Encapsulation of primary salivary gland acinar cell clusters and intercalated ducts (AIDUCs) within matrix metalloproteinase (MMP)-degradable hydrogels to maintain tissue structure and function

Abstract

Progress in the development of salivary gland regenerative strategies is limited by poor maintenance of the secretory function of salivary gland cells (SGCs) in vitro. To reduce the precipitous loss of secretory function, a modified approach to isolate intact acinar cell clusters and intercalated ducts (AIDUCs), rather than commonly used single cell suspension, was investigated. This isolation approach yielded AIDUCs that maintain many of the cell-cell and cell-matrix interactions of intact glands. Encapsulation of AIDUCs in matrix metalloproteinase (MMP)-degradable PEG hydrogels promoted self-assembly into salivary gland mimetics (SGm) with acinar-like structure. Expression of Mist1, a transcription factor associated with secretory function, was detectable throughout the in vitro culture period up to 14 days. Immunohistochemistry also confirmed expression of acinar cell markers (NKCC1, PIP and AQP5), duct cell markers (K7 and K5), and myoepithelial cell markers (SMA). Robust carbachol and ATP-stimulated calcium flux was observed within the SGm for up to 14 days after encapsulation, indicating that secretory function is maintained. Though some acinar-to-ductal metaplasia was observed within SGm, it was reduced compared to previous reports. In conclusion, cell-cell interactions maintained within AIDUCs together with the hydrogel microenvironment may be a promising platform for salivary gland regenerative strategies.

Introduction

Approximately 50,000 patients in the United States and as many as 500,000 people worldwide are diagnosed annually with head and neck cancers¹. Most patients receive standard treatment of radiation therapy or a combination of radiation therapy with chemotherapy^{2, 3}. However, radiation therapy results in atrophy, fibrosis, and degeneration of both major and minor salivary gland tissue, leading to salivary gland hypofunction and xerostomia, which severely affects oral health and quality of life⁴. There are no regenerative therapies to restore salivary gland function after radiation damage, so patients suffer from life-long xerostomia. Palliative treatments for xerostomia include the use of saliva substitutes and sialagogues to lubricate the mouth, however, these treatments do not fully restore or emulate the myriad functions of the salivary gland^{5, 6}. Thus, development of treatments to provide durable recovery of salivary function are in dire need.

Cell-based therapies are a promising approach to restore salivary gland function. Direct injection of primary mouse submandibular gland (SMG) cells^{7, 8}, c-Kit⁺ salivary gland cells (SGCs)^9–11, or mesenchymal stem cells (MSCs)^12–15 into irradiated glands has been shown to partially improve gland function, but restoration of saliva secretion was incomplete and highly variable. Tissue engineering has emerged as a promising strategy to coordinate salivary gland regeneration. Tissue engineering involves the use of a combination of cells, tissue matrix/scaffold and soluble factors that create a 3D microenvironment in which functional constructs assemble and actively interact with the surrounding extracellular matrix (ECM) to restore, maintain, or improve damaged tissues or whole organs. Approaches include use of nanofibers, decellularized extracellular matrix, and hydrogels together with SGCs, growth factors, and other stimuli^16–24. Although a few studies have translated their findings in vivo^25–27, no study to date has demonstrated complete restoration of gland function.

In our previous research, hydrogels formed via norbornene functionalized PEG macromers crosslinked with MMP degradable peptides were used to encapsulate pre-aggregated SGCs that were derived by culturing single SGCs in tissue culture plates for 2 days in vitro and then cultured for up to 2 weeks in hydrogels²². Adult salivary gland cells have been shown to express multiple MMPs^{28, 29}, which can degrade the peptide crosslinker during cellular assembly. Within MMP-degradable PEG hydrogels, cells organize into structures with apicobasal polarity and increase expression of secretory acinar cell markers (e.g., aquaporin5 (Aqp5)). Although these data are promising, secretory marker expression is still reduced compared to native glands²² and secretory acinar cells begin to express duct markers, including cytokeratins³⁰, likely due to cell dissociation-induced loss of cell-cell and cell-matrix interactions and associated stress induced cellular plasticity^{30, 31}. Indeed, survivability was severely compromised when isolated SGCs were immediately encapsulated within PEG hydrogels (e.g., prior to aggregation), suggesting that cell-cell contacts and cell-material interactions contribute to maintaining cell viability²³.

To increase cell viability upon hydrogel encapsulation immediately after isolation and avoid precipitous loss of secretory function during the 2-day aggregate formation, we investigated the isolation of intact acinar cell clusters and intercalated ducts (AIDUCs), an approach which maintains cell-cell contacts during isolation and includes encapsulation. We hypothesized that isolation of AIDUCs versus more rigorous single or multicellular isolation would reduce cell stress and encourage the organization of SGCs into more complex structures that resemble acinar units for use in salivary gland regenerative strategies.

Materials and Methods

Mouse Models:

All experiments were performed on tissue from mice maintained on the C57BL/6J background. Mist1^CreERT2 mice were produced as previously described³². R26^tdTomato(Gt(ROSA)26Sor^{tm9(CAG-tdTomato)Hze}/J) and C57BL/6J mice were purchased from Jackson Laboratories. Standard genotyping was performed. Prior to cell isolation, animals were housed on a 12-hour light/dark cycle with ad libitum access to food and water. Genotyping was conducted prior to weaning. All animal protocols were approved and carried out in accordance with the University of Rochester Medical Center’s University Committee on Animal Resources.

Norbornene-functionalized PEG synthesis.

Four-arm 20 kDa PEG was functionalized with norbornene (PEG-Norb) using N,N’-dicyclohexylcarbodiimide (DCC) coupling as previously described³³. Briefly, norbornene carboxylate (10 meq per PEG hydroxyl), and DCC (5 meq per PEG hydroxyl) were dissolved in dichloromethane (DCM) at room temperature for 30 min until a white precipitate (dicyclohexylurea) formed, which indicated formation of dinorbornene carboxylic acid anhydride. Four-arm PEG, pyridine (5 meq per PEG hydroxyl) and 4-dimethylaminopyridine (DMAP, 0.5 meq per PEG hydroxyl) was dissolved in DCM and then added dropwise to the anhydride solution, which was subsequently septa sealed, vented with a needle, and stirred overnight at room temperature.

The resulting solution was vacuum filtered and the filtrate was precipitated in chilled diethyl ether, collected by vacuum filtration, re-dissolved in 75 mL DCM, and re-precipitated two additional times in chilled diethyl ether. Structure and percent functionalization (>90%) were determined by ¹H-NMR using a Bruker Avance 400MHz spectrometer (¹H NMR[CDCL3]: d = 6.0–6.3 [norbornene vinyl protons, 8H, multiplet]), (3.5–3.9 PEG ether protons, 1817H, multiplet). The final product was dialyzed against distilled water (ddH₂O) for 3 days using 3500 MWCO dialysis tubing (Spectrum Laboratories) and lyophilized.

Peptide synthesis

The MMP-degradable peptide crosslinker GKKCGPQG↓IWGQCKKG was synthesized by standard solid-phase peptide synthesis on FMOC-Gly-Wang resin (EMD) using a Liberty 1 Microwave-Assisted Peptide Synthesizer (CEM) with UV monitoring, as previously described³⁴. All amino acids (AAPPTec) were dissolved in N-methylpyrrolidone (NMP) at 0.2 M. All amino acids were FMOC-protected, with cysteine and lysine side chains protected with trityl and tertbutyloxycarbonyl, respectively. 0.5 M Benzotriazole-N,N,N’,N’-tetramethyl-uronium-hexafluoro-phosphate (Ana-Spec) in dimethylformamide (DMF) was used as activator, 2 M diisopropropylethylamine (Alfa Aesar) in NMP was used as activator base, and 5% piperazine (Alfa Aesar) in DMF was used for deprotection. Peptides were cleaved and de-protected by adding the on-resin 0.25 mmol peptide (0.316 g resin) to a cleavage cocktail composed of 18.5 mL trifluoroacetic acid (TFA, Acros Organic), 0.5 mL triisopropylsilane (Sigma Aldrich), 0.5 mL ddH2O, and 0.5 mL 3,6 dioxa-1,8-octane dithiol and mixing for 2 h. Cleaved peptide was separated from resin via vacuum filtration, purified via precipitation in ice-cold diethyl ether (~180 mL), and collected by centrifugation. The peptide was resuspended in ice-cold diethyl ether, centrifuged twice more, dried under vacuum overnight, dialyzed against ddH₂O using 500 MWCO dialysis tubing (Spectrum Labs) for 48 h, and lyophilized.

Peptide molecular weight was verified using a Bruker AutoflexIII Smartbeam matrix-assisted laser desorption ionization time-of-flight mass spectrometer with 1 mg/mL peptide dissolved in 50:50 acetonitrile: ddH₂O + 1% TFA added in a 1:1 volumetric ratio of 10 mg/mL α-Cyano-4-hydroxycinnamic acid matrix (Tokyo Chemical Industry) dissolved in 50:50 acetonitrile: H2O + 1% TFA. Peptide standards (Bruker Peptide Calibration Standard, 206195) were used for calibration. Peptide purity was analyzed via 205 nm absorbance measured in ddH₂O with an Evolution UV/Vis detector (Thermo Scientific) and compared with predicted values using a previously published methodology^{22, 23}.

The photoinitiator LAP was synthesized in a two-step process, as previously described³³. Dimethyl phenylphosphonite (Acros Organics) was reacted with 2,4,6-trimethylbenzoyl chloride (Sigma-Aldrich) via a Michaelis-Arbuzov reaction. 3.2 g (0.018 mol) of 2,4,6-trimethylbenzoyl chloride was added dropwise to an equimolar amount of continuously stirred dimethyl phenylphosphonite (3.0 g) at room temperature and under argon. After 18 hours’ reaction, a four-fold excess of lithium bromide (6.1 g) in 100 mL of 2-butanone was added to the reaction mixture and then heated to 50 °C. After 10 minutes, a solid precipitate formed. The mixture was cooled to ambient temperature, allowed to rest for four hours and then filtered. The filtrate was washed and filtered 3 times with 2-butanone to remove unreacted lithium bromide, and vacuum dried overnight.

Isolation and culture of AIDUCs

Primary AIDUCs cells were isolated, as previously described, with minor modifications^{35, 36}. Briefly, female C57BL/6 mice aged 8–12 weeks were euthanized according to animal protocols approved by the University of Rochester University Committee for Animal Resources. SMGs were surgically removed, finely minced with a razor blade, and dissociated by incubation in 10 mL Hank’s Balanced Salt Solution (HBSS) (Gibco) supplemented with 0.5% BSA, 1M CaCl₂, 1M MgCl₂, 100 U/ml collagenase Type II (Gibco), and 1 mg/mL hyaluronidase (Sigma) at 37 °C for 60 min with gentle shaking. Cells were then passed through 100 μm filters to remove undigested tissues and then 20 μm filters to remove cell debris and red blood cells. The resulting cellular aggregates, sized 20–100 μm, were re-suspended in 5 mL complete media (Dulbecco’s Modified Eagle Medium (DMEM):F12 [1:1; GIBCO]) supplemented with Glutamine (2 mM, GIBCO), Antibiotic-Antimycotic solution (100 U/mL penicillin, 100 μg/mL streptomycin, GIBCO), N2 supplement (0.5×, Invitrogen), 2.6 ng/mL insulin (Life Technologies), 2 nM dexamethasone (Sigma), 20 ng/mL epidermal growth factor (EGF; Life Technologies), and 20 ng/mL basic fibroblast growth factor (bFGF; Life Technologies) and cultured at 37 °C.

Three days prior to dissection and AIDUC isolation from female Mist1^CreERT2;R26 ^tdTomato mice, mice were administered tamoxifen (0.25 mg/g body weight) to activate the expression of the tdTomato fluorescent protein in acinar cells. AIDUCs were isolated using the same approach described above.

Hydrogel encapsulation of primary AIDUCs

MMP-degradable PEG hydrogels were used to encapsulate AIDUCs (Fig. 1). Hydrogel precursor solutions were made by dissolving lyophilized norbornene-functionalized 4-arm PEG macromer (1.5 mM), MMP-degradable peptide (3 mM), 0.05 wt% of the photoinitiator LAP and 100 μg/ml laminin in PBS. The compressive modulus and swelling ratio of hydrogels were 5.4 ± 0.9 kPa and 2.0 ± 0.1, respectively. The hydrogel precursor solution was then added to freshly-isolated AIDUCs and mixed gently by pipetting until the cells were distributed homogenously in hydrogel suspension. The final cell density in the hydrogel solution was 5×10⁵ cells/mL from which 40 μL of the mixed solution was pipetted into 1 mL syringes with the tips cut off and photopolymerized via exposure to 5 mW/cm², 365 nm UV light for 5 min. Immediately after polymerization, hydrogels were placed in 24-well plates and incubated at 37°C with 5% CO₂. Media was changed daily.

Fig. 1. Schematic representation of hydrogel encapsulation of AIDUCs.

Freshly isolated AIDUCs were mixed with hydrogel precursor solution composed of norbornene-functionalized 4-arm PEG macromer, MMP-degradable peptide and the photoinitiator LAP and 100 μg/ml laminin, and 40 μL of the mixed solution was pipetted into 1 mL syringes with the tips cut off and photopolymerized via exposure to 5 mW/cm², 365 nm UV light for 5 min to form hydrogels.

Cell viability of hydrogel encapsulated AIDUCs

To evaluate the survivability of encapsulated AIDUCs, the LIVE/DEAD assay (Invitrogen) was utilized according to manufacturer’s instructions with imaging using a Olympus FV1000 laser scanning confocal microscope. Prestoblue (Thermo Fisher Scientific) was used to evaluate encapsulated cell metabolic activity to complement LIVE/DEAD imaging. At designated time points, 10% Prestoblue in cell culture media was added to the cell-hydrogel and incubated for 20 min at 37 °C. Metabolic fluorescent conversion was measured with a Biotek plate reader under the excitation wavelength of 560 nm and emission of 590 nm.

Gene expression analysis

To extract RNA from the hydrogels, hydrogels were washed with PBS twice at designated time points and then transferred to 1.5 mL Eppendorf tube filled with ~500 uL 1000 U/ml collagenase II solution. After incubation at 37 °C for ~30 minutes until hydrogels were degraded, the cell pellet was collected via centrifuge at 1500 RPM for 5 min and re-suspended in 1 mL TRIzol (Invitrogen), and stored at −80 °C until RNA extraction. Total RNA was extracted using the OMEGA kit (Omega Bio-Tek). RNA concentration and quality was determined using a Nano-Vue^™ UV-Vis spectrophotometer (GE Healthcare).

cDNA was synthesized using the iScriptTM cDNA synthesis kit (BioRad) according to the manufacturer’s instructions. Quantitative PCR analysis of individual cDNAs was performed on a CFX96^™ Real-Time System (Bio-Rad) using PowerUp SYBR^™ Green Master Mix (Bio-Rad) for genes listed in Supplemental Table 1. Quantitative PCR results were normalized to mouse Rps29 mRNA levels and analyzed using the 2^−⊿⊿CT method.

Immunohistochemistry staining

After culture for 14 days, hydrogels were fixed with 4% paraformaldehyde (PFA) for 30 min at room temperature. Hydrogels were then washed 3 times with PBS and embedded in Optimal Cutting Temperature solution (OCT, Tissue-Tek). Then, the hydrogels were cryosectioned using a CM1860 UV cryostat (Leica) into 10 μm sections and mounted on SuperFrost Plus slides (Fisher Scientific). Staining was then performed as previously described^{22, 36}. The primary and secondary antibodies used are listed in Supplemental Table 2. All samples were imaged using a FluoView FV1000 Laser Scanning Confocal Microscope (Olympus).

5-Ethynyl-2’-deoxyuridine (EdU) staining

At designated time points, SGm that formed within hydrogels were treated with 10 μM EdU (Thermo Fisher Scientific) for 3 h prior to fixation with 4 % PFA for 30 min, washed twice with DPBS and permeabilized with 0.5 vol% of Triton X-100 for 30min. Samples were washed 2× with 3 wt% BSA in DPBS and labeled with the Click-iT® Plus EdU AlexaFluor® 488 imaging kit (Thermo Fisher Scientific) and DAPI (Thermo Fisher Scientific) as a nuclear counterstain. Hydrogel samples were imaged using a FV1000 laser scanning confocal microscope (Olympus).

Measurement of stimuli-responsive intracellular Ca²⁺ flux

To evaluate the function of SGm within hydrogels, intracellular Ca²⁺ flux was measured in response to agonist stimulus as previous described³⁶. To ensure fast diffusion of the ATP and CCh and immediate stimulation of SGm after addition, SGm were isolated from the hydrogel before calcium signaling experiments. Isolation also decreases hydrogel background during intracellular calcium flux imaging. Briefly, SGm were dissected from the hydrogel under microscopy at designated time points and incubated with 5 μM Fura-2 acetoxymethyl ester (AM) for 30 min at 37 °C after washing with PBS. Then 100 nM Carbachol or 100 μM ATP was added to stimulate the spheres for 180 s and the total duration time for Ca²⁺ measurement was 600 s. Intracellular Ca²⁺ flux was recorded using an inverted microscope (Nikon Diaphot 200) equipped with an imaging system (Till Photonics, Pleasanton, CA). The fluorescence ratio of 340 nm over 380 nm was calculated, and all data are presented as the change in fluorescent ratios.

Statistical analysis

All graphs are presented as mean ±standard deviation. At least three biological replicates were included in each group. Statistical significance between two groups was determined using Prism software (GraphPad) with p < 0.05 as the threshold for statistical significance unless otherwise indicated in legends. Statistics are analyzed using one-way ANOVA with Dunnett’s post-hoc test.

Results

Isolation of primary AIDUCs

Our previous work established that primary SGCs dissociated to either single or multiple cellular aggregates exhibit low viability when directly encapsulated within hydrogels^{22, 23}. To counter the loss in viability, cells were previously cultured for 2 days under non-adherent conditions to form aggregates before hydrogel encapsulation. The viability of cells within the aggregates, which included both acinar and duct cells³⁷, was significantly increased, suggesting that cell-cell interactions support survivability. Although the aggregates promoted viability, proliferation, and even exhibited evidence of cellular organization including apico-basolateral organization of acinar markers, there was a loss of acinar cell phenotype during aggregate formation with evidence for acinar transition to ductal phenotypes^{22, 30}.

To maintain differentiated acini, a milder dissociation protocol was adapted to isolate AIDUCs^{35, 36}, and the size distribution of the freshly isolated AIDUCs, analyzed using light microscopy, was 20–100 μm (Fig. S1). While comprised of both acinar and duct cells (Fig. S2A), most cells within AIDUCs are acinar cells, as indicated by the positive staining of acinar cell makers NKCC1 and AQP5 (Fig. S2A–C). Furthermore, cell-cell contacts were also investigated via immunostaining. The positive staining of ZO-1 and E-Cadherin indicated that cell-cell contacts were successfully maintained in AIDUCs (Fig. S2 B,C). Additionally, high cell viability (>95%) was achieved immediately following isolation (Fig. S3 A,B). After 2 days of culture under non-adherent conditions, unencapsulated AIDUC controls also formed aggregates (Fig. S3 C,D).

Formation of SGm within hydrogel microenvironments

AIDUCs were encapsulated in PEG hydrogels immediately after isolation and the viability of encapsulated AIDUCs was analyzed using LIVE/DEAD imaging at days 0, 4, 7, and 14 (Fig. 2A–D). In contrast to the low viability of cells encapsulated directly after isolation in our previous report²³, high cell viability was observed over 14 days post-encapsulation. The same protocol was used to isolate human AIDUCs, which were subsequently encapsulated within PEG hydrogels. Our data show that encapsulated human AIDUCs maintain viability and self-assemble into SGm within the hydrogel (Fig. S4).

Fig. 2. Cell viability of hydrogel encapsulated AIDUCs.

LIVE/DEAD imaging of encapsulated AIDUCs at day 0 (A), day 4 (B), day 7 (C), day 14 (D) and blocked with ZO-1 (E) and E-Cadherin (F) antibody. Scale bars = 800 μm. (G) Quantification of sphere size at different time points after encapsulation. Viability of hydrogel encapsulated AIDUCs (H). ** p < 0.01, **** p < 0.0001 compared with day 0 determined using one-way ANOVA with Dunnett’s correction for multiple comparisons test. N = 4.

To investigate the importance of cell-cell contacts on cell survivability, AIDUC interactions were blocked with antibodies against ZO-1 or E-Cadherin for 30 mins prior to encapsulation, resulting in few live cells at day 4 (Fig. 2E,F), indicating that preservation of cell-cell contacts directly contributes to cell survival after AIDUC encapsulation.

Over the 14-day culture period, there was a significant increase in SGm size. SGm expanded to ~100 ±19 μm by day 7 and continued to increase in size through day 14 to form structures of ~135 ±25 μm (Fig. 2G). Cell metabolic activity, a surrogate for proliferation, also increased 8-fold over 14 days, suggesting that proliferation may play a part in the observed increases in SGm size (Fig. 2H). We further stained for Ki67, which is a nuclear protein present at all stages of cell division (G1, S, G2, and mitosis) but not during the resting stage (G0)³⁸. Abundant nuclear Ki67 was observed among peripheral cells of the SGm, again suggesting that cell proliferation within the SGm may play an important role in tissue development and organization during hydrogel culture (Fig. 3E).

Fig. 3. Cavity formation within AIDUCs involves apoptosis.

A representative phase (A) and fluorescent LIVE/DEAD (B) image of an acinar-like structure in hydrogel after culture for 14 days. Immunostaining of cell apoptosis marker, cleaved caspase 3, in SGm after 7 days (C) and 14 days (D) culture in hydrogels. (E) Proliferation of AIDUCs cultured in hydrogels assessed by Ki67 immunofluorescence staining. DAPI is used to stain nuclei.

To further investigate the underlying mechanism of increased tissue size, imaging was performed. Fluorescent and brightfield images of the SGm at day 14 are shown in Fig. 3A–B, some dead cells (red, Fig. 3B) were observed in the center of the structures.

Selective apoptosis at the center of a developing cell mass is essential to glandular organogensis³⁹. To determine whether our SGm cultures developed in a similar fashion, cleaved caspase 3, an indicator of cell apoptosis, was stained. While the number of dead cells is relatively low in Fig. 3B, positive staining for cleaved-caspase-3, observed at the center of SGm at day 7 and 14, respectively (Fig. 3C, D), confirms the presence of apoptotic cells. Cells located in the interior of the aggregates do not have the same cell-cell contacts as the polarized epithelial surface cells, which may result in cell death. Previous studies by other groups have also observed dead cells and positive caspase-3 staining in the center of the acini-like structures when culturing salivary gland progenitor cells in various hydrogels^{20, 27, 40–43}. Cavities with varying number and size were observed by day 14 (Fig. 3D, E), but they cannot be designated as lumens, as it is unclear how or when they are formed and whether they are polarized differently from the outer cell layer. We speculate that cells on the inside of the spheres may undergo apoptosis because they are excluded from cell-cell contacts found on the outside of the sphere.

SGm formed in hydrogels maintain expression of acinar cell markers

To further evaluate the effect of hydrogel encapsulation on acinar cell phenotypic maintenance, real time PCR was used to determine expression levels of acinar cell-specific markers. Significant decreases in Mist1, Nkcc1, Aqp5, and Ip3r3 expression were observed at day 14 when compared to freshly isolated AIDUCs (day 0) (Fig. 4A–D), consistent with our previous findings^{22, 30, 36}. Specifically, expression was maintained at 1%, 4.7%, 0.22%, and 73%, respectively, compared with day 0 AIDUCs. Additionally, expression of duct and myoepithelial cell markers, cytokeratin 7 (K7) and cytokeratin 5 (K5), showed robust increases at day 14 of 5.9- and 9-fold higher versus day 0 levels (Fig. 4E, F). The myoepithelial marker, smooth muscle actin (Sma), exhibited only 13% expression at day 14 compared to day 0 level (Fig. 4G).

Fig. 4. Hydrogel encapsulated SGm express markers of all three major cell types in the salivary gland (acinar, duct and myoepithelial).

Quantitative PCR was used to measure gene expression of acinar cell markers Mist 1 (A), Nkcc1 (B), Aqp5 (C) and Ip3r3 (D), duct cell markers K7 (E) and K5 (F) and myoepithelial cell marker Sma (G) relative to the reporter gene Rps29 (ribosomal gene). Statistics are relative to Day 0, using one-way ANOVA with Dunnett’s post-hoc test. *p < 0.05, **p < 0.01, *** p < 0.001 and **** p < 0.0001. N=4.

Hydrogel encapsulated SGm express salivary gland cell markers

The salivary gland is composed of acinar, duct, and myoepithelial cells. To characterize the morphology of SGm encapsulated in hydrogel, specific markers of these 3 cell types were examined using immunostaining. NKCC1 is a marker of acinar cells and localized at the basolateral membrane of acinar cells (Fig. 5A,B). NKCC1 remains basolaterally localized in the SGm encapsulated in hydrogels at day 14 (Fig. 5E,F). K7 and K5, which are duct and myoepithelial cell markers, were largely expressed at the periphery of the tissue mimetics (Fig. 5E,G), similar to the gland tissue (Fig. 5A,C). SMA is a myoepithelial cell marker, which showed staining that was peripherally localized in the aggregates (Fig. 5F). These data suggest that all three cell phenotypes responsible for gland function are present in SGm within hydrogels.

Fig. 5. Characterization of SGm with immunostaining at day 14.

Immunochemistry staining of protein expression in intact salivary gland tissue (A-D, I-L) and in SGm (E-H, M-P). Staining for acinar cell markers NKCC1 (A, B, E, F), AQP5 (D, H), IP3R3 (I, M), Duct cell markers K7 (A, E) and K5 (C, G), and myoepithelial cell marker αSMA (B, F), basement membrane proteins, Laminin (K,O) and Collagen IV (L, P), and tight junction protein ZO-1 (J, N). Scale bars = 40 μm.

AQP5, which is the primary water channel expressed on the apical membrane of acinar cells, plays an important role in secretion of fluid into the lumen of ducts by activation of muscarinic acetylcholine receptors. AQP5 staining is apically localized within SGm at day 14 (Fig. 5H). Inositol 1,4,5-triphosphate receptor 3 (IP3R3), a regulator of intracellular Ca²⁺ signaling, localizes at the apical membrane of acinar cells and staining of hydrogel-encapsulated SGm suggested similar spatial expression with luminal dominance (Fig. 5M).

Cell-cell junctions and apicobasal distribution of channel proteins are necessary to establish cell polarity and to maintain functional secretory cells⁴⁴. To further assess epithelial polarization, the spatial distribution of tight junction proteins was examined through immunohistochemistry. Tight junctions (TJs) act as intermembrane barriers, preventing lateral movement of apically and basolaterally localized membrane proteins as well as the leakage of ions. ZO-1, a component of the TJ protein complex, is localized in the apical region of acinar and duct cells. ZO-1 staining suggested tight junction formation and apicobasolateral polarity especially within large, isolated cavities formed by day 14 in hydrogels (Fig. 5N). Basement membrane proteins laminin and collagen IV play important roles in promoting apicobasal polarity⁴⁵ and are basolaterally localized in the gland (Fig. 5K,L) and in SGm at day 14 (Fig. 5O,P). Protein expression shown via immunostaining (Fig. 5 E–H, I and Fig. 6F, H) further supported the observed trends in gene expression (Fig. 4 and Fig. 6A, B).

Fig. 6. Hydrogel encapsulated SGm retain expression of secretory proteins.

Gene expression of the secretory proteins Amy1 (A), Pip (B), Muc5b (C) and Lyz2 (D) determined using qPCR. Immunohistochemical staining of Amylase (E,F) and PIP (G,H) in intact gland and SGm at day 14. Scale bars = 40 μm. All gene expression data is relative to day 0 with ribosomal gene Rps29. Statistics are relative to Day 0, using one-way ANOVA with Dunnett’s post-hoc test. *p < 0.05, and *** p < 0.001. N=4.

SGm encapsulated in hydrogels maintain secretory protein expression

To confirm the secretory function of the encapsulated SGm, the expression of secretory protein markers was analyzed. qPCR was used to detect prolactin-inducible protein (Pip), amylase (Amy1), mucin 5b (Muc5b), and lysozyme 2 (Lyz2). Relative to day 0 isolated AIDUCs, Amy1 expression was decreased to 5% at day 7 but rebounded to 15% by day 14 (Fig. 6A), and Pip expression was decreased to 0.002% at day 7 but increased to 0.2% by day 14 (Fig. 6B). Another secretory protein, Muc5b was decreased to 22% of day 0 controls at day 7 but expression increased to 35% by day 14 (Fig. 6C). Additionally, 24-fold up-regulation of Lyz2 expression was detected at day 14 compared to day 0 tissue (Fig. 6D), which may implicate a hypoxia-related stress response⁴⁶.

Immunohistochemical staining was also used to characterize the expression of secretory proteins. Despite significantly reduced mRNA levels, protein expression of Amy1 (Fig. 6E,F) and PIP (Fig. 6G,H) was observed in intact gland and in day 14 SGm, confirming that secretory marker expression is well maintained in SGm.

Secretory function is maintained in SGm encapsulated in hydrogels

The muscarinic acetylcholine receptor 3 (M3r) is the major neurotransmitter receptor involved in salivary fluid secretion from acinar cells. Acinar cells also express several types of purinergic receptors, including P2Y₂, a G protein-coupled protein receptor⁴⁷. Activation of P2Y receptors leads to an increase in intracellular [Ca²⁺], which induces fluid secretion. The expression of these receptors in SGm was interrogated using qPCR. The expression of M₃r decreased to 15% and 21% of primary AIDUCs at day 7 and day 14, respectively, which is consistent with acinar cell marker expression (Fig. 7A). Immunostaining was used to detect M3R expression in SGm (Fig. 7C). The expression of P2Y₂ showed an increasing trend at both days 7 and 14, however, these increases are not significant due to data variability (Fig. 7B). TMEM16A plays a critical role in encoding the apical Ca²⁺-activated Cl⁻ channel (CaCC) efflux pathway required for Ca²⁺-dependent fluid secretion^{48, 49}. The robust expression of M3R and TMEM16A (Fig. 7D) indicate that the secretory function is successfully maintained in SGm.

Fig. 7. SGm in hydrogels are responsive to stimulation with muscarinic agonist carbachol (CCh) and purinergic agonist ATP.

Gene expression profiles of muscarinic receptor type 3 (M₃r, A) and purinergic receptors (P2y₂, B), in SGm at days 7 and 14. Immunostaining of M₃r (C) and Tmem16A (D) at day 14 after encapsulation. Scale bar = 40 μm. Fluorescent calcium flux images of SGm before (E, I), and after (F, J) stimulation with 100 nM CCh and 100 μM ATP, and removal of stimuli (G, K). Quantitative analysis of Ca²⁺ signaling in SGm stimulated with CCh (H) and ATP (L). Dashed lines indicate that the stimuli were added at 60 s and removed at 240 s. All gene expression data is relative to day 0 with ribosomal gene Rps29. Statistics are relative to Day 0, using one-way ANOVA with Dunnett’s post-hoc test. **** p < 0.0001. N=4.

Calcium signaling plays a central role in saliva secretion and function. Functionality of the hydrogel-cultured SGm was assessed by analyzing intracellular calcium release after stimulation by muscarinic and purinergic agonists. Robust calcium response was observed after addition of the muscarinic agonist carbachol (CCh, Fig. 7E–H, Suppl. Video 1) and purinergic agonist ATP at 60s (Fig. 7I–L, Suppl. Video 2). CCh stimulates the M3 muscarinic acetylcholine receptor on acinar cells, leading to increased intracellular free calcium concentrations ([Ca²⁺]i) while ATP is an agonist of purinergic receptors on both acinar and duct cells. Calcium response gradually diminished after removal of CCh/ATP at 240s (Fig. 7G,K). Overall, robust agonist-stimulated calcium signaling suggested that hydrogel-cultured SGm maintained secretory functions.

Hydrogel encapsulated SGm exhibit acinar-to-ductal metaplasia

To lineage trace acinar cells, the tdTomato reporter was activated by inducing Mist1^CreERT2;R26 ^tdTomato mice with tamoxifen 3 days prior to AIDUC isolation. Freshly isolated AIDUCs were encapsulated in hydrogels and cultured for 14 days in vitro before being fixed with 4% PFA for immunostaining. Both tdTomato⁺ acinar cells and non-labeled duct cells are present in freshly isolated AIDUCs (Fig. 8A), AQP5 is apically located on the tdTomato⁺ acinar cells and the majority of the cells are positive for NKCC1 (Fig. 8B,C). Quantification of the images showed that approximately 94% of acinar lineage cells co-expressed NKCC1.

Fig. 8. Salivary gland acinar cells exhibit plasticity after encapsulation in the hydrogels.

Immunostaining of freshly isolated AIDUCs from Mist1^CreERT2/R26 ^tdTomato mouse with antibody to RFP (A), NKCC1 (B) and AQP5(C). EdU staining to evaluate the proliferation of SGm at day 14 (D). Arrowheads indicate cells from acinar lineage that are co-localized with EdU. Expression of NKCC1 and AQP5 is dysregulated in SGm encapsulated within hydrogels. Immunofluorescence staining of SGm with antibodies to AQP5 (E-H) and NKCC1 (I-L) shows co-localization with tdTomato-positive acinar cells at day 14. Arrowheads indicate cells from acinar lineage that are positive for NKCC1 (L), and AQP5 (H). Acinar lineage cells in encapsulated SGm activate expression of keratins. SGm in hydrogel at day 14 stained with antibodies to tdTomato (RFP) and K7 (M-P). SGm in hydrogel at day 14 stained with antibodies to tdTomato (RFP) and K14 (Q-T). Arrowheads indicate cells from acinar lineage that are co-localized with K7 (P), and K14 (T). Nuclei are stained with DAPI. Scale bar =20 μm.

Cell proliferation contributes to SGm size increase after encapsulation in the hydrogels, as indicated by positive Ki67 staining (Fig. 3E). However, it is still unclear which cells contribute to the proliferation. EdU staining was used to assess the cell proliferation within hydrogel encapsulated SGm (Fig. 8D). While imaging suggested acinar cells (red) persist in culture for up to 14 days and both acinar (arrowhead) and non-acinar cells are proliferating, most of the proliferation occurred in non-acinar cells (Fig. 8D).

To further investigate the fate of acinar cells following hydrogel encapsulation for 14 days, co-expression of tdTomato with AQP5 (Fig. 8E–H) and NKCC1 (Fig. 8I–L) was analyzed using immunofluorescence staining. Approximately 25% acinar lineage cells co-expressed AQP5 (Fig. 8H, arrowheads) while ~67% co-localized with NKCC1⁺ cells (Fig. 8L, arrowheads).

Co-expression of tdTomato with duct cell markers K7 and K14 was also assessed. Immunostaining showed ~67% acinar lineage cells co-localized with K7⁺ cells (Fig. 8M–P, arrowheads). For K14⁺ cells, ~38% acinar lineage cells co-expressed K14 (Fig. 8Q–T, arrowheads). These results indicated that hydrogel encapsulated AIDUCs underwent some acinar-to-ductal metaplasia over time based on the expression of duct cell-specific markers by cells of the acinar cell-lineage.

Discussion

Maintenance or development of functional salivary gland units for transplantation within a biocompatible hydrogel matrix that enables integration into damaged glands is a key step toward creation of a functional tissue-engineered salivary gland. In this study, functional salivary gland tissues were developed ex vivo by entrapping primary AIDUCs within engineered extracellular matrices, instead of pre-formed aggregates or singular SGCs used in our previous studies^{22, 23, 30}. Our data show that AIDUCs encapsulated within MMP-degradable PEG hydrogels form acinar-like units with phenotypic and functional characteristics consistent with the native salivary gland. Hydrogel-encapsulated AIDUCs showed long-term cell viability, maintained expression of acinar-related genes and proteins, and exhibited polarized localization of basement and secretory membrane proteins, as well as cellular organization. Remarkably, hydrogel-cultured SGm also retained the ability to respond to Ca²⁺ agonists.

Cell-cell interactions were found to be particularly critical for survival and function of encapsulated AIDUCs. When cell-cell contacts in freshly isolated AIDUCs were blocked with antibodies against ZO-1 or E-Cadherin, survivability was significantly reduced. Cell-cell interactions occur via specialized adhesion junctions and regulate a variety of functions in multicellular organisms, including survival, differentiation, barrier formation, tissue function and signal transduction^50–52. The absence of cell-cell contacts in dispersed primary SGCs renders them unable to survive, proliferate, self-assemble, and re-aggregate to form functional acinar units²³. By utilizing a milder cell isolation procedure, cell-cell contacts and apicobasolateral organization was well preserved in AIDUCs, contributing to cell viability and proliferation and function^{23, 52, 53}.

Long-term maintenance of the acinar cell phenotype in vitro is a challenge in salivary gland tissue engineering approaches. Previous studies have shown that 3D culture of SGCs in various hydrogels improved maintenance of acinar cell marker expression, including α-amylase, Aqp5, Mist1, and Nkcc1 compared with traditional 2D culture^{22, 24, 54–59}. Nevertheless, significant reductions in acinar marker expression of Mist1, amylase, and Aqp5 are still observed. By encapsulating AIDUCs in our hydrogel system, a 1.1-fold and 2.1-fold increase of Mist1 and Nkcc1 expression level is observed at day 14 compared to our previous studies^{22, 30}. Despite this advantage, the overall acinar gene expression level is still lower than native tissue. However, IHC staining of acinar cell phenotypic and functional protein markers, including NKCC1, AQP5, PIP and Amylase, showed consistent expression with intact glands, which is similar to previous studies using other hydrogel chemistries^{19–22, 27, 30, 54–57} and indicates the importance of the 3D environment of ECM on acinar cell phenotypic maintenance. Calcium signaling plays a central role in saliva secretion and function. SGm removed from hydrogels were tested for CCh and ATP responsiveness using calcium flux analysis. Robust calcium flux was observed upon stimulation with both agonists, indicating successful maintenance of secretory function, similar to previous findings^{24, 59–61}. These data indicate the advantage of well-maintained cell-cell contacts in AIDUCs, as well as the engineered ECM, in maintaining acinar cell phenotype and secretory function.

Acinar-to-ductal metaplasia has been described for both primary cultured pancreatic and salivary gland acinar cells^{30, 31}, and is accompanied with loss of Mist1 expression and increased expression of duct-specific genes^{31, 62, 63}, resulting in cellular disorganization, changes in polarity and loss of secretory function^{63, 64}. Freshly isolated AIDUCs are comprised of both acinar and duct cells, with most cells of acinar origin (>90%, Fig. S2). However, after encapsulation in hydrogels for 14 days, both gene expression data and lineage tracing of Mist1 in encapsulated SGm showed decreased expression levels, and increased expression of duct cell markers K5 (~9 fold) and K7 (~5.9 fold), which is consistent with the phenomenon of acinar-to-ductal metaplasia induced by acute injury or stress^{30, 31, 65}. However, this study reports reduced metaplasia compared to our previous study³⁰, which might be due to the preservation of cell-cell contacts in AIDUCs that relieved stress associated with cell isolation and promoted self-aggregation and tissue function of SGm.

Reduction or inhibition of acinar-to-ductal metaplasia would benefit the long-term in vitro culture of SGCs, maintenance of acinar cell phenotype, and development of cell-transplantation therapies for salivary gland regeneration. Inhibition of Mist1 transcriptional activity in pancreatic acinar cells leads to activation of duct-specific genes, such as K7, K19, and K20⁶³. However, the incidence of acinar-to-ductal metaplasia was significantly attenuated by forced expression of Mist1⁶⁶. Small molecules such as Baicalein and GRP78 have been shown to inhibit the acinar-to-ductal metaplasia process by modulating the inflammatory microenvironment or reducing EGFR expression and blocking downstream RAS and AKT signaling pathways^{67, 68}. Numb, κB-Ras, and Ral GTPases were also reported to regulate acinar-to-ductal metaplasia by inhibiting Notch and EGFR/KRas signaling^{69, 70}. Suppression of acinar-to-ductal metaplasia using functional molecules to block specific signaling pathways, such as EGFR, TGF-β1 receptor, and ROCK may also prove useful to maintain salivary gland acinar cell phenotype and function in vitro^{68, 71–73}.

Further studies can be undertaken to further engineer the hydrogel matrix to enhance maintenance of the acinar cell phenotype and to suppress acinar-to-ductal metaplasia. For example, the mode of hydrogel degradation has a significant impact on the organization and phenotype maintenance of SGm^{22, 23}, MMP-degradable hydrogels enhanced tissue polarization and acinar cell marker expression through cellular remodeling based on controls of hydrolytically-degradable or non-degradable hydrogels²². Using these hydrogels, we have recently developed an engineered salivary gland tissue chip, which can be used for high-throughput and high-content screening of potential matrix cues and soluble cues that may further promote salivary gland acinar cell phenotype maintenance³⁶. The rational design of hydrogel ECM degradable properties is likely to benefit functional outcomes of in vitro cultured SGCs. Numerous ECM proteins, especially basement membrane proteins, including laminin, collagen IV, and perlecan/HSPG2, are implicated in salivary gland development^{40, 45, 74, 75}. SGCs cultured on fibrin hydrogel surfaces in the presence of laminin-111 protein and peptide derivatives, such as YIGSR, A99, RGD, and KP24, promoted salivary gland growth, lumen formation, and expression of secretory proteins^{24, 76–79}. Soluble factors, including neurotrophic factors (neurturin and glial derived growth factor)^80–84 and FGFs (FGF2, FGF7 and FGF10)^{11, 73, 85–87}, have been reported to prevent stress responses and promote salivary gland development and acinar cell phenotype maintenance. Additionally, the study of matrix cues to maintain salivary gland function will be another topic of future investigations.

The in vivo environment is very complex compared to in vitro. Various soluble factors, including growth factors, small molecules, and cytokines, affect fate of the encapsulated cells and degradation of the implanted hydrogel. In our previous research, various hydrogels have been successfully used to transplant cells in vivo for tissue regeneration^88–94, including MMP-degradable and hydrolytically degradable hydrogels, and contributions to tissue regeneration were promising. MMP-degradable PEG hydrogel-based tissue engineered periosteum (TEP) improved bone allograft healing by promoting rapid neurovascularization and hard bone callus formation versus hydrolytically degradable TEP modified allografts⁸⁹. MMP-degradable hydrogels have also been used by others for tissue regeneration and promoted tissue integration after implantation^95–98. Together, these data demonstrate that hydrogels are a promising platform for regeneration of functional salivary gland tissue in vivo. Allograft or xenograft transplantation is currently feasible using our approach. Unfortunately autografts are infeasible, as our current hydrogels do not sustain tissue function commensurate with the time frame over which fractionated irradiation treatments for head and neck cancers occur, as would be required based on tissue isolation prior to treatment/damage. Nevertheless, with continued development it may be possible to support longer term function, enabling an autograft approach

Conclusion

This study demonstrates that cell-cell interactions together with encapsulation within MMP-degradable PEG hydrogels can be used to maintain salivary gland tissue phenotype and function in vitro. We show that encapsulated AIDUCs exhibited long-term cell survivability and reorganized into acinar-like units composed of well-differentiated functional acinar, ductal, and myoepithelial structures. 3D-assembled SGm expressed salivary phenotypic markers and TJ proteins. They also exhibited secretory function as indicated by amylase and Pip staining and intracellular calcium flux, as well as reduced acinar-to-ductal metaplasia compared to our previous studies. These results highlight the utility of the MMP-degradable hydrogel platform, as well as the importance of maintaining cell-cell contacts in AIDUCs for the acinar cell phenotype and functional maintenance. In sum, PEG hydrogels are a promising platform for further investigation into the interactions required for regeneration of functional salivary gland tissue.