Given the incidence of corneal dysfunctions and diseases worldwide and the limited availability of healthy, human donors, investigators are working to generate engineered cellular and acellular therapeutic approaches as alternatives to corneal transplants from human cadavers. These engineered strategies aim to address existing complications with human corneal transplants, including graft rejection, infection, and complications resulting from surgical methodologies. The main goals of these research endeavors are to (1) determine ideal mechanical properties, (2) devise methodologies to improve the efficacy of engineered corneal grafts and cell-based therapies, and (3) optimize transplantation of engineered tissue structures in the eye. Thus, recent innovations have sought to address these challenges through both in vitro and in vivo studies. This review covers recent work aimed at evaluating engineered materials, potential therapeutic cells, and the resulting cell-material interactions that lead to optimal corneal graft properties. Furthermore, we discuss promising strategies in corneal tissue engineering techniques and in vivo studies in animal models.

Sponge-like biomaterials formed from silk fibroin are promising as degradable materials in clinical applications due to their controllable breakdown into simple amino acids or small peptides in vivo. Silk fibroin, isolated from Bombyx mori silkworm cocoons, can be used to form sponge-like materials with a variety of tunable parameters including the elastic modulus, porosity and pore size, and level of nanocrystalline domains. These parameters can be independently tuned during formulation resulting in a wide parameter space and set of final materials. Determining the mechanism and rate constants for biomaterial degradation of these tunable silk materials would allow scientists to evaluate and predict the biomaterial performance for the large array of tissue engineering applications and patient ailments a priori. We first measured in vitro degradation rates of silk sponges using common protein-degrading enzymes such as Proteinase K and Protease XIV. The concentration of the enzyme in solution was varied (1, 0.1, 0.01 U/mL) along with one silk sponge formulation parameter: the level of crystallinity within the sponge. Additionally, two experimental degradation methods were evaluated, termed continuous and discrete degradation methods. Silk concentration, polymer chain length and scaffold pore size were held constant during experimentation and kinetic parameter estimation. Experimentally, we observed that the enzyme itself, enzyme concentration within the bulk solution, and the sponge fabrication water annealing time were the major experimental parameters dictating silk sponge degradation in our experimental design. We fit the experimental data to two models, a Michaelis-Menten kinetic model and a modified first order kinetic model. Weighted, non-linear least squares analysis was used to determine the parameters from the data sets and Monte-Carlo simulations were utilized to obtain estimates of the error. We found that modified first order reaction kinetics fit the time-dependent degradation of lyophilized silk sponges and we obtained first order-like rate constants. These results represent the first investigations into determining kinetic parameters to predict lyophilized silk sponge degradation rates and can be a tool for future mathematical representations of silk biomaterial degradation.

Biomaterials can influence the coordinated efforts required to achieve tissue rehabilitation. Sponge-like silk fibroin scaffolds that include bioactive molecules have been shown to influence tissue repair. However, the mechanisms by which scaffold formulations elicit desired in vivo responses is unclear. Here, acellular silk scaffolds consisting of type I collagen, heparin, and/or vascular endothelial growth factor (VEGF) were used to investigate material fabrication and composition parameters that drive scaffold degradation, cell infiltration, and adipose tissue deposition in vivo. In subcutaneous implants, scaffold degradation was assessed, and results show that the percentage of cells infiltrating the scaffold increased when scaffold formulations contained bioactive molecules. To gain further insight, calculated in vitro enzymatic degradation rates increased with higher enzyme concentrations and theoretical cleavage sites. However, the addition of type I collagen and heparin to the scaffold at relevant concentrations did not change degradation rates, compared to silk alone. These in vitro results are contrary to observations in vivo, where bioactive molecules influence local protein deposition, immune cell infiltration rates, and vascularization. Thus, quantitative in vitro and in vivo evaluations aid in determining the mechanisms by which biomaterials influence tissue repair and support intentional biomaterial design for clinical applications.

ObjectiveWe quantify adipose tissue deposition at surgical sites as a function of biomaterial implantation.Impact StatementTo our knowledge, this study is the first investigation to apply convolutional neural network (CNN) models to identify and segment adipose tissue in histological images from silk fibroin biomaterial implants.IntroductionWhen designing biomaterials for the treatment of various soft tissue injuries and diseases, one must consider the extent of adipose tissue deposition. In this work, we implant silk fibroin biomaterials in a rodent subcutaneous injury model. Current strategies for quantifying adipose tissue after biomaterial implantation are often tedious and prone to human bias during analysis.MethodsWe used CNN models with novel spatial histogram layer(s) that can more accurately identify and segment regions of adipose tissue in hematoxylin and eosin (H&E) and Masson’s Trichrome stained images, allowing for determination of the optimal biomaterial formulation. We compared the method, Jointly Optimized Spatial Histogram UNET Architecture (JOSHUA), to the baseline UNET model and an extension of the baseline model, Attention UNET, as well as to versions of the models with a supplemental “attention”-inspired mechanism (JOSHUA+ and UNET+).ResultsThe inclusion of histogram layer(s) in our models shows improved performance through qualitative and quantitative evaluation.ConclusionOur results demonstrate that the proposed methods, JOSHUA and JOSHUA+, are highly beneficial for adipose tissue identification and localization. The new histological dataset and code for our experiments are publicly available.

Objective. We aim to develop a machine learning algorithm to quantify adipose tissue deposition at surgical sites as a function of biomaterial implantation. Impact Statement. To our knowledge, this study is the first investigation to apply convolutional neural network (CNN) models to identify and segment adipose tissue in histological images from silk fibroin biomaterial implants. Introduction. When designing biomaterials for the treatment of various soft tissue injuries and diseases, one must consider the extent of adipose tissue deposition. In this work, we analyzed adipose tissue accumulation in histological images of sectioned silk fibroin-based biomaterials excised from rodents following subcutaneous implantation for 1, 2, 4, or 8 weeks. Current strategies for quantifying adipose tissue after biomaterial implantation are often tedious and prone to human bias during analysis. Methods. We used CNN models with novel spatial histogram layer(s) that can more accurately identify and segment regions of adipose tissue in hematoxylin and eosin (H&E) and Masson’s trichrome stained images, allowing for determination of the optimal biomaterial formulation. We compared the method, Jointly Optimized Spatial Histogram UNET Architecture (JOSHUA), to the baseline UNET model and an extension of the baseline model, attention UNET, as well as to versions of the models with a supplemental attention-inspired mechanism (JOSHUA+ and UNET+). Results. The inclusion of histogram layer(s) in our models shows improved performance through qualitative and quantitative evaluation. Conclusion. Our results demonstrate that the proposed methods, JOSHUA and JOSHUA+, are highly beneficial for adipose tissue identification and localization. The new histological dataset and code used in our experiments are publicly available.