Effects of Hypocalcemic Vitamin D Analogs in the Expression of DNA Damage Induced in Minilungs from hESCs: Implications for Lung Fibrosis

In our previous work, we evaluated the therapeutic effects of 1α,25-Dihydroxyvitamin D3, the biologically active form of vitamin D, in the context of bleomycin-induced lung fibrosis. Contrary to the expected, vitamin D supplementation increased the DNA damage expression and cellular senescence in alveolar epithelial type II cells and aggravated the overall lung pathology induced in mice by bleomycin. These effects were probably due to an alteration in the cellular DNA double-strand breaks’ repair capability. In the present work, we have evaluated the effects of two hypocalcemic vitamin D analogs (calcipotriol and paricalcitol) in the expression of DNA damage in the context of minilungs derived from human embryonic stem cells and in the cell line A549.


Introduction
DNA damage and cellular senescence underlie the physiopathology associated with idiopathic pulmonary fibrosis (IPF) and other chronic conditions that can evolve with fibrosis. IPF is a form of progressive interstitial pneumonia of unknown etiology with an estimated survival of 3 to 4 years [1]. IPF pathogenesis is the consequence of an excessive matrix deposition leading to tissue scarring and irreversible organ injury, probably due to a persistent input of damage and tissue repair response. It has been reported that cellular senescence is implicated in the tissue repair program, and its occurrence in IPF, unfortunately, has a detrimental role in contrast to other fibrogenic conditions [2][3][4]. Vitamin D and its analogs have been demonstrated to be active in the regulation of fibrosis that characterizes multiple chronic diseases, including pulmonary fibrosis [5][6][7][8]. For instance, the preventive use of vitamin D supplementation was associated with a general improvement in the lung fibrosis symptomatology induced in mice, probably due to its anti-inflammatory effects [8,9]. However, as we have reported in our previous work, the "therapeutic" treatment of mice having bleomycin-induced fibrosis seemed to worsen the pathology: the mice treated with vitamin D showed increased architectural distortion, subpleural scarring and more areas of aberrant reepithelization compared to the controls. These areas were defined by the accumulation of alveolar epithelial type II (ATII) cells harboring high levels of DNA damage in the form of DNA double-strand breaks (DSBs). DSBs were also observed in cells throughout respiratory bronchioles or immersed in alveolar fields. The bulk of the DNA damage was preferably associated with epithelial cells; fibroblasts, however, seemed to be more resistant to DNA damage than epithelial cells [10]. Senescence can be induced prematurely as a result of a persistent DNA damage response (DDR) secondary to oxidative stress that induces DNA double-strand breaks [11]. Indeed, DSBs are potent inducers of cell arrest and a typical hallmark of cell senescence [12]. Our results also showed significantly Int. J. Mol. Sci. 2022, 23, 4921 2 of 18 greater levels of DSBs and cell senescence in epithelial cells than in fibroblasts, and they were consistent with the central hypothesis underlying IPF indicating that epithelial injury and impaired regeneration activate fibroblasts and that cellular senescence induced by persistent epithelial damage may be the origin of aberrant epithelial regeneration and the promotion of fibrosis [13][14][15][16].
Although bleomycin reproduces well many aspects of the general pulmonary fibrosis and some lesions present in IPF, it has never been promoted as an experimental equivalent of IPF. The strength of the bleomycin model consists of the reproducibility and versatility as a model of general fibrosis. In addition, its high efficiency levels inducing DSBs make bleomycin a very interesting model to analyze DNA damage [17].
In the present study, we have evaluated various vitamin D analogs in the context of DNA damage induced by bleomycin. A large amount of vitamin D analogs have been synthesized throughout the years, and the clinical use for secondary hyperparathyroidism, osteoporosis or psoriasis has been approved for many of them [18,19]. The potent effect of vitamin D on intestinal calcium and phosphorus absorption and bone mineral mobilization, often leading to the development of hypercalcemia and hyperphosphatemia, has precluded its therapeutic use for many conditions. The ideal analog would retain vitamin D receptorbinding capacities and have minimal effects on calcium and phosphorus metabolism. Our working hypothesis is that hypocalcemic vitamin D analogs could show a lower incidence in the expression of DNA damage upon a bleomycin insult than the active form of vitamin D.

Results
We have tested our postulated hypothesis in the cell line A549, an immortalized counterpart of ATII cells, in 2D minilungs (lung and alveolar differentiated cells from hESCs arranged in bidimensional cultures) and in 3D minilungs from hESCs (human lung bud organoids embedded in Matrigel TM sandwiches). The lung organoids generated from hESCs have enormous advantages over cell lines or simple primary cultures as they offer an unlimited availability of primary cells, show the complete lung epithelial spectrum and emulate structural and functional features of the original organ.
The exposure of A549 cells to a sublethal bleomycin shock (12 µg/mL for 6 h) induces the expression of DNA damage (DD) foci containing TP53BP1, a reliable marker of DSBs [20][21][22][23][24]. These conditions allow the accurate quantification of DSBs. To avoid a potential effect of bleomycin on the beginning of the transcription triggered by vitamin D or its analogs, we have performed short pre-treatments with these ligands before the addition of bleomycin. In addition, as vitamin D and its analogs show differential binding affinities to their receptor, we assessed increasing amounts of ligands in the dose-response assay using the expression of the gene CYP24A1, a vitamin D target gene, as the response. The results showed that, in the case of A549 cells or 2D minilungs, the amount of 5 nmol/L is adequate as we found no significant differences between vitamin D and the analogs evaluated in terms of CYP24A1 expression. However, in the case of the minilungs embedded in Matrigel TM sandwiches (3D organoids), this low concentration of ligand gave rise to significant differences between vitamin D and the analogs evaluated, probably due to the limited access of the ligands to the organoid cells. The amount of 50 nmol/L, however, resulted in similar responses of vitamin D and the hypocalcemic analogs (Supplementary Figure S1).
After bleomycin treatments, DD foci can be rapidly visualized as discrete foci in a pan-nuclear pattern ( Figure 1A). As previously reported [10], the exposure of A549 cells to vitamin D, in the presence of bleomycin, increased the levels of DD foci, both the percentage of damaged cells and the levels of severely damaged cells harboring more than 20 DD foci per nucleus ( Figure 1B-D) (n = 3; >150 cells were analyzed; p < 0.001). However, the two hypocalcemic vitamin D analogs tested (paricalcitol and calcipotriol) were able to drastically reduce the bulk of the DD expression compared to vitamin D in the presence of bleomycin ( Figure 2C,D; n = 3; >150 cells were analyzed; ANOVA p < 0.001). Figure 2A,B shows that the treatment of A549 cells with vitamin D or its analogs, in the absence of bleomycin, did not alter the low basal level of damage of the cell population as previously described for vitamin D [10] (n = 3; >150 cells were analyzed; ANOVA p < 0.001). In order to reproduce these results in much more reliable models of lung structure and function, we generated minilungs from hESCs as previously described [25,26]. On the one hand, we generated lung airway and epithelial cells arranged in bidimensional cultures (2D minilungs) from the hESC line AND-2 as previously described [25]. Briefly, good hESCs colonies are grown along inactivated MEFs (iMEFs), picked up and passaged to new plates with iMEFs in order to accumulate material for lung differentiation. Figure 3A shows the expression of pluripotency marker SOX-2 in a good AND-2 colony and representative micrographs at various times of the differentiation process: embryoid bodies (EBs), anterior foregut endoderm (AFE), cultures at day 23 (lung progenitors) and at day 60 (differentiated lung airway and alveolar cells). The cultures from day 50 onwards show the expression of representative markers of the lung airway and alveolar cells, illustrating the heterogeneity in cell shape, including the presence of flat cells with a crescent shape morphology and granular and roughly cuboidal-shaped cells, likely corresponding to ATI and ATII cells, respectively ( Figure 3A; d60). Although from day 50 they can be considered mature, the cultures were used for the desired experimentation from day 60 on. Figure 3B shows a RT-qPCR result illustrating the complexity of these cultures (n = 3; >4 organoids per experimental replicate were used; ANOVA p < 0.001). As previously described by us and others [25][26][27][28][29], the differentiation protocol applied here yields cultures enriched in alveolar epithelial cells (ATI and ATII cells). On the other hand, the generation of 3D minilungs implies the formation of nascent organoids in suspension at certain times of the protocol (see Figure 3C and Methods for details) and their final embedding in MatrigelTM sandwiches to reach the desirable state of differentiation characterized by the presence of lung buds more or less branched [lung buds organoids (LBOs)], as previously described [26,29] ( Figure 3D). Figure 3E shows representative micrographs of histochemical analysis (H&E staining) and immunohistochemical analysis with surfactant antibodies performed on LBOs sections. In order to analyze the expression of DNA damage and the effect of vitamin D and its analogs, 2D minilungs were treated with 12.5 µg/mL of bleomycin for 72 h. All the cell types of these complex cultures seemed to be affected equally by the bleomycin treatment. As in the case of the A549 cells, neither vitamin D nor its hypocalcemic analogs altered significantly the basal levels of DD in the absence of bleomycin ( Figure 4A,B) (n = 3; >150 cells were analyzed; ANOVA p < 0.001). As expected, the exposure of these cultures to bleomycin and vitamin D increased the levels of DD foci reached by bleomycin itself ( Figure 4C,D). As in the case of the A549 cells, the treatment with paricalcitol and calcipotriol did not further increase the DD levels reached by bleomycin and seemed to reduce significantly the DD expression induced by bleomycin ( Figure 4C,D, (n = 3; >4 organoids per condition were used and >150 cells were analyzed; ANOVA p < 0.001). Equivalent assays were performed on 3D minilungs embedded into matrigelTM sandwiches. As previously reported, lung buds minilungs are mainly constituted by ATII cells [26,29]. Although to a lesser extent, we found similar results to those obtained in the case of 2D minilungs ( Figure 4E-H) [n = 3; >4 organoids per condition were used and >150 cells were analyzed; ANOVA p > 0.05 (panels 4E,F) and ANOVA p < 0.001 for data represented in panels 4G,H]. This reduction in the extension of damage was probably because the organoids are embedded into matrigel TM sandwiches and the access of bleomycin and the ligands to them is more restricted. Finally, we evaluated, in A549 cells, a continuous cell line counterpart of ATII cells, the expression of γH2AFX marker, which is a reliable marker of DD, cell senescence and aging, as previously described [20,30]. Figure 4I shows the significant increase in the γH2AFX expression levels in the presence of bleomycin and vitamin D compared to the controls and the drastic reduction induced by paricalcitol. Equivalent assays were performed using a battery of commercially available less-hypercalcemic vitamin D analogs, including 22-oxacalcitriol, tacalcitol and vitamin D2. All the vitamin D analogs seemed not to further increase the expression levels of γH2AFX reached in bleomycin-treated cells ( Figure 4J). Moreover, in the case of paricalcitol, a significant reduction in the expression of γH2AFX compared to the bleomycin-treated cells was observed. using a battery of commercially available less-hypercalcemic vitamin D analogs, including 22-oxacalcitriol, tacalcitol and vitamin D2. All the vitamin D analogs seemed not to further increase the expression levels of γH2AFX reached in bleomycin-treated cells ( Figure 4J). Moreover, in the case of paricalcitol, a significant reduction in the expression of γH2AFX compared to the bleomycin-treated cells was observed.

Discussion
Besides its function in the mineral homeostasis and immune system, vitamin D plays a role in multiple chronic diseases involving the respiratory system. Epidemiological studies have suggested a link between vitamin D deficiency and the risk of development of chronic lung diseases, such as asthma, chronic obstructive pulmonary disease (COPD), cystic fibrosis and respiratory infections [31]. This association has led to the notion that vitamin D supplementation might ameliorate the progress of these diseases. Vitamin D supplementation, however, needs to be evaluated carefully as it can be a factor contributing to vitamin-D-mediated hypercalcemia and hypercalciuria [32]. In addition, cases of vitamin D toxicity associated with overdoses due to manufacturing or intake errors have been reported [33]. Moreover, we have reported a detrimental role of vitamin D supplementation in a therapeutic experimental system, very likely associated with an impairment in the cellular DSBs repair capabilities and cell senescence [10].
Vitamin D may affect the progression of fibrosis at different stages: anti-fibrinolytic coagulation cascade, inflammation, fibroblasts activation and on the negative regulation of the renin-angiotensin system. Vitamin D seemed to prevent the experimental lung fibrosis induced by bleomycin [8,9,[34][35][36]. However, in these experimental studies, vitamin D is administered either before or very early after the bleomycin insult, so the effects observed were very likely due to the inherent anti-inflammatory properties of vitamin D. Thus, these studies can be defined as preventive. In addition, various hypocalcemic analogs, such as paricalcitol, calcipotriol and 22-oxacalcitriol, have been demonstrated to be active as anti-fibrotic agents in different experimental systems and types of fibrosis [37][38][39][40][41][42][43][44][45][46][47][48][49]. Vitamin D less-hypercalcemic analogs might provide an alternative to vitamin D supplementation to treat many conditions related to fibrosis. The ideal analog would retain vitamin D receptor-binding capacities and have minimal effects on mineral metabolism.
Bleomycin can induce pulmonary fibrosis and fibrogenic cytokine release by oxidantmediated DNA scission in a variety of animal models. The principal drawbacks of the bleomycin model relate to the rapid lung remodeling and the emphysema-like changes induced [17]. However, it reproduces well many aspects of the general pulmonary fibrosis and some lesions present in IPF, although it should be stated here that bleomycin is not a reliable model of IPF [17]. However, the potential of bleomycin in the induction of DSBs and senescence in many cell types is extraordinary [10,13]. We have made use of this advantage in the present study to explore the influence of various vitamin D hypocalcemic analogs in the context of A549 cells and minilungs generated from hESCs. Our experiment approaches the initial steps of the fibrogenic conditions, i.e., the expression of DNA damage underlying many conditions evolving towards fibrosis.
The generation of human minilungs that share the structural features and some extent of the functionality of the native organ may serve as a system model to emulate the DNA damage inflicted during the course of fibrogenic conditions, such as IPF. Currently, the more efficient protocols to generate airway and alveolar epithelial cells from the direct differentiation of hPSCs are biased to the production of alveolar cells [26,27,29]. We have employed either shocks or continuous exposures of bleomycin. The sublethal bleomycin shocks employed here allow the accurate quantification of DNA damage in the form of DSBs and the observation of subtle differences between the experimental conditions that might otherwise be masked by the extraordinary potential of bleomycin. Bleomycin seems to inflict DNA damage in the form of DSBs in all the epithelial cells equally, even when the cell organization is the form of lung buds embedded in Matrigel TM sandwiches. However, the assembly of organoids into Matrigel TM sandwiches can make difficult the access of bleomycin and ligands to the cells. The reduction in the extent of DNA damage inflicted by bleomycin compared to the 2D minilungs or A549 cells might reflect this fact.
As a continuation of our earlier work [10], our current hypothesis states that hypocalcemic vitamin D analogs could show a lower incidence in the expression of DNA damage upon a bleomycin insult than the active form of vitamin D. The initial results presented here suggest that less-hypercalcemic analogs do not show the deleterious effects observed by vitamin D treatment in the presence of bleomycin and could be an alternative to vitamin D supplementation. In addition, the treatment with such vitamin D analogs could be tested as efficient agents to reduce the bulk of the DD expression underlying multiple diseases that can evolve with DNA damage, fibrosis and aging, such as IPF and other lung interstitial conditions. Future in vivo work in this direction will be necessary.
Cells were fed after 36-48 h, depending on cell density, by removing half the old medium and adding half fresh medium.

Formation of Lung Bud Organoids
In this case, the differentiation process was performed under normoxic conditions from the anteriorization stage on. At day 8, cells were briefly trypsinized into small 3-10 cell clumps and the reaction was halted with stop medium (IMDM medium (BE12-722F) supplemented with 50% fetal bovine serum (FBS; F7524; Sigma-Aldrich, Burlington, MA, USA), 2 mM Glutamax, 1% penicillin-streptomycin). Cells were then centrifuged for 5 min at 850 rpm and washed carefully twice with an excess of SFD medium. The clumps were plated onto low-attachment six-well plates (cat.#3471; Corning, Corning, NY, USA) in branching medium (SFD medium containing 3 µM CHIR99021, 10 ng/mL FGF10, 10 ng/mL KGF, 10 ng/mL BMP4, 50 nM all-trans retinoic acid). These three-dimensional clumps (nascent lung bud organoids) were incubated and fed every other day for approximately 20-25 days. After that, these nascent organoids were embedded into a Matrigel TM sandwich assembled on MW96 wells. 50 µL of Matrigel TM were loaded on the MW96 well and allowed to gel. Nascent organoids were picked up with a wide mouth plastic Pasteur pipette, divided into MW96 wells containing 50% Matrigel TM , diluted in branching media and immediately transferred onto the first layer of Matrigel TM . After solidification of this intermediate layer containing the nascent organoids, 50 µL of Matrigel TM were added on top. Finally, each sandwich containing various organoids was incubated with 50 µL branching media. Medium was changed every 2-3 days. Growing branching structures were easily visualized under the microscope after 1 or 2 weeks. Treatments were performed in minilungs maintained in branching medium as indicated in the corresponding experiments.

Indirect Immunofluorescence of A549 Cells and 2D Minilungs
Cells were seeded in 8-well chambers (cat.#154,534; Thermofisher Scientific, Waltham, MA, USA) at a density of 20,000 cells/well. The following day, the cells were treated as indicated in the corresponding experiments. Immunofluorescence was performed as previously described [20]. Basically, the cells were fixed in 2% PFA in PBS for 10 min at RT and permeabilized with 0.1% Triton X-100 and 0.1% sodium citrate for 5 min at RT. Preparations were washed with PBS and washing solution (PBS/0.25% BSA/0.1% Tween 20), blocked for 30 min with blocking solution (washing solution + 2.5% BSA) and incubated overnight with antibodies against TP53BP1 (1:500; sc-16565; Invitrogen, Waltham, MA, USA). Preparations were then washed with washing solution and incubated with secondary antibodies conjugated with Alexa fluor dyes (488, 546) from Life Technologies (cat.#A-11029, cat.#A-11035) for 1 h at RT. Nuclei were counterstained with DAPI, and samples were mounted with ProLong Diamond (cat.#P36961; Life Technologies, Waltham, MA, USA). Cell images were captured with fluorescence microscope (Zeiss Axio) equipped with a camera (AxiocamMRm) and AxioVision software. DNA damage foci were quantified by counting from >150 cells for each experimental condition. For 2D minilungs, the glass chamber slides were incubated overnight at 4 ºC with human fibronectin in order to plate the differentiated cells. Cultures from day 50 were digested with trypsin, neutralized with stop medium and washed with SFD medium. Approximately 40,000 differentiated epithelial cells per well were plated in the epithelial maturation medium. Cultures were maintained under normoxic conditions for one day before treatments.

Indirect Immunofluorescence of Lung Bud Organoids
Organoids were picked up from the MW96 wells, transferred into a well of a MW12 and fixed with 4% paraformaldehyde (PFA) for 15 min at RT. After that, the organoids were washed three times with PBS for 10 min and incubated overnight at 4 • C with 30% sucrose. The sucrose was exchanged for a solution of 7.5% gelatin/15% sucrose and incubated for 15 min at 37 • C. The organoids were carefully transferred to cryomolds and progressively embedded in various layers of solidified 7.5% gelatin/15% sucrose. These preparations were cut into 10-µm sections in a Leica CM3050 cryostat. The mounted sections were washed with PBS and permeabilized with PBS/1% BSA/0.25% Triton X-100 for 5 min at RT. After that, the sections were washed and blocked for 30 min at RT with blocking solution (PBS-BSA 1%). The sections were incubated for 2 h with antibodies against TP53BP1 (1:500; sc-16565; Invitrogen, Waltham, MA, USA) or the pro surfactant protein C (1:200; ab3785, Merck). Preparations were washed with washing solution and incubated with a secondary antibody conjugated with Alexa fluor dye (546) from Life Technologies (cat.#A-11035, Waltham, MA, USA) for 1 h at room temperature. Nuclei were counterstained with DAPI and samples were mounted with ProLong Diamond (cat.#P36961; Life Technologies, Waltham, MA, USA). Cell images were captured with a fluorescence microscope (Zeiss Axio) equipped with a camera (AxiocamMRm) and AxioVision software. DNA damage foci were counted from >150 cells for each experimental condition.

Conclusions
The bleomycin treatment of cells and organoids might be used to reproduce the bulk of the DNA damage expression underlying multiple conditions that evolve with fibrosis. The treatment with vitamin D less-hypercalcemic analogs does not increase the DNA damage expression reached by the sole treatment with bleomycin. In terms of DNA damage expression, these initial results indicate that the vitamin D analog paricalcitol behaves as a potential DNA damage eraser. This fact might be exploited to set up a therapy for fibrogenic diseases.