S, Bless DM, del Rio AM, Connor NP, Ford CN. Therapeutic potential of growth factors for aging voice. Laryngoscope 2004;114:2161–2167. 17. Gugatschka M, Ainodhofer H, Gruber HJ, et al. Age effects on extracellular matrix production of vocal fold scar fibroblasts in rats. Eur Arch Otorhinolaryngol 2013; 271:1107–1112. 18. Chen CZ, Raghunath M. Focus on collagen: in vitro systems to study fibrogenesis and antifibrosis state of the art. Fibrogenesis Tissue Repair 2009;2:7 doi: 10.1186/1755-1536-2-7. 19. Yamashita M, Bless DM, Welham NV. Morphological and extracellular matrix changes following vocal fold injury in mice. Cells Tissues Organs 2010;192:262–271. 20. Titze IR, Hitchcock RW, Broadhead K, et al. Design and validation of a bioreactor for engineering vocal fold tissues under combined tensile and vibrational stresses. J Biomech 2004;37:1521–1529.
Graupp et al.: MMC for Vocal Fold Fibrosis Research
E209
Rhinological and Otological Society, Inc.
Establishing Principles of Macromolecular Crowding for In Vitro Fibrosis Research of the Vocal Fold Lamina Propria Matthias Graupp, MD; Hans-J€ urgen Gruber, PhD; Gregor Weiss, PhD; Karl Kiesler, MD; Sophie Bachna-Rotter, MD; Gerhard Friedrich, MD; Markus Gugatschka, MD Objectives/Hypothesis: Vocal fold fibrosis represents a major disease burden. Screening of antifibrotic compounds could be facilitated by an in vitro fibrogenesis system. Limitations of existing models might be overcome by implication of the excluded volume effect. Study Design: In-vitro study. Methods: Vocal fold fibroblasts obtained from rats’ lamina propria were cultured in four different settings: in standard medium, under “crowded” conditions by adding inert macromolecules, under external administration of transforming growth factor (TGF)ß-1, and under a combination of both. After 5 days, supernatant and cell layer were collected and analyzed by enzyme-linked immunosorbent assay. Immunofluorescence was additionally performed. Results: Collagen-alpha1(I) deposition increased significantly under crowded conditions and after administration of TGFb21. Amounts of collagen in the cell layer were significantly higher under crowding conditions with TGFb21 compared to administration of TGFb21 alone. Conclusion: Crowding enhanced collagen deposition, resulting in more favorable conditions for studying fibrogenesis. This can be the first step toward developing a robust in vitro model for testing antifibrotic compounds. Key Words: Vocal fold scar, in vitro fibrogenesis, fibrosis, macromolecular crowding. Level of Evidence: NA Laryngoscope, 125:E203–E209, 2015
INTRODUCTION Treatment of fibrotic and scarred vocal folds (VF) is still an unresolved chapter in laryngology. To date, it has not been possible to restore scarred VFs to preinjury status due to the highly complex micro-architecture of the VF, and in particular of the trilayered VF lamina propria. Standard therapy is comprised of conservative speech therapy and surgical medialization procedures that relieve symptoms of dysphonia but cannot restore to preinjury status.1 During the last decade, significant progress occurred in exploring the cellular and biochemical mechanisms of the fibrotic cascade. Basic knowledge of cellular mechanisms of VF injury and scarring has emerged, giving deeper insights and new understandings to the complex interplay between interstitial proteins (e.g., fibronectin, decorin, fibromodulin),
From the ENT University Hospital Graz, Department of Phoniatrics (MATTHIAS G., S.B-R., K.K., G.F., MARKUS G.), Clinical Institute of Medical and Chemical Laboratory Diagnostics (H-J.G.), and the Institute of Cell Biology, Histology and Embryology (G.W.), Medical University Graz, Graz, Austria Editor’s Note: This Manuscript was accepted for publication December 1, 2014. This study received public funding from the City of Graz. The authors have no other funding, financial relationships, or conflicts of interest to disclose. Send correspondence to Markus Gugatschka, Assistant Professor MD, DMSci, Department of Phoniatrics, ENT University Hospital Graz, Medical University Graz, Auenbruggerplatz 26, A-8036 Graz, Austria. E-mail: [email protected]
glycosaminoglycans (e.g., hyaluronic acid [HA]); various extracellular matrix fibers such as collagen, procollagen, elastin) and inflammatory cytokine signaling.2 Despite this increase in knowledge, substantial obstacles to successful development of regenerative therapies remained, one of them being the difficulty of trials in humans. Although in vitro and in vivo studies in animal models of several antifibrotic compounds, such as fibroblast growth factor3 and hepatocyte growth factor (HGF),4 have shown very promising results, no such therapies have been approved for treatment in humans. Likewise, the approval of stem cell therapy is hindered by the risk of malignant transformation. The development of an in vitro fibrogenesis model that mimics VF fibrosis/fibrogenesis and antifibrosis is highly desirable. Such a model would be an invaluable tool for several reasons: It would allow rapid screening of antifibrotic compounds before going into larger expensive clinical trials. A technically well-characterized fibrogenesis model would allow targeting of the fibrotic cascade at different stages that are invisible under routine culture conditions. In addition, in vitro models using VF fibroblasts (VFF) could be a step closer to potentially eliminate animal studies. However, the development of a fibrogenesis model bears several biological and technical obstacles. Perhaps the biggest challenge is to guarantee a sufficient in vitro deposition of collagen and its incorporation into a pericellular matrix. Fibrogenic cells in monolayer culture do not lay down significant amounts of collagen in an
DOI: 10.1002/lary.25103
Laryngoscope 125: June 2015
Graupp et al.: MMC for Vocal Fold Fibrosis Research
E203
appropriate time for testing antifibrotic compounds. This is also due to the fact that cells under culture conditions find themselves in an aqueous environment that is hardly representative of its in vivo microenvironment from where they originally derive. Furthermore, collagen is a central target in fibrosis research, yet there are additional features characteristic for fibrosis such as asmooth-muscle-actin (a-SMA).5 The aforementioned transition of fibroblasts to fibrotic myofibroblasts during fibrogenesis and a survey of production and deposition of noncollagenous extracellular matrix (ECM)-proteins must be considered as well. Many of these obstacles can be overcome by implicating the principles of the excluded volume effect. This biophysical effect is created by additional macromolecules that occupy a given volume, thereby confining other molecules to the remaining space. The biochemical and biophysical implications of this effect have consequences on reaction kinetics and molecular assembly. Macromolecules drive reaction partners into closer collaboration, resulting in improved proteinfolding and protein–protein interactions.6 The confounding factor of cell culture in standard aqueous media is its characteristic difference from the natural, “crowded” state. That means that in living systems biochemical processes proceed in an environment containing a high concentrations of macromolecules (50– 400 mg/ml). The experimental approach of macromolecular crowding (MMC) has been successfully employed in areas other than laryngology. Lareu et al. showed that collagen matrix deposition is dramatically enhanced in vitro when crowded with macromolecules. Both collagen deposition and collagen cross-linking were found to be increased significantly.7 The aim of our study group was to develop an in vitro fibrogenesis model by applying the principles of MMC to laryngeal fibrosis research. In a first step, we proved the transferability of these principles to VFF and compared the production of different ECM components and collagen deposition under conventional cell culture conditions and under crowding conditions.
MATERIALS AND METHODS VFF were obtained from one male Sprague-Dawley rat, aged 5 months. The rat was humanely euthanized by intracardiac injection of ketamine 100 mg/ml after sedation with isoflurane (5%) in an induction chamber with an oxygen flow of 1.5 l/ min. After that, the larynges were excised. All study procedures were conducted according to the Austrian guidelines for animal experiments and were approved by the Austrian Ministry of Science. Under stereo-magnification, epithelium and underlying lamina propria of both VFs were carefully excised, as described earlier.8 Specimens were minced, and fibroblasts were extracted and cultured following established tissue-explant protocols.9 Growth medium consisted of Dulbecco’s modified Eagle’s medium (DMEM) (Life Technologies, Gaithersburg, MD) enriched with 10% fetal calf serum (FCS) and 1% penicillin/ streptomycin (P/S). After confluence of 90%, VFF were passaged using 0.5 % trypsin. Cells at passage 3 were used for the experiment and seeded at a concentration of 5 3 104 cells into 24 well plates (Nalgen Nunc International, NY) in 10% FCS, 5% CO2 at 37 C. After 24 hours, medium was changed to serum-free
Laryngoscope 125: June 2015
E204
medium for starvation. Another 24 hours later, wells were allocated to the different groups by changing medium again (each group consisted of three replicates). The control group was set on DMEM containing 0.5% FCS, 1% P/S, and 100 mM of Lascorbic acid 2-phosphate (as the magnesium salt, a more stable form of ascorbate10). To achieve myofibroblast differentiation, 5 ng/mL21 transforming growth factor beta-1 (TGFb21) (Sigma-Aldrich St. Louis, MO) was added, as described earlier.11 Another group was cultured under “crowded” conditions by adding macromolecules, a mixture of 37.5 mg/mL21 70 kDa Ficoll (Fc) (Sigma-Aldrich St. Louis, MO) with 25 mg/mL21 400 kDa Fc5; and for the last group, TGFb21 and macromolecules were added simultaneously. A specific formulation of Ficoll (Sigma-Aldrich) was used because it proved to show the best results in terms of collagen deposition time and crosslinking.12 After 5 days, supernatants were sampled and immediately stored at 280 C. The cell layer was detached and lysed using cOmplete Lysis M (Roche Applied Sciences, Basel, SUI, Indianapolis, IN) and stored at 280 C. Assessments of collagen-alpha1(I), collagen-III, HA, and fibronectin levels from supernatants and cell layers were carried out separately using enzyme-linked immunosorbent assays (ELISA) kit (USCN Life Science Inc., Houston, TX, USA), PRC for collagen-I-a, rat-fibronectin ELISA (from ALPCO, Salem, NH, USA), Quantikine ELISA for hyaluronan (R&D Systems, Minneapolis, MN, USA), and rat collagen type III ELISA (Qayee-Bio, Shanghai, PRC). Alpha-smooth-muscleactin (a-SMA) is a commonly used marker for myofibroblast formation and an essential part of the cytoskeleton. As such, ELISA was performed only from the cell layer (a-SMA ELISA by Cusabio, Wuhan, PRC). All assays were performed in duplicates (technical duplicates).
Immunofluorescence Fibroblasts were grown on glass chamber slides (Lab-Tek II, Nalgene Nunc International, Naperville, IL, USA) and washed with phosphate buffered saline (PBS) before being harvested. Slides were air-dried overnight and stored at 220 C. Cells were immuno-labeled using the UltraVision LP Detection System (Thermo Scientific, Fremont, CA) according to the manufacturer’s instructions. The following antibodies were diluted in antibody diluent (Dako, Glostrup, Denmark) and applied for 30 minutes at room temperature: Collagen-1 (rabbit IgG, 8.5 mg/ml, Proteintech, Chicago, IL); ACTA-2 (goat IgG, 20 mg/ml, LifeSpan BioSciences, Seattle, WA). Cells were washed three times with PBS followed by incubation with the secondary antibody (Alexa Fluor 555 Donkey Anti-goat IgG, 10 mg/ml, Alexa Fluor 488 Donkey Anti-rabbit IgG, 10 mg/ml; both Life Technologies, Carlsbad, CA) for 30 minutes. Slides were again washed with PBS, mounted with ProLong Gold antifade reagent (Life Technologies, Gaithersburg, MD, USA) and observed with a Leica DM600B fluorescent microscope (Leica, Wetzlar, Germany) connected to an Olympus DP72 digital camera (Olympus, Tokyo, Japan).
Statistical Analysis Differences of the mean were analyzed by paired t tests after proof of normal distribution using Predictive Analytics Software (PASW) statistics 18.0 (SPSS Inc., Chicago, IL, USA). 0.05 was chosen as level for statistical significance. Normal distribution was given in all parameters, as confirmed by the Kolmogorov-Smirnov test. Biological replicates were treated as independent variables, whereas the technical duplicates were averaged.
Graupp et al.: MMC for Vocal Fold Fibrosis Research
TABLE I. Results for Different Parameters in Different Conditions. Collagen Ia (ng/ml) Uncrowded
Supernatant Control TGF P valueA Cell layer
Fibronectin (mg/ml) Crowded
P valueB
Uncrowded
Crowded
P ValueB
379.5 6 23.8
3,152.8 6 94.6
0.05
9.2 6 0.9
13.1 6 1.8
0.124
3,382.8 6 122.5
3,855.5 6 226.9
0.099
15.8 6 1.6
28.3 6 3.2
0.05
0.05
0.199
0.05
0.05
Control
336.1 6 55.0
1,251.8 6 142.9
0.05
0.5 6 0.04
1.3 6 0.1
0.05
TGF P valuea
709.8 6 136.4 0.05
1,303.0 6 136.2 0.839
0.05
0.7 6 0.03 0.05
1.7 6 0.04 0.05
0.05
Hyaluronic Acid (ng/ml)
a-SMA (ng/ml) P Valueb
Uncrowded
Crowded
P Valueb
Uncrowded
Crowded
Control
429.8 6 25.5
1,053.3 6 31.9
0.05
–
–
–
TGF P valuea
1,064.8 6 31.9 0.05
1,295.8 6 58.7 0.05
0.05
– –
– –
–
Control TGF
180.0 6 20.5 323.0 6 45.0
642.5 6 73.1 712.0 6 67.2
0.05 0.05
Below range 7.7 6 0.6
9.33 6 3.8 10.0 6 1.7
– 0.118
P valuea
0.05
0.124
–
0.774
Supernatant
Cell layer
P valuea 5 comparing control and TGF; P valueb 5 comparing uncrowded and crowded; TGF 5 transforming growth factor.
RESULTS Collagen-Alpha1(I) Collagen-alpha1(I) deposition was significantly enhanced under crowded conditions compared to standard culture medium in supernatant (379.5 6 23.8 ng/ml vs. 3152.8 6 94.6 ng/ml; P 0.05) as well as in the cell layer (336.1 6 55.0 ng/ml vs. 1251.8 6 142.9 ng/ml; P 0.05). See Table I for all results. The administration of TGF-b alone significantly increased the deposition of collagen-I-a in the supernatant (379.5 6 23.8 ng/ml vs. 3382.8 6 122.5 ng/ml; P 0.05) and in the cell layer (336.1 6 55.0 ng/ml vs. 709.8 6 136.4 ng/ml; P 0.05) compared to the control. The combination of crowding and TGF-b also significantly increased collagen-alpha1(I) deposition in the supernatant and in the cell layer compared to the uncrowded control (379.5 6 23.8 ng/ml vs. 3855.5 6 226.9 ng/ml; P 0.05; resp. 336.1 6 55.0 ng/ml vs. 1303.0 6 136.2 ng/ml; P 0.05) (see Fig. 1). Moreover, the combination of crowding and TGF-b administration resulted in a significant increase of collagen-alpha1(I) levels in the cell layer compared to TGF-b administration alone (1303.0 6 136.2 ng/ml vs. 709.8 6 136.4 ng/ml; P 0.05) but not when compared to crowding alone (1303.0 6 136.2 ng/ml vs. 1251.8 6 142.9 ng/ml; P 5 0.839) (see Table I). Results of immunocytochemistry correlated to results achieved from ELISA trials. Deposition of collagen-alpha1(I) was more dense Laryngoscope 125: June 2015
under crowded conditions compared to the uncrowded control (see Fig. 2, right column).
Hyaluronic Acid (HA) Crowding led to significantly increased levels of HA in supernatants (429.8 6 25.5 ng/ml vs. 1053.3 6 31.9 ng/ ml; P 0.05) as well as in the cell layer (180.0 6 20.5 ng/ ml vs. 642.5 6 73.1 ng/ml; P 0.05) when compared to the uncrowded control. Administration of TGF-b also enhanced the amount of HA significantly compared to the control under uncrowded conditions in the supernatant (429.8 6 25.5 ng/ml vs. 1064.8 6 31.9 ng/ml; P 0.05) and in the cell layer (180.0 6 20.5 ng/ml vs. 323.0 6 45.0 ng/ml; P 0.05). In addition, the combination of crowding and TGF-b increased HA in supernatant (429.8 6 25.5 ng/ml vs. 1295.8 6 34.6 ng/ml; P 0.05) and the cell layer (180.0 6 20.5 ng/ml vs. 712.0 6 67.2 ng/ml; P 0.05) (see Fig. 3).
Fibronectin (Fn) Fn was significantly increased in the cell layer in all conditions compared to the uncrowded control (uncrowded control 0.5 6 0.04 mg/ml vs. crowding 1.3 6 0.1 mg/ml; P 0.05 vs. uncrowded TGF-ß 0.7 6 0.03 mg/ml; P 0.05 vs. crowding TGF-ß 1.7 6 0.04 mg/ml; P 0.05). There was no significant difference between uncrowded control and crowding in the supernatant Graupp et al.: MMC for Vocal Fold Fibrosis Research
E205
Fig. 1. Deposition of Collagen I-a in different conditions. *P . 05 versus control in supernatant respectively cell layer.
(9.2 6 0.9 mg/ml vs. 13.1 6 1.8 mg/ml; P 5 0.124); whereas administration of TGF-ß significantly increased Fn levels under crowded and uncrowded conditions (9.2 6 0.9 mg/ml vs. 15.8 6 1.6 mg/ml; P 0.05; resp. 9.2 6 0.9 mg/ml vs. 28.3 6 3.2 mg/ml; P 0.05) (see Fig. 4).
crowding and TGF-b (10.0 6 1.7 ng/ml). There was no significant difference between the groups (see Fig. 5).
DISCUSSION
As described above, alpha-smooth muscle actin (aSMA) was determined only from the cell layer after lysing because it is part of the cytoskeleton. In uncrowded control samples, a-SMA was below range, indicating their nonscar condition and expression patterns. Therefore, this group had to be excluded from statistical analysis. In the other groups, a-SMA was detectable, with the lowest value after administration of TGF-b (7.7 6 0.6 ng/ml). Crowding alone resulted in slightly elevated values (9.33 6 3.8 ng/ml), as well as the combination of
Similar to other fields in modern medicine, big hopes were also put in tissue engineering techniques and regenerative medicine in the field of laryngology with a strong focus on VF injury and VF scars. Based on deeper insights in the molecular mechanisms between local and chemo-attracted cells and components of ECM, novel therapeutic attempts have been created such as (stem) cell therapy, implantation of matrices and scaffolds, or growth factor therapy.13 Hirano and coworkers published numerous articles investigating the effects of HGF on normal and scarred VFF in vivo and in vitro.14–16 Despite the clear antifibrotic effect, HGF it is not approved for treatment in humans thus far due to safety reasons.
Fig. 2. Deposition of hyaluronic acid in different conditions. *P .05 versus control in supernatant respectively cell layer.
Fig. 3. Deposition of fibronectin in different conditions. *P . 05 versus control in supernatant respectively cell layer.
Alpha-Smooth Muscle Actin (a-SMA)
Laryngoscope 125: June 2015
E206
Graupp et al.: MMC for Vocal Fold Fibrosis Research
Fig. 4. Deposition of a-SMA in different conditions.
Significant burden arise from the fact that the laryngeal structures are only poorly accessible. As a consequence, most examinations are restricted to simple endoscopy for which often only indirect signs of pathophysiological processes can be detected. Furthermore,
monitoring disease or therapeutic interventions at a histological level is highly problematic; biopsies are virtually impossible under normal conditions because this inevitably leads to deterioration of the delicate structures and thus again to fibrosis. Therefore, the VFF as the central cell type of fibrogenesis/inflammation is practically unapproachable and thus poorly characterized and understood. Of note, there is only one publication in the field that successfully investigated the scar fibroblast (aka. myofibroblast) differentiation in an in vitro model.11 Recent publications by our group showed that age is, among others, an important factor when dealing with VFF. The cultured scar VFF of younger animals produced significantly higher levels of HA, even 3 months after injury, compared to the unaffected contralateral side and to the cultured scar VFF of older animals.17 We observed the same phenomenon under external stimulation for which the cultured sVFF of younger animals responded stronger to HGFstimulation.8 An in vitro test system that mimics laryngeal fibrogenesis might overcome some of the aforementioned
Fig. 5. Immunocytochemistry for a-SMA (left column: a, d, e) and collagen I (right column: b, d, f). Each row represents different conditions: uncrowded control (a, b), crowded control (c, d), and crowded TGF-b (e, f). Scale bar: 100 mm. [Color figure can be viewed in the online issue, which is available at www.laryngoscope.com.]
Laryngoscope 125: June 2015
Graupp et al.: MMC for Vocal Fold Fibrosis Research
E207
problems. Raghunath et al. published the scar-in-a-jar approach based on the principles of MMC.18 They demonstrated that they could resolve several problems with in vitro fibrogenesis by adding inert macromolecules into the culture medium. Within a very short time, this dramatically increased the collagen deposition in the pericellular matrix, with a higher degree of cross-linking compared to earlier models.18 As mentioned earlier, the fibrogenic cells in aqueous monolayer culture do not lay down sufficient amounts of collagen in a useful time for testing antifibrotic compounds. Fibrillar collagen-I is synthesized and exported into the extracellular space as water-soluble procollagen, where it undergoes the removal of the C- and N-terminal propeptides. Furthermore, the deposition of a collagen matrix depends on the conversion of newly synthesized procollagen to collagen in the extracellular space or shortly before its release into the same.6 The supramolecular assembly of collagen into fibrils via the removal of the C-terminal propeptide from the procollagen I molecule is driven by the C-proteinase/bone morphogenetic protein and is enzymatically rate-limited and occurs at a very slow rate.5 This results in an accumulation of unprocessed water-soluble procollagen in the cell culture medium, where it is discarded with every medium change. Therefore, although analytic methods have mostly settled for the measurement of procollagen secreted into culture medium, these conditions barely mimic a fibrotic situation. As another consequence of this slow enzymatic conversion, the effects of C-proteinase inhibitor as an antifibrotic agent are undetectable under routine culture conditions.6 The macromolecules present in the cell occupy a significant part of the total volume of the medium (about 40%), which reduces the accessible volume in the cell. In multicellular organisms, crowding is not confined to cellular interiors but also occurs in the extracellular matrix of tissues.5,7,18 In 2009, Chen et al. presented a MMCmodel for which they tested novel antifibrotic compounds at the epigenetic, posttranscriptional/translational/secretional level using human lung fibroblasts.5 Their setting allowed for the evaluation of several antifibrotic compounds by optical and biochemical quantitation methods. MMC enabled scaffold-free approaches that are coherent enough to be manipulated by the presence of sufficient ECM to stabilize the structures. Ascorbate is another crucial cosubstrate of the enzymes responsible for the posttranslational hydroxylation of prolyl and lysyl residues for rendering the collagen triple helix thermostable and for regular extracellular crosslinking.18 The addition of ascorbic acid, in the form of a magnesium salt, resulted in enhanced production and deposition of collagen on the cell layer. However, this aspect was not reviewed in VF scar models thus far. Our experiments with MMC in cultured rats VFF confirmed the previous results in terms of col-1-a deposition. In supernatant as well as in the cell layer, MMC conditions lead to a four- to nine-fold increase of col-1-a levels compared to the standard medium (control). The same effect was observed after the administration of TGF-b under crowded and uncrowded cell culture settings (see Fig. 1). In the cell layer, we again observed the strong Laryngoscope 125: June 2015
E208
effects of both TGF-b and crowding as reflected by elevated collagen-alpha1(I) levels. The combination of TGF-ß with crowding resulted in significantly higher levels of collagen compared to TGF-ß alone in the cell layer, but not in the supernatant. This finding is important and can be interpreted by the fact that crowding strongly supports the incorporation of collagen into the ECM, which is represented by the cell layer.12 This is desirable for an in vitro fibrogenesis model because considerable amounts of (water-soluble) newly produced collagen in supernatants get lost for analysis with every medium change under conventional culture conditions. Whereas the growth factors seem to influence both compartments (supernatant and cell layer) to a similar extent, crowding works more effectively in the cell layer. These results were confirmed by immunocytochemistry, which showed a considerably more dense deposition of collagen under crowded conditions compared to the uncrowded control. Similar effects were observed when checking for HA levels in supernatants, where the levels of HA were elevated in all settings. In the cell layer, TGF-b and crowding increased HA production. The combination of TGF-b and crowding significantly increased the levels of HA compared to the uncrowded control. These findings suggest a possible role of the crowding for formation of several ECM components in vitro (e.g., scaffold-free ECM production). Fibronectin is a glycoprotein that functions as an adhesion molecule for cell-to-cell and cell-to-ECM interaction. Yamashita et al. demonstrated the increased abundance of fibronectin in a postinjury mouse model.19 Interestingly, the Fn levels were not elevated in supernatants compared between the crowded and uncrowded conditions. The administration of TGF-ß, and hence the conversion of VFF into myofibroblasts, led to increased Fn levels under both uncrowded and crowded conditions. a-SMA, as an important marker for myofibroblast differentiation, was increased under the administration of TGF-b, as well as under crowding conditions compared to the uncrowded controls. Applying the principles of MMC to VF fibrogenesis might open new perspectives for the understanding and treatment of chronic VF scars. However, considerable limitations exist. VFF behave differently under static conditions in terms of gene/protein expression.20 In addition, we cannot yet integrate other cell types such as mononuclear, which play decisive roles in vivo, into our in vitro model. Likewise, the influence and effects of the overlying epithelium after injury or disruption must be considered when further developing this MMC approach. Every in vitro model is just an approximation and simulation of in vivo reactions. In the special case of the larynx, which is difficult to access and very sensitive to trauma, the presented in vitro fibrogenesis model can substantially improve the characterization of the fibrotic cascade and the development and testing of antifibrotic compounds. The next steps must include a nearer morphological characterization of the newly built collagen structures, the testing of established and novel antifibrotic agents, and a transition to human cells. Graupp et al.: MMC for Vocal Fold Fibrosis Research
CONCLUSION Our study demonstrated that MMC is a suitable approach to enhance collagen-alpha1(I) deposition in cultured rats’ VFF. Collagen-alpha1(I) deposition increased significantly under crowded conditions in supernatant (up to 9-fold) as well as in the cell layer (up to 4-fold). Levels of numerous other important ECM components (HA, fibronectin, collagen-III) were also found to be increased. This experiment was the first step to establish MMC as a valuable in vitro fibrogenesis tool for VF scar research. This model has the potential for fast and effective in vitro collagen deposition and may provide a valid alternative to animal experiments.
Acknowledgments The authors would like to thank Prof. Michael Raghunath MD, PhD for fruitful discussion and training in macromolecular crowding; Richard Ackbar, PhD, for proofreading and discussing the manuscript; and Anita Leitner, MTA, and Sabine Pailer, MTA, for laboratory analyses.
BIBLIOGRAPHY 1. Friedrich G, Dikkers FG, Arens C, et al. Vocal fold scars: current concepts and future directions. Consensus report of the phonosurgery committee of the European laryngological society. Eur Arch Otorhinolaryngol 2013; 270:2491–507. doi: 10.1007/s00405-013-2498-9. 2. Allen J. Cause of vocal fold scar. Curr Opin Otolaryngol Head Neck Surg 2010;18:475–480. 3. Hirano S, Kishimoto Y, Suehiro A, Kanemaru S, Ito J. Regeneration of aged vocal fold: first human case treated with fibroblast growth factor. Laryngoscope 2008;118:2254–2259. 4. Hirano S, Bless DM, Nagai H, et al. Growth factor therapy for vocal fold scarring in a canine model. Ann Otol Rhinol Laryngol 2004;113:777– 785. 5. Chen CZ, Peng YX, Wang ZB, et al. The Scar-in-a-Jar: studying potential antifibrotic compounds from the epigenetic to extracellular level in a single well. Br J Pharmacol 2009;158:1196–1209.
Laryngoscope 125: June 2015
6. Lareu RR, Arsianti I, Subramhanya HK, Yanxian P, Raghunath M. In vitro enhancement of collagen matrix formation and crosslinking for applications in tissue engineering: a preliminary study. Tissue Eng 2007;13:385–391. 7. Lareu RR, Subramhanya KH, Peng Y, et al. Collagen matrix deposition is dramatically enhanced in vitro when crowded with charged macromolecules: the biological relevance of the excluded volume effect. FEBS Lett 2007;581:2709–2714. 8. Graupp M, Kiesler K, Friedrich G, et al. Vocal fold fibroblast response to growth factor treatment is age dependent: results from an in vitro study. J Voice 2014;28:420–423. doi: 10.1016/j.jvoice.2013.11.005. 9. Thibeault SL, Li W, Bartley S. A method for identification of vocal fold lamina propria fibroblasts in culture. Otolaryngol Head Neck Surg 2008; 139:816–822. 10. Hata R, Senoo H. L-ascorbic acid 2-phosphate stimulates collagen accumulation, cell proliferation, and formation of a three-dimensional tissuelike substance by skin fibroblasts. J Cell Physiol 1989;138:8–16. 11. Vyas B, Ishikawa K, Duflo S, Chen X, Thibeault SL. Inhibitory effects of hepatocyte growth factor and interleukin-6 on transforming growth factor-beta1 mediated vocal fold fibroblast-myofibroblast differentiation. Ann Otol Rhinol Laryngol 2010;119:350–357. 12. Chen C, Loe F, Blocki A, Peng Y, Raghunath M. Applying macromolecular crowding to enhance extracellular matrix deposition and its remodeling in vitro for tissue engineering and cell-based therapies. Adv Drug Deliv Rev 2011;63:277–290. 13. Gugatschka M, Ohno S, Saxena A, Hirano S. Regenerative Medicine of the Larynx. Where are we Today? A Review. J Voice 2012;26:670.e7–e13. doi: 10.1016/j.jvoice.2012.03.009. 14. Kishimoto Y, Hirano S, Kitani Y, et al. Chronic vocal fold scar restoration with hepatocyte growth factor hydrogel. Laryngoscope 2010;120:108– 113. 15. Ohno T, French LC, Hirano S, Ossoff RH, Rousseau B. Effect of hepatocyte growth factor on gene expression of extracellular matrix during wound healing of the injured rat vocal fold. Ann Otol Rhinol Laryngol 2008; 117:696–702. 16. Hirano S, Bless DM, del Rio AM, Connor NP, Ford CN. Therapeutic potential of growth factors for aging voice. Laryngoscope 2004;114:2161–2167. 17. Gugatschka M, Ainodhofer H, Gruber HJ, et al. Age effects on extracellular matrix production of vocal fold scar fibroblasts in rats. Eur Arch Otorhinolaryngol 2013; 271:1107–1112. 18. Chen CZ, Raghunath M. Focus on collagen: in vitro systems to study fibrogenesis and antifibrosis state of the art. Fibrogenesis Tissue Repair 2009;2:7 doi: 10.1186/1755-1536-2-7. 19. Yamashita M, Bless DM, Welham NV. Morphological and extracellular matrix changes following vocal fold injury in mice. Cells Tissues Organs 2010;192:262–271. 20. Titze IR, Hitchcock RW, Broadhead K, et al. Design and validation of a bioreactor for engineering vocal fold tissues under combined tensile and vibrational stresses. J Biomech 2004;37:1521–1529.
Graupp et al.: MMC for Vocal Fold Fibrosis Research
E209
