4.5.1. ICC and Western Blot of iSGECs in Monolayer Culture

The cytokeratins KRT8, KRT18, and KRT19 are cytoplasmic proteins expressed in the salivary gland epithelium and are characterization markers for ductal and acinar cells [29,30]. We first characterized iSGECs by KRT8/18/19 protein expression and found iSGEC-nSS1 expressed all three epithelial markers more intensely and uniformly than iSGEC-pSS1 and -nSS2 (Table 2, Supplementary Figure S3). KRT19 protein localized within cells containing a larger cytoplasm and were dispersed throughout iSGEC-nSS2 cultures, whereas smaller cells demonstrated a low expression or a lack thereof (Supplementary Figure S3). KRT18 was the most uniform and ubiquitously expressed cytokeratin among iSGECs, whereas KRT8 exhibited focal protein expression in iSGEC-nSS2 (Supplementary Figure S3). Heterogenous expression of KRT19 and KRT8 could indicate multiple stages of differentiation among cells [26]. Overall, the expression of KRT8, KRT18, and KRT19 indicates these cell lines to be of ductal origin and reflects other long-term cultures of SGECs in both mice and humans [5,30,31].

Acinar cell markers, AQP5 and AMY1A, were expressed in all three iSGECs, and highest in both iSGEC-pSS1 and -nSS2. Within these cultures, cells in close contact expressed the highest levels of AQP5, unlike AMY1A where intercellular contact or proximity to other cells did not appear to have an influence. We observed an increase in AQP5 protein expression in both p-14 and p-80 iSGEC-nSS2 cultures when supplemented with 1.2 mM Ca2<sup>+</sup>. Increased AQP5 protein was determined by ICC and replicated by Western blot using whole-cell lysates (Supplementary Figure S3, Figure 6A). ZO-1 expression was low in all iSGECs when accessed by ICC and confirmed by Western blot in whole-cell lysates. iSGEC-nSS1 exhibited the lowest expression of ZO-1 and CDH-1 and higher expression of VIM in monolayer culture, which reiterates the observed mesenchymal phenotype [25].

**Figure 6.** *Cont*.

**Figure 6.** Protein expression of acinar and epithelial cell markers byWestern blot in iSGEC-nSS2. **Legend.** (**A**) Protein expression of acinar (AQP5) (**left**) and epithelial (ZO-1) (**middle**) markers in whole-cell lysate of monolayer cultured iSGEC-nSS2. (**left**) Expression of AQP5 in iSGEC-nSS2 whole-cell lysate compared to HMC-3A, which was used as a (+) control for AQP5 in Western blots. (**right**) After stimulation with 10 μM epinephrine for 45 min ((+) Epi) cell culture supernatant was replaced, and cells were grown for 72 h before media were collected and concentrated. Cells without stimulation by media supplementation with 10μM epinephrine ((−) Epi) served as the control. Secreted α-amylase was detected in both stimulated ((+) Epi) and unstimulated ((−) Epi) cultures. 10μM epinephrine stimulated ((+) Epi) iSGEC-nSS2 cultures exhibited higher levels of α-amylase in cell culture supernatant than unstimulated cultures after 72 h. Western blots were normalized to vinculin expression in whole-cell lysate of their respective sample. (**B**,**C**) Media were harvested after stimulation with 10μM epinephrine for 45 min and amylase activity measured by colorimetric assay. (**B**) Supernatant of monolayer cultured iSGECs treated with 10μM ((+) Epi) had significantly higher α-amylase activity compared to untreated cells ((−) Epi). (C) Comparison of early (p-14) and late (p-80) passaged iSGEC-nSS2 secretion of α-amylase into media when cultured on matrigel. Results are listed as fold-change over the untreated control. Significance of differences between means was determined by Student's *t*-test and *p*-values corrected using the Holm–Šidák multiple comparisons test (alpha = 0.05). Error bars represent mean +/− SEM. \* *p* < 0.05.

In the case of iSGEC-nSS2, α-SMA was not detected in either p-14 or p-80 monolayer cultures as compared to iSGEC-pSS1 and -nSS1, where it was expressed at very low levels (Supplementary Figure S3). Pringle et al. demonstrated low α-SMA expression in cultured salivary gland stem cells (SGSCs) likely originating from intercalated and striated ductal cells, which could explain these results [3]. Vimentin is a marker for mesenchymal cells within the salivary glands [32,33]. Vimentin was expressed sporadically in iSGEC-nSS2 monolayer culture and could indicate a small population of mesenchymal cells or epithelial cells reverting to a mesenchymal-like state within cultures (Supplementary Figure S3).

In all three iSGEC lines, α-amylase was secreted into the cell culture medium after β-adrenergic stimulation by epinephrine (Figure 6B). Additionally, iSGEC-nSS2 early (p-14) and late passages (p-80) secreted α-amylase into cell culture media with and without β-adrenergic stimulation (Figure 6B). iSGECs-nSS2 p-80 secrete α-amylase, demonstrating retention of pro-acinar cell characteristics after significant cellular expansion. Moreover, iSGEC-nSS2 cells cultured on matrigel formed spheroid structures and secreted α-amylase in response to 10 mM epinephrine stimulation (Figure 6C). An overall increase in α-amylase secretion when compared to unstimulated cells was observed in iSGEC-nSS2 and could indicate a stricter regulation of secretion within the differentiated cells.

#### 4.5.2. 3D-Matrigel Spheroid Culture

Matrigel contains a mixture of basement membrane and extracellular matrix proteins, which have been shown to promote SGEC and salivary gland stem cell (SGSC) differentiation [3,11,34,35]. SGECs plated on matrigel form spheroids resembling a basic acinus structure, composed of differentiated

acinar and ductal cells [3,34,35]. iSGECs began to form spheroids within 24 h of plating on matrigel (2 mg/mL) and were cultured for a total seven days. ICC–IF protein expression analysis of iSGEC-nSS2 spheroid cultures showed high localization of acinar and epithelial cell markers ZO-1, AQP5, and AMY1A within acinus-like spheroid structures (Figure 7). Cytokeratins KRT8, KRT18, and KRT19 were all expressed in iSGEC-nSS2 cultured on matrigel (Figure 7). We observed greater levels of KRT8 expression within interior cells of spheroids similar to previously described high expression in luminal cells [36]. At a matrigel concentration of 2 mg/mL, some cells formed monolayer sheets across the matrigel that expressed KRT18 and KRT19 at high levels (Figure 7). α-SMA protein expression was localized to myoepithelial-like cells located on the spheroid exteriors.

mRNA expression analysis by qRT-PCR of iSGEC-nSS2 matrigel cultures demonstrated an increase in pro-acinar cell expression of AQP5 compared to monolayer cultures (Figure 8A,C). Spheroids from late passaged iSGEC-nSS2 expressed AMY1A and AQP5 at levels higher than early (p-14) cultures while expressing ANO1 at significantly lower (Figure 8D). Most characterization markers in iSGEC-nSS2 (p-14 and p-80) cultured on matrigel were differentially expressed (Figure 8E–G). Two consistent differentially expressed genes were VIM and KRT19, where VIM increased in matrigel cultures and KRT19 decreased. Similar to monolayer cultures on plastic, p-80 expressed NANOG at higher and KRT19 at lower levels, indicating a potential dedifferentiation when cultured for extended periods. However, p-80 exhibited the same increase in AQP5 mRNA expression as p-14 iSGEC-nSS2 cells on matrigel, retaining their ability to differentiate.

#### **5. Discussion**

This is the first study to establish immortalized primary cultures of xerostomic female patients using LSG biopsy tissue. We generated three novel composite iSGEC lines, with one capable of growing over 100 passages and maintaining expression of characterization markers.

Characterization of iSGEC-nSS2 indicates these cells to be of an undifferentiated ductal cell origin due to their high mRNA expression of KRT5 and relatively spotty distribution of KRT19 protein by ICC during monolayer culture on plastic [37,38]. When grown on matrigel, expression of KRT19 protein appeared more uniform over cells spread across the dish as a monolayer and is a marker expressed in both acinar and ductal cells [29,39]. Although a low concentration of 2 mg/mL was used for the entirety of this study, we tested whether higher at concentrations of matrigel, those monolayer cells would remain. At a concentration of (4.5 mg/mL) those monolayer cells fully incorporated into spheroid structures.

Isolated acinar cells undergo dedifferentiation into ductal-like cells when explanted into cell culture plates [17]. Dedifferentiation of rat acinar cells can be halted by the inhibition of Src and p-38 pathways, allowing the cells to retain high cytoplasm complexity and size in culture [17]. Understanding the interactions between ductal and acinar cells within the salivary gland epithelium could provide more insight into acinar cell regeneration and population maintenance. Additionally, SGECs have demonstrated the ability to trans-differentiate into acinar-like cells in matrigel cultures [10,40]. Pathways regulating trans-differentiation would likely provide viable therapeutic targets, for example, with small-molecule inhibitors, bypassing the need for stem-cell implantation [41]. iSGEC-nSS2 spheroids resembled acinus-like structures with cells high in AQP5 expression and exterior surrounding myoepithelial-like cells expressing α-SMA (Figure 7). Changes in protein expression exemplified by a lack of α-SMA in iSGEC-nSS2 monolayer cultures and high expression in matrigel indicate cell differentiation during spheroid formation.

**Figure 7.** Immunofluorescence detection of salivary epithelium markers in matrigel cultured iSGEC-nSS2 cells. **Legend.** Detection of salivary epithelium markers (**red**) by immunofluorescence in matrigel cultures of iSGEC-nSS2 p-80 after 7 days. Spheroids expressed acinar cell markers (AQP5, AMY1A), tight-junction protein (ZO-1), and had clearly defined myoepithelial cells (α-SMA). Cells growing on the surface of the matrigel and spheroids expressed salivary epithelial markers KRT8, KRT18, and KRT19 proteins. Vimentin was expressed low and sparsely among cells and did not indicate regular staining patters. Scale bar represents 50 μm. DNA (**blue**) is highlighted with DAPI.

**Figure 8.** *Cont*.

**Figure 8.** Changes in mRNA expression of early (p-14) and late (p-80) iSGEC-nSS2 cultures grown on matrigel. **Legend.** mRNA expression of early (p-14) and late (p-80) passaged iSGEC-nSS2 grown on either uncoated (monolayer) or matrigel coated plates by qRT-PCR (ΔCT). Changes in gene expression are indicated with arrows: increased (**red**), decreased (**green**), and no change (**yellow**). (**A**–**C**) AQP5 expression was the only acinar cell marker to increase in both p-14 (**B**) and p-80 (**C**) when cultured on matrigel (final: 2mg/mL) coated plates. In p-80 cells, matrigel had a greater effect on AQP5 expression compared to p-14. (**D**) Over extended passages, iSGEC-nSS2 expressed both AQP5 and AMY1A higher, whereas ANO1 decreased in expression over time. (**E**–**H**) The mRNAs of characterization markers CLDN1, CDH1, VIM, and KRT19 were differentially expressed in both p-14 and p-80 cultures, respectively, when grown as a monolayer or matrigel coated plates. (**F**–**G**) Only the ductal cell marker, KRT19, was consistently downregulated by matrigel. (**F**–**G**) Vimentin (VIM) can be an indicator of dedifferentiation in SGECs and was increased (**red**) in both p-14 and p-80 matrigel cultures. Results calculated by Student's *t*-test and *p*-values corrected by the Holm-Šídák multiple comparisons test (\* *p* < 0.05). Error bars represent mean +/− SEM.

The KRT5 transcript was found highly expressed in our iSGEC cultures similar to non-transformed SGECs, which is in agreement with reported protein expression by progenitor and ductal basal cells within the salivary gland epithelium [5]. Under normal homeostatic conditions, KRT5-positive cells give rise to intercalated and striated ductal populations [42]. However, during radiation-induced salivary gland injury KRT5-expressing ductal cells are capable of differentiation into acinar cells [42]. KRT5-positive ductal cells demonstrate higher robustness than acinar cells to radiation induced salivary gland damage and represent a favorable target in restoring acinar cell populations within damaged salivary glands [38]. Importantly, SV40Lt immortalization has been shown to not affect the differentiation potential of immortalized progenitor cells [43]. iSGEC-nSS2 could be a suitable model for understanding the cellular factors associated with trans-differentiation potential due to their high KRT5 mRNA expression.

Early (p-14) and late (p-80) passaged iSGEC-nSS2 cells expressed AQP5 at higher levels in spheroid cultures (Figure 8B,C), possibly indicating a greater extent of differentiation and therefore displaying increased heterogeneity. However, the expression of characterization marker KRT19 decreased while VIM expression increased in iSGEC-nSS2 cells cultured on matrigel. Higher VIM expression may indicate an increase in stemness properties by dedifferentiation, along with reduced KRT19 expression (differentiated ductal cell marker), which could also indicate an increase in stemness [26,44]. Similarly, changes in VIM and KRT19 expression were observed in long term 2D cultured (p-80) cells (Figure 4D). Overall, dedifferentiation of iSGEC-nSS2 cells in both 2D and 3D cultures is likely associated with their high growth and differentiation potentials when cultured on matrigel. Last, iSGEC-nSS2 could offer an in vitro alternative for salivary gland developmental research due to its dedifferentiation properties.

Direct introduction of an AQP1 expressing AAV vector into salivary glands demonstrated favorable clinical trial results where participants experienced a subjective decrease in xerostomia symptoms [45]. Although the original rationale behind AQP1 AAV therapy was to increase water permeability in ductal cells, it was later found that acinar cells were responsible for the expression of AQP1 and saliva secretion [46]. As previously stated, KRT5 protein expressing ductal progenitor cells are more abundant and can differentiate into acinar cells after injury [38]. Spheroid cultures obtained via iSGEC-nSS2 differentiation could serve as a suitable model to investigate how KRT5-positive progenitor cells possibly contribute to the generation of AQP1-expressing acinar after salivary gland damage and AQP1 AAV therapy.

## **6. Limitations**

SGECs express AQP5 mRNA and protein under established culturing conditions, although being commonly identified of ductal origin [5,12]. Moreover, a spontaneously immortalized mouse cell line expressing mixed ductal and myoepithelial protein markers when cultured as a monolayer showed mRNA expression of acinar markers in monolayer and 3D culture [40]. Markers of differentiated ductal cells include KRT8 and KRT19 among others [26]. The lack of uniform expression by either KRT8 and KRT19 in monolayer cultures of iSGEC-nSS2 indicates a potential progenitor cell population, which is capable of differentiation into acinar-like and ductal-like cells when cultured on matrigel. Following culture on matrigel, iSGEC-nSS2 cells could be replated onto plastic tissue culture plates and further grown, thereby potentially expanding the differentiated cell populations for further analysis. To determine the extent of differentiation among acinar-like cells, immunocytochemistry targeting the terminally differentiated acinar cell marker, MIST1, should be employed. Moreover, dual staining of acinar and differentiated ductal cell markers to better identify and characterize the iSGEC-nSS2 cell line could be required to determine the extent of heterogeneity within the monolayer cultured cells.

Further research is needed to analyze the presence of tight junctions and their location throughout the apical portion of the plasma membrane, an important feature for polarized secretion by acinar cells [47]. Previous studies have demonstrated trans-epithelial resistance and polarized secretion of α-amylase within monolayer SGEC cultures [5]. Among the several cell mechanisms dictating epithelial polarization and directional secretion, tight junction formation is a critical factor [47]. Techniques better visualizing the location of tight junctions within iSGEC-nSS2 spheroids should be employed to better determine the extent of polarization and direction secretion of the high AQP5-expressing acinar cells. Furthermore, ZO-1 was expressed at low levels in monolayer iSGEC cultures indicated by immunocytochemistry and its presence confirmed by Western blot. Spheroids were uncut during the immunocytochemistry procedure, which could have impacted the interior binding of the ZO-1 antibody or the visualization of the 3D architecture.

When salivary glands are damaged, acinar cells revert to a ductal-like state in mouse models [48]. Similar to the phenomenon observed in mice, rat acinar cells when extracted and cultured on plastic revert to a ductal like state, which can be reversed by inhibition of p-38 and/or Src pathways [16,17]. Due to the outlined characteristics of iSGEC-nSS2, populations of differentiated acinar–ductal cells and ductal–progenitor cells could be present. Overall expression of AQP5 protein within iSGEC-nSS2 monolayer cultures localized within the plasma membrane and cytoplasm. Confirmation of AQP5 protein was further demonstrated byWestern blot and exhibited a level of protein expression comparable to HMC-3A, which expressed AQP5 mRNA highest among all cell lines tested. When transitioned to culture on matrigel, AQP5 protein expression appeared to increase substantially within spheroid formations by ICC–IF. The increase in AQP5 expression among cells in matrigel culture could be a result of either de-differentiated acinar cell populations or progenitor-like cell populations (KRT5-positive) further differentiating into acinar cells. Before iSGEC-nSS2 use in developing therapies for increasing AQP5 expressing cells within the salivary gland, further cell sorting would need to be employed to ensure a homogenous cell population to the extent possible for reproducibility in drug assays.

In addition, age, gender, and associated hormonal changes account for possible factors affecting expression profiles of iSGECs, impacting their biological properties. Moreover, acinar cells of lacrimal glands, acinar progenitor cells of salivary glands, and saliva DNA of pSS patients may have shortened telomeres, which could be more prominent with older age [3,49,50]. Extensive cell passaging due to age combined with chronically shortened telomeres could lead to chromosomal instability and changes in DNA expression. Short Tandem Repeat (STR) profiling would be useful in the future to determine the extent of chromosome stability when passaged for extended periods and to establish a unique identifier for iSGEC-nSS2.

Although spheroids were formed by iSGEC-nSS2 in both early and late cultured cells, lumen formation needs to be further visualized, which could be useful in determining the extent of differentiation and cell types present. Furthermore, a panel of cell lines derived from different types of xerostomic patients (i.e., nSS "sicca" or radiation-induced) may be needed to develop effective treatments.
