Decellularized squid mantle scaffolds as tissue‐engineered corneal stroma for promoting corneal regeneration

Abstract Corneal blindness is a worldwide major cause of vision loss, and corneal transplantation remains to be the most effective way to restore the vision. However, often there is a shortage of the donor corneas for transplantation. Therefore, it is urgent to develop a novel tissue‐engineered corneal substitute. The present study envisaged the development of a novel and efficient method to prepare the corneal stromal equivalent from the marine biomaterials‐squid. A chemical method was employed to decellularize the squid mantle scaffold to create a cell‐free tissue substitute using 0.5% sodium dodecyl sulfate (SDS) solution. Subsequently, a novel clearing method, namely clear, unobstructed brain imaging cocktails (CUBIC) method was used to transparent it. Decellularized squid mantle scaffold (DSMS) has high decellularization efficiency, is rich in essential amino acids, and maintains the regular fiber alignment. In vitro experiments showed that the soaking solution of DSMS was non‐toxic to human corneal epithelium cells. DSMS exhibited a good biocompatibility in the rat muscle by undergoing a complete degradation, and promoted the growth of the muscle. In addition, the DSMS showed a good compatibility with the corneal stroma in the rabbit inter‐corneal implantation model, and promoted the regeneration of the corneal stroma without any evident rejection. Our results indicate that the squid mantle can be a potential new type of tissue‐engineered corneal stroma material with a promising clinical application.

blind people worldwide. Among them, corneal blindness can be restored by corneal transplantation. However, only 1 in 70 patients can ultimately undergo the graft transplantation. 3 The cornea can be divided into five layers from anterior to posterior: epithelial layer, Bowman's membrane, stromal layer, Descemet membrane, and endothelial layer. 4 The corneal stroma layer is approximately 500-μm thick, accounting to a 90% of the entire corneal thickness, and consists of nearly 200 layers of regularly arranged collagen fibers. Unlike the epithelial layer, which has a strong regenerative capacity, the stromal layer is usually thought to lose its original crosslinked structure due to changes in the diameter and gaps between the fibers of collagen formed by tissue repair after stromal injury, resulting in scarring and affecting vision. 5 Currently, corneal transplants have evolved from full-thickness penetrating keratoplasty (PKP) to partial lamellar keratoplasty (ALK). Compared to PKP, ALK has the advantages of fewer intraoperative complications, keeping the eye as intact as possible, and less possibility of postoperative graft rejection. 6 Therefore, the development of corneal stromal equivalents needs an urgent study.
In recent decades, the amniotic membrane (AM), the acellular porcine corneal stroma (APCS), and hydrogel have made rapid progress in the research and development of corneal tissue engineering, 7-10 and a substantial amount of the research work has already reached the clinical practice [11][12][13][14][15] However, the application of AM has many challenges, owing to its limited sources, potential infectious diseases, rejection after transplantation and ethical issues. 16 The main difficulty faced in the use of the porcine-derived xenograft as a donor in corneal transplantation is the immune reaction after transplantation, resulting in loss of transparency, neovascularization, and rejection. As a new functional material, hydrogels offer good biocompatibility, biodegradability, high water absorption, and water retention properties, which leads to its wide application in biomedical fields, including tissue engineering, drug delivery systems, wound dressings, and so on. 17,18 Although its physical, chemical, and mechanical properties are certain.
It cannot simulate the composition and spatial configuration of the natural corneal stroma. 19 Finally, a variety of natural and synthetic biomaterials have been used in corneal tissue engineering, still no material can perfectly satisfy all the characteristics of a tissueengineered cornea. Therefore, it is urgent to develop a novel tissueengineered corneal material.
Corneal graft materials need to possess several important properties, such as superior transparency, low immunogenicity, and excellent biocompatibility. The main component of the corneal stroma is collagen. Meanwhile, marine collagen has been found to possess an amino acid composition similar to the human type I collagen, 20 as well as has advantages over mammalian collagen because of its excellent biocompatibility, low immunogenicity, low production cost, and lack of risk of transmission human susceptible viruses. 21,22 Currently, marine collagen has made significant advances in the fields of medical tissue engineering, drug delivery systems, and food medical and nutraceutical products. [23][24][25] Squid mantles are rich in collagen (up to 29.3%) and mainly consist of type I collagen, and are abundant in essential amino acids. 26 Interestingly, the squid mantle consists of epidermis and endothelium and a muscle layer sandwiched between the two, with a structural composition similar to that of the cornea. 27 Squid is known to possess great application prospects in biomedicine and various tissue engineering fields. For example, squid chitin pen was used in the development of cartilage tissue engineering, which could promote the regeneration of rabbit cartilage. 28 Squid outer skin was used to produce industrial-grade films for medical usage. 29 Decellularized squid mantle was used as urethral reconstruction materials. 30 Based on the above findings, we speculated that squid mantle could be a potential corneal graft material. To the best of our knowledge, relevant research of squid in corneal tissue engineering has not been reported.
In the present study, we explored the feasibility of utilizing the squid mantle as a corneal stromal substitute in corneal transplantation (Scheme 1). First, the squid mantle was subjected to decellularization and transparency, and then the histological structure and chemical composition were further characterized. Subsequently, the in vivo and in vitro toxicity was evaluated. Finally, we implanted the decellularized squid mantle scaffold (DSMS) into the cornea of rabbit and observed its biocompatibility with the cornea. This study established a novel material for the construction of tissue-engineered corneal stroma, which could possess potential applications in the field of cornea regeneration.

| Transparency and characterization of DSMS
The material used for corneal transplantation should be transparent.
Compared with the other six groups, the DFT and DT groups exhibited a better transparency (Figures 1a and S1). The translucency of the squid mantle was found to be improved after the decellularization and transparency processes.
Cryogenic transmission electron microscopy results (Figure 1b) show that the eight groups of materials exhibited a 3D grid-like structure of acellular functional ECM. The untreated control group (Ctrl) presented a three-dimensional structure with interwoven fibers and mesh, and the fibers were neatly arranged. Compared with the Ctrl group, the fiber structure of other groups was damaged to a certain extent. The fiber diameter of the DFT group was shorter than that of the Ctrl group, and the fiber pore size was larger than that of the Ctrl group, which was consistent with the view that the decellularization could damage the cytoskeleton. Besides, the fiber diameter of the DFT group was longer than that of the DT group, and the structure was more stable than that of the D group.
We examined the amino acid content in the eight groups of the squid mantle. The results (Figure 1c) showed that the squid mantle contained 18 amino acids. Glycine is the smallest amino acid, and the presence of glycine in every third amino acid residue is a key requirement for the superhelix structure of collagen. 31 The stability of the helix is directly proportional to the amino acid content. The results ( Figure 1d) indicated that the glycine and proline content decreased after the treatment, compared to the control group, but the content of the DFT group was higher than that of the DT group, indicating that the structure of the DFT group was more stable than that of the DT group.

| Histological staining of DSMS
The surface texture of the squid mantle was smooth and dense. To further explore its internal structure, the paraffin sections of the tissue were stained with different histological staining methods. In the hematoxylin-eosin (H&E) staining (Figure 2a), myogenic fibers (red) in the Ctrl group were correctly arranged and nuclei (blue) in the fibers were clearly visible. Myogenic fibers in the DFT, FT, T, TD, and D groups were irregular compared with the Ctrl group, but many nuclei were visible in the DFT and DT groups. Myogenic fibers in the DT group were blurred and it was difficult to differentiate the fibrous tissue. The tissue gap in the FTD group was moderately extended, compared with that of the Ctrl group. In Sirius Red staining (Figure 2b), myogenic fibers (red) were clearly seen in the Ctrl, DFT, T, FTD, TD, and D groups, but the myogenic fibers in DT and FT groups were blurred and indistinguishable, and some small fragments appeared.
Meanwhile, in Masson staining (Figure 2c), the squid mantle tissues of the Ctrl group were closely connected with the myocollagen (blue) and myogenic fibers (red), and it was seen that the myocollagen was evenly distributed in the tissues (showing a purple color). The color of the other groups was between blue and red, and we speculated that it could be due to a change in the ratio between the myocollagen and myogenic fibers due to the decrease in the binding between them or a decrease in the myocollagen in the tissues after different treatments of the squid mantle.

| Effects of decellularization of squid mantle using different methods
To verify the efficiency of different method of decellularization, the squid mantle was subjected to staining with DAPI for evaluating the nuclear number (Figure 3a). The results showed that the acellular effect in DFT and DT groups was significantly higher than that of the Ctrl group ( Figure 3b). Furthermore, the experimental results of the quantitative analysis of DNA content were consistent with that of the DAPI staining ( Figure 3c). In the present study, triton X-100 was one of the main components of the CUBIC clearing method. As a result, the double washing by SDS and triton X-100 could better remove the nuclei. This shows that the proposed method was very efficient in the decellularization process. Figure S2 shows that the Ctrl and D groups exhibited deformation at 12 h, and the phosphatebuffered saline (PBS) solution became significantly turbid. On the third day, the tissue was swollen, which then gradually degraded and became smaller in size. In the DT, T, and TD groups (without fixation), the tissue size began to slowly decrease at 12 h, but basically remained round. The DFT, FT, and FTD groups maintained relatively stable and regular shapes from beginning to end, indicating that fixation made the tissues more compact and firmer. In particular, PBS in the DFT group remained clear all the time, indicating that the DFT group had the strongest resistance to collagenase.

S C H E M E 1
Overview of the development of DSMS for corneal regeneration. DSMS, decellularized squid mantle scaffold.

| In vitro and in vivo biocompatibility assessment of DSMS
In vitro Cell Counting Kit-8 (CCK-8) tests of the leaching liquor from DSMS were performed to observe the effects of DSMS on the proliferation and growth of corneal cells. It was observed that there were no significant differences in the corneal epithelial cell viability between all groups compared with the baseline (Figure 4b). Furthermore, the live-dead staining assay was also performed to discriminate between the live and dead cells. In accordance with the CCK-8 tests, there was no significant increase in the number of dead cells in each group when compared to the Ctrl group ( Figure S4). The results indicated that DSMS was biocompatible with the corneal cells. removal. 33 We measured the residual SDS content of the materials, and found the SDS content was lower in the DFT group when compared to the D group ( Figure S3). An exacerbation of inflammation due to the mechanical irritation of the double suture certainly cannot be ruled out. Besides, the size of DSMS transplanted into muscle in this study was much larger than the size of DSMS transplanted into the cornea. As a result, we presume that these factors led to a strong inflammatory reaction at the graft site; however, after 2 weeks, the inflammation gradually subsided. In the fourth week sample, the inflammatory cells disappeared, especially in the DFT and DT groups, and the peripheral muscle structure also recovered like the normal muscle. In particular, the implants were degraded in the control group, leading to a loss of the normal muscle structure, and became vacuolated. All rats survived during the postoperative follow-up. Our results indicate that although DSMS are heterogeneous biomaterials, they have excellent histocompatibility.
The wound-healing assay showed that the distance between the cell scratches from each group was equal to that of the control group at every time point (Figure 4c,d). It indicated that DSMS was nontoxic to the cells and could promote their normal proliferation and migration. Besides, immunofluorescence staining with ZO-1 was further conducted, and the results showed that DSMS did not cause any phenotypic changes ( Figure S5).

| Postoperative observation by slit lamp
Rabbit corneal interlamellar implantation was performed to evaluate the condition of cornea after DSMS implantation. Based on the above results, the DSMS using DFT method was found to be better for the corneal stromal implantation graft in New Zealand rabbits. Figure 5b shows that in the sham operation group, the cornea remained transparent throughout the whole follow-up period, and corneal edema recovered rapidly, when compared to the normal cornea ( Figure 5a).
In the DSMS implantation group, the cornea seemed obscure at the beginning, which could be due to the absorption of water from the material during the preoperative preservation (Figure 5c

| Postoperative observation by in vivo confocal microscopy
In vivo confocal microscopy was performed to assess the morphology of the implants and all cellular layers of the surrounding host corneal tissue. 34 The results from Figure 7a   Masson study was performed to stain the collagen fibers. Masson staining results showed that the corneal stromal fibers in the sham operation group were arranged neatly, and no scar was observed. In the implantation group, it was observed that the rabbit corneal stromal fibers were arranged neatly, and the graft was closely fitted to the stroma, indicating that DSMS exhibited a good histocompatibility with cornea ( Figure 8b). and no corneal protrusion was caused due to the transplantation.
The were noted, when compared to normal corneas.

| DISCUSSION
Although varieties of natural and synthetic polymer scaffolds are widely used for tissue-engineered cornea, the original scaffolds of decellularized animals are more valuable due to their excellent biocompatibility, biodegradability, and low immunogenicity. 35 13 Furthermore, the original porcine corneal stroma material also has several problems, for instance, long reproduction period, high production cost, the risk of zoonotic diseases, and religious restrictions. Meanwhile, synthetic polymer scaffolds like hydrogel show a massive potential for tissue-engineered cornea in animal studies; however, a further validation for safety is needed before clinical application. Therefore, it is necessary to develop a novel tissue-engineered corneal stromal substitute, to enrich the selectivity of materials.
Our findings provide evidence that the intracorneal stromal implantation of a decellularized marine-derived colloidal material is a safe and feasible process, wherein the DSMS transplants not only remained relatively clear, but also promoted the corneal stromal regeneration along with enhancing the thickness of the corneal stroma. In recent years, the combination of tissue engineering and marine biomaterials has provided rich resources for the preparation of artificial organs or tissue substitutes. 39 Ueda et al. reported that decellularized and decalcified fish scale-derived collagen matrix met the basic features to act as artificial corneas to replace the donor corneas. 40 This showed a great potential of non-corneal-derived marine materials in the field of corneal tissue engineering.
The cornea is the only transparent organ in the body and being the main component structure of the refractive system, results in a high demand for the transparency of the scaffolds. Multiple materials based on decellularization methods did not achieve the ideal transparency for artificial corneas according to the literature. 41,42 Therefore, improving the transparency, while ensuring a low immunogenicity of the scaffold is indeed a challenging task. 15 Tissue clearing technology was originally created to make the biological samples transparent, to observe the 3D visualization structure of tissues, and to provide a new platform for further exploration of human diseases. 43 DSMS was prepared by using a combination of the SDS decellularization method and the modified CUBIC method. 44 The CUBIC method used water-soluble reagents for clearing the tissue.
Although the transparent effect was not as potent as the hydrophobic method, the former has a higher level of biocompatibility, biosafety, and protein preservation. The hydrophilic reagent allows the protein to form hydrogen bonds with surrounding water molecules, which helps to preserve the three-dimensional structure of the tissue components. 45 Our data showed that the CUBIC technique resulted in transparent squid mantles, which further demonstrated the feasibility of using a graft material of non-corneal source. To our knowledge, this is the first report of combining decellularization and clearing methods and using them to make tissue-engineered corneal stromal equivalents. However, it has to be admitted that even the mildest method may cause some damage to the tissue structure. According to our findings, the damage to the squid mantles can be reduced after the modification and fixation process.
Collagen materials have been widely used as scaffold materials in the field of tissue engineering and collagen-based materials have been implanted in humans in clinical trials, owing to its excellent biocompatibility, biodegradability, low immunogenicity, and cell adhesion. 14,32,46-49 While most common collagen is derived from terrestrial animals, over the past few years, there has been an increasing interest in other sources of collagen due to the risk of zoonotic disease transmission. 21 Studies have shown that marine collagen has certain advantages over the terrestrial collagen, such as abundance in source and no risk of disease transmission. It has also been demonstrated that squid-derived collagen showed a less antigenic response than the terrestrial collagen to support cell growth. 50 However, a potential limitation is the immunogenicity of the decellularized material may arise due to the incomplete removal of the cellular components and DNA during the decellularization process. Minute quantity residues of cellular components or chemicals can trigger an immune response. So far, most of the tissue engineering products from animals (pigs or squid) have not been genetically modified to neutralize genes that could trigger an immune reaction in humans and some immunogenicity is expected, especially if the corneal immune amnesty has been compromised by the initial disease (since rabbits employed in this study were healthy, but in case of humans, they usually suffer from corneal disease).

| Preparation of DSMS
The whole raw squid was obtained and the head, chitin pen, transparent strip, and viscera inside the squid were removed. The outer membrane was peeled off. Deuterium-depleted water (DDW) was used to clean the mantle several times at 4 C. Next, the mantle was cut to Reagent required for 0.5% SDS method: 5 g of sodium dodecyl sulfate (SDS) powder was dissolved in 1000 mL DDW.
ScaleCUBIC-1 (also known as reagent-1): 125 g urea (Solarbio) and 156 g 80% Quadrol were added in 144 g deionized water with heating and stirring, dissolved fully and then cooled to room temperature (RT), then added 75 g of Triton X-100 (Solarbio), dissolved fully and then cooled to RT. 43

| Transparency test
Each set of samples was cut with a 5-mm trephine and placed on an A-card to observe the transparency and imaged by a digital camera.

| Cryo-transfer transmission electron microscopy
The samples were swiftly frozen and fractured by liquid nitrogen under high vacuum conditions. Subsequently, the samples were subjected to sublimation and conductivity spraying under vacuum at À90 C. The samples were then transferred to a scanning electron microscope cooling stage (down to À160 C) by a freeze transfer system in the next step. The microstructure of each group of samples was observed by cryogenic transmission scanning at an accelerating voltage of 3.0 kV.

| DNA extraction
Each group of the squid mantle was placed in an oven at 60 C and dehydrated overnight. The next day, 1000 mg of the sample was weighed and transferred into tubes. First, 500 μL DNA extraction and protease K solution were added into each tube and digested in a 55 C water bath for 1 h. Then, an equal volume of 500 μL phenol/ chloroform (1:1) mixture was added into tubes, mixed for 10 times, and the samples were centrifuged at 12,000 rpm for 10 min at 4 C.
The supernatant was drained gently into the new tubes and phenol/ chloroform/isoamyl alcohol solution (25:24:1 added in equal volumes) was added and centrifuged at 12,000 rpm for 10 min at 4 C. The supernatant was poured off carefully, 75% alcohol was added for precipitation, and centrifuged at 12,000 rpm for 5 min, at 4 C two times. The remaining alcohol in the EB tubes was volatilized.
Then, the DNA content was measured using a spectrophotometer (ND-ONE-A1501015).

| Wound healing assay
For wound healing assays, the ibidi Culture-Insert 2 Well (ibidi Gmbh) was placed in each well of 12-well plate, and eight inserts were used for each treatment, each container separated by a 500-μm wall. 51

| Vibrating microtome
The samples for animal study were prepared as follows: The instru-

| Rabbit corneal interlaminar implantation
Eighteen male New Zealand White rabbits (2.5-3.0 kg) were selected for this study. In three normal controls, three rabbits underwent sham operation on the right eye (without implantation), six rabbits received the DSMS (thickness = 300 μm) implantation, and six rabbits received the DSMS (thickness = 350 μm) implantation. Each rabbit received the procedure only in the right eye. All rabbits were acclimatized to the laboratory conditions at the animal center of Xiamen University for 2 weeks before surgery. Animals were anesthetized with intra-muscular injection of xylazine hydrochloride injection

| H&E and Masson staining of cornea
The rabbits were killed at 4 and 8 weeks after the operation and the eyeballs were fixed in 4% paraformaldehyde (Sigma-Aldrich) for a week. Subsequently, the corneal tissue was embedded in paraffin and sliced. Samples were stained with H&E (Auragene) and Masson, and the slices were observed under a light microscope (Nikon).

| Immunofluorescence staining
The steps for fluorescent staining are as follows: Paraffin sections were dewaxed with xylene and dehydrated with gradient alcohol con-

| Statistical analysis
All data were expressed as mean ± standard deviation (SD). According to the normality of the data distribution, using GraphPad Prism 9.0 software, the statistical significance was evaluated by a one-way analysis of variance through Tukey's post hoc test and unpaired two-tailed Student's t test. A p value less than 0.05 was considered to be statistically significant.

| CONCLUSION
In conclusion, we constructed a novel tissue- writingreview and editing (lead).