Laminin-111-derived peptide conjugated fibrin hydrogel restores salivary gland function

Hyposalivation reduces the patient quality of life, as saliva is important for maintaining oral health. Current treatments for hyposalivation are limited to medications such as the muscarinic receptor agonists, pilocarpine and cevimeline. However, these therapies only provide temporary relief. Therefore, alternative therapies are essential to restore salivary gland function. An option is to use bioengineered scaffolds to promote functional salivary gland regeneration. Previous studies demonstrated that the laminin-111 protein is critical for intact salivary gland cell cluster formation and organization. However, laminin-111 protein as a whole is not suitable for clinical applications as some protein domains may contribute to unwanted side effects such as degradation, tumorigenesis and immune responses. Conversely, the use of synthetic laminin-111 peptides makes it possible to minimize the immune reactivity or pathogen transfer. In addition, it is relatively simple and inexpensive as compared to animal-derived proteins. Therefore, the goal of this study was to demonstrate whether a 20 day treatment with laminin-111-derived peptide conjugated fibrin hydrogel promotes tissue regeneration in submandibular glands of a wound healing mouse model. In this study, laminin-111-derived peptide conjugated fibrin hydrogel significantly accelerated formation of salivary gland tissue. The regenerated gland tissues displayed not only structural but also functional restoration.


Synthesis of peptide or DyLight 680 conjugated fibrinogen
Peptide and DyLight 680 conjugated fibrinogen were prepared as previously described [21,24]. Briefly, two L1 derived peptides (A99: CGGALRGDN-amide, YIGSR: CGGADPGYIGSR GAA-amide) were synthesized on a peptide synthesizer using Fmoc solid-phase peptide. In order to create a thiol-reactive fibrinogen, the primary amine groups were activated with 7.2 equivalents of sulfo-LC-SPDP for 1 h at room temperature. Then, the excess cross linker and its by-products were removed by dialysis (molecular weight cut-off (M.W.C.O) = 3.5 kDa). For peptide conjugation, LC-SPDP activated fibrinogen was reacted with 2 equivalents of peptide for 18 h at room temperature. Finally, peptide conjugated fibrinogen was dialyzed against ultrapure water (M.W.C.O. = 3.5 kDa) and filtered using a 0.22 μm syringe filter. Then, the product was lyophilized. To monitor fibrin hydrogels stability, fluorescently conjugated fibrinogen was also produced. Lyophilized fibrinogen was reacted with DyLight 680 for 1 h at room temperature. Non-reacted reagent was removed by dialysis (M.W.C.O = 3.5 kDa). Then, the product was filtered using a 0.22 μm syringe filter and lyophilized. All fibrinogens were stored at −80˚C until use. The reactions were monitored and confirmed using thin-layer chromatography, UV-Vis spectrum and static light scattering data, as described previously [21]. Based on these data, six peptides and five DyLight 680 were conjugated to a fibrinogen molecule. In addition, the final percent yields for YIGSR conjugated fibrinogen, A99 conjugated fibrinogen, and DyLight 680 conjugated fibrinogen were 83.32%, 86.73%, and 78.15%, respectively.

Rheological parameters
The rheological properties are highly dependent on the macromolecular structure, which in turn affects the mechanical of hydrogels [25]. To this end, rheological measurements were performed on a stress-controlled AR 2000ex rheometer (TA Instruments, Crawley, UK) using a stainless-steel cone and plate geometry (4˚cone angle; 20 mm cone diameter and truncation height of 114 μm) at 37˚C with a continuous strain value of 0 to 100. One hundred microliter of mixture was applied to the bottom plate, and a solvent trap filled with water was used to prevent sample evaporation. Then, the modulus of elasticity (G') and the strain (%) were recorded for 5 min.

Animals
Female C57BL/6 mice 5-7-weeks-old, weighing 17-22 g, were purchased from the Jackson Laboratory (Bar Harbor, ME). Then, 28 mice were randomly distributed into 4 groups: untreated, FH treated, L 1p -FH treated and sham surgery control. Animals were housed in cages in a room with a controlled environment (12-hour day/night cycles) and provided with a standard pellet diet and water. All animal management, anesthesia, and surgeries followed the protocol (protocol number: 17-05001) approved by the Institutional Animal Care and Use Committee (IACUC) at the University of Utah. Carprofen (5 mg/kg/day) was used to manage incision-induced pain after surgery.
Surgical procedure and scaffold stability C57BL/6 mice were anesthetized with 3% isoflurane with an oxygen flow rate set at 2.0 L/ min. Then, a skin incision of approximately 1 cm in length was made along the anterior surface of the neck ( Fig 1A). Subsequently, mSMGs were exposed and the surgical wounds were created using a 3-mm diameter biopsy punch ( Fig 1B). To determine the effects of FH, 20 μL  of this scaffold was added at the surgical wounds where a coverslip was placed underneath to prevent leakage ( Fig 1C). The different scaffolds used in the experiments are listed in Table 1. Finally, the skin incision was sutured and post-surgical studies at day 20 were performed ( Fig 1D).
To monitor scaffold stability in vivo, FH were conjugated with Dylight 680. Fluorescent intensity of the scaffold was monitored in a Xenogen IVIS 100 Bioluminescent Imager (University of Utah, Center for Quantitative Cancer Imaging) at post-surgery day 1, 3, 8 and 20. Fluorescent images were acquired with a filter set using excitation/emission at 692/712 nm.

Measurement of body weight
In order to monitor post-surgery body mass, mice were weighed at the start of each experiment and data was collected for 20 days. Statistical significance was assessed by two-way ANOVA (p < 0.01) and Dunnett's post-hoc test for multiple comparisons to the untreated group.

Salivary secretion rate
Mice were anesthetized with ketamine (100 mg/kg) and xylazine (5 mg/kg), and injected with pilocarpine (10 mg/kg) via intraperitoneal injection to stimulate saliva secretion. Then, whole saliva was collected and measured using a 200 μl pipette. Statistical significance was assessed by one-way ANOVA (p < 0.01) and Dunnett's post-hoc test for multiple comparisons to the untreated group.

Saliva protein composition
To determine the saliva composition of each condition, 15 μg of saliva protein from each group were fractionated by SDS-PAGE. Saliva samples were denatured at 95˚C for 5 min in a sample loading buffer. The denatured samples were loaded onto the Mini-PROTEAN TGX precast electrophoresis gel (Any kD™, Bio-Rad) and subjected to electrophoresis in 25 mM Tris/192 mM Glycine buffer with 0.1% SDS (w/v) at 100 V for 70 min. The electrophoresis gel was fixed in a solution of 25% ethanol, 15% formaldehyde, 60% water for 1 h and re-fixed with 50% methanol, 40% water, 10% glacial acetic acid for overnight. For general protein staining, the gel was stained with 0.25% Coomassie Brilliant Blue R-250 in 50% (v/v) methanol, 10% (v/v) glacial acetic acid for 1 h and destained overnight in 20% (v/v) methanol and 10% (v/v) acetic acid. For mucin staining, the fixed gel was stained with 0.5% Alcian Blue 8GX in 2% (v/v) acetic acid for 1 h. Then, the gel was destained overnight in 20% (v/v) methanol and 10% (v/v) acetic acid. Protein images of gels were captured using a Chemi Doc mp imaging system (Bio-Rad). ImageJ was used to perform the image analysis. All statistical analyses were performed with GraphPad Prism 6. Histological studies mSMG tissues were immersed in 10% formalin at room temperature for one day, dehydrated in serial ethanol solutions (50%, 70% and 100% for 2 h each), embedded in paraffin wax and cut into 7 μm sections. mSMG sections from each group were deparaffinized with xylene and rehydrated with serial ethanol solutions (100%, 70% and 50%) and distilled water. Then, hematoxylin and eosin (H&E) and picrosirius red staining were performed, and tissue sections were subjected to a blind histopathological analysis using a Leica DMI6000B inverted microscope (Leica Microsystems, Wetzlar, Germany) as well as an Olympus BX53 Light Microscope (Olympus America, Center Valley, PA). In addition, the ratio of acinar structures to ductal structures was analyzed using ImageJ and GraphPad Prism 6.

Confocal microscopy
Deparaffinized sections were incubated in sodium citrate buffer (10 mM sodium citrate, 0.05% Tween 20, pH 6.0) at 95˚C for 30 min for antigen retrieval. Then, sections were washed with distilled water, and permeabilized with 0.1% Triton X-100 in PBS at room temperature for 45 min. Sections were blocked in 5% goat serum in PBS for 1 h at room temperature and incubated for overnight at 4˚C with primary antibody solution as described in Table 2. The following day, tissue sections were washed three times with PBS and incubated with secondary antibody solution ( Table 2) for 1 h at room temperature. Sections were then washed three times with PBS and counter-stained with TO-PRO-3 iodide at room temperature for 15 min (1:1000 dilution). Finally, tissue samples were analyzed using a confocal Zeiss LSM 700 microscope using a 40× objective.

Proliferation assay
Confocal images of Ki67 stained tissue samples were captured at 40× magnification using a confocal Zeiss LSM 700. The number of Ki67 positive cells was counted using ImageJ software. Statistical significance was assessed by one-way ANOVA (p < 0.01) and Dunnett's post-hoc test for multiple comparisons to the sham control group.

Scaffold mechanical property and stability
Rheology measurements showed that L 1p -FH display a significant decrease in elasticity as compared to unconjugated FH thereby making a softer structure (Fig 2A). Moreover, the  fluorescent intensity of FH at day 3 ( Fig 2C) was similar with the post-surgery day 1 group ( Fig  2B). However, the fluorescent intensity of the FH at day 8 ( Fig 2D) was approximately 6 times lower as compared to post-surgery day 1 or day 3 groups (Fig 2B and 2C). Moreover, the fluorescent intensity of FH at day 20 ( Fig 2E) was undetectable. These results suggest successful attachment of FH scaffold in the wounded tissue (i.e., high stability) and likely degradation over time in vivo.

Measurement of body weight
Since saliva is important in eating, swallowing, and digestion, we sought to determine the ability of the mice to eat by measuring body weight at different times after surgery. We found no significant weight difference between untreated mice and FH alone treated mice (Fig 3). However, mice treated with L 1p -FH had similar weights as the sham control group, which were significantly higher as compared to untreated mice and FH alone treated mice (p < 0.01).

Saliva flow rate
As shown in Fig 4, animals with no scaffold (untreated) or with FH alone displayed a significant decrease in saliva secretion rates (44% vs sham). In contrast, mice treated with L 1p -FH showed a significant increase in saliva secretion rates as compared to untreated and FH alonetreated mice. Moreover, L 1p -FH-treated mice showed increased saliva flow rates (75%) to levels close to sham controls (open incision but no surgical wound).

Saliva protein composition
We analyzed the protein composition in the stimulated saliva using SDS-PAGE. The total protein ( Fig 5A) and mucin ( Fig 5B) composition of the saliva from untreated and FH alone group showed clearly different patterns compared to the saliva from sham control group. The untreated group displayed decreased proline rich protein (15 kDa~30 kDa) and cystatin (10 kDa) levels. In addition, animals with no scaffold or with FH alone displayed a significant decrease in MUC7 (p < 0.0001). However, the protein patterns of the L 1p -FH treated group showed comparable protein patterns to sham control ( Fig 5C). Moreover, the ratio of MUC5B and MUC7 in the saliva slightly differed between sham and L 1p -FH treated group (p = 0.0111).
These results indicate that the L 1p -FH treated SMG could produce a similar quality of saliva as compared to sham controls.

Histopathological studies
To determine whether L 1p -FH promoted tissue regeneration of mSMG surgical wounds in vivo, mSMG tissue sections were stained with H&E and picrosirius red. As shown in Fig 6, mSMG surgical wounds covered with L 1p -FH displayed organized round acinar (red arrows) and ductal structures (yellow arrows) (Fig 6E) with organized collagen formation (Fig 6F). In contrast, wounded mSMG treated with no scaffold and FH alone formed disorganized collagen and failed to form organized round structures (Fig 6A-6D). A single blind histopathological analysis revealed that all mSMG treated with FH in general showed no differences in

L 1p -FH applied to mSMG improved saliva secretion over untreated and FH alone-treated mice.
Mice were anesthetized and stimulated with pilocarpine at day 20. Then, saliva was collected for 5 min. Data represent the means ± SD of n = 5 mice per condition and statistical significance was assessed by one-way ANOVA (p < 0.01) and Dunnett's post-hoc test for multiple comparisons to the untreated group. * = significant difference from the untreated group; n.s. = no significant difference from the untreated group. proliferation rates between the different groups and lack of evidence for cellular atypia. Additionally, the ratio of acinar and ductal structures is comparable to sham controls ( Fig 6I).

Proliferation assay
Our previous studies demonstrated that L 1p -FH treated tissues displayed a significantly higher number of Ki67 positive cells as compared with untreated and FH treated controls on postsurgery day 8 [24]. However, L 1p -FH, FH alone -treated tissues as well as no-scaffold controls showed no significant differences as compared to sham controls on post-surgery day 20 ( Fig  8). Moreover, our statistical analyses showed no significant difference in cell proliferation between the different groups. L1p-FH restores SG function

Discussion
Recently, we published a study utilizing a mouse model for wounded mSMG to demonstrate that L1 peptide conjugated FH contributes to tissue regeneration [24]. In the previous study, regenerated tissue showed a strong expression of structural protein cell markers; however, the signals for functional cell markers were weak indicating incomplete wound healing. We believe this was due to a short monitoring time-frame. Therefore, we decided to test whether L 1p -FH could restore both tissue structure organization and function when treatments lasted for longer periods. In fact, the monitoring of L 1p -FH over time indicates that even though L 1p -FH was significantly softer than FH (Fig 2A), it can still successfully attach to wounded SMG, is stable for several days and the gel is degraded over time in vivo (Fig 2B-2E). Moreover, our results indicated that L 1p -FH are able to almost completely restore functional markers in the wounded mSMG as compared to previous studies [24]. While the mechanisms by which this process occurs are not fully understood, previous studies have shown that L1 peptides bind to α 3 β 1 , α 6 β 1 , α 6 β 4 and α 7 β 1 integrins, while fibrinogen binds to αIIbβ 3 and α 5 β 1 integrins [26][27][28]. These integrins are present in epithelial tissues and their activation induces cell migration, proliferation, and adhesion [29][30][31][32].
The main clinical concern in patients with salivary gland damage is the reduction of salivary flow [33]. Moreover, salivary gland defects are highly related to weight loss as saliva is necessary for eating, chewing, swallowing and digesting [34,35]. Therefore, in this study we L1p-FH restores SG function demonstrated that animals with untreated with untreated SMG wounds not only showed significant reduction of saliva flow (Fig 4) but also displayed significant weight loss (Fig 3), while treatment with L 1p -FH completely abolished these problems. These results suggest that L 1p -FH was effective in promoting salivary gland function. To prevent misinterpretation of the data shown above, we were able to use a sham control group that was comparable to healthy mice that showed no weight loss or signs of infection (data not shown).
Mucins are high molecular weight glycoproteins secreted by salivary glands. The best studied mucins in the human saliva are a) MUC5B (a larger salivary gland mucin) which is present in submandibular gland (SMG) secretions and is thought to be related to the perception of dry mouth (i.e., by retaining moisture in the mucosa) and b) MUC7 (a smaller salivary gland mucin) appears to have a role in preventing bacterial attachment to the enamel and mucosal surfaces [36]. Moreover, MUC7 is less abundant in the gel-phase, making saliva less viscous and therefore contributing to the natural saliva rheology [37]. MUC7 appears to be undetectable in saliva from minor glands; however, it is present in saliva from SMG and sublingual glands [38]. In this study, we found that animals with no scaffold or with FH alone displayed a significant increase of MUC5B protein levels and a significant decrease MUC7 protein levels as compared to the sham control group while treatment with L 1p -FH restored MUC5B and MUC7 expression to levels comparable to sham controls (Fig 5). This result indicates that L1 peptides are able to restore saliva component and saliva rheology. L1p-FH restores SG function Collagen acts as a structural scaffold in epithelial tissues, also controls many cellular functions (e.g. cell migration, differentiation, and synthesis of proteins) [39,40]. Therefore, collagen organization plays a critical role in the wound healing process [41,42]. Our results showed that L 1p -FH was able to promote organized cell structure and collagen formation (Fig 6). We speculate that these effects are due to L 1p -FH's ability to bind and attract stem/progenitor cells through interactions with integrins described above, and thus accelerating wound healing. However, when wound healing is incomplete or excessive it may can cause unwanted side effects. Specifically, it was reported that inflammation pathways during wound healing may promote growth and survival of cancer stem cells [43]. However, our results showed no differences in cell proliferation between the studied groups (Fig 8). Moreover, a single blinded histopathological study described an absence of cellular atypia. Together these results indicate that the regeneration process using L 1p -FH is controlled and could be used safely in other species.
Regarding the expression of functional cell markers, we found a strong aquaporin 5 signal in the regenerated mSMG (Fig 7C). Since this protein transports water across the cell membrane during fluid secretion, its expression is also essential for SMG regeneration [44,45]. In addition, TMEM16A (apical chloride transporter), Na + /K + -ATPase (basolateral antiport transporter) were expressed in the regenerated mSMG ( Fig 7G). These proteins are responsible for maintaining the proper ionic composition for saliva secretion and are also critical for SMG functioning [46,47]. Finally, we detected PECAM-1 and β-Tubulin III signal in the regenerating gland (Fig 7K), indicating formation of blood and nerves, respectively, which have been shown to regulate SMG development [48]. Our results indicate that L 1p -FH are able to promote SMG regeneration in vivo and are consistent with previous studies [21,24,49]; however, here we found better regeneration endpoints due to longer treatments times. These studies are highly significant as they could offer viable path forward for advancing the treatment of hyposalivation.

Conclusion
In summary, we demonstrated that FH modified with L1 peptides facilitated salivary gland tissue healing. Our results suggest that L 1p -FH is suitable for in vivo applications (as it is both biodegradable and biocompatible and significantly accelerates formation of new tissue as compared to FH alone or no scaffold). The regenerated gland tissues displayed not only structural but also functional similarities to normal gland tissues. As stated, the increased duration of the current experiment allowed for additional developments to occur and for more definitive data to emerge. The current results suggest that L1p-FH may have activated multiple cellular processes that contributed to tissue regeneration.