Lyophilized human fibrinogen, tris base, ethylenediaminetetraacetic acid (EDTA), pilocarpine, isoproterenol, goat serum, hydrochloric acid, hematoxylin, eosin Y solution, Tween® 20, calcium chloride (CaCl₂) and ε-aminocaproic acid (εACA) were purchased from MilliporeSigma (Burlington, MA). Rabbit anti-zonula occludens 1 (ZO-1) antibody, rabbit anti-induced nitric oxide synthase (iNOS) antibody, Alexa Fluor 488 conjugated anti-rabbit IgG secondary antibody, Alexa Fluor 568 conjugated anti-rabbit IgG secondary antibody and Alexa Fluor 568 conjugated anti-mouse IgG secondary antibody were purchased from Invitrogen (Carlsbad, CA). Rabbit anti-transmembrane Protein 16A (TMEM16A) antibody and mouse anti-intercellular adhesion molecule (ICAM-1) antibody were purchased from Abcam (Cambridge, MA). Rabbit anti-vascular cell adhesion molecule 1 (VCAM-1) antibody and rabbit Arginase-1 (Arg-1) antibody were purchased from Cell Signaling Technology (Danvers, MA). Mouse anti-Na⁺/K⁺-ATPase antibody was purchased from Santa Cruz Biotechnology (Dallas, TX). Mouse anti-E-cadherin antibody was purchased from BD Biosciences (San Jose, CA). Phosphate buffered saline (PBS), DyLight™ 680 NHS-ester, 4′,6-diamidino-2-phenylindole (DAPI), Triton X-100, sodium citrate, xylene and ethanol were purchased from Thermo Fisher Scientific (Waltham, MA). Ketamine and xylazine were purchased from VetOne (Boise, ID). Insulin syringes (28G) were purchased from BD (Franklin Lakes, NJ). Peptides were synthesized by University of Utah DNA/Peptide synthesis core facility, as previously described (Nam et al., 2016; Nam et al., 2017a; Nam et al., 2017b).

Female 6-week-old C57BL/6J mice weighing ∼17–20 g were purchased from Jackson Laboratory (Bar Harbor, ME). Power analysis was performed to determine mouse numbers using G^*Power 3.1.9.7 software (Heinrich-Heine-Universität Düsseldorf, Düsseldorf, Germany; http://www.gpower.hhu.de/). All calculations were conducted using a significance level of 0.05 with 95% power. Then, 105 mice were randomly distributed into three groups to receive the following treatments: non-irradiated (40 mice), irradiated without L_1p-FH injection (40 mice), and irradiated while also receiving the L_1p-FH injection (25 mice), comprising treatment groups 1–3, respectively. All animal usage, anesthesia and surgeries were conducted with the approval of the University of Utah Institutional Animal Care and Use Committee (IACUC) in compliance with the ARRIVE guidelines.

Salivary gland tissue damage is a late degenerative response observed after radiation therapy (Wu and Leung, 2019; Jasmer et al., 2020). To confirm L_1p-FH regenerative effects in a more clinically relevant animal model, a widely accepted head and neck irradiated mouse model was used for this study (Deasy et al., 2010; Varghese et al., 2018). Briefly, mice were anesthetized with ketamine (100 mg/kg) and xylazine (5 mg/kg) solution administered intraperitoneally with the head and neck area positioned over the 1 cm slit of a customized lead shield, thereby protecting other areas of the body from radiation. SMGs then received a single 15 Gy radiation dose using a JL Shepherd ¹³⁷Cs irradiator (Figure 1A). Animals were allowed to recover for 3 days and received hydrogel treatment soon after, as detailed below.

Peptides and DyLight 680 conjugated fibrinogen were prepared, as previously described (Nam et al., 2017a; Nam et al., 2017b). Briefly, two Laminin-1 peptides (A99 and YIGSR) were synthesized on a peptide synthesizer. Peptides were then conjugated to the fibrinogen using sulfo-LC-SPDP and cysteine residue in peptides. In addition, fibrinogen was chemically labeled with a fluorescent dye through NHS ester of DyLight 680. Finally, laminin-1 peptide conjugated fibrinogens and DyLight 680 labeled fibrinogen were dialyzed against ultrapure water, lyophilized, and stored at −80°C until use. L_1p-FH were prepared similar to previous studies (Nam et al., 2017a; Nam et al., 2017b) except for the use of exogenous thrombin (thereby preventing rapid polymerization inside the syringe) as follows: YIGSR-conjugated fibrinogen (1.2 mg/ml), A99-conjugated fibrinogen (1.2 mg/ml), DyLight 680 conjugated fibrinogen (0.1 mg/ml), CaCl₂ (2.5 mM) and εACA (2 mg/ml) were mixed in a tris buffered saline (TBS) solution. Polymerization of L_1p-FH was confirmed from fluorescence in the SMG of randomly selected mice (Figure 1C).

C57BL/6J mice were anesthetized with 3% isoflurane using an oxygen flow rate set at 2.0 L/min, and 10 μL of freshly mixed L_1p-FH solution was transdermally injected using insulin syringe (G 28) to irradiated mouse SMGs at post-radiation day 3. L_1p-FH effects were studied at days 8 and 30. Using thrombin prior transdermal injection causes rapid polymerization of L_1p-FH which clogs the needle. To overcome this issue, the mixture was applied in a liquid form using endogenous thrombin for internal polymerization. To confirm scaffold implantation in vivo, FH was labeled with DyLight 680 and quantified within dissected glands using a Bio-Rad Chemi-Doc™ MP imaging system (Figure 1C).

SMGs were fixed in 10% formalin at room temperature overnight, dehydrated in 70% ethanol solution, embedded in paraffin wax and cut into 3 μm sections. Sections were then deparaffinized with xylene and rehydrated with serial ethanol solutions (100%, 95% 80, 70 and 50%, v/v) and distilled water. For hematoxylin and eosin staining, the rehydrated sections were stained with hematoxylin for 5 min, washed with distilled water for 5 min, tap water for 5 min and distilled water for 2 min. Next, slides were stained with eosin for 30 s, washed with tap water for 5 min and distilled water for 2 min. Finally, hematoxylin and eosin stained gland sections were dehydrated with 95 and 100% ethanol (v/v), cleared in xylene and mounted with a xylene-based mounting medium. As for Masson’s trichrome staining, the rehydrated sections were re-fixed in Bouin’s solution at 60°C for 1 h then washed with running tap water for 10 min and distilled water for 5 min. Next, sections were stained with Weigert’s iron hematoxylin solution for 10 min then washed with running warm tap water for 10 min and distilled water for 5 min. For cytoplasm staining, sections were incubated with Biebrich scarlet acid fuchsine solution for 5 min and washed three times with distilled water for 2 min. Regarding collagen staining, sections were incubated in phosphotungstic/phosphomolybdic acid for 15 min, stained with aniline blue solution for 5 min and washed three times with distilled water for 2 min. Stained sections were then differentiated in 1% acetic acid solution for 1 min and washed two times with distilled water for 2 min. Finally, Masson’s trichrome stained sections were dehydrated with serial ethanol solutions (95 and 100%), cleared in xylene and mounted with a xylene-based mounting medium. Finally, the samples were analyzed using a Leica DMI6000B (Leica Microsystems, Wetzlar, Germany) to determine tissue morphology.

For antigen retrieval, the rehydrated and fixed tissue sections were incubated in Tris-EDTA buffer [10 mM Tris, 1 mM EDTA, 0.05% (v/v) Tween® 20, pH 9.0] for ZO-1 and E-cadherin or with sodium citrate buffer [10 mM sodium citrate, 0.05% (v/v) Tween® 20, pH 6.0] for TMEM16A, Na⁺/K⁺-ATPase, iNOS, Arg-1, VCAM-1 and ICAM-1 at 95°C for 30 min. Next, samples were permeabilized with 0.1% (v/v) triton X-100 in PBS at room temperature for 45 min. Specimens were then blocked in 5% (v/v) goat serum in PBS for 1 h at room temperature and incubated at 4°C with the following primary antibodies overnight: rabbit anti-ZO-1, mouse anti-E-cadherin, rabbit anti-TMEM16A, mouse anti-Na⁺/K⁺-ATPase, rabbit anti-VCAM-1 or mouse anti-ICAM-1. At that time, sections were incubated with anti-rabbit Alexa Fluor 488 and anti-mouse Alexa Fluor 568 secondary antibodies in 5% goat serum at room temperature for 1 h followed by 300 nM DAPI staining at room temperature for 5 min. For M1 and M2 marker staining, specimens were blocked in 3% (w/v) bovine serum albumin (BSA) in PBS for 1 h at room temperature and incubated with primary antibodies (rabbit anti-iNOS or rabbit anti-Arg-1) at 37°C for 1 h. Then, sections were incubated with anti-rabbit Alexa Fluor 568 in 3% BSA at room temperature for 1 h followed by 300 nM DAPI staining at room temperature for 5 min. Finally, specimens were analyzed using a STELLARIS Confocal Microscope (Leica Microsystems, Wetzlar, Germany).

M1 and M2 macrophage cells were determined using ImageJ. Specifically, the color threshold was set to isolate the colocalized signal of nuclei and M1 (Figure 4, white arrows)/M2 (Figure 4, red arrows) positive cells, which were counted and normalized by area. Statistical significance was assessed using one-way ANOVA (*p < 0.01) and Dunnett’s post-hoc test for multiple comparisons to group 2 (irradiated with no L_1p-FH injection at day 30).

Mice were anesthetized with ketamine (100 mg/kg) and xylazine (5 mg/kg) followed by intraperitoneal injection with pilocarpine (25 mg/kg) and isoproterenol (0.5 mg/kg). Then, whole saliva was collected using a micropipette for 5 min and flow rate was calculated using the following formula: Saliva flow rate =   Stimulated saliva  ( µL ) Body weight of mouse  ( g )  x collection time  ( 5 min )

Experimental data were analyzed using one-way ANOVA and Dunnett’s post hoc test for multiple comparisons to the non-irradiated group 1 at day 30. All values represent means ± SD (n = 5), where p values <0.01 were considered statistically significant. Finally, these calculations were performed using GraphPad Prism 6.

To investigate whether L_1p-FH could restore irradiated SMG structure and function, C57BL/6J mice were subjected to a single radiation treatment as described in Materials and Methods (Figure 1A). Mice treated with a single 15 Gy radiation dose displayed a significant reduction in saliva flow rates as compared to non-irradiated controls (i.e., from 1.43 to 0.80 μL/g/min, n = 5, p < 0.01) in the first 8 days and remained steady thereafter until day 30 (Figure 1B). These results demonstrated that the radiation dose utilized here caused significant loss of salivary secretory function and can thus be used as a head and neck irradiated preclinical model, consistent with previous studies (Lombaert et al., 2008; Varghese et al., 2018; Weng et al., 2018).

Our previous studies showed the biocompatibility of L_1p-FH with host tissue when surgically implanted in a wounded mouse model (Nam et al., 2017a; Nam et al., 2017b). To avoid an open wound surgery, we attempted to deliver the L_1p-FH to irradiated mouse SMG via transdermal injection as described in Materials and Methods. For these experiments, we used a fluorescently labeled hydrogel using DyLight 680 and successfully implanted L_1p-FH in irradiated mouse SMG via transdermal injection (Figure 1C, white arrows).

Our previous studies showed that L_1p-FH promoted tissue repair in a wounded SMG mouse model (Nam et al., 2017a; Nam et al., 2017b; Nam et al., 2019b). To determine whether these effects occur in the head and neck irradiated mouse model, we randomly distributed mice in three groups and applied this scaffold as follows: non-irradiated, irradiated without L_1p-FH injection and irradiated that received the L_1p-FH injection, comprising treatment groups 1–3, respectively (see Material and Methods section). As shown in Figure 2, group 1 (non-irradiated glands) displayed intact lobules where the parenchyma was separated by areas of thin connective tissue at days 8 (Figures 2A,B) and 30 (Figures 2C,D). As for cytologic features, serous acini cells showed a typical pyramidal shape with basophilic cytoplasm and basal nuclei. In contrast, mucous cells showed a pale cytoplasm with flat basilar nuclei, intercalated ducts were lined by cuboidal and/or flat cells, striated ducts showed cuboidal to low columnar cells and granular convoluted ducts were lined by tall columnar cells containing intracytoplasmic eosinophilic granules. Together, these features indicate that the non-irradiated glands in group 1 showed the morphology of a healthy epithelium. In contrast, group 2 (irradiated with no L_1p-FH injection) demonstrated glandular parenchyma separated by thicker connective tissue strands, ductal areas with ectasia, intraluminal depositions and increased presence of fibrosis when compared to controls (Figures 2E,F). Furthermore, tissue damage was even more severe at day 30 (Figures 2G,H), where SMG showed an extensive disruption of the lobular architecture as indicated by the replacement of acini and ducts with sheets of vacuolated cells, adipocytes and fibrosis. Together, these results indicated that irradiated glands with no L_1p-FH injection (group 2) dramatically lost epithelial integrity. Remarkably, mice in group 3 (irradiated with L_1p-FH injection) recovered many of the features of healthy glands. For instance, we observed the presence of serous acinar units with organized ductal structures surrounded by thin connective tissue strands similar to the non-irradiated group 1 at both days 8 (Figures 2I,J) and 30 (Figures 2K,L). These changes indicate that group 3 (irradiated glands treated with L_1p-FH) had a morphology consistent with a healthy salivary gland epithelium and results in this section indicate that L_1p-FH is a suitable scaffold for promoting epithelial integrity in irradiated SMG.

To determine whether L_1p-FH maintained epithelial polarity in an irradiated mouse model, we stained the SMG sections with the apical tight junction marker ZO-1 and basolateral marker E-cadherin. As shown in Figure 3A, group 1 (non-irradiated glands) displayed apical ZO-1 (green) and basolateral E-cadherin (red) after 30 days. However, in group 2 (irradiated glands with no L_1p-FH injection), a mild residual ZO-1 signal was detected at day 8 (Figures 3B,F, blue solid line), and a weaker ZO-1 signal was expressed at day 30 (Figure 3F, blue dotted line), together with ZO-1 disorganization (Figure 3C), thereby indicating loss of epithelial polarity. In contrast, group 3 (irradiated glands treated with L_1p-FH) showed apical ZO-1 and basolateral E-cadherin signals both at days 8 (Figure 3D) and 30 (Figure 3E), indicating that the scaffold treatment helps to maintain epithelial polarity (Figure 3F, red line and red dotted line). Regarding the presence of functional markers, group 1 (non-irradiated SMG) showed apical TMEM16A (Figure 3G, green) and basolateral Na⁺/K⁺-ATPase localization (Figure 3G, red) at day 30, consistent with a healthy salivary epithelium. In contrast, group 2 (irradiated glands with no L_1p-FH injection) showed a moderate TMEM16A signal (Figure 3L, blue solid line) at day 8 (Figure 3H, green) and weaker TMEM16A signal (Figure 3L, blue dotted line) at day 30 (Figure 3I, green). Interestingly, group 3 (irradiated glands treated with L_1p-FH) expressed strong apical TMEM16 (Figures 3J,K, green; Figure 3L, red line and red dotted line) and basolateral Na⁺/K⁺-ATPase similar to non-irradiated glands, thus suggesting that L_1P-FH treatment helps to maintain epithelial polarity and preserve ion transport expression, both of which are critical for saliva secretion.

Our previous studies indicated that treatment with L_1p-FH promoted macrophage polarization in a wounded SMG female mouse model (Brown et al., 2020). To determine whether similar effects occur in an irradiated mouse model, we identified the presence of M1 and M2 subtypes within the SMG using macrophage-specific antibodies (i.e., iNOS and Arg-1, corresponding to M1 and M2, respectively). As shown in Figures 4A,F, group 1 (non-irradiated glands) expressed iNOS-positive cells with approximately 0.94 macrophages per 100,000 µm (Sroussi et al., 2017). In contrast, group 2 (irradiated glands with no L_1p-FH injection) showed a significant increase in M1 macrophages (approximately 28.65 iNOS-positive cells) at day 30 (Figures 4C,F). Notably, group 3 (irradiated glands treated with L_1p-FH) showed a significant decrease of M1 macrophages (approximately 5.92 iNOS-positive cells) at day 30 (Figures 4E,F) compared to group 2. Regarding the presence of M2 markers, group 2 (irradiated glands with no L_1p-FH injection) expressed Arg-1-positive cells with approximately 5.92 macrophages at day 30 (Figures 2I,L), which is not a significant difference from group 1 (Figures 4G,L, 2.60 macrophages). Interestingly, group 3 (irradiated glands treated with L_1p-FH) expressed a significant increase of Arg-1-positive cells at day 30 (approximately 11.37 macrophages, Figure 4K,L). Together, these results indicate that L_1p-FH causes a decrease in M1 macrophages together with an increase in M2 macrophages in SMG following radiation treatment.

Our previous studies indicate that treatment with L_1p-FH enhances saliva secretion in a wounded SMG mouse model (Nam et al., 2017a; Nam et al., 2017b; Nam et al., 2019b). To determine whether similar effects occur in an irradiated mouse model, we treated irradiated SMG with a transdermal injection of L_1p-FH as described in Materials and Methods. As shown in Figure 5, group 1 (non-irradiated glands) showed intact saliva flow rates (i.e., 1.43 μL/g/min), as expected. In contrast, group 2 (irradiated untreated glands) exhibited a significant reduction in saliva flow rates (i.e., 0.80 μL/g/min, n = 5, p < 0.01). Notably, group 3 (irradiated glands treated with L_1p-FH) showed a significant increase of saliva flow rates (1.32 μL/g/min, n = 5, p < 0.01) at day 30, thereby demonstrating that L_1p-FH restores saliva secretion after radiation treatment.