Open Access
Add to BibSonomy
Abstract
Glaucoma filtration surgery (GFS) commonly fails due to excessive fibrosis. As collagen structure aberrations is implicated in adverse fibrotic progression, this study aims to uncover collagen organization alterations during postoperative scarring. Via quantitative second harmonic generation/ two photon excitation multiphoton imaging, we reveal the scar development and phenotype in the mouse model of conjunctival scarring. We also show that multiphoton imaging corroborated the collagen ultrastructure anomaly characteristic of the SPARC-/- mouse postoperative conjunctiva. These data improve our understanding of postoperative conjunctival scarring and further enhance the utility of this model for the development of anti-fibrotic therapeutics for GFS.
© 2020 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
1. Introduction
Scar formation is the final outcome of the wound healing process after any surgery in post-fetal mammals. As opposed to scarless healing, also known as tissue regeneration, which is unique to fetuses in mammals and where regrowth of complete suites of tissue structure is achievable [1], normal scarring results in collagen being deposited without recreating the original architecture. Scar formation follows a phasic and well-structured wound healing process comprising of three sequential and overlapping stages: inflammation, proliferation (scar deposition) and remodeling (scar maturation) [2]. Excessive or pathological scarring is thought to arise when this process is disrupted, often due to exacerbation or prolongation of one or more of these phases, and leading generally to an imbalance between the production and breakdown of collagen, the primary component of a scar [3]. However, new discoveries now indicate that the collagen organization is equally critical in progressive fibrosis [4]. The implication is that measuring whole collagen content alone is insufficient, and that collagen organization should also be considered when determining adverse fibrotic progression.
Glaucoma filtration surgery (GFS), or trabeculectomy, is the most commonly performed procedure for ensuring high quality management of intraocular pressure (IOP) in glaucoma [5]. This incisional procedure involves the removal of a small piece of the corneoscleral tissue at the limbus creating a sclerostomy that provides a new drainage route for aqueous humor to exit the eye, bypassing the conventional passageway through the trabecular meshwork. GFS has a tendency to fail when subconjunctival and episcleral fibrosis develop over the sclerostomy site [6]. Even with adjunctive anti-fibrotic therapy, commonly achieved through anti-metabolites such as mitomycin C, clinical studies have reported failure rates that were above 50% at 2 to 5 years [7–10]. Ongoing therapeutic investigations aimed mainly at collagen reduction has also remained largely unsuccessful at the clinical level [11]. We speculate that new therapeutic strategies involving the targeting collagen structure may be necessary to improve anti-fibrotic therapy in GFS. To this end, although altered collagen organization has been recognized to contribute to surgical failure [12], the collagen structure in the GFS scar is poorly understood.
The primary goal of this study is to unravel the postoperative scar organization in the conjunctiva, and this was achieved by means of the mouse model of conjunctival scarring [13]. This mouse model has been validated to produce a similar response as GFS patients to the anti-fibrotic adjunct mitomycin-C [14], and scarring is definitively associated with increase in collagen production in the postoperative tissue [15]. To assess the structure of the collagen matrix, we employed label-free multiphoton imaging based on non-linear optical second harmonic generation/ two photon excitation fluorescence (SHG/ TPE) microscopy. This technology, combined with quantitative analysis techniques, has not only been demonstrated to be a promising and useful diagnostic tool for multiple disease conditions, including ocular pathologies [16,17], but also as an important tool for assessing scar formation and fibrotic progression [18]. SHG exploits the non-centrosymmetric, tightly packed arrangement of amino acids in collagen to provide label-free, highly specific imaging of fibrillar collagen at micrometer resolution [19]. Although the conjunctivas of animal models have been imaged by multiphoton technology in both normal [20,21] and wound healing [22] contexts, collagen structural parameters in this tissue have not been quantitated. In this study, progressive changes in postoperative conjunctival collagen structure and morphology in the SHG images were quantified using a morphology-based quantification algorithm [23,24]. We also measured the postoperative collagen structure changes in the conjunctiva of the secreted protein, acidic and rich in cysteine knockout (SPARC-/-) mouse. SPARC is a matricellular protein implicated in regulating collagen assembly, and this is evident in the SPARC-/- mouse which demonstrates collagen defects [13,25–27]. Through the SPARC-/- mouse, our secondary goals were to validate the reproducibility and sensitivity of SHG to reveal the architecture of collagen in the conjunctiva, as well as to uncover new collagen characteristics that may serve as potential targets for anti-fibrotic drug development. Taken together, we report here detailed features in the postoperative collagen framework that deviated from the normal, unoperated tissue, which in turn, provide insights into postoperative conjunctival scar development, scar phenotype, as well as potential new structure-based targets for a multi-faceted approach to anti-fibrotic therapy for GFS.
2. Methods
2.1. Mouse model of conjunctival scarring
All experiments with animals were approved by the Institutional Animal Care and Use Committee (IACUC) and treated in accordance with the Association for Research in Vision and Ophthalmology (ARVO) Statement on the Use of Animals in Ophthalmic and Vision Research. For evaluation of progressive scar development, C57BL/6 mice were used. SPARC-/- and its congenic wild-type (WT) 129SVE mice were originally obtained from the Benaroya Research Institute at Virginia Mason (Seattle, WA) [28]. All mice were bred in the SingHealth Experimental Medicine Centre (Singapore). Both male and female mice with age ranging from 8 to 10 weeks old were used. Experimental surgery was performed as described previously [13]. Fucithalmic ointment (Leo Pharmaceutical Products, Ballerup, Denmark) was instilled at the end of the procedure to prevent infection. At the indicated time points, mice were euthanized by intraperitoneal injection of pentobarbitol sodium, at 100 mg/kg body weight, before the eyes were enucleated for preparation of cryosections. We collected both the operated and contralateral unoperated eyes of each mouse for quantitative multiphoton analyses.
2.2. Preparation of cryosections and histochemistry
Each enucleated eye ball was fixed with PBS-buffered 4% paraformaldehyde and then placed in a slurry of optimal cutting temperature (OCT) compound in cryomold before freezing in dry ice and storage in a −80 °C freezer until ready for sectioning using the Microm HM550 (Carl Zeiss Ltd). 5 µm thick serial cryosections stained with haematoxylin and eosin (H&E) or Picrosirius Red were visualized as described previously [29]. Unstained sections were scanned by multiphoton imaging system described below.
2.3. Multiphoton image acquisition system
Images of unstained cryosections were acquired on a fully automated, programmable, multiphoton imaging platform (Genesis 200, HistoIndex Pte Ltd, Singapore). Laser excitation was carried out at 780 nm, and forward-scatter two-photon excitation (TPE) and second harmonic generation (SHG) signals were detected simultaneously using dedicated photomultiplier tubes for each channel at 550 nm, and 390 nm respectively and by using a dichroic mirror (450DCLP, Omega) to separate TPE from SHG. Images were acquired at 20x with 512 × 512-pixel resolution with 2x frame averaging feature, and each image had a dimension of 200 × 200 µm2. A bandpass filter with centre wavelength at 550 nm and bandwidth of 88 nm was set in front of the TPE photodetector to reveal tissue structure via nicotinamide adenine (phosphorylated) dinucleotide [NAD(P)H] which emits at (400-600 nm) [30,31]. The laser beam with horizontal polarization was used for excitation. The FibroIndex software (HistoIndex Pte Ltd., Singapore) was used to analyze the region of interest (ROI) in the images (https://www.histoindex.com/product-and-services/). For more detail of the procedure please refer to FibroIndex user manual. Distribution of collagen fiber morphometric traits, including area, density, texture, reticulation, thickness, length, straightness, as well as distribution relative to tissue morphology, were described by normalized quantitative fibrosis parameters (Table 1). Five independent mice were examined for each time point, and five independent cryosections from each eye was imaged for quantitation.
Table 1. List of parameters quantified by the software FibroIndexa.
2.4. Statistical analysis
All data are expressed as mean ± standard deviation (SD). Where operated and unoperated values at each time point were compared, the significance of differences was determined by the two-tailed Student’s t-test. Where operated values across all 4 time points were compared, the significance of differences was determined by one-way ANOVA with Bonferroni adjustments using SPSS statistics. Statistical significance was defined as p<0.05.
3. Results
3.1. Comparison of the postoperative conjunctiva visualized by histochemical staining and stain-free multiphoton imaging
We first investigated the specificity of the stain-free SHG/TPE system in detecting collagen organization against histochemical staining with hematoxylin and eosin (H&E) and Picrosirius Red. H&E allows visualization of cell nuclei in blue, and the extracellular matrix (ECM) in pink. Under the light microscope, H&E revealed several outstanding features of the postoperative blebs at progressive time points from day 2 to day 21 after surgery (Fig. 1, left panel). First, the postoperative bleb size appeared to shrink with time. The change in size is expected and may be explained as follows. The day 2 postoperative tissue is expected to be largest as the tissue may be expanded due to fluid filtration as a result of the surgical creation of a subconjunctival pocket with little resistance to fluid outflow into the subconjunctival space. With time, we expect the tissue to shrink as a result of a combination of diminishing inflammation, the start of new collagen formation [29], increasing tissue contraction as part of the wound healing process, as well as less fluid egressing out of the eye as resistance to outflow increases. Secondly, it is apparent that ECM fibers have progressively morphed from weakly stained, discontinuous, and disorderly strands, most striking on day 2, to strongly-stained, unbroken, and organized strands by day 21 which are more similar to the unoperated conjunctival appearance. These observations suggest that collagen fibers had undergone dramatic changes in architecture and structure during wound healing.
Fig. 1. Comparison of multiphoton imaging to histochemical staining of collagen by hematoxylin and eosin (H&E) and Picrosirius Red in serial frozen sections of the postoperative mouse conjunctiva. Left panel, collagen is stained pink by eosin in H&E. Middle panel, Picrosirius red interacts with collagen fibers under polarized light to reveal a spectrum of colours including yellow, red and green. Right panel, the SHG signals (green) map the distribution of collagen fibers while the endogenous TPE signals (red) reveal the morphology of the tissue from extracellular and cellular chromophores. S, sclera. Scale bars = 100 µm.
Next, we visualized the collagen matrix using one of the gold standards in collagen imaging, namely linear polarized light microscopy. Linear polarized light microscopy performed in association with Picrosirius Red staining, reveals collagen bundles as red, orange, yellow, or green fibers against a black background [32]. We observed that collagen fibers were indeed disrupted on days 2 and 7 post-surgery, appearing particularly scattered and weakly birefringent. On day 14, there was apparent recovery of the collagen network with more steadily birefringent collagen strands, but these had adopted an uncharacteristic wave-like pattern that rose up vertically towards the epithelium or vice versa. Finally, on day 21 post-surgery, there was evident stabilization of the collagen matrix with solid, continuous, and strongly birefringent collagen strands that appeared in the main to run parallel with the sclera (Fig. 1, middle panel).
Finally, we examined the collagen matrix using multiphoton imaging which has also been considered a gold standard for imaging collagen in tissues [31,33]. Multiphoton imaging is advantageous over Picrosirius red staining as the latter’s reproducibility is limited by the variability in the staining protocols and in the induced birefringence, and also angle of observation [34–36]. SHG is induced by biomolecular structures, particularly fibrillar collagen, whereas most extracellular TPE images mainly arise from intrinsic fluorophores such as NAD(P)H, flavins, retinoids, lipofuscin, elastin, and others [30,31]. The green SHG signals clearly showed collagen fibril distribution embedded within the postoperative conjunctival matrix containing other intrinsic cellular/ other structural fluorescence that were revealed as red TPE signals (Fig. 1, right panel). The SHG signals reiterated the above histochemical detections with high fidelity by revealing progressively more organized collagen network, with seemingly increasing density in the packing of collagen fibers within the postoperative tissues with time. Overall, SHG/ TPE imaging was comparable to conventional histochemical light microscopy in detecting changes in the collagen matrix.
3.2. Quantitative measurements of collagen parameters
Originally developed for evaluating liver fibrosis, Fibroindex computes and provides quantitation of fine details of collagen distribution and architectural features, including morphology such as collagen width, length, perimeter, area, and orientation (Table 1) [24]. As a detailed approach to uncover normal and advancing scar pathology in the postoperative mouse conjunctiva, we analyzed both operated and contralateral unoperated conjunctival tissues at each time point, as well as compared operated tissues across the time points. The ROI included the entire operated conjunctival area and excluded the sclera (Fig. 2). Since SHG signals reflect the total collagen content [37], and TPE signals represent the amount of other tissue constituents [38], we also quantitatively analyzed the ratio of SHG and TPE signals for specific parameters so as to determine potential changes in the balance of collagen with respect to other tissue components.
Fig. 2. Image processing for quantitative analysis of collagen structure in the conjunctiva. (A) Left, multiphoton multimodal image of a sample mouse conjunctiva after experimental surgery: SHG signal in green color revealed the collagen fibers while TPE signal in red color revealed other tissue constituents; middle, separated SHG image; right, separated TPE image. (B) Left, the region of interest (ROI), centered on the conjunctiva and excluding the sclera, was used for the quantification procedure; right, a white mask area was applied on the ROI, consisting of both SHG and TPE, as part of the image processing. S, sclera.
3.2.1. Quantitative changes in collagen area
First, we evaluated collagen distribution within the postoperative tissues. Collagen distribution may be quantified as collagen area (CA), or collagen area ratio (CAR) (Table 1). We measured significant increase in CA in the postoperative tissue relative to the unoperated on days 2 (1.79-fold) and 7 (1.88-fold) after surgery, indicating collagen area increase at these time points (Fig. 3(A)). The same profile was measured with respect to tissue area measurements (Fig. 3(B)). The particular downward transitions between day 7 and day 14 CA and TA were significant, and together with the areas from day 14 onwards being similar to the unoperated tissue, the suggestion is a possible switch to scar maturation at later time points from day 14 onwards. However, as CA measurement is necessarily dependent on the size of the specimen, which may be highly variable between samples, it may be technically more reliable to examine CA as a ratio of another size-dependent but distinct parameter. Indeed, when examined in relation to the total area of interest, collagen area ratio (CAR) values indicated that collagen area was significantly higher than the unoperated exclusively on day 7 post-surgery (Fig. 3(C)). This greater collagen area expansion with respect to the total area evaluated refines the identification of intense collagen deposition as occurring at 1 week post-surgery, suggesting that this time point is at the proliferative phase. Hence, SHG area signals may quantitatively indicate active collagen deposition and at the same time, provide inference to the progression towards scar maturation.
Fig. 3. Progressive alteration of postoperative collagen areas. Related parameters for collagen area that were significantly different between operated and unoperated are shown. (A) collagen area (CA), (B) tissue area (TA), (C) collagen area ratio (CAR). Left panels, comparisons of operated levels against contralateral, unoperated control levels at the indicated time points. *, p<0.05; fold changes between means are indicated if significant. Right panel, the data for operated tissues in the left panels were replotted here to reveal changes in operated levels across the time points. *, p<0.05 (Bonferroni-adjusted); significant changes between two adjacent time points are indicated. Each symbol represents one eye from one mouse of a total of five mice measured at each time point (N=5). C = contralateral, unoperated control conjunctiva; O = operated conjunctival tissue.
3.2.2. Quantitative changes in collagen density
Collagen fiber density (CFD) is defined as the sum of the SHG pixel intensities within the collagen area. In this study, the CFD was significantly lower than unoperated on days 2 and 7, but became significantly higher on days 14 and 21 by mean 1.30- and 1.24-folds respectively (Fig. 4(A)). Taking into account that at earlier time points, collagen areas were exceedingly high, relatively lower CFDs in the operated tissues may be expected, whereas in the later time points, the opposite was the case. By this interpretation, the scars at later time points contain more densely packed collagen fibers within an area that had seemingly returned to the unoperated level (Fig. 3(A)). Notably, a sharp and significant deviation in CFD between days 7 and 14 again suggests a switch in scar maturation between these time points. On the other hand, tissue density (TD), defined as sum of TPE pixel intensities within the tissue area, was only found to be significantly elevated on day 14 post-surgery (Fig. 4(B)). This suggests that other tissue content only exceeded the tissue area on day 14 compared to unoperated. When we evaluated the CFD:TD ratios, we found that only the day 21 postoperative tissue demonstrated significantly higher CFD:TD ratio compared to the unoperated counterpart (Fig. 4(C)). When postoperative density values for all these density parameters were examined across the time points, it is not only obvious that densities generally increase with time up to day 14, but that this was especially significant in the transition from day 7 to day 14 (Fig. 4), suggesting that dramatic changes in collagen content coupled with collagen area are occurring between these two time points. In contrast, it appears that some stabilization in these characteristics have taken place in the period from day 14 to day 21, suggesting the transition to a different phase of wound healing during these times.
Fig. 4. Progressive alteration of postoperative collagen and tissue intensity. Parameters for collagen and matrix intensity that were significantly different between operated and unoperated are shown. (A) collagen fiber density (CFD), (B) tissue density (TD), (C) CFD:TD ratio.
3.2.3. Quantitative changes in collagen texture and orientation
Texture analysis of the SHG images uncovers the surface property of the collagen matrix. The collagen texture contrast (CTCon) parameter measures the definition and the degree of texture depth of the groove pattern of the SHG signal. A contrast value that approaches zero is indicative of a collagen matrix that has a less defined fibrillar contrast structure. As the operated tissues demonstrated lower contrast values compared to unoperated tissue at all time points, it would appear that the collagen fibers in the postoperative tissues at all time points featured reduced fine structure as compared to those in the unoperated tissue, more so in the earlier time points which were reduced by at least 1.5-fold, whereas by day 21, the contrast value was reduced by only 1.15-fold (Fig. 5(A)). Overall, the collagen texture analyses suggest that within the first week of recovery, the collagen matrix featured dramatic reduction in fine structure, with gradual improvement in fiber definition from the second week of wound healing, but never fully recovering to the unoperated level. We may therefore deduce that the mature scar featured a collagen matrix that contained less defined collagen structure compared to the unoperated counterpart.
Fig. 5. Progressive alteration of postoperative collagen texture and orientation. Parameters for collagen texture and orientation that were significantly different between operated and unoperated are shown. (A) collagen texture contrast (CTCon), (B) collagen fiber orientation variation.
The degree of alignment of the final collagen matrix has been implicated in determining the amount of scarring and final quality of the scar [39]. In the postoperative conjunctiva, we detected a significant difference in collagen orientation variation only on day 14 by an increase of 1.08-fold above the normal, unoperated value (Fig. 5(B)). This concurs with the histological staining patterns of uncharacteristic wave-like collagen strands in the matrix. Hence, it appears that collagen orientation became significantly more scattered on day 14 post-surgery but this was temporary as it returned to normal by day 21.
3.2.4. Quantitative changes in collagen reticulation
Reticulation of the fibrillar collagen network and collagen density are known independent determinants of tissue stiffness [40–42]. Interestingly, there was significant increase in collagen branching in the days 2 and 7 postoperative tissue by 1.34- and 1.29-folds respectively when reticulation was calculated as a function of the length of the collagen network (CRI) (Fig. 6(A)). When analyzed as a function of collagen area (CARD), significant increase in branching by 1.15-fold above normal was determined only on day 14 post-surgery (Fig. 6(B)). The most likely reason for the lack of increase in CARD on days 2 and 7 is the dramatic difference in CAs which were dramatically larger than normal in the two earlier time points, and so down-rendering the reticulation values. The relatively small significant increase in CARD on day 14, also reflected in the higher, albeit non-significant CRI, may signify that increased reticulation may still exist at this time point, when CFD is coincidentally high, but the overall trend is for diminishing reticulation towards normal as time progressed. Together, the data suggest that the postoperative tissue may be characterized by increased collagen reticulation for 1-2 weeks post-surgery. On the other hand, a different branching profile was observed with respect to other tissue structures, where significant increase in branching by 1.12-fold was only detected on day 14 post-surgery when measured as TRI (Fig. 6(D)). This coincidental similarity in profile with tissue density (Fig. 4(B)) may be a reflection of enhanced production or activity of other tissue constituents at this time point. Overall, although the increase in branching in the collagen network may signify a stiffer [43] postoperative tissue within the first week post-surgery, this likelihood is greatly diminished by the ostensibly reduced collagen densities during this period (Fig. 4(A)).
Fig. 6. Progressive alteration of postoperative collagen reticulation. Related parameters for collagen and matrix reticulation that were significantly different between operated and unoperated are shown. (A) collagen reticulation index (CRI), (B) collagen area reticulation density (CARD), (C) tissue reticulation index (TRI).
3.2.5. Quantitative changes in collagen fiber dimensions
As collagen thickness distribution is highly heterogeneous at the micrometer scale, with the diameter of the collagen fibrils and fibers ranging from a few tenth of nm to a few µm [44], we evaluated collagen thickness in the conjunctiva according to the lower quartile, median, and upper quartile thickness distribution. Whereas the thickness of the collagen fibers in the day 2 postoperative tissue were not significantly different from the unoperated counterpart, the postoperative tissues at all later time points generally appeared to feature thinner collagen fibers in all three thickness categories (Fig. 7(A)).
Fig. 7. Progressive alteration of postoperative collagen fiber dimensions. Parameters for collagen fiber dimensions that were significantly different between operated and unoperated are shown. (A) collagen fiber thickness, (B) collagen fiber length, and (C) collagen straightness. Each collagen fiber property was measured with respect to the 25th quartile, median, and 75th quartile distribution. *, p<0.05; fold changes between two means are indicated if significant. C = contralateral, unoperated control conjunctiva; O = operated conjunctival tissue.
When collagen lengths were similarly evaluated, we found that collagen fibers in the days 14 and 21 postoperative tissues were generally significantly shorter than the unoperated counterpart across the lower, median, and upper quartile length distributions (Fig. 7(B)).
Interestingly, with regards to collagen fiber straightness, collagen fibers in the days 2 and 7 postoperative tissues were generally significantly less straight than the unoperated counterpart across the lower, median, and upper quartile straightness distributions (Fig. 7(C)). However, there did not appear to be any significant deviation in collagen fiber straightness from normal at days 14 and 21 post-surgery.
Overall, collagen fibers in the first week post-surgery were less straight than normal, while collagen fibers from the second week of wound healing onwards were thinner and shorter compared to fibers in the unoperated, normal conjunctiva.
3.3. Multiphoton imaging of the postoperative SPARC-/- conjunctiva
We have previously established via electron microscopy that deficiency in the SPARC protein resulted in the deposition of reduced and thinner collagen fibers in the postoperative conjunctiva [13]. The SPARC-/- mouse is therefore a useful model for validating the capacity of quantitative multiphoton technology to provide accurate quantitative description of collagen parameters and measure changes arising from the ultrastructure level.
Both histochemical staining with hematoxylin and eosin (H&E) and Picrosirius Red immediately revealed ostensibly fewer collagen fibers in the SPARC-/- postoperative conjunctiva on both days 7 and 14 (Fig. 8, left and middle panels). As before, multiphoton imaging produced similar observations, particularly with conspicuously weaker SHG signals in the SPARC-/- postoperative conjunctiva (Fig. 8, right panel).
Fig. 8. Visualization of postoperative WT and SPARC-/- collagen matrix by histochemical staining and multiphoton microscopy. Serial sections were stained by hematoxylin and eosin (H&E) or Picrosirius Red followed by viewing with brightfield or polarized microscopy respectively while unstained sections were viewed via SHG/ TPE. S, sclera. Scale bars = 100 µm.
3.4. Quantitative changes in postoperative SPARC-/- collagen organization
As we have established that days 7 and 14 in C57BL/6 WT mice were the time points when most collagen parameters significantly deviated between operated tissues, these time points were chosen for further evaluation of the postoperative SPARC-/- conjunctiva after experimental surgery. When compared against its congenic 129SVE WT postoperative counterpart, we found that CAR, CFD and CTCon were all significantly reduced in the SPARC-/- postoperative tissue at both time points (Fig. 9(A, (B), (C))). The lower collagen content at both time points in the operated SPARC-/- conjunctiva compared to WT is in agreement with our previous molecular findings [13]. A 1.05-fold significant increase in collagen fiber orientation variation was detected in the day 7 but not day 14 SPARC-/- operated tissue (Fig. 9(D)). Curiously, the SPARC-/- postoperative conjunctiva featured increased reticulation compared to the WT counterpart at both time points, as demonstrated by the significant elevation of CRI (Fig. 9(E)). Again, given the large decrease in collagen density in the SPARC-/- operated tissue at both time points, it is unlikely that the increased reticulation detected implies increased tissue stiffness. Most importantly, multiphoton imaging captured significantly reduced collagen fiber thickness in the postoperative SPARC-/- conjunctiva, in agreement with previous findings of this phenomenon in the mouse model (Fig. 9(F)) [13,27]. Moreover, we found that previously uncharacterized collagen fiber properties including length and straightness were further diminished in the SPARC-/- operated conjunctiva compared to WT scar (Fig. 9(G, (H))). Overall, deficiency in SPARC likely leads to a mature postoperative scar that features greater reductions in collagen density, fine structure, fiber thickness, length and straightness compared to WT scar. These alterations in collagen properties are potential contributors to the delayed scarring response observed previously [13].
Fig. 9. Alteration of collagen characteristics in the SPARC-/- conjunctiva after experimental surgery. Parameters that were significantly different between SPARC-/- and WT at the respective time points are shown. (A) collagen area ratio (CAR), (B) collagen fiber density (CFD), (C) collagen texture contrast (CTCon), (D) collagen fiber orientation variation, (E) collagen reticulation index (CRI), (F) mean collagen fiber thickness, (G) mean collagen fiber length, (H) mean collagen fiber straightness. *, p<0.05; fold changes between two means are indicated where significant.
4. Discussion
In this study of postoperative conjunctival scar progression by means of quantitative label-free multiphoton microscopy, we have collected SHG emission in the forward direction which allows the fibrillar nature of the collagen matrix, features of collagen bundles, including size, diameter, and their directionality, to be revealed [45]. As backward signals were not recorded, information concerning the submicron structure of the collagen architecture is not available. In spite of this shortcoming, this method was sufficient to demonstrate that perturbation of collagen homeostasis in the operated tissue involved evolving abnormality in collagen fiber organization. Based on their quantitative deviations from the normal tissue, as well as striking deviation between time points, we have identified parameters that may identify distinct stages of scar development as well as describe scar pathology. In particular, by combining previously reported molecular data [15] with the current SHG measurements, it is now possible to stage conjunctival scar development in the mouse model of conjunctival scarring (Fig. 10). The proliferative phase of wound healing is defined by the activation of hypertrophic fibroblasts responsible for producing the “scar tissue” which comprises of a dense, cellular, collagenous connective tissue matrix [46]. In other words, this phase is characterized by active collagen production and deposition. Indeed, we have previously shown that progressive fibrotic development in the mouse model of conjunctival scarring may be characterized by increased expression of type I collagen which is the predominant component of the ECM and connective tissues [15,29]. We have also found, through RNA sequencing, that other collagen genes, particularly Col8a1, Col11a1 and Col8a2, were highly upregulated in the mouse model [15]. However, in contrast to the non-fibrillar type VIII collagen, type I collagen was the most consistently upgraded at both mRNA and protein levels, suggesting that fibrillar type I collagen is the most stable biomarker of the postoperative conjunctival scar. As highest type I collagen transcription was measured on day 7, it is reasonable to infer that the proliferative phase of wound healing occurs at this time point [15,29]. In support, SHG imaging of fibrillar collagen, particularly efficient for type I collagen [47], detected an increase in the area of the postoperative tissue occupied by collagen (CAR) on day 7 post-surgery. However, the increase in CAR was not accompanied by an increase in collagen fiber density (CFD), which was detected only from day 14 onwards. As CFD is a measure of SHG signal intensity while CAR is related to SHG signal area, it is likely that the increase in CAR on day 7 was a reflection of an increase in well-dispersed deposition of fine, immature, delicate fibers over an expanded tissue area, all of which were apparent in both H&E and Picrosirius Red histochemistry. Day 7 therefore likely marks the beginning of the proliferative phase of wound healing that is characterized by the formation of a provisional, loose and highly hydrated wound matrix/ granulation tissue that allows cellular invasion and/ proliferation and repair to occur. The CAs from day 14 onwards have seemingly dwindled to the normal size of the unoperated tissue, suggesting that tissue contraction that marks the end of the proliferative phase, has occurred by day 14. Sustained production and/ or stability of collagen protein within a smaller CA would explain the detection of increased CFD only from day 14 onwards. An increase of other tissue components, detected through the TPE signals, measured as TD, was also noted on day 14, supporting the formation of a dense, cellular scar tissue at this juncture. Hence, in the postoperative conjunctiva, the proliferative stage very likely occurs within the first 2 weeks post-surgery.
Fig. 10. Collagen attributes in postoperative versus unoperated, normal conjunctiva in the mouse model of conjunctival scarring. Schematic of postoperative tissue size, Col1a1 mRNA induction [15], and collagen characteristics based on SHG measurements are collectively combined to facilitate staging of scar development and profiling of the mature scar phenotype. Diminishing sizes of the postoperative area together with changes in collagen density (represented by increasing colour density) are shown schematically. -, no significant change from unoperated, normal conjunctiva. Black arrow = direction of significant change relative to unoperated control. Green/ blue arrow in parentheses = direction of significant change in postoperative SPARC-/- relative to postoperative WT on day 7/ day 14 respectively.
The remodeling or wound resolution phase involves the activation of processes that attempt to restore “normality” to the injured tissue. In this phase, biological processes that are activated include the shrinking of collagen content by suppressing its production and increasing its degradation. This phase concludes with the maturation of the injury-induced type I collagen that determines scar integrity and strength. In the mouse model of conjunctival scarring, collagen transcript induction was indeed observed to take a sharp downward turn from day 14 onwards [15]. We may also draw on knowledge from skin wound healing to further identify the remodeling or scar maturation phase. In normal skin, type I collagen predominates, but this status is altered during the early stages of wound healing when increase in type III collagen expression overtakes that of type I collagen, with return to the normal ratio only when the scar matures [48]. Thus, increased ratio of type III to type I collagen may identify an immature scar. In the mouse postoperative conjunctiva, the induction of type III collagen trended towards being higher than type I collagen in the first 2 weeks, with the gap being largest on day 2 by about 2.5-fold, and closing in progressively till day 21, when the induction became significantly lower than that of type I collagen by about 2.1-fold [15]. Assuming that conjunctival wound healing is analogous to that of skin, these molecular data suggest that scar maturation in the mouse postoperative conjunctiva have occurred by day 21. In this regard, molecular assessments are useful complements to SHG, as the latter has its limitations, including difficulties in obtaining imageable SHG signals from type III collagen [49]. On the other hand, SHG/ TPE measurements did provide other indications, such as return of the density of other tissue components (TD) to preoperative level on day 21, as would be anticipated of the resolution phase. The putative mature conjunctival scar on day 21 is, however, clearly different from normal tissue, and may be distinguished from the unoperated tissue by a significant increase in collagen content (CFD), amongst several other characteristics (Fig. 10).
Errors occurring in the wound resolution phase have been implicated in excessive wound healing leading to hypertrophic or chronic dermal scarring with persistent granulation tissue. Collagen characteristics that define the mature scar may therefore potentially serve as focal points for scar prognosis, diagnosis, and even anti-fibrotic drug development. As SHG is dependent on the crystalline triple-helix structure of collagen for generation of harmonic signal, the persistent reduction of collagen texture (CTCon) signals in the scar tissues over time was a sign that the scar tissue being formed was significantly different from normal.
The degree of alignment of the final collagen matrix has been implicated in determining the level of scarring. In fact, orientation of the fibrous matrix is often considered to be the most significant difference between normal and scar tissue in the skin [50]. Early studies of skin incisions, as well as sophisticated modeling systems of skin wound healing, have all demonstrated that the orientation of collagen fibers were most likely influenced by the direction of movement of fibroblasts that synthesize them, and that variation of fiber orientation is related to the degree of randomness of cell paths [51,52]. While a fairly random orientation is the norm in normal skin, increased collagen alignment is often associated with scarring in the skin [39, 50–54]. It was also suggested that if fiber orientation was disorganized during the proliferative phase, then reorganization during the final remodeling phase would occur to establish the final more aligned orientation [55]. In the mouse postoperative conjunctiva, we noticed a marginal (less than 10%) but significant increase in fiber orientation variation on day 14, which may reflect the distinct collagen pattern seen in both the histochemical images and multiphoton scans at this time point. The existence of an apparent pattern suggests that some fibroblasts which were laying down intense collagen tracks were not migrating randomly, but rather appeared to have been drawn from the episclera towards the conjunctival epithelium or vice versa. However, by day 21, the fiber orientation variation was no longer significantly different from normal, affirming that remodeling has occurred to reorganize the collagen orientation to the preoperative level. Although we did not detect a deviation from normal in collagen orientation in the putative mature scar, likely due to this being a model of normal scarring, there remains the possibility that collagen alignment may alter dramatically when pathological excessive scarring occurs in the human conjunctiva.
While collagen orientation defines the network organization, the stability of this arrangement is conferred by reticulation, or branching of the fibers, which in turn contributes to the mechanical resilience of the tissue. The self-assembly of collagen into fibrillar networks is encoded in the collagen molecule, and under physiologically relevant conditions, fibrils branch and cross-link readily to form three-dimensional networks [40,43]. However, the degree of collagen reticulation is not easily assessable, as it is challenging for imaging technology to unambiguously distinguish a branched or cross-linked pair in an apparent junction of two fibers from an entangled or obliquely-overlapping pair. With this caveat in mind, it is possible that the increased reticulation observed in the early time points may be confused with either disrupted or newly-formed fibers crossing over one another in the freshly-damaged day 2 or proliferative stage day 7 tissue respectively. In other words, the enhanced reticulation observed in the early time points may not be indicative of bona fide collagen branching. On the other hand, the relatively marginal increase in collagen branching on day 14 may more likely represent enhanced branching/ cross-linking as it not only coincided with increased branching in other tissue fibers at this time point, as indicated by TRI, but also coincided with features that suggest a more mature network. From this perspective, collagen reticulation measurements based on imaging may be more accurately suited to describing the collagen network from day 14 onwards, with earlier time points likely presenting other confounding factors for this parameter. In any case, alteration in collagen reticulation is not apparent in the putative mature conjunctival scar.
Collagen fiber dimensions are known to be altered in scars. For instance, collagen in skin scars graduated towards thinner and thicker fibers in hypertrophic and keloid scars respectively [54]. In the wounded ligament, the persistence of small collagen fibril diameters in the scar matrix has been described [55]. In the postoperative conjunctiva, we too detected thinner collagen fibers in the tissue from day 7 onwards, and this appears likely to be a long-term defect as the phenotype persisted till at least day 21 in the putative mature scar. The collagen fiber lengths demonstrate a similar profile as fiber thickness across the time points, with significantly shorter fibers being particularly dominant from day 14 onwards. Although the straightness measurements recorded reductions in the days 2 and 7 matrix, likely reflecting the nature of disrupted and newly-formed fibers respectively, this feature did not appear to deviate from normal in the putative mature scar. Hence, in terms of fiber dimensions, the putative mature scar on day 21 may be described as being comprised of thinner and shorter collagen fibers.
When probed further on the capacity of SHG to report on the known abnormal collagen fiber thickness in the postoperative scar of the SPARC-/- mouse [13], we have provided validation for this technology as an accurate method for measuring postoperative conjunctival scar pathology and predicting alterations in collagen ultrastructure. SHG indicated that collagen fiber thickness in the SPARC-/- operated tissue were 40-60% of that in the WT. This corroborates with our previous finding that collagen fibrils in the SPARC-/- conjunctiva was approximately 50% that of WT when measured by electron microscopy [13]. Hence, phenotypic differences detectable by SHG analysis are associated with differences in collagen ultrastructure. Simultaneously, SHG has also revealed other hitherto unknown collagen abnormalities in the SPARC-/- mouse that may contribute to the delayed scarring observed in the postoperative conjunctiva [13]. The deviation in SPARC-/- scar phenotype, in order of greatest to least difference from the WT scar, are CFD, CRI, fiber thickness, CAR, CTCon, fiber length and straightness. These differences in collagen qualities may potentially serve as useful targets for therapeutic intervention to delay scar progression. Further work is required to determine whether these characteristics associated with SPARC-/- are prognostic and/ or diagnostic for delayed scar development and/ or reduced scarring respectively.
Taken together, multiphoton imaging has provided crucial characterization of the dynamic alterations that occur in the collagen structure of the conjunctiva after surgery. For an even more detailed understanding of the scar structural phenotype, further work includes the use of polarization-resolved SHG imaging which has been demonstrated to facilitate the extraction of more defined structural data regarding the macrostructure of the fiber bundles in fibrotic collagen such as fiber packing, molecular orientation and degree of organization [18,34]. This study therefore highlights the value of multiphoton technology in GFS. Beyond the direct application of SHG as a tool for evaluating anti-fibrotic drug candidates in animal models of GFS, recent advances in in vivo imaging of the anterior segment of the human eye [56] suggests great promise for this technology to be harnessed as an objective and accurate diagnostic method for early signs of pathological scar progression in patients who have undergone GFS.
5. Conclusions
We have used the potential of intrinsic SHG signals to extract quantitative biomorphologic features on collagen-related changes, and thereby revealed the unique hallmarks of scar development in the postoperative conjunctiva. The mature postoperative conjunctival scar in the mouse model of conjunctival scarring may be summarized as consisting of a denser than normal assembly of thinner and shorter collagen fibers. By combining the present data with previous molecular assessments, we are able to define the stages of scar progression, with the proliferative phase occurring between days 7 and 14, and scar maturation occurring by day 21. As this is a model of normal conjunctival scarring, we expect that the collagen characteristics associated with distinct wound healing phases may potentially be applied as “biomarkers” for anti-fibrotic therapeutic investigations, such as in evaluations of anti-fibrotic candidate agents for the capacity to modulate both collagen density/ content and collagen organization. New findings of collagen structural deviations in the SPARC-/- mouse may provide additional collagen qualities that may serve as targets for anti-fibrotic drug development. We envisage that anti-fibrotic therapeutic approaches that combine the reduction of collagen load with simultaneous targeting of collagen structure will be more effective than remedies that target either one of these aspects alone.
Funding
National Medical Research Council (NHIC-I2D-1503041, NMRC/CG/015/2013, NMRC/CSA-SI/0001/2015); SERI-Lee Foundation Pilot Grant (R1493/76/2017).
Acknowledgments
The authors thank Dr Gideon Ho (Histoindex Pte Ltd) and Ms. Serene Lek (Clinnovate Health Pte Ltd) for their help in supporting and coordinating the project.
Disclosures
The authors declare no conflicts of interest.
References
1. B. J. Larson, M. T. Longaker, and H. P. Lorenz HP, “Scarless fetal wound healing: a basic science review,” Plast. Reconstr. Surg. 126(4), 1172–1180 (2010). [CrossRef]
2. C. D. Marshall, M. S. Hu, T. Leavitt, L. A. Barnes, H. P. Lorenz, and M. T. Longaker, “Cutaneous Scarring: Basic Science, Current Treatments, and Future Directions,” Adv. Wound Care 7(2), 29–45 (2018). [CrossRef]
3. M. Xue and C. J. Jackson, “Extracellular Matrix Reorganization During Wound Healing and Its Impact on Abnormal Scarring,” Adv. Wound Care 4(3), 119–136 (2015). [CrossRef]
4. M. G. Jones, O. G. Andriotis, J. J. Roberts, K. Lunn, V. J. Tear, L. Cao, K. Ask, D. E. Smart, A. Bonfanti, P. Johnson, A. Alzetani, F. Conforti, R. Doherty, C. Y. Lai, B. Johnson, K. N. Bourdakos, S. V. Fletcher, B. G. Marshall, S. Jogai, C. J. Brereton, S. J. Chee, C. H. Ottensmeier, P. Sime, J. Gauldie, M. Kolb, S. Mahajan, A. Fabre, A. Bhaskar, W. Jarolimek, L. Richeldi, K. M. O’Reilly, P. D. Monk, P. J. Thurner, and D. E. Davies, “Nanoscale dysregulation of collagen structure-function disrupts mechano-homeostasis and mediates pulmonary fibrosis,” eLife 7, e36354 (2018). [CrossRef]
5. R. N. Weinreb, T. Aung, and F. A. Medeiros, “The pathophysiology and treatment of glaucoma: a review,” JAMA 311(18), 1901–1911 (2014). [CrossRef]
6. G. L. Skuta and R. K. Parrish 2nd, “Wound healing in glaucoma filtering surgery,” Surv. Ophthalmol. 32(3), 149–170 (1987). [CrossRef]
7. C. Tan, P. T. Chew, W. L. Lum, and C. Chee, “Trabeculectomy - success rates in a Singapore hospital,” Singapore Med. J. 37(5), 505–507 (1996).
8. F. M. Sharif and S. Selvarajah, “The outcome of trabeculectomy for primary glaucoma in adult patients in UKM,” Med. J. Malaysia 52(1), 17–25 (1997).
9. N. M. Radcliffe, “Trabeculectomy revision as a treatment for failed trabeculectomy,” Glaucoma Today Fall, 25–28 (2010).
10. S. K. Law, S. P. Modjtahedi, A. Mansury, and J. Caprioli, “Intermediate-term comparison of trabeculectomy with intraoperative mitomycin-C between Asian American and Caucasian glaucoma patients: a case-controlled comparison,” Eye 21(1), 71–78 (2007). [CrossRef]
11. G. Holló, “Wound Healing and glaucoma surgery: modulating the scarring process with conventional antimetabolites and new molecules,” Dev. Ophthalmol. 59, 80–89 (2017). [CrossRef]
12. O. Yamanaka, A. Kitano-Izutani, K. Tomoyose, and P. S. Reinach, “Pathobiology of wound healing after glaucoma filtration surgery,” BMC Ophthalmol. 15(S1), 157 (2015). [CrossRef]
13. L. F. Seet, R. Su, V. A. Barathi, W. S. Lee, R. Poh, Y. M. Heng, E. Manser, E. N. Vithana, T. Aung, M. Weaver, E. H. Sage, and T. T. Wong, “SPARC deficiency results in improved surgical survival in a novel mouse model of glaucoma filtration surgery,” PLoS One 5(2), e9415 (2010). [CrossRef]
14. L. F. Seet, W. S. Lee, R. Su, S. N. Finger, J. G. Crowston, and T. T. Wong, “Validation of the glaucoma filtration surgical mouse model for antifibrotic drug evaluation,” Mol. Med. 17(5-6), 557–567 (2011). [CrossRef]
15. L. F. Seet, L. Z. Toh, S. W. L. Chu, S. N. Finger, J. L. L. Chua, and T. T. Wong, “Upregulation of distinct collagen transcripts in post-surgery scar tissue: a study of conjunctival fibrosis,” Dis. Models Mech. 10(6), 751–760 (2017). [CrossRef]
16. J. M. Bueno, F. J. Ávila, and M. C. Martínez-García, “Quantitative analysis of the corneal collagen distribution after in vivo cross-linking with second harmonic microscopy,” BioMed Res. Int. 2019, 1–12 (2019). [CrossRef]
17. F. J. Ávila, P. Artal, and J. M. Bueno, “Quantitative discrimination of healthy and diseased corneas with second harmonic generation microscopy,” Trans. Vis. Sci. Tech. 8(3), 51 (2019). [CrossRef]
18. L. Mostaço-Guidolin, N. L. Rosin, and T. L. Hackett, “Imaging Collagen in Scar Tissue: Developments in Second Harmonic Generation Microscopy for Biomedical Applications,” Int. J. Mol. Sci. 18(8), 1772 (2017). [CrossRef]
19. P. J. Campagnola, A. C. Millard, M. Terasaki, P. E. Hoppe, C. J. Malone, and W. A. Mohler, “Three-dimensional high-resolution second-harmonic generation imaging of endogenous structural proteins in biological tissues,” Biophys. J. 82(1), 493–508 (2002). [CrossRef]
20. S. W. Teng, H. Y. Tan, J. L. Peng, H. H. Lin, K. H. Kim, W. Lo, Y. Sun, W. C. Lin, S. J. Lin, S. H. Jee, P. T. So, and C. Y. Dong, “Multiphoton autofluorescence and second-harmonic generation imaging of the ex vivo porcine eye,” Invest. Ophthalmol. Visual Sci. 47(3), 1216–1224 (2006). [CrossRef]
21. O. Masihzadeh, T. C. Lei, D. A. Ammar, M. Y. Kahook, and E. A. Gibson, “A multiphoton microscope platform for imaging the mouse eye,” Mol. Vis. 18, 1840–1848 (2012).
22. V. E. Tovell, A. H. Dahlmann-Noor, P. T. Khaw, and M. Bailly, “Advancing the treatment of conjunctival scarring: a novel ex vivo model,” Arch. Ophthalmol. 129(5), 619–627 (2011). [CrossRef]
23. D. C. Tai, N. Tan, S. Xu, C. H. Kang, S. M. Chia, C. L. Cheng, A. Wee, C. L. Wei, A. M. Raja, G. Xiao, S. Chang, J. C. Rajapakse, P. T. So, H. H. Tang, C. S. Chen, and H. Yu, “Fibro-C-Index: comprehensive, morphology-based quantification of liver fibrosis using second harmonic generation and twophoton microscopy,” J. Biomed. Opt. 14(4), 044013 (2009). [CrossRef]
24. S. Xu, Y. Wang, D. C. S. Tai, S. Wang, C. L. Cheng, Q. Peng, J. Yan, Y. Chen, J. Sun, X. Liang, Y. Zhu, J. C. Rajapakse, R. E. Welsch, P. T. C. So, A. Wee, J. Hou, and H. Yu, “qFibrosis: a fully-quantitative innovative method incorporating histological features to facilitate accurate fibrosis scoring in animal model and chronic hepatitis B patients,” J. Hepatol. 61(2), 260–269 (2014). [CrossRef]
25. P. Puolakkainen, A. D. Bradshaw, T. R. Kyriakides, M. Reed, R. Brekken, T. Wight, P. Bornstein, B. Ratner, and E. H. Sage, “Compromised production of extracellular matrix in mice lacking secreted protein, acidic and rich in cysteine (SPARC) leads to a reduced foreign body reaction to implanted biomaterials,” Am. J. Pathol. 162(2), 627–635 (2003). [CrossRef]
26. T. J. Rentz, F. Poobalarahi, P. Bornstein, E. H. Sage, and A. D. Bradshaw, “SPARC regulates processing of procollagen I and collagen fibrillogenesis in dermal fibroblasts,” J. Biol. Chem. 282(30), 22062–22071 (2007). [CrossRef]
27. H. Ho, H. M. Htoon, G. H. Yam, L. Z. Toh, N. C. Lwin, S. Chu, Y. S. Lee, T. T. Wong, and L. F. Seet, “Altered Anterior Segment Biometric Parameters in Mice Deficient in SPARC,” Invest. Ophthalmol. Visual Sci. 58(1), 386–393 (2017). [CrossRef]
28. K. Norose, J. I. Clark, N. A. Syed, A. Basu, E. Heber-Katz, E. H. Sage, and C. C. Howe, “SPARC deficiency leads to early-onset cataractogenesis,” Invest. Ophthalmol. Visual Sci. 39(13), 2674–2680 (1998).
29. L. F. Seet, S. N. Finger, S. W. Chu, L. Z. Toh, and T. T. Wong, “Novel insight into the inflammatory and cellular responses following experimental glaucoma surgery: a roadmap for inhibiting fibrosis,” Curr. Mol. Med. 13(6), 911–928 (2013). [CrossRef]
30. F. Aptel, N. Olivier, A. Deniset-Besseau, J. M. Legeais, K. Plamann, M. C. Schanne-Klein, and E. Beaurepaire, “Multimodal nonlinear imaging of the human cornea,” Invest. Ophthalmol. Visual Sci. 51(5), 2459–2465 (2010). [CrossRef]
31. W. R. Zipfel, R. M. Williams, R. Christie, A. Y. Nikitin, B. T. Hyman, and W. W. Webb, “Live tissue intrinsic emission microscopy using multiphoton-excited native fluorescence and second harmonic generation,” Proc. Natl. Acad. Sci. U. S. A. 100(12), 7075–7080 (2003). [CrossRef]
32. L. C. Junqueira, G. Bignolas, and R. R. Brentani, “Picrosirius staining plus polarization microscopy, a specific method for collagen detection in tissue sections,” Histochem. J. 11(4), 447–455 (1979). [CrossRef]
33. X. Chen, O. Nadiarynkh, S. Plotnikov, and P. J. Campagnola, “Second harmonic generation microscopy for quantitative analysis of collagen fibrillar structure,” Nat. Protoc. 7(4), 654–669 (2012). [CrossRef]
34. M. Strupler, A. M. Pena, M. Hernest, P. L. Tharaux, J. L. Martin, E. Beaurepaire, and M. C. Schanne-Klein, “Second harmonic imaging and scoring of collagen in fibrotic tissues,” Opt. Express 15(7), 4054–4065 (2007). [CrossRef]
35. R. Lattouf, R. Younes, D. Lutomski, N. Naaman, G. Godeau, K. Senni, and S. Changotade, “Picrosirius red staining: a useful tool to appraise collagen networks in normal and pathological tissues,” J. Histochem. Cytochem. 62(10), 751–758 (2014). [CrossRef]
36. C. R. Drifka, A. G. Loeffler, K. Mathewson, G. Mehta, A. Keikhosravi, Y. Liu, S. Lemancik, W. A. Ricke, S. M. Weber, W. J. Kao, and K. W. Eliceiri, “Comparison of picrosirius red staining with second harmonic generation imaging for the quantification of clinically relevant collagen fiber features in histopathology samples,” J. Histochem. Cytochem. 64(9), 519–529 (2016). [CrossRef]
37. E. Brown, T. McKee, E. diTomaso, A. Pluen, B. Seed, Y. Boucher, and R. K. Jain, “Dynamic imaging of collagen and its modulation in tumors in vivo using second-harmonic generation,” Nat. Med. 9(6), 796–800 (2003). [CrossRef]
38. S. Zhuo, J. Chen, B. Yu, X. Jiang, T. Luo, Q. Liu, R. Chen, and S. Xie, “Nonlinear optical microscopy of the bronchus,” J. Biomed. Opt. 13(5), 054024 (2008). [CrossRef]
39. J. M. Rawlins, W. L. Lam, R. O. Karoo, I. L. Naylor, and D. T. Sharpe, “Quantifying collagen type in mature burn scars: a novel approach using histology and digital image analysis,” J. Burn Care Res. 27(1), 60–65 (2006). [CrossRef]
40. K. A. Jansen, A. J. Licup, A. Sharma, R. Rens, F. C. MacKintosh, and G. H. Koenderink, “The role of network architecture in collagen mechanics,” Biophys. J. 114(11), 2665–2678 (2018). [CrossRef]
41. C. F. Whittington, M. C. Yoder, and S. L. Voytik-Harbin, “Collagen-polymer guidance of vessel network formation and stabilization by endothelial colony forming cells in vitro,” Macromol. Biosci. 13(9), 1135–1149 (2013). [CrossRef]
42. S. Lin and L. Gu, “Influence of crosslink density and stiffness on mechanical properties of type I collagen gel,” Materials 8(2), 551–560 (2015). [CrossRef]
43. J. L. Bailey, P. J. Critser, C. Whittington, J. L. Kuske, M. C. Yoder, and S. L. Voytik-Harbin, “Collagen oligomers modulate physical and biological properties of three-dimensional self-assembled matrices,” Biopolymers 95(2), 77–93 (2011). [CrossRef]
44. D. J. Hulmes, “Building collagen molecules, fibrils, and suprafibrillar structures,” J. Struct. Biol. 137(1-2), 2–10 (2002). [CrossRef]
45. F. Légaré, C. Pfeffer, and B. R. Olsen, “The role of backscattering in SHG tissue imaging,” Biophys. J. 93(4), 1312–1320 (2007). [CrossRef]
46. T. J. Shaw and P. Martin, “Wound repair at a glance,” J. Cell Sci. 122(18), 3209–3213 (2009). [CrossRef]
47. P. Campagnola, “Second harmonic generation imaging microscopy: applications to diseases diagnostics,” Anal. Chem. 83(9), 3224–3231 (2011). [CrossRef]
48. T. Hayakawa, Y. Hashimoto, Y. Myokei, H. Aoyama, and Y. Izawa, “Changes in type of collagen during the development of human post-burn hypertrophic scars,” Clin. Chim. Acta 93(1), 119–125 (1979). [CrossRef]
49. G. Cox, E. Kable, A. Jones, I. Fraser, F. Manconi, and M. D. Gorrell, “3-dimensional imaging of collagen using second harmonic generation,” J. Struct. Biol. 141(1), 53–62 (2003). [CrossRef]
50. D. Wolfram, A. Tzankov, P. Pülzl, and H. Piza-Katzer, “Hypertrophic scars and keloids - a review of their pathophysiology, risk factors, and therapeutic management,” Dermatol. Surg. 35(2), 171–181 (2009). [CrossRef]
51. M. Laufer, C. Ashkenazi, D. Katz, and M. Wolman, “Orientation of collagen in wound healing,” Br. J. Exp. Pathol. 55(3), 233–236 (1974).
52. B. D. Cumming, D. L. McElwain, and Z. Upton, “A mathematical model of wound healing and subsequent scarring,” J. R. Soc., Interface 7(42), 19–34 (2010). [CrossRef]
53. J. C. Dallon, J. A. Sherratt, and P. K. Maini, “Mathematical modelling of extracellular matrix dynamics using discrete cells: fiber orientation and tissue regeneration,” J. Theor. Biol. 199(4), 449–471 (1999). [CrossRef]
54. P. D. Verhaegen, P. P. van Zuijlen, N. M. Pennings, J. van Marle, F. B. Niessen, C. M. van der Horst, and E. Middelkoop, “Differences in collagen architecture between keloid, hypertrophic scar, normotrophic scar, and normal skin: An objective histopathological analysis,” Wound Repair Regen. 17(5), 649–656 (2009). [CrossRef]
55. C. B. Frank, “Ligament structure, physiology and function,” J. Musculoskelet. Neuronal Interact. 4(2), 199–201 (2004).
56. F. J. Ávila, A. Gambín, P. Artal, and J. M. Bueno, “In vivo two-photon microscopy of the human eye,” Sci. Rep. 9(1), 10121 (2019). [CrossRef]
