Fibrin hydrogels fortified with FGF-7/10 and laminin-1 peptides promote regeneration of irradiated salivary glands

Abstract

Ionizing radiation, commonly used for head and neck cancer treatment, typically damages the salivary glands, resulting in hyposalivation. The development of treatments to restore this lost function is crucial for improving the quality of life for patients suffering from this condition. To address this clinical need, we have developed an innovative hydrogel by chemically conjugating laminin-1 peptides (A99 and YIGSR) and growth factors, FGF-7 and FGF-10, to fibrin hydrogels. Our results demonstrate that FGF-7/10 and laminin-1 peptides fortified fibrin hydrogel [enhanced laminin-1 peptides fibrin hydrogel (Ep-FH)] promotes salivary gland regeneration and functionality by improving epithelial tissue organization, establishing a healthy network of blood vessels and nerves, while reducing fibrosis in a head and neck irradiated mouse model. These results indicate that fibrin hydrogel-based implantable scaffolds containing pro-regenerative signals promote sustained secretory function of irradiated salivary glands, offering a potential alternative treatment for hyposalivation in head and neck cancer patients undergoing radiation treatment. These unique findings emphasize the potential of fibrin hydrogel-based implantable scaffolds enriched with pro-regenerative signals in sustaining the secretory function of irradiated salivary glands and offer a promising alternative treatment for addressing hyposalivation in head and neck cancer patients undergoing radiation therapy.

1. Introduction

According to the American Cancer Society, head and neck cancer (HNC) affects a significant number of individuals, with over 82,000 new cases expected each year [1, 2]. Radiation treatment is a crucial approach modality for HNC patients, offering a chance for survival. However, one of the significant drawbacks of ionizing radiation exposure is the destruction of salivary glands leading to hyposalivation thereby adversely affecting the quality of life for these patients [3]. Current treatments for hyposalivation only provide temporary relief and include the use of medications such as pilocarpine and cevimeline to stimulate saliva secretion from residual acinar cells, as well as the use of saliva substitutes [4–10]. While these palliative measures offer some benefits, they do not address the root of cause of hyposalivation. Therefore, it is imperative to develop innovative therapeutic interventions that can restore salivary gland function in HNC patients.

Multiple experimental therapies are currently being developed including the use of stem cells, organ culture transplantation, cell sheets, bioprinting and bioengineered scaffolds. Regarding stem cells, previous studies showed that c-kit⁺ stem cells can be expanded ex vivo to restore salivary gland function; however, it is still unclear how they incorporate to the host tissues as well as their long-term effects [11, 12]. Likewise, adipose tissue-derived mesenchymal stem cells have been shown to enhance salivary gland secretory function by preventing cell death after radiation exposure; but the specific mechanisms are yet to be elucidated [13, 14]. As for organ culture transplantation, previous studies showed cultured embryonic salivary cells can be transplanted in vivo [15] but again, studies demonstrating their long-term outcomes are still lacking. Another emerging approach is the use of salivary gland cell sheets which promote cell differentiation and tissue integrity in a wounded mouse submandibular gland (SMG) model, yet the main challenge facing this technology is the need to standardize cell composition within the sheets to achieve greater reproducibility [16, 17]. Bioprinting strategies have also shown promise in assembling multiple cell types into salivary gland organotypic cultures; however, this technology does not yet mimic the salivary gland native architecture (e.g., cell polarity and organization [18, 19]) and in vivo studies demonstrating efficacy are lacking [20]. Regarding scaffolds other than the Fibrin Hydrogels (FH), various biomaterials have been shown to promote cell growth and attachment but the degree of structural organization (e.g., the presence of hollow multi-lumen formation, cell polarity and functionality), has been modest [21–26]. According to recent research, hyaluronic acid-based scaffold has been demonstrated to serve as promising support structures for tissue regeneration using human stem/progenitor cells in an immunosuppressed mini swine model [27]. However, the incorporation of human cells remains a challenge to be addressed in the clinic, as it can still provoke immune responses. In general, these technologies offer the potential for more permanent solutions to hyposalivation due to head and neck radiation exposure.

Regarding FH, previous studies have demonstrated that laminin-1 peptides (L_1p) conjugated FH (i.e., L_1p-FH) induce acinar-like formation in vitro [28]. Moreover, L_1p-FH promote healing in a wounded SMG mouse model [29, 30], and in head and neck irradiated mouse model [31]. In addition, combination of vascular endothelial growth factor (VEGF) and fibroblast growth factor 9 (FGF-9) promoted salivary epithelial tissue regeneration, vascularization, neurogenesis and healing in a wounded SMG mouse model [32]. However, the long-term sustainability of the therapeutic effects of these FH-based scaffolds may be compromised when applied to radiation exposure models due to residual fibrosis and inadequate formation of blood vessels and nerves within the regenerated tissue. To address these challenges, we adopted a bioengineering strategy centered on harnessing the potential of the FGF receptor 2b (FGFR2b) signaling pathway, known for its role in promoting tissue regeneration, angiogenesis, and neural cell differentiation [33–36], thereby playing a critical role in salivary gland development and sustainability (i.e., for tissue homeostasis) [37, 38]. Based on these considerations, our hypothesis centered on the activation of FGFR2 signaling through the incorporation of fibroblast growth factors 7 and 10 (FGF-7 and FGF-10, respectively). We envisioned that this approach would not only enhance epithelial tissue regeneration but also promote the formation of blood vessels and nerve networks. Note that in this study, Ep-FH was administered three days after radiation exposure, deliberately chosen to coincide with the gradual release of growth factors facilitated by the slow degradation of the hydrogel, aligning with the progression of fibrosis before its initiation. Through the use of Ep-FH, we aimed to promote effective and sustained regeneration of damaged salivary glands, thereby creating an advanced FH-based scaffold that surpasses the capabilities of previous prototypes [33–36]. Ultimately, our goal was to achieve superior treatment outcomes for hyposalivation resulting from radiation exposure.

2. Materials and methods

2.1. Materials

Calcium chloride (CaCl₂), ethylenediaminetetraacetic acid (EDTA), glycerol, goat serum, hematoxylin and eosin Y solution, Luria broth, lyophilized human fibrinogen, pilocarpine, Terrific broth, tert–amyl alcohol, tribromoethanol, Tris base, ε-aminocaproic acid (εACA), lysozyme from chicken egg white and Origami^™ B(DE3) Competent Cells were obtained from Millipore-Sigma (Burlington, MA). Alexa Fluor 488 conjugated anti-rabbit IgG/Alexa Fluor 568 conjugated anti-mouse IgG secondary antibodies were obtained from Invitrogen (Carlsbad, CA). Picro-Sirius Red Stain Kit, rabbit anti-aquaporin 5 (AQP5), rabbit anti-Ki67, mouse anti-cytokeratin 7 (CK7) and mouse anti-β-tubulin III antibodies were obtained from Abcam (Cambridge, MA). Bluing reagent, 6-diamidino-2-phenylindole (DAPI), kanamycin sulfate, phosphate buffered saline (PBS), sodium chloride, sodium citrate, triton X-100 and xylene were obtained from Thermo Fisher Scientific (Waltham, MA). Ethanol was obtained from Decon Labs (King of Prussia, PA). Rabbit anti-CD31 was obtained from Novus Biologicals (Centennial, CO). pET28-MBP-TEV [39] expression vector was obtained from Addgene (Watertown, MA). Insulin syringes (28 G) were obtained from BD (Franklin Lakes, NJ). A99 and YIGSR peptides were synthesized by the University of Missouri Peptide Synthesis Core.

2.2. Fusion protein preparation

The codon-optimized cDNAs of FGF-7 and FGF-10 were individually cloned into the pET28-MBP-TEV expression vector, which contains an MBP-tag and a kanamycin resistance selectable marker [40, 41]. These cloned vectors were subsequently introduced into the Origami (DE3) strain of Escherichia coli (E. coli). Two 3 mL Luria broth bacterial starter cultures containing MBP-FGF-7 or MBP-FGF-10 were incubated in a shaker at 37 °C for 16 h for expansion. Next, the starter cultures were added to 400 mL of Overnight Express Terrific broth supplemented with 100 μg/mL of kanamycin and 1 % glycerol (v/v) using a 2.8 L beveled flask. The solution was incubated in a shaker at 22 °C, 250 rpm for 24 h, followed by centrifugation at 4000 r.c.f for 15 min to collect the cell pellet, which was stored at −80 °C for a minimum of 6 h before lysis. Subsequently, the cell pellet was lysed using 50 mL of lysis buffer [50 mM Tris, 500 mM NaCl, 1 mg/mL lysozyme from chicken egg, 2 % Triton X-100 (v/v), 2 % glycerol (v/v), pH= 7.2] at room temperature, with stirring at 200 rpm for 45 min. The lysate was then sonicated at 50 % amplitude for 1 min and 40 s on ice with 10 s on and 30 s off cycles. After sonication, the lysate was centrifuged at 4 °C, 15,000 r.c.f for 15 min and the solution was passed through an MBP-Trap column to separate the MBP-tagged fusion protein from the cell lysate. Finally, the purified protein was stored at −80 °C in the presence of 0.1 % BSA (w/v) until it was needed.

2.2.1. Protein sequence of MBP-FGF-7

MGSSHHHHHHSSGLVPRGSHMKIEEGKLVIWINGDKGYNGLAEVGKKFEKDTGIKVTVEHPDKLEEKFPQVAATGDGPDIIFWAHDRFGGYAQSGLLAEITPDKAFQDKLYPFTWDAVRYNGKLIAYPIAVEALSLIYNKDLLPNPPKTWEEIPALDKELKAKGKSALMFNLQEPYFTWPLIAADGGYAFKYENGKYDIKDVGVDNAGAKAGLTFLVDLIKNKHMNADTDYSIAEAAFNKGETAMTINGPWAWSNIDTSKVNYGVTVLPTFKGQPSKPFVGVLSAGINAASPNKELAKEFLENYLLTDEGLEAVNKDKPLGAVALKSYEEELAKDPRIAATMENAQKGEIMPNIPQMSAFWYAVRTAVINAASGRQTVDEALKDAQTNSSSNNNNNNNNNNLGIEGRGGSLVPRGSGGGSNQEQVSPGGGMHKWILTWILPTLLYRSCFHIICLVGTISLACNDMTPEQMATNVNCSSPERHTRSYDYMEGGDIRVRRLFCRTQWYLRIDKRGKVKGTQEMKNNYNIMEIRTVAVGIVAIKGVESEFYLAMNKEGKLYAKKECNEDCNFKELILENHYNTYASAKWTHNGGEMFVALNQKGIPVRGKKTKKEQKTAHFLPMAIT*

2.2.2. Protein sequence of MBP-FGF-10

MGSSHHHHHHSSGLVPRGSHMKIEEGKLVIWINGDKGYNGLAEVGKKFEKDTGIKVTVEHPDKLEEKFPQVAATGDGPDIIFWAHDRFGGYAQSGLLAEITPDKAFQDKLYPFTWDAVRYNGKLIAYPIAVEALSLIYNKDLLPNPPKTWEEIPALDKELKAKGKSALMFNLQEPYFTWPLIAADGGYAFKYENGKYDIKDVGVDNAGAKAGLTFLVDLIKNKHMNADTDYSIAEAAFNKGETAMTINGPWAWSNIDTSKVNYGVTVLPTFKGQPSKPFVGVLSAGINAASPNKELAKEFLENYLLTDEGLEAVNKDKPLGAVALKSYEEELAKDPRIAATMENAQKGEIMPNIPQMSAFWYAVRTAVINAASGRQTVDEALKDAQTNSSSNNNNNNNNNNLGIEGRGGSLVPRGSGGGSNQEQVSPGGGSWKWILTHCASAFPHLPGCCCCCFLLLFLVSSVPVTCQALGQDMVSPEATNSSSSSFSSPSSAGRHVRSYNHLQGDVRWRKLFSFTKYFLKIEKNGKVSGTKKENCPYSILEITSVEIGVVAVKAINSNYYLAMNKKGKLYGSKEFNNDCKLKERIEENGYNTYASFNWQHNGRQMY VALNGKGAPRRGQKTRRKNTSAHFLPMVVHS*

2.3. Animals

Female 8-week-old C57BL/6 J mice, weighing approximately 18 g, were obtained from Jackson Laboratory (Bar Harbor, ME). To determine the appropriate number of mice per group for saliva flow rate analysis to achieve a statistical significance level of at least p < 0.05, power analysis was performed using G*Power 3.1.9.7 software [effect size = 0.82, α err prob = 0.05, power (1−β err prob = 0.95, number of groups = 3), which indicated that 9 mice were required per group. For tissue analysis, 5 mice per group were used (3 time points, number of groups = 3). In this study, a total of 72 mice were randomly assigned to one of three groups as follows: a non-irradiated group (n = 24), an irradiated group without FGF-7/10 and laminin-1 peptides fortified fibrin hydrogel [enhanced laminin-1 peptides fibrin hydrogel (Ep-FH)] injection (n = 24), and an irradiated group receiving the Ep-FH injection (n = 24). All animal handling, anesthesia, and treatments were conducted in compliance with ARRIVE guidelines and received prior approval from the University of Missouri-Columbia Animal Care and Use Committee (ACUC). The animals were maintained in a facility with a 12-hour light/dark cycle and were managed by a veterinarian in an animal room.

2.4. Head and neck irradiated mouse model

To assess the regenerative effect of Ep-FH on salivary gland tissue damage induced by radiation exposure, a widely accepted head and neck irradiated mouse model was employed [42–45]. Specifically, mice were anesthetized with 2.5 % isoflurane (100 mL/min), and a custom-designed lead shield was used to protect against radiation exposure. The external layer of the shield was made of 3 mm thick lead, while the internal layer was made of 3 mm thick aluminum. The shield featured a single slit measuring 1 cm, allowing for targeted radiation exposure to the neck region while protecting other areas of the body. Next, a single 15 Gy dose of radiation was administered to the head and neck area using a Multi-Rad225 machine (Precision, Madison, CT) and mice were allowed to recover for 3 days before receiving treatment with Ep-FH.

2.5. Hydrogel preparation

L_1p-FH was prepared as previously described [28]. Briefly, two L_1p peptides, A99 and YIGSR, each containing a cysteine residue, were synthesized using a peptide synthesizer. Next, fibrinogen was activated using sulfo-LC-SPDP and then activated fibrinogen was conjugated with the synthesized peptides. The resulting peptide-conjugated fibrinogens were purified, lyophilized, and stored at −80 °C until use. Then, YIGSR-conjugated fibrinogen (1.25 mg/mL), A99-conjugated fibrinogen (1.25 mg/mL), CaCl₂ (2.5 mM), εACA (2 mg/mL), MBP-FGF-7 (100 ng/mL), and MBP-FGF-10 (100 ng/mL) were mixed in Tris-buffered saline (TBS). Note that the MBP domain of the fusion proteins contains a thrombin cleavage site (i.e., LVPRGS), which allows endogenous thrombin to cleave off the MBP-tag. Consequently, the activated fibrin-binding domain NQE-QVSP is exposed, and can be enzymatically incorporated into fibrin through endogenous factor XIII. Therefore, upon transdermal injection, the solution forms a stable hydrogel with conjugated L_1p and growth factors immediately at the injection site. By employing this approach, the burst effect that may occur when using commercially available FGF-7 and FGF-10 directly mixed and injected can be reduced [46]. The release of the growth factors is gradual as the hydrogel degrades thereby resulting in sustained effects in the injected area [46].

2.6. Transdermal injection

C57BL/6 J mice were anesthetized using 2.5 % isoflurane (100 mL/min) and transdermal injections were administered to irradiated mouse SMGs (both left and right sides) using an insulin syringe (28 G) containing 10 μL of freshly mixed Ep-FH solution on day 3 post-radiation exposure, as mentioned above. There was no addition of exogenous thrombin and factor XIII in this study given that fibrinogen and growth factors can be polymerized with the animal’s endogenous enzymes [31]. Finally, saliva and tissue samples were collected at days 30, 60, and 90 post-radiation exposure to investigate the effects of Ep-FH.

2.7. Histochemical analysis

SMGs were fixed in 10 % formalin (w/v) at room temperature overnight. The fixed tissues were then dehydrated in a series of ethanol solutions (starting with 70 % ethanol (v/v), followed by 100 % ethanol), embedded in paraffin wax, and cut into 5 μm tissue sections. Next, sections were deparaffinized using xylene and rehydrated using a series of ethanol solutions (100 %, 95 %, 80 %, 70 %, and 50 %; v/v) and distilled water. For hematoxylin and eosin staining, tissue sections were treated with hematoxylin for 5 min and rinsed twice with distilled water to remove any excess stain. Bluing reagent was then applied to the tissue sections with a 10 s incubation followed by rinsing twice with distilled water. Then, the slides were dipped in 100 % ethanol, excess ethanol was removed, and eosin Y solution was applied to the tissue sections for 3 min. Slides were then rinsed, dehydrated using absolute ethanol, cleared, and mounted in a xylene-based mounting solution. To perform Picro-Sirius Red staining, tissue samples were stained with Picro-Sirius Red solution for 1 h, quickly rinsed in two changes of 0.5 % acetic acid (v/v) solution and rinsed with 100 % ethanol. Following this step, slides were dehydrated using absolute ethanol, cleared, and mounted in a xylene-based mounting solution. Finally, stained slides were analyzed using a Leica DMI6000B microscope (Leica Microsystems, Wetzlar, Germany) under bright field imaging (H&E and Picro-Sirius Red staining). In order to identify the regions of interest, fluorescence imaging was applied to these samples followed by artificial intelligence analysis as depicted in Fig. 4 [47, 48] while statistical analysis was performed with GraphPad Prism 6 using one-way (Fig. 4e) and two-way analysis of variance (ANOVA; Fig. 4a–d, f–h) and Dunnett’s post hoc test for multiple comparisons (p < 0.05, n = 4 mice/group).

Fig. 4.

Treatment with Ep-FH significantly prevents fibrosis in irradiated salivary glands. Results from Picro-Sirius Red staining were quantified using the AVIA software as detailed in Materials and Methods. Statistical analysis was performed with GraphPad Prism 6 using one-way (for Fig. 4e) and two-way analysis of variance (ANOVA; for Fig. 4a–d, f–h) and Dunnett’s post hoc test for multiple comparisons (p < 0.05, n = 4). White scale bars represent 1 mm.

2.8. Confocal analysis

Rehydrated tissue sections were subjected to antigen retrieval by incubation in sodium citrate buffer in a pressure cooker for 20 min. After this step, the sections were thoroughly washed with distilled water to remove any remaining buffer. Sections were then permeabilized with 0.1 % Triton X-100 (v/v) in PBS at room temperature for 45 min. To prevent non-specific binding of the primary antibody, sections were blocked with 5 % goat serum (w/v) in PBS for 1 h at room temperature. The primary antibody solutions were diluted in blocking buffer as described in Table 1, added to the sections, and incubated overnight at 4 °C. The next day, the sections were washed five times with PBS to remove any unbound primary antibody. To visualize the bound primary antibody, the sections were incubated with the appropriate secondary antibody solution, as described in Table 1, for 1 h at room temperature. The sections were then washed five times with PBS to remove any unbound secondary antibody. Next, the sections were counterstained with DAPI (5 μM) at room temperature for 5 min to visualize nuclei. Finally, all specimens were analyzed using a STELLARIS confocal microscope (Leica Microsystems, Wetzlar, Germany).

Table 1.

Antibody dilutions and suppliers for immunofluorescence.

Antibody	Company (Catalog number)	Dilution
Rabbit anti-aquaporin 5	Abcam (ab78486)	200
Mouse anti-cytokeratin 7	Abcam (ab9021)	300
Rabbit anti-CD31	Novus Biologicals (NB100-2284)	100
Mouse anti-β-tubulin III	Abcam (ab7751)	100
Rabbit anti-Ki67	Abcam (ab15580)	200
Alexa Fluor 488 conjugated anti-rabbit IgG	Invitrogen (A11008)	500
Alexa Fluor 568 conjugated anti-mouse IgG	Invitrogen (A11031)	500

2.9. Deep-learning image analysis

The distribution of collagen deposition (e.g., fibrosis) as well as vascular and neural connectivity were analyzed using a commercial artificial intelligence-based image analysis software (Aivia 11.01, Leica Microsystems, Wetzlar, Germany) [49]. For these studies, a deep learning model utilizing transfer learning was employed. The first step involved machine training for capturing SMG fluorescent images. Once the machine learned and recognized patterns, new algorithms were created and applied to all images captured within the Aivia software. Next, algorithms automatically processed the images and extracted relevant information regarding collagen distribution, blood vessel and neural connectivity. Finally, ImageJ software (Fiji Distribution, Version 1.5p) was used to evaluate the reliability of image analysis.

2.10. Saliva flow rate measurements

Mice were anesthetized using Avertin (300 mg/kg) followed by an intraperitoneal injection of pilocarpine (2.5 mg/kg) to stimulate saliva secretion. Subsequently, mice were positioned with their abdomen facing upwards at a 45-degree angle to allow saliva collection. One minute after the administration of pilocarpine, saliva was collected for 20 min using a 200 μL pipette. Saliva was then centrifuged at 12,000 rpm for 10 min. After this step, the supernatant was transferred to a new tube and the volume of the collected saliva measured. Next, the saliva flow rate was calculated using the formula (below):

$$\mathrm{S}\mathrm{a}\mathrm{l}\mathrm{i}\mathrm{v}\mathrm{a}\mathrm{f}\mathrm{l}\mathrm{o}\mathrm{w}\mathrm{r}\mathrm{a}\mathrm{t}\mathrm{e}=\frac{\text{Stimulated}\text{saliva}\text{volume}\left(\mu \mathrm{L}\right)}{\text{Body}\text{weight}\text{of}\text{mouse}\left(\mathrm{g}\right)\times \text{Collection}\text{time}\left(20\mathrm{m}\mathrm{i}\mathrm{n}\right)}$$

Data were analyzed with GraphPad Prism 6 using one-way ANOVA and Dunnett’s post hoc test for multiple comparisons between groups and expressed as means ± SD (n = 9) with *p < 0.05; **p < 0.01; ***p < 0.001, ****p < 0.0001 indicating a significant difference from non-irradiated control group (Supplemental Fig. 1) or irradiated without Ep-FH treatment (Fig. 8).

Fig. 8.

Treatment with Ep-FH increases saliva secretion after radiation exposure. Saliva was collected for 20 min following pilocarpine stimulation (2.5 mg/kg). Results represent data from n = 9 mice per condition and data presented as means ± SD. Statistical significance was assessed using one-way ANOVA, followed by Dunnett’s post-hoc test for multiple comparisons to the irradiated group. Symbols are defined as follows: circles represent D30, squares represent D60, and triangles represent D90. Symbol notation is as follows: black for (non-IR), white for (IR-untreated), and gray for (IR-Ep-FH treated) with *p < 0.05; **p < 0.01; ****p < 0.0001 indicating significant differences compared to irradiated control group.

2.11. Statistical analysis

Tissue analysis was conducted with 3 to 5 mice, while saliva flow measurements involved 9 mice. The data are reported as the means ± SD. Data from different groups were analyzed by one-way or two-way ANOVA followed by Dunnett’s post hoc test for multiple comparisons. All statistical analysis was done using GraphPad Prism 6. Statistical significance and their corresponding significance levels (*p < 0.05; **p < 0.01; ***p < 0.001, ****p < 0.0001) were reported for each result.

3. Results

3.1. Generation of fusion proteins that preserve their native structure

To maintain function and stability, it is crucial for fusion constructs to resemble the protein from which they were derived [50]. To achieve these properties, various techniques can be employed, such as using flexible linkers or connecting proteins at structurally less critical regions [51]. In this study, the structure of the fusion protein was predicted using the software ColabFold v1.5.2 [52]. The analysis and visualization of the protein structure were performed using PyMOL v2.5.5 [53]. Our results showed no predicted structural changes between the fusion constructs and native FGF-7 and FGF-10 (Fig. 1). Specifically, both the conjugated proteins (Fig. 1, red) and native proteins (Fig. 1, yellow) showed similar 3D structures, thereby indicating the likelihood of functional preservation.

Fig. 1.

(a) Schematic representation of the fusion proteins. (b) Predicted 3D structure of the MBP-fused (green), NQEQVSP conjugated (activated) and native FGF-7 and FGF-10. The structure of the fusion protein was predicted using ColabFold v1.5.2, with the resulting model indicating that the NQEQVSP-conjugated proteins (red) are likely to resemble the native structure of FGF-7 and FGF-10 (yellow).

3.2. Ep-FH preserved epithelial integrity after radiation exposure

To determine the effects of FGF-7/10 and laminin-1 peptides fortified fibrin hydrogel (Ep-FH) on SMG morphology, we obtained hematoxylin and eosin tissue sections and visualized with a Leica DMI6000B microscope as detailed in Materials and Methods. Our results showed that non-irradiated SMG display lobules separated by thin strands of connective tissue (Fig. 2 and Supplemental Fig. 2). Moreover, we observed pyramidal cell clusters with round nuclei consistent with acinar cells as well as cuboidal and columnar cells with a round nuclei consistent with ductal structures (Fig. 2 and Supplemental Fig. 2). Finally, the granular convoluted ducts were rich in eosinophilic secretory granules (Fig. 2 and Supplemental Fig. 2). Together, the morphology of the non-irradiated SMG described above was consistent with healthy tissue, as expected. In contrast, untreated irradiated glands exhibited a progressive loss (i.e., from day 30–90 post-radiation exposure) of lobular architecture, characterized by fatty replacement (Fig. 2 and Supplemental Fig. 2, white arrows) and interstitial fibrosis (Fig. 2 and Supplemental Fig. 2, red arrows). Also, irradiated SMG without Ep-FH displayed progressive acinar cell atrophy (Fig. 2 and Supplemental Fig. 2, green arrows) and vacuolization (Fig. 2 and Supplemental Fig. 2, blue arrows) together with a loss of secretory granules, intraductal microliths and periductal inflammatory cells (Fig. 2 and Supplemental Fig. 2, yellow arrows). These alterations were more pronounced at day 90 post-radiation exposure. Collectively, the morphology of irradiated glands was consistent with severe damage over time, as expected. Interestingly, SMG treated with Ep-FH maintained lobular architecture, recovered acinar and ductal morphology and showed decreased inflammatory infiltrates closely mimicking the non-irradiated group. These results suggest that Ep-FH preserved the structural integrity, composition and morphology of irradiated SMG.

Fig. 2.

Treatment with Ep-FH preserves epithelial integrity when applied after radiation exposure. Hematoxylin and eosin staining was performed and tissue morphology was analyzed using a Leica DMI6000B microscope. Arrows indicate the following: white arrows for fatty replacement, red arrows for interstitial fibrosis, green arrows for acinar cell atrophy, blue arrows for vacuolization and yellow arrows for inflammatory cells. Black scale bars represent 1 mm and white scale bars represent 500 μm. SMG indicates submandibular gland, and SLG indicates sublingual gland, respectively.

3.3. Ep-FH reduced fibrosis after radiation exposure

To further investigate the effects of Ep-FH on SMG fibrosis, we performed Picro-Sirius Red staining, a specific dye staining for collagen (Fig. 3). Our results showed that non-irradiated SMG displayed thin collagen fibers between parenchymal lobules and periductal areas, as expected. In contrast, irradiated SMG without Ep-FH injection showed a progressive interstitial fibrosis with exaggerated collagen deposition (Fig. 3, blue arrows) in periductal and periacinar areas together with loss of the lobular architecture that was more severe at day 90 post-radiation exposure. Remarkably, irradiated SMG treated with Ep-FH showed thin fibers of collagen located between parenchymal lobules and in periductal areas resembling the non-irradiated group and therefore indicating reduction of fibrosis. To further analyze these results, the amount of collagen deposition was quantified using the AVIA software and ImageJ. As shown in Fig. 4, treatment with Ep-FH resulted in significant reduction of collagen deposition, suggesting that Ep-FH reduced fibrosis dramatically in irradiated SMG.

Fig. 3.

Treatment with Ep-FH prevents fibrosis in irradiated salivary glands. Picro-Sirius Red staining was performed and tissue morphology was analyzed using a Leica DMI6000B microscope. Fibrosis is indicated by blue arrows in the images. Black scale bars represent 1 mm.

3.4. Ep-FH maintained blood vessel and neuronal structure after radiation exposure

To investigate the impact of Ep-FH on the recovery of vascular and neuronal damage induced by radiation exposure in a head and neck irradiated mouse model, CD31 and β-tubulin III antibodies were used to detect endothelial and neuronal tissues, respectively, and analyzed using the AIVIA software, as described in Materials and Methods. As shown in Figs. 5 and 6, non-irradiated SMG displayed continuous green and red solid line staining indicating intact blood vessels and nerves respectively. In contrast, irradiated SMG without Ep-FH injection displayed scattered CD31 and discontinued β-tubulin III, especially at later times (60 and 90 days post-radiation exposure), suggestive of blood vessel and neuron damage. Notably, treatment with Ep-FH retained the blood vessel density at 60 days post radiation exposure to similar level as compared to non-irradiated control; as well as maintained nerve continuity after 30 days suggesting that Ep-FH restored SMG angiogenesis and neurogenesis in irradiated SMG.

Fig. 5.

Treatment with Ep-FH promotes CD31 and β-tubulin III expression in irradiated salivary glands. Blood vessels and neuronal structures were detected with CD31 (highly expressed on the surface of endothelial cells; green) and β-tubulin III (microtubule exclusively found in neurons; red) antibodies, respectively. Nuclei were counter-stained in blue with DAPI and images were analyzed using STELLARIS confocal microscope. Images represent n = 3 mice/group, where scale bars = 100 μm.

Fig. 6.

Treatment with Ep-FH promotes neuronal and vascular connections in irradiated salivary glands. The images from Fig. 5 were processed with the use of AIVIA software to render 3D constructs. Images represent n = 3 mice/group. The scale bars represent 25 μm.

3.5. Ep-FH maintained AQP5 and CK7 expression after radiation exposure

To investigate the impact of Ep-FH on restoration of disrupted epithelial integrity caused by radiation exposure, AQP5 and CK7 antibodies were used to detect acinar and ductal structures, respectively, within the SMG sections and analyzed using confocal microscopy, as described in Materials and Methods. As shown in Fig. 7, non-irradiated SMG displayed apical green and basolateral red staining indicative of organized intact acinar and ductal structures, respectively. In contrast, untreated irradiated SMG injection displayed focal areas lacking expression of AQP5 and CK7 (Fig. 7, white arrows), especially at later times (Day 90 IR), suggestive of epithelial tissue damage. Notably, irradiated SMG with Ep-FH injection retained acinar and ductal markers within the SMG comparable to normal levels after 60 days. These results indicate that treatment with Ep-FH restored epithelial integrity in irradiated SMG.

Fig. 7.

Treatment with Ep-FH maintains AQP5 and CK7 expression in irradiated salivary. Acinar and ductal structures were detected with AQP5 (green) and CK7 (red) antibodies, respectively. Nuclei were counterstained in blue with DAPI and images were analyzed using STELLARIS confocal microscope. Images represent n = 3 samples, where scale bars = 25 μm.

3.6. Ep-FH increased saliva secretion after radiation exposure

To explore the potential synergistic effects of FGF-7 and FGF-10 binding to L_1p-FH, we initially administered each FGF-7 and FGF-10 separately in combination with FH and then evaluated saliva flow on day 30 post-radiation exposure (Supplemental Fig 3). When the growth factors were administered individually, they did not yield significant results. However, when combined, saliva flow rate was similar to L_1p-FH group. Next, to investigate the impact of Ep-FH on saliva secretion over an extended period of time, saliva flow rates were measured up to 90 days. As shown in Fig. 8, non-irradiated cohorts exhibited intact saliva flow rates (0.77 μL/g/min). In contrast, irradiated mice showed a significant reduction in saliva flow rates (0.43 μL/g/min) after 30 days, which continued to decrease up to 90 days (0.22 μL/g/min), which is approximately 28.5 % of the saliva secretion when compared to the non-irradiated group. Notably, irradiated mice treated with Ep-FH displayed a significant increase in saliva flow rates (0.63 μL/g/min) after 30 days, which was maintained at high levels up to 90 days (0.48 μL/g/min), demonstrating that a one-time transdermal treatment Ep-FH after radiation exposure restored saliva secretion to 62 % of the saliva secretion observed in the non-irradiated group.

4. Discussion

Saliva secretion is a highly regulated process where proper vascularization and innervation is critical for maintaining tissue homeostasis [54, 55]. Specifically, blood vessels surrounding the salivary glands supply oxygen and nutrients to this organ while nerves provide the electrical signals that regulate saliva secretion via specific neurotransmitters (e.g., acetylcholine and noradrenaline). Unfortunately, radiation treatment, commonly used to eliminate cancer cells in the head and neck area, also harms the remaining healthy cells in salivary glands by increasing cytotoxicity, inflammation and immune responses [56]. These events result in the collapse of both blood vessels and nerves thereby reducing irrigation and innervation into the gland, ultimately leading to hyposalivation [57, 58]. Moreover, radiation exposure can also induce excessive fibrosis in the salivary glands, further compromising their structure and function [59]. The excessive fibrosis can further exacerbate damage of blood vessels and nerves, leading to a severe impairment in saliva secretion [60, 61]. While there are emerging approaches to restore salivary gland function in head and neck irradiated patients, there are no suitable therapies at this time. To this end, the current study provides a new approach using local application of a FH-based scaffold in irradiated mouse SMG that restored both structure and functionality and suggesting potential clinical applications.

Our previous studies demonstrated that salivary cells grown on FH alone form fibroblast-like monolayers lacking polarity and 3D organization [28, 29]. In contrast, salivary cells grown on a modified FH matrix containing immobilized laminin-1 peptides (i.e., L_1p-FH) formed functional 3D salivary cell clusters [28, 29]. To confirm whether this new scaffold was useful for in vivo applications, we developed a SMG wounded mouse model with results demonstrating that treatment with L_1p-FH induced formation of newly organized salivary tissue [29, 30]. This model was further expanded to head and neck irradiated mouse model where we confirmed that treatment with L_1p-FH improves saliva secretion by day 30 post-radiation exposure [31] when compared with animals treated with no L_1p-FH. Taken together, these results indicated that damaged salivary glands can regenerate when treated with L_1p-FH; however, the long-term sustainability of these effects was uncertain due to inadequate blood vessels and nerve formation. To address these limitations, here we were able to introduce FGF-7 and FGF-10 into L_1p-FH to restore long-term structure and function of irradiated SMG.

Previous studies in embryonic mouse SMG indicate that both FGF-7 and FGF-10 bind to FGFR2b via a regulatory network that influences morphogenesis during the development of this organ [62, 63]. Specifically, FGF-7 appears to promote epithelial cell proliferation in a paracrine manner, resulting in proliferation and budding, while FGF-10 causes elongation of the duct behind the proliferating tip [62]. Similarly, we showed that controlling the mode of FGF-7 and FGF-10 presentation within FH controlled migration (FGF-7) and proliferation (FGF-10) of SG cells, promoting branching morphogenesis in vitro [46]. Consistent with these studies, local administration of FGF-7 in head and neck irradiated mice prevented SG hypofunction [64]. Regarding FGF-7 and FGF-10′ effects in other cell types, previous results demonstrated that FGF-7 induces angiogenesis and protects barrier function in endothelial cells [65]. Moreover, exogenous FGF-7 was shown to promote wound healing in multiple epithelial tissues including skin [66–68]. As for FGF-10, previous studies indicate that this growth factor is not only highly expressed in neurons but is also involved in neurogenesis and differentiation [69, 70]. Moreover, several studies revealed that FGF-10 is upregulated in the nervous system after injury [71]. Interestingly, FGF-10 also diminishes neural inflammation in rats with peripheral nerve injury [36]. Finally, FGF-7 and FGF-10 also reduce fibrosis by various mechanisms including ERK1/2, AKT and PLCγ signaling which are involved in the regulation of fibrosis [72]. Given the overwhelming influence of FGF-7 and FGF-10 in mouse SMG morphogenesis, angiogenesis and neurogenesis, the current study expanded the use of these growth factors to adult tissues and demonstrated that immobilized FGF-7 and FGF-10 (i.e., using MBP-FGF-7 and MBP-FGF-10 within L_1p-FH scaffold) have a synergistic effect on promoting epithelial integrity as well as vascular and nerve cell repair, with the two latter effects not previously seen when using L_1p-FH alone [31]. To this end, the salivary gland regeneration was significantly improved by local Ep-FH injection, and salivary secretion was maintained at a constant level as well as significantly higher when compared with irradiated mice without Ep-FH local injection at all time points. These results are significant as they demonstrate the first use of Ep-FH in irradiated glands to restore their form and function for an extended period of time (i.e., 90 days).

In summary, we engineered FGF-7 and FGF-10 and immobilized them into L_1p-FH. The resulting hydrogel, Ep-FH, restored irradiated salivary gland functionality by enhancing epithelial tissue organization, promoting the development of a healthy network of blood vessels and nerves as well as reduction of fibrosis. Note that the tissue damage observed in 15 Gy irradiated SMG at 90 days in mice, closely resembles the human condition (i.e., SMG exposed to 72 Gy cumulative doses after 6 years) highlighting the significance of this study [73]. Moreover, we did not observe significant differences in cell division between Ep-FH-treated mice when compared to non-irradiated healthy controls, indicating a low likelihood of tumorigenesis (Supplemental Fig. 4). As such, the maintenance of tissue integrity and saliva secretion in mice suggests a remarkable potential for clinical applications. Future studies will encompass experiments involving male mice to investigate whether sexual dimorphism influences the performance of this hydrogel. In addition, we will utilize flow cytometry and inhibition studies to identify and characterize specific cell populations as well as to uncover the underlying signaling pathways activated by Ep-FH. Furthermore, we plan to enhance our approach by combining the Ep-FH scaffold with polymeric microparticles and administering injections at multiple time points. This approach will enable the controlled release of pro-angiogenic and pro-innervation growth factors in a sequential manner, closely mimicking the in vivo physiology. Finally, we anticipate that this approach will make a substantial contribution to improving the functional recovery of salivary glands following radiation treatment.

Statement of significance.

Radiation therapies used to treat head and neck cancers often result in damaged salivary gland, leading to severe dryness of the oral cavity. In this study, we engineered FGF-7 and FGF-10 and immobilized them into L_1p-FH. The resulting hydrogel, Ep-FH, restored irradiated salivary gland functionality by enhancing epithelial tissue organization, promoting the development of a healthy network of blood vessels and nerves as well as reduction of fibrosis.