Biocompatibility of Subcutaneously Implanted Plant-Derived Cellulose Biomaterials

There is intense interest in developing novel biomaterials which support the invasion and proliferation of living cells for potential applications in tissue engineering and regenerative medicine. Decellularization of existing tissues have formed the basis of one major approach to producing 3D scaffolds for such purposes. In this study, we utilize the native hypanthium tissue of apples and a simple preparation methodology to create implantable cellulose scaffolds. To examine biocompatibility, scaffolds were subcutaneously implanted in wild-type, immunocompetent mice (males and females; 6–9 weeks old). Following the implantation, the scaffolds were resected at 1, 4 and 8 weeks and processed for histological analysis (H&E, Masson’s Trichrome, anti-CD31 and anti-CD45 antibodies). Histological analysis revealed a characteristic foreign body response to the scaffold 1 week post-implantation. However, the immune response was observed to gradually disappear by 8 weeks post-implantation. By 8 weeks, there was no immune response in the surrounding dermis tissue and active fibroblast migration within the cellulose scaffold was observed. This was concomitant with the deposition of a new collagen extracellular matrix. Furthermore, active blood vessel formation within the scaffold was observed throughout the period of study indicating the pro-angiogenic properties of the native scaffolds. Finally, while the scaffolds retain much of their original shape they do undergo a slow deformation over the 8-week length of the study. Taken together, our results demonstrate that native cellulose scaffolds are biocompatible and exhibit promising potential as a surgical biomaterial.


Introduction
The development of novel biomaterials for tissue engineering strategies is currently under intense investigation [1][2][3]. Biomaterials are being developed for the local delivery of therapeutic cells to target tissues [4,5], the regeneration of damaged or diseased tissues [6][7][8][9] or the replacement of whole organs [10][11][12][13][14][15]. In their most general form, biomaterials provide a three-cellulose constructs. The cellulose scaffolds were employed for in vitro 3D culture of NIH3T3 fibroblasts, mouse C2C12 muscle myoblasts and human HeLa epithelial cells. Our previous work revealed that these cells could adhere, invade and proliferate within the cellulose scaffolds and retain high viability even after 12 continuous weeks of culture.
Our previous work opens the question of in vivo biocompatibility [27]. Therefore, the objective of this study is to characterize the response of the body to apple-derived cellulose scaffolds. Macroscopic (~25mm 3 ) cell-free cellulose biomaterials were produced and subcutaneously implanted in mouse model for 1, 4 and 8 weeks. Here, we assess the immunological response of immunocompetent mice, deposition of extracellular matrix on the scaffolds and evidence of angiogenesis (vascularization) in the implanted cellulose biomaterials. Notably, although a foreign body response was observed immediately post-implantation, as expected for a surgical procedure, by the completion of the study only a low immunological response was observed with no fatalities or noticeable infections whatsoever in all animal groups. Surrounding mouse cells were also found to invade the scaffold, mainly activated fibroblasts, and deposit a new extracellular matrix. As well, the scaffold itself was able to retain much of its original shape and structure over the 8-week study. Importantly, the scaffolds clearly had a pro-angiogenic effect, resulting in the growth of functional blood vessels throughout the implanted biomaterial. Taken together, our work demonstrates that we can easily produce 3D cellulose scaffolds that are biocompatible, becoming vascularized and integrated into surrounding healthy tissues.

Material and Methods Animals
All experimental procedures were approved by the Animal Care and Use Committee of the University of Ottawa. Wild-type C57BL/10ScSnJ mice (males and females; 6-9 weeks old; n = 7 mice for each group) were purchased from The Jackson Laboratory (Bar Harbor, Maine, USA) and bred in our facilities. All animals were kept at constant room temperature (±22°C) and humidity (*52%). They were fed a normal chow diet and kept under a controlled 12 hours light/dark cycle.

Cellulose scaffold preparation
As described previously [27], McIntosh Red apples (Canada Fancy) were stored at 4°C in the dark for a maximum of two weeks. In order to prepare apple sections, the fruit was cut with a mandolin slicer to a uniform thickness of 1.14±0.08mm, measured with a Vernier caliper. Only the outer (hypanthium) tissue of the apple was used. Slices containing visible ovary-core tissue were not used. The slices were then cut parallel to the direction of the apple pedicel into squares segments of 5.14±0.21mm in length and with an area of 26.14±1.76mm 2 . Apple tissue was decellularized using a well-established protocol [14] for removal of cellular material and DNA from tissue samples while leaving behind an intact and three-dimensional scaffold. Individual apple tissue samples were placed in sterilized 2.5ml microcentrifuge tubes and 2ml of 0.1% sodium dodecyl sulphate (SDS; Sigma-Aldrich) solution was added to each tube. Samples were shaken for 48 hours at 180 RPM at room temperature. The resultant cellulose scaffolds were then transferred into new sterile microcentrifuge tubes, washed and incubated for 12 hours in PBS (Sigma-Aldrich). The cellulose scaffold were sterilized by incubation in 70% ethanol for 1 hour and then repeatedly washed 12 times with PBS. The samples were then kept in PBS. At this point, the samples were immediately used or stored at 4°C for no more than 2 weeks.

Cellulose implantation
The mice were anesthetized using 2% Isoflurane USP-PPC (Pharmaceutical partners of Canada, Richmond, ON, Canada) with the eyes protected the application of ophthalmic liquid gel (Alco Canada In., ON, Canada). The mouse back hairs were shaved with the underlying skin cleaned and sterilized using ENDURE 400 Scrub-Stat4 Surgical Scrub (chlorhexidine gluconate, 4% solution; Ecolab Inc., Minnesota, USA) and Soluprep (2% w/v chlorhexidine and 70% v/v isopropyl alcohol; 3M Canada, London, ON, Canada). Animal hydration was maintained, via subcutaneous injection (s.c) of 1 ml of 0.9% sodium chloride solution (Hospira, Montréal, QC, Canada). Throughout the surgical procedures all strict sterility measures were upheld for survival surgeries. To implant the scaffolds, two 8mm incisions were cut on the dorsal section of each mouse (upper and lower). Two cellulose scaffold samples were separately and independently implanted into each mouse. The incisions were then sutured using Surgipro II monofilament polypropylene 6-0 (Covidien, Massachusetts, USA) and transdermal bupivicaine 2% (as monohydrate; Chiron Compounding Pharmacy Inc., Guelph, ON, Canada) was topically applied to the surgery sites to prevent infection. Additionally, buprenorphine (as HCL) (0.03mg/ml; Chiron Compounding Pharmacy Inc. Guelph, ON, Canada) was administrated s. c. as a pain reliever. All animals were then carefully monitored for the following 3 days by animal care services and received additional treatment of the same pharmacological treatments.

Scaffold resections
At 1, 4 and 8 weeks after scaffold implantation, the mice were euthanized using CO 2 inhalation. After blood collection, the dorsal skin was carefully resected and immediately immersed in PBS solution. The skin sections containing cellulose scaffolds were then photographed, cut and fixed in 10% formalin for at least 48 hours. The samples were then kept in 70% ethanol before being embedded in paraffin by the PALM Histology Core Facility of the University of Ottawa.

Histological analysis
Serial 5μm thick sections were cut, beginning at 1 mm inside the cellulose scaffold, and stained with hematoxylin-eosin (H&E) and Masson's trichrome. For immunocytochemistry, heat induced epitope retrieval was performed at 110°C for 12 min with citrate buffer (pH 6.0). Anti-CD31/PECAM1 (1:100; Novus Biologicals, NB100-2284, Oakville, ON, Canada), anti-alpha smooth muscle actin (1:1000, ab5694, abcam, Toronto, ON, Canada) and anti-CD45 (1:3000; ab10558, abcam, Toronto, ON, Canada) primary antibodies were incubated for a hour at room temperature. Blocking reagent (Background Sniper, Biocare, Medical, Concorde, CA, USA) and detection system MACH 4 (Biocare Medical, Concord, CA, USA) were applied according to company specifications. For the evaluation of cell infiltration, extracellular matrix deposition and vascularisation (angiogenesis), micrographs were captured using Zeiss MIRAX MIDI Slide Scanner (Zeiss, Toronto, Canada) equipped with 40x objective and analysed using Pannoramic Viewer (3DHISTECH Ltd., Budapest, Hungary) and ImageJ software. The scoring of inflammation was evaluated by a pathologist. The scoring was subjectively assigned by qualitative analysis of the magnitude of the total foreign response as well, the cell population proportions within the foreign response.

Quantification of cellulose volume fraction
To quantitatively determine the average volume of the scaffolds occupied by cellulose H&E images of bare scaffolds were processed according to the following protocol. At least 5 regions of the interest (ROI) of approximate 900μm 2 were identified in a given H&E image of bare cellulose. In total we analyzed n = 5 H&E images, for a total of 39 ROIs. The area occupied by the cellulose was determined by creating a binary segmented image through thresholding. As each H&E stained section was 5μm thick, this allowed us to determine the total volume occupied by cellulose compared to the gross volume of the entire ROI. All image processing was performed with ImageJ 1.47v.

Scanning electron microscopy (SEM)
The structure of cellulose was studied using a scanning electron microscopy. Globally, scaffolds were dehydrated through successive gradients of ethanol (50%, 70%, 95% and 100%). Samples were then gold-coated at a current of 15mA for 3 minutes with a Hitachi E-1010 ion sputter device. SEM imaging was conducted at voltages ranging from 2.00-10.0 kV on a JSM-7500F Field Emission SEM (JEOL, Peabody, MA, USA).

Statistical analysis
All values reported here are the average ± standard deviations. Statistical analyses were performed with one-way ANOVA by using SigmaStat 3.5 software (Dundas Software Ltd, Germany). A value of p < 0.05 was considered statistically significant.

Scaffold Preparation
Cellulose scaffolds were prepared from apple tissue using a modified decellularization technique we have previously described [27]. All scaffolds were cut to a size of 5.14±0.21 x 5.14 ±0.21 x 1.14±0.08mm (Fig 1A), decellularized and prepared for implantation ( Fig 1B). The scaffolds appear translucent after decellularization due to the loss of all plant cellular material and debris. The removal of apple cells was also confirmed with histological observation ( Fig  1C) and scanning electron microscopy ( Fig 1D). Analysis and quantification of the histological images reveals an average cellulose cell wall thickness of 4.04±1.4μm, and that the cellulose only occupies 16.9±3.0% of the total volume of the entire scaffold. The acellular cellulose scaffolds appear to maintain their shape very well and their structure is likely capable of being invaded by nearby cells after implantation in an animal model.

Implantation of Cellulose Scaffolds
Two independent skin incisions (8mm) were produced on the back of each mouse to create small pouches for the biomaterial implantation (Fig 2A). One cellulose scaffold ( Fig 2B) was implanted in each subcutaneous pouch. Throughout the study, there were no cases of mice exhibiting any pain behaviour that may have been induced by the cellulose scaffold implantation and none of the mice displayed visible inflammation or infection. The cellulose scaffolds were resected at 1 week, 4 weeks and 8 weeks after their implantation and were photographed to measure the change in scaffold dimensions (Fig 2D-2F). At all-time points, healthy tissue can be observed surrounding the cellulose scaffold with the presence of blood vessels, that are proximal or in direct contact, and the scaffolds retain their square shape. The pre-implantation scaffold had an area of 26.3±1.98mm 2 and it was observed to slowly decrease as function of the implantation time, based on the scaffold area that is visible to the naked eye on the skin ( Fig  2G). At 8 weeks post-implantation, the scaffold dimensions reach a near plateau measurement of 13.82±3.88mm 2 demonstrating an approximate 12mm 2 (48%) change over the course of this study.

Biocompatibility and cell infiltration in plant derived cellulose scaffolds
Scaffold biocompatibility and cell infiltration was examined with H&E staining of fixed cellulose scaffolds at 1, 4 and 8 weeks following their implantation (Fig 3). The global views of longitudinal section of representative cellulose scaffolds are shown in Fig 3A-3C. The scaffolds are implanted under the muscular layer of the dermis. Interstitial fluids, stained in pink, can be seen throughout the implanted scaffold, in contrast to a non-implanted scaffold (Fig 1C). Within the global view it was observed that the scaffold maintains its general shape throughout the study. In Fig 3D-3F, a magnified section of the perimeter of the scaffold is shown at each post-implantation time points. At 1 week, the dermis tissue surrounding implant displays  symptoms of an acute moderate to severe immune response (qualitative study performed by a pathologist) (Fig 3D). As well a dense layer of cells can be seen infiltrating into the cellulose scaffolds. The population of cells within the scaffold at 1 week consist mainly of granulocytes, specifically; polymorphonuclear (PMN) and eosinophils (Fig 3D). There is also a population of dead cells and apparent cell debris. Importantly, all of these observations are completely consistent with an expected acute foreign body reaction that follows implantation [84][85][86]. At the 4 week point we observed a stark difference in both the surrounding epidermis tissue and in the cell population migrating into the cellulose scaffold (Fig 3E). The epidermal tissue surrounding the cellulose scaffold has a decreased immune response, now scored as mild to low. The population of cells within the epidermis surrounding scaffolds now contain higher levels of macrophages and lymphocytes (Fig 3E). This is an anticipated characteristic of the foreign body are presented (G). The pre-implantation scaffold had an area of 26.30±1.98mm 2 . Following the implantation, the area of the scaffold declined to 20.74±1.80mm 2 after 1 week, 16.41±2.44mm 2 after 4 weeks and 13.82 ±3.88mm 2 after 8 weeks. The surface area of the cellulose scaffold has a significant decrease of about 12mm 2 (48%) after 8 weeks implantation (* = P<0.001; n = 12-14).
doi:10.1371/journal.pone.0157894.g002 reaction to an implanted biomaterial, demonstrating the scaffold cleaning process [84][85][86]. There is also an increase in the population of multinucleated cells within the interior of the scaffold as part of an inflammatory response (Fig 3E). Finally, 8 weeks post-implantation, the immune response apparent at 1 and 4 weeks has completely disappeared (Fig 3F), with the epidermis tissue now appearing normal. In fact, the epidermis tissue in contact with the cellulose scaffold contains the same structures as normal epidermis tissue. In the cellulose scaffold perimeter there is now a lower density of cells due to the decreased inflammation and notably, there are no fragmented dead cells present. Instead, the population of cells now contain an elevated level of macrophages, multinucleated cells and active fibroblasts, identified through morphological analysis (H&E staining). The active fibroblasts (appearing spindle shaped), can be observed migrating from the surrounding epidermis into the cellulose scaffold. In fact, fibroblasts were found throughout the cellulose scaffold. These results demonstrate that by 8 weeks post-implantation the cellulose scaffold has been accepted by the host. In parallel with the H&E inflammation analysis, we performed anti-CD45 staining to evaluate the level of inflammation throughout the scaffold and surrounding dermis tissue (Fig 3G-3I). It is clear that the inflammation throughout the dermis and within the scaffold is elevated after 1 week. However, the amount of leukocytes significantly decreases in the surrounding dermis and scaffold over the implantation time reaching a near basal level at 8 weeks. At each time point the majority of infiltrated cells can be observed along the periphery of the cellulose scaffold. However, individual cells that have migrated from the periphery can also be observed within the center of the scaffold.

Extracellular Matrix Deposition in the Cellulose Scaffolds
The presence of active fibroblasts led us to question if the cellulose scaffold was acting as a substrate for the deposition of new extracellular matrix. This was determined using Masson's Trichrome staining of fixed cellulose scaffolds slides at each time point following implantation (Fig 4). At 1-week post-implantation, the histological study shows the absence of collagen structures inside the collagen scaffold (Fig 4A, 4D and 4G). After 4-weeks, small amounts of collagen begin to be deposited inside the scaffold (Fig 4B, 4E and 4H) and by 8-weeks, large amounts of collagen are clearly visible within many scaffold cavities (Fig 4C, 4F and 4I). The presence of active fibroblasts identified through morphology (H&E staining, spindle shaped) and anti-alpha smooth muscle actin staining (data not shown) are completely consistent with the large degree of collagen deposits observed at 8-weeks. The complexity of the deposited collagen network is highlighted in Fig 4I, where individual collagen fibers within the collagen matrix are visible. This is in contrast to the characteristic high density, thick, cable-like organization of collagen found in scar tissue.

Vascularization of the Cellulose Scaffolds
Capillaries ranging from 8 to 25μm were also identified within the scaffolds as early as 1 week post-implantation. At 4 week and 8 week post implantation, blood vessels and capillaries can be observed extensively within the scaffold and the surrounding dermal tissue. We observed blood vessels presence on the cellulose scaffold and in surrounding dermis in the macroscopic photos taken during the resection (Fig 5A). Multiple cross sections of blood vessels, with the presence of red blood cells (RBCs), are identified within 4 weeks of scaffold implantations ( Fig  5B; H&E stain). The same results are obtained 8 weeks after implantation where capillaries with RBC and endothelial cells are clearly seen (Fig 5C; Masson's Trichrome). This data prompted us to verify the presence of endothelial cells in capillary structures found within the scaffold with anti-CD31 staining (Fig 5D).

Discussion
In this study, our objective was to examine the in vivo biocompatibility of acellular cellulose scaffolds derived from apple hypanthium tissue. To this end, acellular cellulose scaffolds were subcutaneously implanted within immunocompetent mice to establish their biocompatibility. Our data reveals that the implanted scaffolds demonstrate a low inflammatory response, promote cell invasion and extracellular matrix deposition, and act as a pro-angiogenic environment. Remarkably, none of the mice in this study died or demonstrated any symptoms of implant rejection such as edema, exudates or discomfort during the course of this research indicative of a successful implantation of the cellulose scaffolds. The implanted scaffolds are composed of a porous network of cavities in which the original host plant cells resided [69]. This architecture efficiently facilitates transfer of nutrients throughout the plant tissue. As we have shown here and in our previous study, the apple tissues are easily decellularized [27]. This simple treatment changes the appearance of the hypanthium tissue so that it becomes transparent, as a result of the removal of cellular materials.
Several important conclusions emerge from the current study. First, after implantation, the scaffolds are rapidly infiltrated with host cells, which begin with inflammatory cells. Consistent with previous findings, the immune response of the host animals followed a well-known timeline [84][85][86][87][88], ultimately demonstrating biocompatibility. As expected, the cell population within the scaffold after 1 week post-implantation are mainly granulocytes, specifically; polymorphonuclear (PMN) and eosinophils, constituting a clear inflammatory response. The production of a provisional matrix around the scaffold was also observed resulting in an inflamed appearance in the tissue surrounding the scaffold [84][85][86][87][88]. This is not unexpected and is the result of the foreign material as well as a response to the surgical procedure [84][85][86][87][88]. Four weeks post implantation, the population of cells within the scaffold have evolved and are now lymphocytes, monocytes, macrophages, foreign body multinucleated cells as well as scattered eosinophils. Typical with chronic inflammation, the cellular debris present in the provisional matrix at 1 week, is now being cleared by the host immune system [84][85][86][87][88]. At 8 weeks, the cellulose scaffold is now void of all provisional matrix and cellular debris and low levels of macrophages and foreign body multinucleated cells are still visible within the scaffold. Consistent with the immune response within the cellulose scaffold, the surrounding tissue is observed to return to its original physiology. In fact, at 8 week implantation the surrounding tissue is nearly similar to control tissue. Although the immune response and inflammation at 8 weeks is low, low levels of macrophages can be observed within the scaffold. Although traditionally associated with inflammation, macrophages have beneficial roles consistent with our findings. Specifically, macrophages are also known to secrete growth and pro-angiogenic factors, ECM proteins and pro-fibrogenic factors that actively regulate the fibro-proliferation and angiogenesis in tissue repair and regeneration [86]. Regardless, the vast population of cells within the scaffold after 8 weeks are now reactive fibroblasts. These cells are altering the microenvironment of the scaffold through the secretion of a new collagen extracellular matrix. Importantly the new matrix displays a remarkably low density compared, suggestive of regeneration as opposed to the characteristic high density, cable-like organization of collagen found in scar tissues [89]. In the Fig 3D-3F, the majority of cells can be observed infiltrating primarily along the periphery of the cellulose scaffold. As the cellulose scaffold is composed of interconnecting cell wall cavities it is expected that it would be more difficult to migrate into the central portion of the scaffold. Limited cell infiltration is an issue with all novel biomaterials and as such researchers are constantly modifying the physical and chemical properties of established biomaterials to enhance the cell infiltration [90][91][92]. Now that we have confirmed the biocompatibility of the scaffold, in future studies we will being to systematically modify the structure and mechanics of the scaffolds to enable optimal cell infiltration.
Our data also demonstrates that the scaffolds are pro-angiogenic, which is critical to ensuring blood transport from the surrounding tissue [93]. As with native tissue, limited blood supply to the scaffold will result in ischemia and potentially necrosis. Interestingly, it was demonstrated that bioceramics with small pore diameters limits blood vessel diameter in vivo. The porous structure of the cellulose scaffolds is due to overlapping cavities with an average internal cross-sectional area of 0.013±0.007mm 2 (as determined from H&E images). There was a large distribution in the size of the cavities with diameters ranging from 100-300μm with minimal interconnection distance of 4.04±1.4μm. As such, the low cell wall fiber volume-fraction of 16.9±3.07% of the cellulose scaffolds are consistent with the promotion of blood vessel formation [94]. At 4 and 8 week implantation time we observed blood vessels formation within the cellulose scaffold. Interestingly, blood vessel formation are apparent in various biomaterials at similar implantation times [95,96]. Taken together, the cellulose scaffold now appears to completely void of the provisional matrix and fully accepted as a subcutaneous implant. We also observed a decrease in the scaffold area over time, but it does not appear that the cellulose scaffold is in the processes of degradation. Rather, the change in area is due to the collapse of the cell wall cavities on the perimeter of the scaffold resulting from the active movement of the mouse. Active biological degradation is not expected to be possible as mammals lack the appropriate enzymes to digest plant-synthesized cellulose [97,98]. Moreover, the highly crystalline form of cellulose that is found in plant tissues is also known to be resistant to degradation in mammals [98]. Alternatively, it has been demonstrated that in vivo cellulose based implants can be chemically activated in order to be more easily degraded [99]. Most importantly however, highly crystalline forms of cellulose have some of the lowest reported immunological responses [98].
A large variety of clinically approved biomaterials are used to treat specific conditions within patients [1]. Such biomaterials can be derived from human and animal tissues, synthetic polymers, as well as materials such as titanium and ceramics [1,2,26,49,50,53,54,56,74,76,94,96,[100][101][102][103][104][105][106][107][108][109][110]. However, these approaches are not without disadvantages that arise from concerns about the source, production costs and/or widespread availability [48]. There is currently an intense interest in developing resorbable biomaterials that will degrade in vivo and only act as a temporary scaffold that will promote and support the repair or regeneration of damaged/diseased tissue [49]. Although this is an ideal scenario, newly formed structures are also found to collapse as the scaffold degrade [53,64,[111][112][113]. Moreover, the products of degradation can also be found to have toxic or undesirable sideeffects [53,114,115]. For example, the reconstruction of the ear has become a well-known challenge in tissue engineering. Early studies have employed scaffolds in the shape of an ear that are produced from animal or human derived cartilage [53,58,59,61,63,64]. However, after implantation and eventual scaffold degradation, the ear is often found to collapse or deform [60][61][62]. Recent strategies have now opted to create biological composite materials composed of both a titanium frame embedded in a biological matrix [53]. Therefore, there is still a clear need for non-resorbable, yet inert and biocompatible, scaffolds persists in the field of tissue and organ engineering.
We suggest that plant-derived cellulose biomaterials offer one potential approach for the production of implantable scaffolds. This approach is complementary to bacterial cellulose strategies which have demonstrated clear utility as well [66,[69][70][71]73,80,83,108,110,[116][117][118][119]. However, plant derived materials are cost effective to produce and are extremely straightforward to prepare for implantation, exhibit clear biocompatibility, an ability to retain their shape while supporting the production of natural extracellular matrix and most importantly, the promotion of vascularization. In our previous work we have shown that the scaffolds can also be functionalized with proteins prior to culture in vitro. Such work will also be conducted in the future in order to explore the use of scaffold surface functionalization with growth factors and matrix proteins to promote the invasion of specific cell types, further minimize the early immune response and promote maximal vascularization. Moreover, the cellulose scaffolds can easily be formed into specific shapes and sizes, offering an opportunity to create new tissues with specific geometrical properties. Although there are numerous new avenues of research to follow, we have been able to demonstrate that acellular cellulose scaffolds are biocompatible in vivo in immunocompetent mice and might be considered as a new strategy for tissue regeneration.