In human, lung aging is characterized by a progressive lung function decline, with decreased forced expiratory volume over forced vital capacity ratio (FEV1/FVC), decreased elastic recoil and a “senile emphysema” comparable to chronic obstructive pulmonary disease (COPD) emphysema, but devoid of inflammation and alveolar septa destruction (Brandsma et al., 2017; Navarro and Driscoll, 2017; Bowdish, 2019). Static compliance also raises with age in human (Janssens et al., 1999; Birch et al., 2018). However, in healthy elderly, these functional changes might only be noticeable during exercise (Bowdish, 2019). Remodeling of the parenchyma and airspace enlargement are accompanied by a decrease in elastin content and an increase in collagen content, partly explaining the loss of elastic recoil and the increased stiffness of the lung with age; however, loss of elastic recoil is also a consequence of the reduction in surface tension due to the increased airspace volume (Janssens et al., 1999; Brandsma et al., 2017; Navarro and Driscoll, 2017). Immunosenescence occurring in aging is the result of a decline in the immune system, objectified by a higher susceptibility to infections, auto-immune diseases or cancer and the development of a chronic basal mild inflammatory state called “inflammaging” by Franceschi and colleagues (Franceschi et al., 2000; Brandsma et al., 2017; Bowdish, 2019). Human studies being hard to realize and standardize, more complete studies on lung aging were undertaken in rodents. They ended up with similar findings, especially an increase in static compliance and total inspiratory capacity as well as an increase in lung volume and airspace volume in old mice. However, they allowed a more precise description of structural/microstructural evolution of the lung with age (Yamamoto et al., 2003; Huang et al., 2007; Elliott et al., 2016; Kling et al., 2017; Schulte et al., 2019; Veldhuizen et al., 2019).

Tenascin-C (TNC) is a glycoprotein of the extracellular matrix (ECM) which is highly expressed during organogenesis, but very little or not at all thereafter, except in tissues submitted to high tensile stress or in some stem cell niches. TNC is transiently de novo re-expressed in adults during tissue repair following injury or inflammation, or continuously in chronic pathological situations like cancer (Chiquet-Ehrismann and Chiquet, 2003; Chiquet-Ehrismann, 2004; Midwood and Orend, 2009; Chiquet-Ehrismann et al., 2014). As a matricellular protein, TNC binds to cell surface receptors, ECM proteins, soluble factors or pathogens, to modulate cell adhesion, proliferation, migration and differentiation properties in an autocrine and paracrine manner (Chiquet-Ehrismann and Chiquet, 2003; Hsia and Schwarzbauer, 2005; Midwood et al., 2011, 2016). TNC was shown to be an important component of lung development, with an impairment of lung branching morphogenesis (early lung development) and delayed classical and continued alveolarization processes (late lung development) occurring in TNC-deficient mice (Roth-Kleiner et al., 2004; Schittny, 2018; Mund and Schittny, 2020). However, a compensatory mechanism, maybe through other members of the tenascin family resulted in a catch-up of the morphometric measurements of the lung at 3 months of age (Mund and Schittny, 2020). It is likely, that part of the developmental role of TNC is mediated by α8β1 integrin. Indeed, α8β1 integrin recognizes TNC and α8-integrin subunit-deficient mice showed a very similar pulmonary phenotype as TNC-deficient mice (Cremona et al., 2020).

In adult lung, a weak expression of TNC was consistently reported in airway smooth muscle and basal cell layers, while expression in the parenchyma was described as very faint and inconsistent by some authors (Koukoulis et al., 1991; Natali et al., 1991; Kaarteenaho-Wiik et al., 2001; Liesker et al., 2009; Lofdahl et al., 2011). The role of TNC in normal aging was explored essentially in tissues submitted to high tensile stress such as the skin (Choi et al., 2020), cardiac muscle (Sato and Shimada, 2001), ligament/tendon (Fujii et al., 1993; Veronesi et al., 2015; Ribitsch et al., 2020) or even cartilage (Gruber et al., 2002, 2011). In the lung, a single report comparing 24 to 8 months old decellularized mouse lung scaffolds showed that aging was linked to an increase in collagen and collagen-related proteins, and that this increase was preceded by an increase in TNC and osteopontin expression and associated with an increase in the number of senescent cells (Calhoun et al., 2016). A recent review suggested that TNC might, together with other matricellular proteins, contribute to cellular senescence by acting on the cell cycle, apoptosis and the pool of secreted proteins in several tissues among which the lung (Blokland et al., 2020).

General aging hallmarks were proposed to be nine: genomic instability, telomere attrition, epigenetic alterations, loss of proteostasis, deregulated nutrient sensing, mitochondrial dysfunction, cellular senescence, stem cell exhaustion and altered intercellular communication (Lopez-Otin et al., 2013). More recently, Meiners et al. (2015) proposed a tenth hallmark of aging specific for lung aging, which was “dysregulation of the extracellular matrix”. Major chronic lung diseases are exacerbated by aging hallmarks, which put elderly people at higher risk of developing chronic lung diseases such as idiopathic pulmonary fibrosis (IPF), COPD or lung cancer (Meiners et al., 2015; Brandsma et al., 2017; Navarro and Driscoll, 2017). TNC is highly expressed in such chronic lung diseases as COPD, bronchopulmonary dysplasia (BPD), respiratory distress syndrome (RDS), IPF or asthma (Laitinen et al., 1997; Kaarteenaho-Wiik et al., 2002; Lofdahl et al., 2011; Estany et al., 2014; Yasuda et al., 2018). In this paper, we aimed at deciphering whether TNC could play a role in lung aging. We found that TNC knockout (TNC KO) animals at 18 months of age presented an increased lung volume, parenchymal volume, total airspace volume and septal surface area, compared to wild-type (WT) counterparts, with no change in functional pressure-volume curves. Increased morphometric parameters were accompanied by an increase in cell proliferation, essentially in the parenchyma. Apoptotic cells were not detected and senescence was only observed in bronchial epithelial cells and alveolar macrophages in both genotypes. We conclude that old TNC knockout mice exhibit exaggerated structural hallmarks of lung aging, compared to WT counterparts.

Human lung aging is described as a functional decline, with increased residual volume (RV) and functional residual capacity (FRC), reduced regeneration capacity and structurally, airspace enlargement, defined as “senile emphysema” in opposition to the COPD-related inflammatory emphysema. However, human lung evaluations were always fragmentary and better functional and structural analyses of lung aging were provided by animal studies with longitudinal survey of lung morphometry/microstructure and functional parameters in controlled ventilated animals at different ages (Yamamoto et al., 2003; Huang et al., 2007; Elliott et al., 2016; Schulte et al., 2019; Veldhuizen et al., 2019). We aimed at understanding the potential role of TNC in this process.

In the present study, we showed that at 18 months of age, lung volume, parenchymal volume, airspace volume and septal surface area were one-third higher in TNC KO mice, compared to WT counterparts, in males and females. Schulte et al. (2019) recently demonstrated in a very extensive stereological analysis of mouse lungs at 3, 6, 12, 18, and 24 months of age that normal lung aging was accompanied by an increase in lung volume, parenchymal volume, total airspace, ductal airspace and alveolar airspace, occurring between 12 and 18 months. They also showed that septal surface area was concurrently increased (Schulte et al., 2019). Therefore, all the age-related structural modifications described in normal aging in mice were even higher in TNC-deficient mice. These results could either mean that the lung volume is enlarged but the alveoli network is identical (and then the mesh size is larger) (distention of the lung as in lung aging), or that the volume is enlarged but the mesh size is identical (and then the alveoli network is bigger) in TNC KO lungs, compared to their WT counterparts. The length of the free septal edges represents the total length of all alveolar entrance rings. Accordingly, the unchanged length of the free septal edges observed favors the assumption of a conserved mesh size in both genotypes, but in a larger volume. Calculation of a similar Lm in both genotypes strengthened this hypothesis. We previously showed that TNC inactivation led to perturbation of the branching morphogenesis and delays in both phases of alveolarization (classical and continued) during lung development. This was accompanied by a temporary reduction of the total surface area and the total length of the free septal edge, but a catch-up was reached at the age of 3 months, where both TNC KO and WT counterparts had identical lung volume, length of the free septal edge, and septal surface area (Roth-Kleiner et al., 2004; Mund and Schittny, 2020). Consequently, our results implied that somewhere between 3 and 18 months of age, a modification in lung metabolism materialized by structural modifications occurred in TNC KO animals.

An increase in pulmonary volumes and surface area with identical network mesh size could indicate either an enhanced tissue production or a reduced tissue destruction. To figure out the mechanisms underlying these changes, we analyzed proliferation, apoptosis/necrosis and autophagy in the TNC KO and WT lungs, and showed that apoptosis/necrosis markers expression was not altered in old TNC KO lungs, compared to WT lungs (apoptosis was almost undetected in TNC KO and WT lungs). Both autophagy markers LC3bII and p62 proteins were up-regulated in old TNC KO lungs, but as already stated above, activation of autophagy should be related to an increased LC3bII protein but a decreased p62 protein expression. Therefore, autophagy was probably not stimulated in old TNC KO lungs. However, cell proliferation markers PCNA and ki67 were 2 to 3-fold higher in old TNC KO lungs, compared to WT counterparts. Moreover, this was exclusively observed in the parenchymal compartment, which favors the hypothesis of an increased tissue production rather than a decreased cell destruction in old TNC KO lungs, compared to WT lungs. Among the cells present in the parenchyma are alveolar epithelial type I and type II cells (AEC1 and AEC2), immune cells, fibroblasts (and myofibroblasts), and pericytes and endothelial cells forming the capillaries. Proliferation rate and density of AEC1 were previously shown to be decreased in old, compared to young adult mice, contrary to AEC2 (alveolar stem cells and progenitors for AEC1); AEC1 being terminally differentiated cells with no proliferation capacity, the authors concluded that the decrease was probably related to the rate of AEC2 to AEC1 differentiation (Watson et al., 2020). In another study, AEC2 cell density, but not cell number per alveolus, was reported to decrease with age (Schulte et al., 2019). Fibroblast proliferation was also shown to decrease in aged lungs, with a decreased regeneration capacity and a terminal differentiation in myofibroblast (Paxson et al., 2011). Aging is overall linked to a decrease in proliferation capacity, while proliferation was somehow increased or less decreased in old TNC KO lungs, compared to WT counterparts in the present study. Increased stereological measurements observed in normal aging are thought to result more from the loss of supporting tissue rather than a difference in cell proliferation. Therefore, if TNC KO lungs show exaggerated stereological modifications compared to normal aging, they do not come from the same mechanism. It is interesting to note that only short isoforms of TNC mRNA were found to be expressed in adult lungs at all ages tested (3, 6 or 24 months of age) [(Gremlich et al., 2020); data not shown]. Short TNC isoforms are known to promote cell attachment and differentiation, while long isoforms are expressed during organogenesis or in cancer, and promote cell migration and proliferation (Chiquet, 1992; Giblin and Midwood, 2015). The overall increased or less decreased proliferation could then result from the absence of the short TNC isoforms in TNC KO.

Structural modifications of the lung were accompanied by an increase in total collagen content. We had previously shown in 3 months old mice, that TNC inactivation induced a thickening of the collagen layer around small airways compared to WT lungs (Gremlich et al., 2020). In 18 months old lungs, collagen staining around small airways was similar in both genotypes, but total collagen content was doubled, suggesting that the increased collagen was located in the parenchyma, in a diffuse manner. Normal aging is already known to be a condition with increased collagen production and decreased elastin content, although collagen cross-linking or fiber size is also altered (Brandenberger and Muhlfeld, 2017; Sicard et al., 2018). In the present study, collagen content was even higher in the TNC KO lungs, compared to the normal aged WT lungs. Biomechanical properties of lung parenchyma were extensively studied, notably integrating the contribution of collagen and elastin (Suki and Bates, 2011; Suki and Bartolak-Suki, 2015; Bou Jawde et al., 2020). In aging, ECM and parenchymal stiffness were shown to be increased, in part due to an increase in collagen cross-linking occurring with age, and the increased compliance observed in lung aging suggested to result from other modifications like an increase in alveoli size/airspace dimensions. In the present study, we did not find any functional difference between genotypes, despite an increase in collagen content; Lm and total length of the free septal edges values were also similar, although a trend toward higher values in TNC KO lungs, could potentially be perceived (Figures 2G,H).

Senescent cells were shown to accumulate in the lung with aging, which results in increased inflammation, stem cell dysfunction and senescence of adjacent cells; this also favors the development of chronic lung diseases (Campisi, 2016; Brandenberger and Muhlfeld, 2017; Birch et al., 2018; Parikh et al., 2019). In old TNC KO and WT lung samples, senescence-associated β-galactosidase (SA-β-gal) activity was detected in epithelial cells of the conducting airways, and in cells scattered throughout lung parenchyma and morphologically identified as alveolar macrophages. SA-β-gal activity in epithelium of the conducting airways was too variable among samples, to allow any correlation with the genotype. In old mouse lungs, TNC was shown to co-localize with senescent cells, but also to appear ahead of the increase in senescent cells number occurring in aging; moreover, senescent cells were also shown to secrete TNC (Calhoun et al., 2016). It could then be interesting to further characterize senescence appearance in old TNC KO and WT lungs to potentially strengthen the role of TNC in this process suggested in a recent review (Blokland et al., 2020). SA-β-gal staining was also observed in alveolar macrophages in both genotypes in a similar way. However, Hall et al. demonstrated that a proportion of the macrophages expressing SA-β-gal (and p16INK4a, CDK4/6 inhibitor, in their study) were not bona fide senescent cells, but a particular class of macrophages with a physiological and reversible polarization. These macrophages retained their phagocytic capability, contrary to bona fide senescent cells. They nevertheless accumulated in aged organisms and probably participated in the “inflammaging” process but not through a true senescent phenotype (Hall et al., 2016, 2017). TNC was recently proposed to have a role in innate immune system regulation. It was shown to induce the secretion of several cytokines (IL-6, IL-8 or TNFα) during inflammation or tissue repair, in macrophages, dendritic cells or even fibroblasts, and to promote macrophage migration and influence M1/M2 macrophage polarization according to the context (Shimojo et al., 2015; Sha et al., 2017; Abbadi et al., 2018; Marzeda and Midwood, 2018; Wang et al., 2018; Kimura et al., 2019). In a model of cardiac dysfunction, TNC was even shown to induce inflammation and monocyte/macrophage recruitment, while the effect was abolished in TNC-deficient mice (Abbadi et al., 2018). Therefore, studying more in depth alveolar macrophages in TNC KO and WT mice might be of interest in the context of aging and inflammaging as well.

Mechanistically, TNC can bind to many different types of proteins to activate downstream effectors and mediate its effect on cell proliferation, migration or differentiation: ECM proteins, cell surface receptors or soluble proteins such as proteases or growth factors for example (Humphries et al., 2006; Midwood and Orend, 2009; Midwood et al., 2016). Transforming growth factor β (TGFβ) was shown to be one of the major modulator of TNC expression, and Toll-like receptor 4 (TLR4) and integrins, major cell surface receptors binding TNC [reviewed in (Chiovaro et al., 2015; Tucker and Chiquet-Ehrismann, 2015; Yoshida et al., 2015; Bhattacharyya et al., 2016; Midwood et al., 2016; Marzeda and Midwood, 2018; Yalcin et al., 2020)]. We had previously shown that several TGFβ and TLR4 pathway intermediates were tremendously up-regulated at the mRNA level in 3 months old TNC KO lungs, compared to WT lungs (5–20-times fold increase) (Gremlich et al., 2020). In the present study, mRNA expression of TGFβ and TLR4 pathway intermediates were roughly similar in lungs of both genotypes, meaning that TNC inactivation effect on cell proliferation was not going through an over-activation of these pathways. However, TGFβR3 mRNA expression was slightly decreased and TLR4 mRNA slightly increased. TGFβR3 is a co-receptor, which binds to TGFβR ligands to regulate their interaction with TGFβR1/2. In the lung, during TGFβ-induced myofibroblast differentiation, TGFβR3 expression was shown to be inhibited concomitantly with the increase in α-SMA and pro-collagen type I expression; conversely, if TFβR3 was reintroduced, differentiation was inhibited (Ahn et al., 2010). Knowing that TNC induces myofibroblast differentiation, it is surprising that the only TGFβ superfamily member with a modified mRNA expression in TNC KO lungs is TGFβR3, and that its negative regulation rather favors myofibroblast differentiation. However, total myofibroblast population was overall not affected by TNC inactivation in the present study, as shown by the flow cytometry experiment. As for TLR4 mRNA expression, the small increase detected in TNC KO lungs could possibly be an attempt of compensation for TNC inactivation the same way as in 3 months old lungs (Gremlich et al., 2020). Molecular pathways by which TNC was able to enhance cell proliferation through integrins binding in pancreatic cancer cells have recently been described in two publications. Cai et al. (2018) showed that TNC controlled cell cycle progression by acting on the AKT/FOXO1 pathway, resulting in p27kip1 gene (a cyclin-dependent kinase inhibitor) down-regulation and cyclin D1 gene up-regulation. Both effects allow the cells to pass the G1/S restriction point and enter the S phase of DNA replication, by promoting cyclin D-CDK4/6 complex formation (Suryadinata et al., 2010). Shi et al. (2015) also showed that TNC promoted apoptosis resistance through the ERK/NF-kB pathway, which ended in anti-apoptotic Bcl-2 and Bcl-xL gene expression stimulation. In the present study, neither cyclin D1 nor p27kip1 mRNA expression were altered in old TNC KO lungs, compared to WT counterparts. However, mRNA expression of Bcl-xL, but not Bcl-2, was reduced in TNC KO lungs, in accordance with the literature. This could indicate that TNC inactivation might additionally modulate cell proliferation through an increase in apoptosis in old TNC KO lungs. However, we could almost not detect any apoptosis in 18 months old TNC KO or WT lungs.

Lung function decline with age has been well described in mice, and is characterized by an increase in static compliance (Cst) and inspiratory capacity (IC) (Elliott et al., 2016; Kling et al., 2017; Schulte et al., 2019; Veldhuizen et al., 2019). In the present study, old TNC-deficient mice did not show any difference in basal Cst or IC, compared to their WT counterparts, even if males tended to have higher basal values in TNC-deficient mice than in WT mice (only significant for IC). As for HTV ventilation, female and male old TNC KO and WT mice showed superimposed curves, at least during the first 45 min for the males. Indeed, old males were less resistant to mechanical ventilation than females, showed greater standard deviations between individuals, and had functional parameters values that tended to fall after 45 min of HTV ventilation (especially in WT animals). Surprisingly, old TNC KO and WT mice had similar basal Cst and IC values as TNC KO mice at 3 months of age. Indeed, we previously showed that young adult TNC KO mice had higher Cst and IC values than their WT counterparts (Gremlich et al., 2020; Supplementary Figure 3). This means that 3 months old TNC KO mice already exhibited the Cst and IC observed in normal aging, as a premature lung aging phenotype, knowing that these functional changes were described to occur between 3 and 6 months of age in a very extensive study on mouse lung aging (Schulte et al., 2019).

Using flow cytometry, we showed that the percentage of α-SMA + cells was very marginally increased in 18 months old TNC KO lungs compared to WT counterparts. Quantitative analysis of α-SMA protein expression in whole lung protein extracts also resulted in identical expression of the protein in old TNC KO and WT lungs. Qualitative analysis of α-SMA by immunohistochemistry showed that α-SMA staining appeared similar in old TNC KO and WT lung sections, with a fragmented staining around bronchioles and blood vessels, but no staining in the parenchyma, as was already the case in young adult mice at 3 months of age (Gremlich et al., 2020). Flow cytometry analyses, however, revealed that the percentage of SMA + desmin + cells was very much higher in old TNC KO lungs than in WT lungs, and that these cells were essentially non-proliferative cells. Non-proliferative SMA + vimentin + cells were conversely decreased in TNC KO lungs, compared to WT lungs, but the major part of SMA + vimentin + cells were proliferative cells, present in the same proportion in both genotypes. Myofibroblasts are normally not present in adults, but appear de novo during repair processes or morphogenesis by differentiation from several types of cells, like fibroblasts, pericytes or even mesenchymal progenitor cells (MOC) (Phan, 2008, 2012). With age, lung myofibroblasts were shown to gradually lose their de-differentiation capacity, adopt a senescent phenotype and become resistant to apoptosis (Paxson et al., 2011; Yanai et al., 2015; Kato et al., 2020). TNC promotes myofibroblast differentiation by inducing SMA and collagen type I mRNA expression, but TNC mRNA expression is also induced during myofibroblast differentiation through TGFβ pathway (Imanaka-Yoshida et al., 2001; Tamaoki et al., 2005; Pegorier et al., 2010; Marzeda and Midwood, 2018; Katoh et al., 2020). Therefore, TNC inactivation could be in part responsible for the lower proportion of myofibroblasts in the non-proliferative pool of SMA + cells. As for SMA + desmin + cells, a general thickening of vascular walls was described in aging, due to VSMC proliferation, in addition to perturbation in elastin and collagen deposition (Xu et al., 2017). VSMC were shown to shift from a contractile phenotype (differentiated cells) to a synthetic phenotype (dedifferentiated cells) with age, with increased migration from the tunica media into the tunica intima of blood vessels, increased proliferation and increased ECM synthesis (Monk and George, 2015; Lacolley et al., 2018; Sicard et al., 2018). With respect to airway wall thickening, it was previously thought to occur with a higher frequency in old subjects; however, a more recent report using high-resolution computed tomography could show that airway wall thickness was rather decreased between 20 and 80 years, with a good correlation factor (Telenga et al., 2017). Our flow cytometry experiment did not allow us to distinguish between VSMC and ASM cells. However, an increased VSMC population in TNC KO lungs would indicate an exaggeration of the vascular wall thickening already observed in normal aging. Further studies looking more precisely at the blood vessels in the distal lung might maybe shed some light on the contribution of VSMC to the increased SMA + desmin + population cited above.

A schematic representation of TNC inactivation effects in lung aging is presented in Figure 9. Old TNC KO mice showed morphometric results compatible with exaggerated features of lung aging, such as increased lung volume, parenchymal volume, total airspace volume and septal surface compared to their WT counterparts, already shown to have increased volume and surface values compared to young mice (Schulte et al., 2019). This was associated with an increased collagen content in TNC KO lungs compared to WT lungs, a feature already existing in normal aging. Analysis of PV curves showed that basal Cst and IC values were identical in old TNC KO and WT mice, but were identical to 3 months old TNC KO mice values, themselves higher than their WT counterparts (Gremlich et al., 2020). Increased Cst and IC were previously shown to be changes occurring with normal lung aging between 3 and 6 months of age, while structural alterations essentially occurred between 12 and 18 months of age (Schulte et al., 2019). Collectively, several differences observed were compatible with exaggerated features of lung aging in TNC-deficient animals. As a weak expression of TNC was consistently reported in adult lung (Koukoulis et al., 1991; Natali et al., 1991; Kaarteenaho-Wiik et al., 2001; Liesker et al., 2009; Lofdahl et al., 2011), these effects might either still be inherited from the developmental alterations observed in fetal or neonatal life (Roth-Kleiner et al., 2004, 2014; Mund and Schittny, 2020), or result from the basal low TNC expression remaining at all ages. Further experiments on TNC might anyhow show an improvement of lung aging with TNC supplementation.

The original contributions presented in the study are included in the article/Supplementary Material, further inquiries can be directed to the corresponding author.

The animal study was reviewed and approved by the Animal Ethics Committee of the Federal Food Safety and Veterinary Office, and the Veterinary Service of the Canton Bern.

EY and JS: breeding and handling of the mice. EY, JS, SG, and TC: harvesting of the lungs. EY: embedding of the tissue and histological stainings. SG, TC, and JS: morphological observations. SG: immunohistochemistry, western blots, senescence assay, apoptosis assay, RT-qPCR, and writing the first draft. TC and SG: lung function tests. KF: flow cytometry. FC: stereology. JS, MR-K, and SG: conceive the study. All authors read and approved the manuscript.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.