Advertisement

Epilipidomics of Senescent Dermal Fibroblasts Identify Lysophosphatidylcholines as Pleiotropic Senescence-Associated Secretory Phenotype (SASP) Factors

  • Author Footnotes
    14 These authors contributed equally to this work.
    Marie-Sophie Narzt
    Footnotes
    14 These authors contributed equally to this work.
    Affiliations
    Christian Doppler Laboratory on Biotechnology of Skin Aging, Vienna, Austria

    Department of Dermatology, Medical University of Vienna, Vienna, Austria

    Ludwig Boltzmann Institute for Experimental and Clinical Traumatology, AUVA Research Center, Linz and Vienna, Austria
    Search for articles by this author
  • Author Footnotes
    14 These authors contributed equally to this work.
    Vera Pils
    Footnotes
    14 These authors contributed equally to this work.
    Affiliations
    Department of Dermatology, Medical University of Vienna, Vienna, Austria

    Department of Biotechnology, University of Natural Resources and Life Sciences Vienna, Vienna, Austria
    Search for articles by this author
  • Christopher Kremslehner
    Affiliations
    Christian Doppler Laboratory on Biotechnology of Skin Aging, Vienna, Austria

    Department of Dermatology, Medical University of Vienna, Vienna, Austria

    Christian Doppler Laboratory for Skin Multimodal Imaging of Aging and Senescence, Vienna, Austria
    Search for articles by this author
  • Ionela-Mariana Nagelreiter
    Affiliations
    Christian Doppler Laboratory on Biotechnology of Skin Aging, Vienna, Austria

    Department of Dermatology, Medical University of Vienna, Vienna, Austria

    Christian Doppler Laboratory for Skin Multimodal Imaging of Aging and Senescence, Vienna, Austria

    Center for Brain Research, Department of Molecular Neurosciences, Medical University of Vienna, Vienna, Austria
    Search for articles by this author
  • Markus Schosserer
    Affiliations
    Department of Biotechnology, University of Natural Resources and Life Sciences Vienna, Vienna, Austria

    Christian Doppler Laboratory for Skin Multimodal Imaging of Aging and Senescence, Vienna, Austria

    Austrian Cluster for Tissue Regeneration, Vienna, Austria
    Search for articles by this author
  • Emilia Bessonova
    Affiliations
    Department of Biotechnology, University of Natural Resources and Life Sciences Vienna, Vienna, Austria
    Search for articles by this author
  • Alina Bayer
    Affiliations
    Department of Biotechnology, University of Natural Resources and Life Sciences Vienna, Vienna, Austria
    Search for articles by this author
  • Raffaela Reifschneider
    Affiliations
    Department of Biotechnology, University of Natural Resources and Life Sciences Vienna, Vienna, Austria
    Search for articles by this author
  • Lucia Terlecki-Zaniewicz
    Affiliations
    Christian Doppler Laboratory on Biotechnology of Skin Aging, Vienna, Austria

    Department of Biotechnology, University of Natural Resources and Life Sciences Vienna, Vienna, Austria
    Search for articles by this author
  • Petra Waidhofer-Söllner
    Affiliations
    Institute of Immunology, Center of Pathophysiology, Infectiology and Immunology, Medical University of Vienna, Vienna, Austria
    Search for articles by this author
  • Michael Mildner
    Affiliations
    Department of Dermatology, Medical University of Vienna, Vienna, Austria
    Search for articles by this author
  • Erwin Tschachler
    Affiliations
    Department of Dermatology, Medical University of Vienna, Vienna, Austria
    Search for articles by this author
  • Maria Cavinato
    Affiliations
    Institute for Biomedical Aging Research, University of Innsbruck, Austria

    Center for Molecular Biosciences Innsbruck, Innsbruck, Austria
    Search for articles by this author
  • Sophia Wedel
    Affiliations
    Institute for Biomedical Aging Research, University of Innsbruck, Austria

    Center for Molecular Biosciences Innsbruck, Innsbruck, Austria
    Search for articles by this author
  • Pidder Jansen-Dürr
    Affiliations
    Institute for Biomedical Aging Research, University of Innsbruck, Austria

    Center for Molecular Biosciences Innsbruck, Innsbruck, Austria
    Search for articles by this author
  • Lucia Nanic
    Affiliations
    Ruder Boskovic Institute, Division of Molecular Biology, Laboratory for Molecular and Cellular Biology, Zagreb, Croatia
    Search for articles by this author
  • Ivica Rubelj
    Affiliations
    Ruder Boskovic Institute, Division of Molecular Biology, Laboratory for Molecular and Cellular Biology, Zagreb, Croatia
    Search for articles by this author
  • Abdoelwaheb El-Ghalbzouri
    Affiliations
    Department of Dermatology, Leiden University Medical Center, Leiden, Netherlands
    Search for articles by this author
  • Samuele Zoratto
    Affiliations
    Christian Doppler Laboratory for Skin Multimodal Imaging of Aging and Senescence, Vienna, Austria

    Institute of Chemical Technologies and Analytics, TU Wien, Vienna, Austria
    Search for articles by this author
  • Martina Marchetti-Deschmann
    Affiliations
    Christian Doppler Laboratory for Skin Multimodal Imaging of Aging and Senescence, Vienna, Austria

    Austrian Cluster for Tissue Regeneration, Vienna, Austria

    Institute of Chemical Technologies and Analytics, TU Wien, Vienna, Austria
    Search for articles by this author
  • Johannes Grillari
    Affiliations
    Christian Doppler Laboratory on Biotechnology of Skin Aging, Vienna, Austria

    Ludwig Boltzmann Institute for Experimental and Clinical Traumatology, AUVA Research Center, Linz and Vienna, Austria

    Department of Biotechnology, University of Natural Resources and Life Sciences Vienna, Vienna, Austria

    Austrian Cluster for Tissue Regeneration, Vienna, Austria
    Search for articles by this author
  • Florian Gruber
    Correspondence
    Correspondence: Florian Gruber, Department of Dermatology, Medical University of Vienna, Leitstelle 7J Währinger Gürtel 18-20, 1090, Vienna, Austria.
    Affiliations
    Christian Doppler Laboratory on Biotechnology of Skin Aging, Vienna, Austria

    Department of Dermatology, Medical University of Vienna, Vienna, Austria

    Christian Doppler Laboratory for Skin Multimodal Imaging of Aging and Senescence, Vienna, Austria
    Search for articles by this author
  • Ingo Lämmermann
    Affiliations
    Christian Doppler Laboratory on Biotechnology of Skin Aging, Vienna, Austria

    Department of Biotechnology, University of Natural Resources and Life Sciences Vienna, Vienna, Austria
    Search for articles by this author
  • Author Footnotes
    14 These authors contributed equally to this work.
Open ArchivePublished:December 14, 2020DOI:https://doi.org/10.1016/j.jid.2020.11.020
      During aging, skin accumulates senescent cells. The transient presence of senescent cells, followed by their clearance by the immune system, is important in tissue repair and homeostasis. The persistence of senescent cells that evade clearance contributes to the age-related deterioration of the skin. The senescence-associated secretory phenotype of these cells contains immunomodulatory molecules that facilitate clearance but also promote chronic damage. Here, we investigated the epilipidome—the oxidative modifications of phospholipids—of senescent dermal fibroblasts, because these molecules are among the bioactive lipids that were recently identified as senescence-associated secretory phenotype factors. Using replicative- and stress- induced senescence protocols, we identified lysophosphatidylcholines as universally elevated in senescent fibroblasts, whereas other oxidized lipids displayed a pattern that was characteristic for the used senescence protocol. When we tested the lysophosphatidylcholines for senescence-associated secretory phenotype activity, we found that they elicit chemokine release in nonsenescent fibroblasts but also interfere with toll-like receptor 2 and 6/CD36 signaling and phagocytic capacity in macrophages. Using matrix-assisted laser desorption/ionization Fourier transform ion cyclotron resonance mass spectrometry imaging, we localized two lysophosphatidylcholine species in aged skin. This suggests that lysophospholipids may facilitate immune evasion and low-grade chronic inflammation in skin aging.

      Graphical abstract

      Abbreviations:

      DPPC (di-palminoyl-sn-glycero-3-phosphorylcholine), FB (fibroblast), HODE (hydroxyoctadecadienoic acid), LysoPC (lysophosphatidylcholine), LysoPPC (1-palmitoyl-2-lyso-sn-glycero-3-phosphorylcholine), LysoSPC (1-stearoyl-2-lyso-sn-glycero-3-phosphorylcholine), MALDI-FTICR-MSI (matrix-assisted laser desorption/ionization Fourier transform ion cyclotron resonance mass spectrometry imaging), OxPL (oxidized phospholipid), PAPC (1-palmitoyl-2-arachidonoyl-sn-glycero-3-phosphorylcholine), PC (phosphatidylcholine), PD (population doubling), PL (phospholipid), PLPC (1-palmitoyl-2-linoleoyl-sn-glycero-3-phosphorylcholine), SAPC (1-stearoyl-2-arachidonoyl-sn-glycero-3-phosphorylcholine), SASP (senescence-associated secretory phenotype), SIPS (stress-induced premature senescence), TLR (toll-like receptor)

      Introduction

      Cellular senescence is a state of arrested proliferation and altered metabolism (
      • Muñoz-Espín D.
      • Serrano M.
      Cellular senescence: from physiology to pathology.
      ) that is triggered by endogenous and exogenous stressors, telomere shortening, and aging. Senescent cells accumulate with age in various tissues (
      • Burton D.G.
      • Krizhanovsky V.
      Physiological and pathological consequences of cellular senescence.
      ), including in the skin (
      • Demaria M.
      • Desprez P.Y.
      • Campisi J.
      • Velarde M.C.
      Cell autonomous and non-autonomous effects of senescent cells in the skin.
      ). In age-related vascular (
      • Minamino T.
      • Miyauchi H.
      • Yoshida T.
      • Ishida Y.
      • Yoshida H.
      • Komuro I.
      Endothelial cell senescence in human atherosclerosis: role of telomere in endothelial dysfunction.
      ) and metabolic diseases (
      • Katsuumi G.
      • Shimizu I.
      • Yoshida Y.
      • Minamino T.
      Vascular senescence in cardiovascular and metabolic diseases.
      ) senescent cells accumulate. Their persistent presence causes chronic inflammation and tissue damage, especially in aged individuals (
      • Childs B.G.
      • Baker D.J.
      • Kirkland J.L.
      • Campisi J.
      • van Deursen J.M.
      Senescence and apoptosis: dueling or complementary cell fates?.
      ), and their clearance from the tissue reduces aging-associated disorders and prolongs the lifespan by improving the function of multiple organs (
      • Baker D.J.
      • Childs B.G.
      • Durik M.
      • Wijers M.E.
      • Sieben C.J.
      • Zhong J.
      • et al.
      Naturally occurring p16(Ink4a)-positive cells shorten healthy lifespan.
      ) in animal models. Senescent cells release a mix of molecules termed the senescence-associated secretory phenotype (SASP) that are immunomodulatory and affect proliferation and motility of nonsenescent cells (
      • Coppé J.P.
      • Patil C.K.
      • Rodier F.
      • Sun Y.
      • Muñoz D.P.
      • Goldstein J.
      • et al.
      Senescence-associated secretory phenotypes reveal cell-nonautonomous functions of oncogenic RAS and the p53 tumor suppressor.
      ;
      • Salama R.
      • Sadaie M.
      • Hoare M.
      • Narita M.
      Cellular senescence and its effector programs.
      ). SASP facilitates the removal of senescent cells by the immune system during development and aging (
      • Egashira M.
      • Hirota Y.
      • Shimizu-Hirota R.
      • Saito-Fujita T.
      • Haraguchi H.
      • Matsumoto L.
      • et al.
      F4/80+ macrophages contribute to clearance of senescent cells in the mouse postpartum uterus.
      ;
      • Ritschka B.
      • Storer M.
      • Mas A.
      • Heinzmann F.
      • Ortells M.C.
      • Morton J.P.
      • et al.
      The senescence-associated secretory phenotype induces cellular plasticity and tissue regeneration.
      ). Persisting SASP, however, promotes chronic inflammation and tissue damage (
      • Young A.R.
      • Narita M.
      SASP reflects senescence.
      ).
      The aging of the skin and its resulting functional impairment reduce the quality of life and predispose for emergence malignancies, most prominently skin cancers. Besides intrinsic aging, driven by time and the individual’s genetic background, the skin is exposed to multiple extrinsic stressors, most prominently UVR, that promote aging. Consequently, aged skin is prone to accumulation of senescent cells, which may comprise up to 50% of all cells within the dermis (
      • Dimri G.P.
      • Lee X.
      • Basile G.
      • Acosta M.
      • Scott G.
      • Roskelley C.
      • et al.
      A biomarker that identifies senescent human cells in culture and in aging skin in vivo.
      ;
      • Ressler S.
      • Bartkova J.
      • Niederegger H.
      • Bartek J.
      • Scharffetter-Kochanek K.
      • Jansen-Dürr P.
      • et al.
      p16INK4A is a robust in vivo biomarker of cellular aging in human skin.
      ). In aging, the dermal skin compartment undergoes changes in structure and cellular composition. The deeper, reticular layer of the dermis extends, whereas the upper papillary dermis, and with it the residing papillary dermal fibroblasts (FBs), recedes. Interestingly, FBs of the reticular dermis display markers of cellular senescence, and after prolonged culture, papillary FBs acquire a reticular phenotype (
      • Janson D.
      • Saintigny G.
      • Mahé C.
      • El Ghalbzouri A.
      Papillary fibroblasts differentiate into reticular fibroblasts after prolonged in vitro culture.
      ). In an organotypic skin model, reticular FBs, unlike papillary FBs, did not support normal stratification of epidermal keratinocytes (
      • Mine S.
      • Fortunel N.O.
      • Pageon H.
      • Asselineau D.
      Aging alters functionally human dermal papillary fibroblasts but not reticular fibroblasts: a new view of skin morphogenesis and aging.
      ), which we also found when introducing senescent FBs (
      • Weinmüllner R.
      • Zbiral B.
      • Becirovic A.
      • Stelzer E.M.
      • Nagelreiter F.
      • Schosserer M.
      • et al.
      Organotypic human skin culture models constructed with senescent fibroblasts show hallmarks of skin aging.
      ). Conversely, certain natural compounds that extend the replicative lifespan and papillary phenotype of FBs support normal stratification (
      • Lämmermann I.
      • Terlecki-Zaniewicz L.
      • Weinmüllner R.
      • Schosserer M.
      • Dellago H.
      • de Matos Branco A.D.
      • et al.
      Blocking negative effects of senescence in human skin fibroblasts with a plant extract.
      ). This effect of senescent FBs on epidermal barrier formation is likely a result of the signaling of senescent cells via SASP, which contains, besides cyto- and chemokines, microRNA-containing extracellular vesicles shuttling from FBs to keratinocytes (
      • Terlecki-Zaniewicz L.
      • Lämmermann I.
      • Latreille J.
      • Bobbili M.R.
      • Pils V.
      • Schosserer M.
      • et al.
      Small extracellular vesicles and their miRNA cargo are anti-apoptotic members of the senescence-associated secretory phenotype.
      ,
      • Terlecki-Zaniewicz L.
      • Pils V.
      • Bobbili M.R.
      • Lämmermann I.
      • Perrotta I.
      • Grillenberger T.
      • et al.
      Extracellular vesicles in human skin: cross-talk from senescent fibroblasts to keratinocytes by miRNAs.
      ) and bioactive phospholipids (PLs) in senescent melanocytes (
      • Ni C.
      • Narzt M.S.
      • Nagelreiter I.M.
      • Zhang C.F.
      • Larue L.
      • Rossiter H.
      • et al.
      Autophagy deficient melanocytes display a senescence associated secretory phenotype that includes oxidized lipid mediators.
      ).
      Lipids of the skin and their chemical or enzymatic modifications (termed the epilipidome [
      • Ni Z.
      • Goracci L.
      • Cruciani G.
      • Fedorova M.
      Computational solutions in redox lipidomics - current strategies and future perspectives.
      ]) emerge as factors that have important functions beyond metabolism, biosynthesis of membranes, or as skin barrier components. High-resolution mass spectrometric techniques have identified, to our knowledge, previously unreported bioactive lipids and uncovered their contribution to skin development, to the regulation of skin inflammation, and to aging promoting stress (
      • Gruber F.
      • Kremslehner C.
      • Narzt M.S.
      The impact of recent advances in lipidomics and redox lipidomics on dermatological research.
      ;
      • Kendall A.C.
      • Koszyczarek M.M.
      • Jones E.A.
      • Hart P.J.
      • Towers M.
      • Griffiths C.E.M.
      • et al.
      Lipidomics for translational skin research: a primer for the uninitiated.
      ). The lipids are synthesized by enzymes (
      • Nicolaou A.
      • Pilkington S.M.
      • Rhodes L.E.
      Ultraviolet-radiation induced skin inflammation: dissecting the role of bioactive lipids.
      ;
      • Niki E.
      Lipid oxidation in the skin.
      ) or result directly from ROS-mediated oxidation (
      • Ademowo O.S.
      • Dias H.K.I.
      • Milic I.
      • Devitt A.
      • Moran R.
      • Mulcahy R.
      • et al.
      Phospholipid oxidation and carotenoid supplementation in Alzheimer's disease patients.
      ;
      • Gruber F.
      • Bicker W.
      • Oskolkova O.V.
      • Tschachler E.
      • Bochkov V.N.
      A simplified procedure for semi-targeted lipidomic analysis of oxidized phosphatidylcholines induced by UVA irradiation.
      ). We have identified that abrogation of autophagy changed the epilipidome of keratinocytes (
      • Song X.
      • Narzt M.S.
      • Nagelreiter I.M.
      • Hohensinner P.
      • Terlecki-Zaniewicz L.
      • Tschachler E.
      • et al.
      Autophagy deficient keratinocytes display increased DNA damage, senescence and aberrant lipid composition after oxidative stress in vitro and in vivo.
      ;
      • Zhao Y.
      • Zhang C.F.
      • Rossiter H.
      • Eckhart L.
      • König U.
      • Karner S.
      • et al.
      Autophagy is induced by UVA and promotes removal of oxidized phospholipids and protein aggregates in epidermal keratinocytes.
      ) and promoted senescence of melanocytes including a lipid SASP (
      • Ni C.
      • Narzt M.S.
      • Nagelreiter I.M.
      • Zhang C.F.
      • Larue L.
      • Rossiter H.
      • et al.
      Autophagy deficient melanocytes display a senescence associated secretory phenotype that includes oxidized lipid mediators.
      ;
      • Zhang C.F.
      • Gruber F.
      • Ni C.
      • Mildner M.
      • Koenig U.
      • Karner S.
      • et al.
      Suppression of autophagy dysregulates the antioxidant response and causes premature senescence of melanocytes.
      ). Oxidized PLs (OxPLs) comprise many known biologically active mediators that relay responses to external and internal stimuli and are capable of paracrine and systemic signaling. Originating from components of cell membranes, OxPLs may act as immunomodulatory lipid whiskers (
      • Greenberg M.E.
      • Li X.M.
      • Gugiu B.G.
      • Gu X.
      • Qin J.
      • Salomon R.G.
      • et al.
      The lipid whisker model of the structure of oxidized cell membranes.
      ) or as danger-associated molecular patterns (
      • Matt U.
      • Sharif O.
      • Martins R.
      • Knapp S.
      Accumulating evidence for a role of oxidized phospholipids in infectious diseases.
      ;
      • Miller Y.I.
      • Choi S.H.
      • Wiesner P.
      • Fang L.
      • Harkewicz R.
      • Hartvigsen K.
      • et al.
      Oxidation-specific epitopes are danger-associated molecular patterns recognized by pattern recognition receptors of innate immunity.
      ) present on cells or cell-released vesicles. Here, we investigated whether cellular senescence of normal dermal FBs would change their epiphospholipidome. We identified lysophosphatidylcholines (LysoPCs) as the most highly regulated species in replicative- and stress-induced premature cellular senescence. We also found that the relative content in LysoPCs (and eight other investigated OxPL species) of microdissected reticular fibroblasts was higher than in papillary FBs from the same donors. We furthermore found that LysoPCs reduced both uptake of dextrane and toll-like receptor (TLR) 2– and TLR6-mediated signaling in macrophages. Finally, we localized both identified LysoPC species prominently in aged human dermal tissue by mass spectrometric imaging. Thus, we propose that elevation of local levels of lysophospholipids and other senescence-associated lipid signaling mediators may, beyond a reported proinflammatory role of these lipids, allow senescent FBs to evade clearance by the innate immune system.

      Results

      Long-term culture of normal human dermal FBs from three donors led to replicative exhaustion and stop of cell division (Figure 1a). FBs developed a phenotype typical of cellular senescence as evidenced by characteristic cell morphology (Figure 1b), expression of p21 mRNA (Figure 1c), and positive staining for senescence-associated β-galactosidase (Figure 1d) and did not undergo cell death to a noticeable extent. These hallmarks of senescence were analyzed late in the replicative lifespan when population doubling (PD) stopped, which occurred with a wide donor-dependent variability. Samples for biochemical and lipidomic analyses were taken at early, middle, and late culture time points as indicated (Figure 1a, arrows).
      Figure thumbnail gr1
      Figure 1Epilipidomics of FBs in replicative senescence identify highly elevated LysoPCs. Human dermal FBs isolated from three different donors underwent a replicative senescence protocol. (a) Diagram charts PDs of FBs from three donors; arrows indicate sampling time points. (b) Representative cell micrographs of dermal FBs at early (E), mid-culture (M), and late (RS) timepoints over replicative lifespan. Bar = 100 μm. (c) Relative mRNA expression of p21 was quantified by using qPCR normalized to GAPDH. (d) Dot blot showing percentage of β-galactosidase–positive cells of early, mid-culture, and replicatively senescent FBs. n = 3; error bars indicate ± SD, asterisks indicate significant differences (∗P < 0.05; ∗∗P < 0.01) determined by one-way ANOVA. (e–h) OxPC species were quantified with HPLC MS/MS in lipid extracts from dermal fibroblasts at early, middle, and end stages of replicative lifespan. Abundance of lipid species was determined by integration of peak areas and normalization to peak area of intrinsic DPPC. (e) Heatmap representation of oxidized phospholipid species normalized peak areas. Columns represent stages replicative lifespan; rows show normalized relative log abundance (left) and fold change (right) of individual OxPC species (color code: normalized relative log abundance from low [orange] to high [green]; fold change: increase [red], decrease [blue]). (f–h) Dot blots showing abundance of the respective species of replicative senescent FBs normalized to DPPC. (f) LysoPPC. (g) LysoSPC. (h) PLPC. n = 3; error bars ± SD. Asterisks represent significant differences (∗P < 0.05; ∗∗P < 0.01) determined by one-way ANOVA. DPPC, 1,2-dipalmitoyl-sn-glycero-3-phosphocholine; E, early; FB, fibroblast; LysoPC, lysophosphatidylcholine; LysoPPC, 1-palmitoyl-2-lyso-sn-glycero-3-phosphorylcholine; LysoSPC, 1-stearoyl-2-lyso-sn-glycero-3-phosphorylcholine; M, mid-culture; MS/MS, tandem mass spectrometry; OxPC, oxidized phosphatidylcholine; PD, population doubling; PLPC, 1-palmitoyl-2-linoleoyl-sn-glycero-3-phosphorylcholine; RS, replicative senescence; SA-β-Gal, senescence-associated β-galactosidase.
      We next performed HPLC–tandem mass spectrometry–based targeted analysis of an array of native and oxidized phosphatidylcholine (PC) lipid species as in
      • Gruber F.
      • Bicker W.
      • Oskolkova O.V.
      • Tschachler E.
      • Bochkov V.N.
      A simplified procedure for semi-targeted lipidomic analysis of oxidized phosphatidylcholines induced by UVA irradiation.
      , yielding the epiphospholipidomes at three stages of replicative senescence. In the cellular lipid extracts, several oxidized species were elevated relative to di-palminoyl-sn-glycero-3-phosphorylcholine (DPPC, m/z 734) as a saturated, nonoxidized, internal control species. The overall highest abundance and highest-fold induction was observed for the lysophospholipids 1-palmitoyl-2-lyso-sn-glycero-3-phosphorylcholine (LysoPPC, m/z 496; Figure 1f) and 1-stearoyl-2-lyso-sn-glycero-3-phosphorylcholine (LysoSPC m/z 524; Figure 1g). The nonoxidized PL 1-palmitoyl-2-linoleoyl-sn-glycero-3-phosphorylcholine (PLPC, m/z 758) was not significantly changed (Figure 1h). We observed a significant increase in the arachidonic acid–containing unoxidized PL species 1-palmitoyl-2-arachidonoyl-sn-glycero-3-phosphorylcholine (PAPC, m/z 782) and 1-stearoyl-2-arachidonoyl-sn-glycero-3-phosphorylcholine (SAPC, m/z 810), and the hydroperoxides of PLPC, SLPC, and SAPC; SAPC-OH; and species we proposed as isoprostaglandin J2-PC and PC 18:1, 18:2-OH were significantly elevated (Figure 1e). A separate analysis of the cell culture supernatants showed that, in those, the ratio of LysoPC species to DPPC was not affected by senescence (Supplementary Figure S1a–c). However, because cultured senescent FBs release sixfold the amount of extracellular vesicles into the culture medium compared with quiescent cells (
      • Terlecki-Zaniewicz L.
      • Lämmermann I.
      • Latreille J.
      • Bobbili M.R.
      • Pils V.
      • Schosserer M.
      • et al.
      Small extracellular vesicles and their miRNA cargo are anti-apoptotic members of the senescence-associated secretory phenotype.
      ), we alternatively normalized the amount of secreted lipids to the number of cultured cells that had yielded the supernatants. This quantification showed that the overall secretion of PLs (LysoPCs and DPPC) per cell was elevated in senescent cells, whereas the ratio of LysoPCs to DPPC was, in contrast to intracellular levels, comparable in early passage and senescent cells (Supplementary Figure S1d–f). Comparison of selected species shows that also PAPC, SAPC, PAPC-OOH, PLPC-OOH, and isoprostaglandin J2-PPC, all significantly induced in replicatively senescent fibroblasts, were not elevated in the supernatants (Supplementary Figure S2 a–l).
      Next, we investigated whether other methods for induction of cellular senescence in vitro would also lead to similarly massive changes in lysophospholipids and other members of the epilipidome. We therefore applied two different stress-induced premature senescence (SIPS) protocols to cultured normal fibroblasts. The first protocol was based on repeated treatment with hydrogen peroxide (H2O2) and the second one was chemotherapy- (doxorubicin) induced senescence, both protocols followed by an 11- or 7-day, respectively, recovery period. Both protocols resulted in a senescent phenotype (Figure 2a and c), arrest of cell division (doxorubicin SIPS, Supplementary Figure S3), and induction of senescence markers (H2O2 SIPS, Figure 2b and d, Supplementary Table S1). In the cellular lipid extracts, doxorubicin induced a significant elevation of the lysophospholipid species (Figure 2h and i) and of numerous other full-length and chain-shortened OxPLs (Figure 2e). The changes to the lipidome after H2O2 SIPS were more subtle. Although a borderline significant increase in lysophospholipids also was observed, we found a significant increase of the nonoxidized arachidonoyl PC and of a species we proposed as isoprostaglandin J2-PC. Also, PLPC-OH (esterified hydroxyoctadecadienoic acid [HODE]) was significantly elevated. For SIPS FBs, a separate epilipidomic analysis of the cell culture supernatants was performed (Supplementary Figures S4 and S5). Here, LysoPCs were not or only borderline significantly elevated relative to DPPC (H2O2, Supplementary Figure S4b and c; doxorubicin, Supplementary Figure S4g and h), but overall secretion when normalized to cell numbers of both was significantly elevated (Supplementary Figure S4d–f and i–k, respectively). Here, we found that, relative to DPPC, the native PLPC was elevated in H2O2 SIPS, whereas the arachidonate-containing native PC, PLPC-OOH, SLPC-OOH, and PAPC-OH were significantly elevated in the supernatant of doxorubicin-senescent cells. We found LysoPPC and LysoSPC also elevated in the cellular lipidomes of FBs that had undergone a UVB senescence protocol (
      • Cavinato M.
      • Koziel R.
      • Romani N.
      • Weinmüllner R.
      • Jenewein B.
      • Hermann M.
      • et al.
      UVB-induced senescence of human dermal fibroblasts involves impairment of proteasome and enhanced autophagic activity.
      ) (Supplementary Figure S7).
      Figure thumbnail gr2
      Figure 2Epilipidomics of SIPS human dermal FBs identify differential lipid signatures. Human dermal fibroblasts were treated with H2O2 and doxorubicin protocols for SIPS. (a, c) Cell micrographs showing NT and SIPS-treated FBs. (a) H2O2. (c) Doxorubicin. Bar = 200 μm. (b, d) p21 mRNA expression was quantified relative to GAPDH using qPCR. (e) Heatmaps representing the normalized relative log abundance and fold change of oxidized phospholipid species of NT and SIPS-treated FBs (color code: normalized relative log abundance from low [orange] to high [green]; fold change: increase [red], decrease [blue]). (f–i) Dot blots showing values of (f, h) LysoPPC and (g, i) LysoSPC of SIPS FBs normalized to DPPC. n = 3; error bars indicate ± SD; asterisks represent significant differences (∗P < 0.05; ∗∗P < 0.01) determined by Student’s t-test. Doxo, doxorubicin; DPPC, 1,2-dipalmitoyl-sn-glycero-3-phosphocholine; FB, fibroblast; fold chg., fold change; H2O2, hydrogen peroxide; LysoPPC, 1-palmitoyl-2-lyso-sn-glycero-3-phosphorylcholine; LysoSPC, 1-stearoyl-2-lyso-sn-glycero-3-phosphorylcholine; norm. abund., normalized abundance; NT, nontreated; rel., relative; SIPS, stress-induced premature senescence.
      The human dermal compartment can be divided into an outer papillary layer with higher cellularity and the inner reticular layer containing phenotypically distinct FB lineages (
      • Korosec A.
      • Frech S.
      • Gesslbauer B.
      • Vierhapper M.
      • Radtke C.
      • Petzelbauer P.
      • et al.
      Lineage identity and location within the dermis determine the function of papillary and reticular fibroblasts in human skin.
      ). In aging, the papillary dermis becomes increasingly atrophic, which involves transdifferentiation, and papillary FBs develop a more reticular phenotype (
      • Lämmermann I.
      • Terlecki-Zaniewicz L.
      • Weinmüllner R.
      • Schosserer M.
      • Dellago H.
      • de Matos Branco A.D.
      • et al.
      Blocking negative effects of senescence in human skin fibroblasts with a plant extract.
      ;
      • Mine S.
      • Fortunel N.O.
      • Pageon H.
      • Asselineau D.
      Aging alters functionally human dermal papillary fibroblasts but not reticular fibroblasts: a new view of skin morphogenesis and aging.
      ). We investigated the intracellular epilipidome in FBs that had been microdissected and short-time cultured (Figure 3a) from papillary and reticular dermis and found that lysophospholipids (by trend), PLPC-OH, 1-palmitoyl-2-glutaroyl-sn-glycero-3-phosphocholine, 1-palmitoyl-2-azelaoyl-sn-glycero-3-phosphocholine, isoprostaglandin J2-PC, SAPC, and the hydroperoxides of PAPC and PLPC were elevated in reticular compared with papillary FBs (Figure 3b–d).
      Figure thumbnail gr3
      Figure 3The epilipidomics of microdissected papillary and reticular FBs identify specific OxPL patterns. (a) Micrographs of FBs isolated out of the papillary and reticular human dermis from three donors. Bar = 200 μm. (b) Heatmaps representing the normalized relative log abundance and fold change of OxPL species of papillary and reticular FBs (color code: normalized relative log abundance from low [orange] to high [green]; fold change: increase [red], decrease [blue]). (c, d) Dot blots showing relative amount of (c) LysoPPC and (d) LysoSPC of papillary and reticular FBs normalized to intrinsic DPPC. Significances of differences were determined by Student’s t-test. DPPC, 1,2-dipalmitoyl-sn-glycero-3-phosphocholine; FB, fibroblast; fold chg., fold change; HDF, human dermal fibroblast; LysoPPC, 1-palmitoyl-2-lyso-sn-glycero-3-phosphorylcholine; LysoSPC, 1-stearoyl-2-lyso-sn-glycero-3-phosphorylcholine; OxPL, oxidized phospholipid; pap, papillary; ret, reticular; rel., relative.
      Long-term cell culture can cause massive changes in cell identity, differentiation, gene regulation, and lipid content (
      • Waaijer M.E.C.
      • Gunn D.A.
      • van Heemst D.
      • Slagboom P.E.
      • Sedivy J.M.
      • Dirks R.W.
      • et al.
      Do senescence markers correlate in vitro and in situ within individual human donors?.
      ). Therefore, we compared the lipidomes of hTERT-immortalized FBs (at >100 PDs) to FBs from the same donor at culture passage (PD 26) or close to replicative exhaustion at PD 50 (Figure 4a and d). The hTERT-immortalized cells did not enter replicative senescence, in contrast to the nonimmortalized FBs from the same donor, which at PD 50 displayed elevated p21 mRNA expression (Figure 4b) and senescence-associated β-galactosidase activity (Figure 4c). Replicative senescence of these cells again led to massive accumulation of LysoPCs and most species that had been identified in the initial replicative senescence experiments (Figure 4d). By contrast, the hTERT FBs did not show significant elevation of lysophospholipids as compared with the PD 26 cells (Figure 4e and f), demonstrating that the epilipidomic changes in replicative senescence of FBs were indeed because of the senescent phenotype and not a consequence of long-term culture.
      Figure thumbnail gr4
      Figure 4Telomerase immortalization inhibits accumulation of LysoPC in long-term culture. (a) Micrographs of hTERT-immortalized FBs (PD >100) and from nonimmortalized FBs from the same donor at early (E, PD2 6) and late (RS, PD 50) replicative age. Bar = 400 μm. (b) Relative mRNA expression of p21 was quantified by using qPCR normalized to GAPDH. (c) Diagram shows percentage of β-galactosidase–positive cells of early PD (E, PD 26), replicative senescent (RS, PD 50), and hTERT-transfected (T , PD >100) FBs (n=10). (d) Heatmaps representing the normalized relative log abundance and fold change of oxidized phospholipid species of early PD (E), replicative senescent (RS), and hTERT-transfected (T) fibroblasts (color code: normalized relative log abundance from low [orange] to high [green]; fold change: increase [red], decrease [blue]). Dot blots showing relative amount of (e) LysoPPC and (f) LysoSPC of E, RS, and hTERT-transfected FBs normalized to intrinsic DPPC. For (b), (d), (e), (f): n = 3; error bars indicate ± SD, asterisks indicate significant differences between PD 26 (E) and PD 50 (RS) (∗P < 0.05; ∗∗P < 0.01), dollar signs indicate significant differences between PD 50 (RS) and hTERT (PD >100) ($P < 0.05; $$P < 0.01), determined by one-way ANOVA. DPPC, 1,2-dipalmitoyl-sn-glycero-3-phosphocholine; E, early; FB, fibroblast; LysoPC, lysophosphatidylcholine; LysoPPC, 1-palmitoyl-2-lyso-sn-glycero-3-phosphorylcholine; LysoSPC, 1-stearoyl-2-lyso-sn-glycero-3-phosphorylcholine; PD, population doubling; rel., relative; RS, replicative senescent; SA-β-Gal, senescence-associated β-galactosidase; T, hTERT-transfected.
      We next investigated whether LysoPCs might be members of SASP and might contribute to SASP responses in nonsenescent cells, such as inflammatory signaling and cyto- and chemokine secretion (
      • Nelson G.
      • Kucheryavenko O.
      • Wordsworth J.
      • von Zglinicki T.
      The senescent bystander effect is caused by ROS-activated NF-κB signalling.
      ;
      • Shimizu R.
      • Kanno K.
      • Sugiyama A.
      • Ohata H.
      • Araki A.
      • Kishikawa N.
      • et al.
      Cholangiocyte senescence caused by lysophosphatidylcholine as a potential implication in carcinogenesis.
      ). First, we tested whether early PD (PD 15) dermal FBs would react to exposure to LysoPCs by secreting proinflammatory cyto- and chemokines and found, only at high doses of 50 μg/ml of LysoPPC and LysoSPC, a low-grade yet significant elevation of IL- 8 protein secretion after 24 hours (Figure 5a–f).
      Figure thumbnail gr5
      Figure 5LysoPCs inhibit signaling through the senescent cell PRRs TLR2/6 CD36 in monocytes and phagocytosis of dextran in macrophages. Human dermal FBs were treated with indicated doses of (a, d) LysoPPC, (b, e) LysoSPC, or (c, f) DPPC for 4 hours. The supernatants were harvested and Luminex analysis was performed for (a–c) IL-6 and (d–f) IL-8 independently; n = 4. (g) Gene expression analysis of U937 cells pretreated with 10 μM LysoPPC and further treated with 1 μg/ml FSL-1 was performed. Graph shows the fold change of the top 10 regulated genes compared with control; n = 4. (h, i) THP-1 and U937 cells incubated with PMA (100 mM) for 72 hours were treated with either OxPAPC (25 μg/ml), LysoPPC (25 μg/ml), or DPPC (25 μg/ml) or stayed untreated to serve as control and were further incubated with AF647-dextran at 37 °C or 4 °C, respectively. Dot blots show mean intensity of AF647-dextran uptake of (h) THP-1 cells and (i) U937 cells analyzed by FACS; n = 3. Error bars indicate ± SD, asterisks represent significant differences (∗P < 0.05; ∗∗P < 0.01) determined by two-way ANOVA and one-way ANOVA for the dextran uptake. ctrl, control; DPPC, 1,2-dipalmitoyl-sn-glycero-3-phosphocholine; FB, fibroblast; LysoPC, lysophosphatidylcholine; LysoPPC, 1-palmitoyl-2-lyso-sn-glycero-3-phosphorylcholine; LysoSPC, 1-stearoyl-2-lyso-sn-glycero-3-phosphorylcholine; OxPAPC, oxidized 1-palmitoyl-2-arachidonoyl-sn-glycero-3-phosphorylcholine; PMA, phorbol 12-myristate 13-acetate; PRR, pattern recognition receptor; TLR, toll-like receptor.
      OxPLs can have agonistic or antagonistic effects on monocytes and monocyte-derived macrophages (
      • Bochkov V.
      • Gesslbauer B.
      • Mauerhofer C.
      • Philippova M.
      • Erne P.
      • Oskolkova O.V.
      Pleiotropic effects of oxidized phospholipids.
      ) and these, besides NK cells, are crucial for elimination of senescent cells (
      • Kale A.
      • Sharma A.
      • Stolzing A.
      • Desprez P.Y.
      • Campisi J.
      Role of immune cells in the removal of deleterious senescent cells.
      ). Lipids recognized by CD36 accumulate in senescent cells (
      • Li X.M.
      • Salomon R.G.
      • Qin J.
      • Hazen S.L.
      Conformation of an endogenous ligand in a membrane bilayer for the macrophage scavenger receptor CD36.
      ), and TLR2/6- and CD36-mediated signaling has been implicated in senescence (
      • Jin H.
      • Zhang Y.
      • Liu D.
      • Wang S.S.
      • Ding Q.
      • Rastogi P.
      • et al.
      Innate immune signaling contributes to tubular cell senescence in the Glis2 knockout mouse model of nephronophthisis.
      ). The interplay of these scavenger receptors is required for uptake of the TLR2/6 agonist FSL1 (
      • Abe T.
      • Shimamura M.
      • Jackman K.
      • Kurinami H.
      • Anrather J.
      • Zhou P.
      • et al.
      Key role of CD36 in toll-like receptor 2 signaling in cerebral ischemia.
      ). We investigated whether lysophospholipids would affect (TLR2/6-mediated) gene regulation in U937 monocytes. Surprisingly, LysoPPC did not lead to significant changes in the transcriptome of the cells 3 hours after exposure, whereas the TLR2/6 ligand FSL1 significantly regulated 119 genes, among which those with functions in inflammatory signaling and dendritic cell maturation (Supplementary Figure S8a and b) were significantly enriched. Coincubation of FSL1 with LysoPPC significantly reduced FSL1-mediated CXCL10 (IP-10) and TNFAIP6 mRNA expression, whereas the other targets of FSL1 were not affected (Figure 5g).
      Next, we investigated whether the presence of LysoPPC would promote or inhibit the phagocytic capacity of monocyte-derived macrophages. We used U937 and THP-1 monocytes that had been differentiated to a macrophage phenotype with a 72-hour phorbol 12-myristate 13-acetate exposure standard protocol (
      • Tsuchiya S.
      • Kobayashi Y.
      • Goto Y.
      • Okumura H.
      • Nakae S.
      • Konno T.
      • et al.
      Induction of maturation in cultured human monocytic leukemia cells by a phorbol diester.
      ) and then measured the uptake of fluorescently labeled dextran in the presence or absence of 1-hour pre-exposure to LysoPC, autoxidized PAPC, and DPPC. Both LysoPPC and autoxidized PAPC, which contains LysoPPC (
      • Watson A.D.
      • Leitinger N.
      • Navab M.
      • Faull K.F.
      • Hörkkö S.
      • Witztum J.L.
      • et al.
      Structural identification by mass spectrometry of oxidized phospholipids in minimally oxidized low density lipoprotein that induce monocyte/endothelial interactions and evidence for their presence in vivo.
      ), led to a significant reduction of dextran uptake, as compared with controls or DPPC-exposed cells (Figure 5h and i; Supplementary Figure S9a–c).
      Finally, we performed an exploratory experiment where we applied matrix-assisted laser desorption/ionization Fourier transform ion cyclotron resonance mass spectrometry imaging (MALDI-FTICR-MSI) after well-established sample preparation (
      • Holzlechner M.
      • Bonta M.
      • Lohninger H.
      • Limbeck A.
      • Marchetti-Deschmann M.
      Multisensor imaging-from sample preparation to integrated multimodal interpretation of LA-ICPMS and MALDI MS imaging data.
      ) on human skin from abdominal body lifts from a young (31 years) and an advanced age (63 years) female donor. We performed imaging in the positive ion mode at 40-μm and 10-μm spatial resolution. By comparing these with the consecutive sections that were H&E-stained (Figure 6a), we could align the ion images tentatively assigned for native LysoPPC, LysoSPC, and DPPC in the whole section at 40-μm resolution (Figure 6a–d). High mass resolution and accuracy allows for tentative analyte assignment, but isobaric lipids cannot clearly be distinguished (for isobaric species, see Supplementary Table S2). However, our tentative assignment is corroborated by the fact that other isobaric compounds such as carnitines will not be detected under the given experimental conditions, and sodium adduct ions are not very likely because the tissue is carefully washed and the matrix applied by sublimation. Although HPLC–tandem mass spectrometry was performed using the PC-specific fragment ion, we here cannot fully rule out codetection of isobaric phosphatidylethanolamines, which are less abundant PL species in FBs (
      • Tavasoli M.
      • Lahire S.
      • Reid T.
      • Brodovsky M.
      • McMaster C.R.
      Genetic diseases of the Kennedy pathway for phospholipid synthesis [e-pub ahead of print].
      ). Therefore, in support of our in vitro lipidomics data, both LysoPPC and LysoSPC appeared to display a stronger diffuse signal throughout the dermis in the aged sample, whereas localized strong signals for all investigated PLs were likely associated with regions of higher dermal cellularity that we cannot identify or correlate properly at this resolution. The second higher resolution experiment (10 μm) confirmed the strong diffuse dermal staining for LysoPCs, especially in the aged skin (Figure 6e–g, asterisks). In the skin of both young and old donors, signals of comparable relative intensities for PLs were detected in the clearly discernible epidermal area (Figure 6e–g, arrowheads). Overall, it can be said that ion abundances suggest relatively higher concentrations for LysoPPC than LysoSPC in aged skin (Supplementary Table S3 and Supplementary Figure S10). Further studies will correlate the distribution of the lipids with the localization of both senescent cells and cells of the immune system (
      • Holzlechner M.
      • Strasser K.
      • Zareva E.
      • Steinhäuser L.
      • Birnleitner H.
      • Beer A.
      • et al.
      In situ characterization of tissue-resident immune cells by MALDI mass spectrometry imaging.
      ). The comparison of just two samples does not allow general conclusions on quantitative distribution of the lipid species; however, we could localize potential SASP factors that are, to our knowledge, previously unreported within human skin using molecular imaging.
      Figure thumbnail gr6
      Figure 6MALDI-FTICR-MSI of human skin cryosections localizes LysoPPC and LysoSPC in human skin. Tissue sections (10 μm thickness) of abdominal body lift material biopsies from a female 31-year-old and a female 63-year-old patient were prepared and serial sections were mounted on glass slides for H&E staining and on indium tin-oxide–covered glass slides for MSI. (a) H&E staining. (b–d) 40-μm resolution positive mode MALDI-FTICR-MSI of (b) LysoPPC, (c) LysoSPC, and (d) DPPC. (e–g) 10-μm resolution positive mode MALDI-FTICR-MSI images of (e) LysoPPC, (f) LysoSPC, and (g) DPPC in biopsies from 31-year-old and 63-year-old skin overlaid over brightfield scan of the matrix-sublimated tissue. Bar = 500 μm (a–d). Bar = 200 μm (e–g). Color scheme bars indicate the TIC normalized relative intensity distribution for each ion from 0% (black) to 100% (yellow). DPPC, 1,2-dipalmitoyl-sn-glycero-3-phosphocholine; LysoPPC, 1-palmitoyl-2-lyso-sn-glycero-3-phosphorylcholine; LysoSPC, 1-stearoyl-2-lyso-sn-glycero-3-phosphorylcholine; MALDI-FTICR-MSI, matrix-assisted laser desorption/ionization Fourier transform ion cyclotron resonance mass spectrometry imaging; TIC, total ion count.

      Discussion

      Timely clearance of senescent cells from tissue is required to limit aging-associated tissue damage that would be induced by persistent sources of SASP factors. Although protein components of SASP and their localization to the cell surface, exosomes, or extracellular space are already meticulously investigated and catalogued, (
      • Basisty N.
      • Kale A.
      • Jeon O.H.
      • Kuehnemann C.
      • Payne T.
      • Rao C.
      • et al.
      A proteomic atlas of senescence-associated secretomes for aging biomarker development.
      ), other molecule classes that might fulfill similar or complementary functions in senescent cell signaling to the immune system are barely investigated. NK cells, macrophages, and neutrophils express pattern recognition receptors that detect senescent cells (
      • Sagiv A.
      • Krizhanovsky V.
      Immunosurveillance of senescent cells: the bright side of the senescence program.
      ) and help to eliminate them (
      • Kale A.
      • Sharma A.
      • Stolzing A.
      • Desprez P.Y.
      • Campisi J.
      Role of immune cells in the removal of deleterious senescent cells.
      ), but increasing age (
      • Karin O.
      • Agrawal A.
      • Porat Z.
      • Krizhanovsky V.
      • Alon U.
      Senescent cell turnover slows with age providing an explanation for the Gompertz law.
      ) or conditions such as type 1 diabetes (
      • Thompson P.J.
      • Shah A.
      • Ntranos V.
      • Van Gool F.
      • Atkinson M.
      • Bhushan A.
      Targeted elimination of senescent beta cells prevents type 1 diabetes.
      ) impair the clearance. Here, we have found that senescent normal dermal FBs accumulate biologically active OxPL species, most prominently LysoPCs. The known biological activities of these PLs on monocyte-derived cells including macrophages suggests that they might present novel pleiotropic factors of SASP of dermal FBs.
      In aging, ROS-mediated oxidized lipids accumulate in skin FBs (
      • Frescas D.
      • Roux C.M.
      • Aygun-Sunar S.
      • Gleiberman A.S.
      • Krasnov P.
      • Kurnasov O.V.
      • et al.
      Senescent cells expose and secrete an oxidized form of membrane-bound vimentin as revealed by a natural polyreactive antibody.
      ;
      • Jørgensen P.
      • Milkovic L.
      • Zarkovic N.
      • Waeg G.
      • Rattan S.I.
      Lipid peroxidation-derived 4-hydroxynonenal-modified proteins accumulate in human facial skin fibroblasts during ageing in vitro.
      ;
      • Lizardo D.Y.
      • Lin Y.L.
      • Gokcumen O.
      • Atilla-Gokcumen G.E.
      Regulation of lipids is central to replicative senescence.
      ). Elevated cyclooxygenase-2 and 5-lipoxygenase activity (
      • Catalano A.
      • Rodilossi S.
      • Caprari P.
      • Coppola V.
      • Procopio A.
      5-Lipoxygenase regulates senescence-like growth arrest by promoting ROS-dependent p53 activation.
      ;
      • Zdanov S.
      • Bernard D.
      • Debacq-Chainiaux F.
      • Martien S.
      • Gosselin K.
      • Vercamer C.
      • et al.
      Normal or stress-induced fibroblast senescence involves COX-2 activity.
      ) and synthesis of leukotrienes was found in senescent lung FBs (
      • Wiley C.D.
      • Brumwell A.N.
      • Davis S.S.
      • Jackson J.R.
      • Valdovinos A.
      • Calhoun C.
      • et al.
      Secretion of leukotrienes by senescent lung fibroblasts promotes pulmonary fibrosis.
      ). Linoleic acid hydroxides (HODE) have been found among metabolites elevated in senescent FBs (
      • James E.L.
      • Lane J.A.
      • Michalek R.D.
      • Karoly E.D.
      • Parkinson E.K.
      Replicatively senescent human fibroblasts reveal a distinct intracellular metabolic profile with alterations in NAD+ and nicotinamide metabolism.
      ). Leukotriene and HODE synthesis by phospholipases and lipoxygenases yields, as a by-product, lysophospholipids; however, selected HODE can be synthesized by lipoxygenases on PL-esterified linoleic acid. Epilipidomic analysis, as we performed here, of the lipid danger-associated molecular patterns is a prerequisite for elucidating their genesis and for later attributing specific activities to single species (
      • Gruber F.
      • Kremslehner C.
      • Narzt M.S.
      The impact of recent advances in lipidomics and redox lipidomics on dermatological research.
      ;
      • Mauerhofer C.
      • Philippova M.
      • Oskolkova O.V.
      • Bochkov V.N.
      Hormetic and anti-inflammatory properties of oxidized phospholipids.
      ;
      • O'Donnell V.B.
      • Aldrovandi M.
      • Murphy R.C.
      • Krönke G.
      Enzymatically oxidized phospholipids assume center stage as essential regulators of innate immunity and cell death.
      ), and MALDI-FTICR-MSI is a powerful technology for label-free identification, verification, and colocalization by multimodal imaging not only of lipids but many other molecules relevant in skin biology. The elevation of LysoPC and several minor species could be prevented by telomerase overexpression, indicating that the changes to the epilipidome were not the consequence of long-term culture. H2O2 SIPS yielded only borderline elevation of LysoPCs but elevation of nonoxidized arachidonate PCs, PLPC-OH (HODE-PC), and PC-esterified isoprostaglandin J2 intracellularly. In contrast, doxorubicin-induced SIPS led to a massive cell-associated lipid peroxidation signature, as demonstrated by high levels of PL-OOH and the downstream autoxidation products, which is in line with a report of lipid peroxidation after doxorubicin-mediated ROS generation (
      • Hrelia S.
      • Fiorentini D.
      • Maraldi T.
      • Angeloni C.
      • Bordoni A.
      • Biagi P.L.
      • et al.
      Doxorubicin induces early lipid peroxidation associated with changes in glucose transport in cultured cardiomyocytes.
      ), and apparently is a consequence of senescence rather than a direct effect, as the treatment had been terminated a week earlier. In the supernatants, the hydroperoxides of SLPC and PLPC and PAPC-OH were found elevated relative to DPPC and without normalizing to cell numbers, making them top candidates for further studies on PL SASP factors. H2O2 SIPS affected PL synthetic genes in FBs, as a reanalysis of earlier published RNA sequencing data showed, including induction of several phospholipases (Supplementary Table S1) and the functional contribution of those to the lipid SASP is under investigation. Both SIPS protocols and our UVB senescence protocol induced telomere shortening–independent (
      • Gorbunova V.
      • Seluanov A.
      • Pereira-Smith O.M.
      Expression of human telomerase (hTERT) does not prevent stress-induced senescence in normal human fibroblasts but protects the cells from stress-induced apoptosis and necrosis.
      ) senescence, and the replicative senescence protocol only showed elevation of oxidized PC species late in the protocol. Therefore, it seems that the stress- and damage-induced lipid species are not causative for senescence. Rather, these protocols yield footprints of bioactive lipids with some specificity regarding the cause of senescence induction in the cell. Both the hydroperoxide SIPS and the reticular phenotype of FBs have in common that the LysoPC increase is less prominent, but that an elevation of esterified isoprostanoids and HODE (PLPC-OH) was observed intracellularly. The acquisition of a reticular phenotype of dermal FBs is further characterized by increased 1-palmitoyl-2-glutaroyl-sn-glycero-3-phosphocholine and 1-palmitoyl-2-azelaoyl-sn-glycero-3-phosphocholine, two PC-esterified dicarboxylic acid species that were not observed elevated in the senescent FBs. Of note, treatment with free azelaic acid can reduce a SIPS phenotype in FBs (
      • Briganti S.
      • Flori E.
      • Mastrofrancesco A.
      • Kovacs D.
      • Camera E.
      • Ludovici M.
      • et al.
      Azelaic acid reduced senescence-like phenotype in photo-irradiated human dermal fibroblasts: possible implication of PPARγ.
      ).
      The composition of the oxidized lipids that are secreted does differ somewhat from the cell-associated one and points to cellular retention (or reuptake) of the species, especially of LysoPCs, and this suggests that further studies will need to address this gradient in the lipid SASP and its consequences on clearance of senescent cells from the tissue. Especially as low plasma LysoPC levels, including LysoPPC, have been negatively correlated with mitochondrial function in a recent longitudinal aging study (
      • Semba R.D.
      • Zhang P.
      • Adelnia F.
      • Sun K.
      • Gonzalez-Freire M.
      • Salem Jr., N.
      • et al.
      Low plasma lysophosphatidylcholines are associated with impaired mitochondrial oxidative capacity in adults in the Baltimore Longitudinal Study of Aging.
      ), the tissue and systemic distribution of the various species appear to be relevant. High-resolution imaging mass spectrometry combined with other imaging modalities will be one valuable tool to address these questions. It will also identify whether the in vitro generated lipid SASP factors can be observed in vivo, an important question given that protein SASP only partially overlaps (
      • Waldera Lupa D.M.
      • Kalfalah F.
      • Safferling K.
      • Boukamp P.
      • Poschmann G.
      • Volpi E.
      • et al.
      Characterization of skin aging-associated secreted proteins (SAASP) produced by dermal fibroblasts isolated from intrinsically aged human skin.
      ), and also whether the generally elevated secretion of PLs by senescent cells is reflected in the microenvironment of tissue-resident senescent cells. Finally, also other lipid classes will be (co)detectable using MSI, which will add tissue localization to recently reported functions of age-regulated sphingolipids in mediating senescence (
      • Trayssac M.
      • Hannun Y.A.
      • Obeid L.M.
      Role of sphingolipids in senescence: implication in aging and age-related diseases.
      ;
      • Wennberg A.M.V.
      • Schafer M.J.
      • LeBrasseur N.K.
      • Savica R.
      • Bui H.H.
      • Hagen C.E.
      • et al.
      Plasma sphingolipids are associated with gait parameters in the Mayo Clinic study of aging.
      ).
      We have identified that autoxidized PAPC can interfere with the signaling of lipopolysaccharide via the pattern recognition receptor TLR4/CD14 (
      • Bochkov V.N.
      • Kadl A.
      • Huber J.
      • Gruber F.
      • Binder B.R.
      • Leitinger N.
      Protective role of phospholipid oxidation products in endotoxin-induced tissue damage.
      ). There is evidence that TLR2/6, together with the lipid receptor CD36 (
      • Jimenez-Dalmaroni M.J.
      • Xiao N.
      • Corper A.L.
      • Verdino P.
      • Ainge G.D.
      • Larsen D.S.
      • et al.
      Soluble CD36 ectodomain binds negatively charged diacylglycerol ligands and acts as a co-receptor for TLR2.
      ;
      • Hoebe K.
      • Georgel P.
      • Rutschmann S.
      • Du X.
      • Mudd S.
      • Crozat K.
      • et al.
      CD36 is a sensor of diacylglycerides.
      ), may be the pattern recognition receptors that are most involved in the sensing of epitopes connected to cellular aging, senescence, and severe cellular stress (
      • Jin H.
      • Zhang Y.
      • Liu D.
      • Wang S.S.
      • Ding Q.
      • Rastogi P.
      • et al.
      Innate immune signaling contributes to tubular cell senescence in the Glis2 knockout mouse model of nephronophthisis.
      ;
      • Karuppagounder V.
      • Giridharan V.V.
      • Arumugam S.
      • Sreedhar R.
      • Palaniyandi S.S.
      • Krishnamurthy P.
      • et al.
      Modulation of macrophage polarization and HMGB1-TLR2/TLR4 cascade plays a crucial role for cardiac remodeling in senescence-accelerated prone mice.
      ;
      • Witztum J.L.
      CEP is an important and ubiquitous oxidation specific epitope recognized by innate pattern recognition receptors.
      ). Papillary FBs differ from reticular FBs in their adipogenic differentiation potential (
      • Korosec A.
      • Frech S.
      • Gesslbauer B.
      • Vierhapper M.
      • Radtke C.
      • Petzelbauer P.
      • et al.
      Lineage identity and location within the dermis determine the function of papillary and reticular fibroblasts in human skin.
      ), and CD36 is expressed on reticular but not on papillary dermal FBs and can induce senescence (
      • Chong M.
      • Yin T.
      • Chen R.
      • Xiang H.
      • Yuan L.
      • Ding Y.
      • et al.
      CD36 initiates the secretory phenotype during the establishment of cellular senescence.
      ). We observed impairment of TLR2/6-mediated signaling through presence of lysophospholipids, together with impairment of the phagocytic capacity in the presence of LysoPC and autoxidized PAPC, suggesting that phagocytes that come in contact with increased localized amounts of LysoPC in or around senescent cells may have reduced capacity to sense these via pattern recognition receptor and to phagocytose the cells. Our results underline that oxidized PAPC and lysophospholipids can partially interfere with gene regulation elicited by binding of the specific agonist FSL-1 to the TLR2/6 receptor. Of note, the TLR2/6 agonism-induced gene CXCL10, which itself was recognized as a SASP factor (
      • Perrott K.M.
      • Wiley C.D.
      • Desprez P.Y.
      • Campisi J.
      Apigenin suppresses the senescence-associated secretory phenotype and paracrine effects on breast cancer cells.
      ), was significantly repressed by the presence of LysoPPC during agonist treatment. CXCL10 serum levels are increased in aging (
      • Antonelli A.
      • Rotondi M.
      • Fallahi P.
      • Ferrari S.M.
      • Paolicchi A.
      • Romagnani P.
      • et al.
      Increase of CXC chemokine CXCL10 and CC chemokine CCL2 serum levels in normal ageing.
      ), and one could expect that this mechanism could contribute to immune evasion, because reduction of CXCL10 would dampen a potential T helper type 1 response.
      Bioactive lipids attract phagocytes and orchestrate engulfment of dying cells (
      • Medina C.B.
      • Ravichandran K.S.
      Do not let death do us part: 'find-me' signals in communication between dying cells and the phagocytes.
      ;
      • Quinn M.T.
      • Parthasarathy S.
      • Steinberg D.
      Lysophosphatidylcholine: a chemotactic factor for human monocytes and its potential role in atherogenesis.
      ). Our data indicate that LysoPC, in combination with other secreted OxPL species, could combine a chemotactic “find me” signal, amplified by IL-8, with a “don’t touch me” mechanism to evade clearance by the immune system, a feature of persistent senescent cells in aged tissues (
      • Demaria M.
      • Desprez P.Y.
      • Campisi J.
      • Velarde M.C.
      Cell autonomous and non-autonomous effects of senescent cells in the skin.
      ). An interesting parallel is observed in the atherosclerotic plaque, where macrophages exposed to high local doses of OxPLs take these up, become foam cells, and turn senescent with a chronic low-grade inflammatory phenotype themselves (
      • Childs B.G.
      • Baker D.J.
      • Wijshake T.
      • Conover C.A.
      • Campisi J.
      • van Deursen J.M.
      Senescent intimal foam cells are deleterious at all stages of atherosclerosis.
      ). We have described a macrophage subpopulation with decreased phagocytic and chemotactic capacity (termed MOx) within murine aortic atherosclerotic lesions (
      • Kadl A.
      • Meher A.K.
      • Sharma P.R.
      • Lee M.Y.
      • Doran A.C.
      • Johnstone S.R.
      • et al.
      Identification of a novel macrophage phenotype that develops in response to atherogenic phospholipids via Nrf2.
      ) that is inducible by OxPLs. The roles are, however, not restricted to immunomodulation, and we and others have identified potential roles of LysoPC and other oxidized PC species in senescence or FB differentiation that include wound healing, which is both promoted by transient senescent cells but impaired in aging (
      • Demaria M.
      • Ohtani N.
      • Youssef S.A.
      • Rodier F.
      • Toussaint W.
      • Mitchell J.R.
      • et al.
      An essential role for senescent cells in optimal wound healing through secretion of PDGF-AA.
      ). We have found a potential role of exosomal OxPLs (some of which are also present in SASP) in tissue regeneration after wounding (
      • Beer L.
      • Zimmermann M.
      • Mitterbauer A.
      • Ellinger A.
      • Gruber F.
      • Narzt M.S.
      • et al.
      Analysis of the secretome of apoptotic peripheral blood mononuclear cells: impact of released proteins and exosomes for tissue regeneration.
      ), and LysoPCs have been shown to disturb collagen synthesis and promote fibrosis (
      • Tseng H.C.
      • Lin C.C.
      • Hsiao L.D.
      • Yang C.M.
      Lysophosphatidylcholine-induced mitochondrial fission contributes to collagen production in human cardiac fibroblasts.
      ).
      Together, these data suggest that ROS-mediated metabolic and enzymatic mechanisms contribute to fundamental changes of the FB lipidome in cellular senescence. Among these are oxidation-specific lipid epitopes (
      • Binder C.J.
      • Papac-Milicevic N.
      • Witztum J.L.
      Innate sensing of oxidation-specific epitopes in health and disease.
      ) or danger-associated molecular patterns that qualify as FB SASP components. Thus, lipid metabolites might contribute to the detrimental low-grade chronic inflammation associated with persistent senescent cells and their SASP. However, they might in addition exert an inhibitory activity on phagocytosis and, in consequence, the ability of macrophages to clear the senescent cells. If such a “look, don’t touch” activity of SASP might also be present in vivo remains to be elucidated. However, our first application of MALDI-FTICR-MSI of (lipid) SASP factors hints at a potentially increased presence of LysoPCs in aged human skin.

      Materials and Methods

      Cell culture and tissue

      Human dermal FBs isolated from female donors between the ages of 49 and 65 were obtained from Evercyte (Vienna, Austria), and site-matched papillary and reticular human dermal FBs were isolated from the dermis of healthy donors by the Department of Dermatology of the Leiden University Medical Center (Leiden, The Netherlands) according to article 467 of the Dutch Law on Medical Treatment Agreement as described (
      • Janson D.
      • Saintigny G.
      • Mahé C.
      • El Ghalbzouri A.
      Papillary fibroblasts differentiate into reticular fibroblasts after prolonged in vitro culture.
      ), and their identity was confirmed by measurement of the expression levels of three papillary and three reticular mRNA markers described previously (
      • Lämmermann I.
      • Terlecki-Zaniewicz L.
      • Weinmüllner R.
      • Schosserer M.
      • Dellago H.
      • de Matos Branco A.D.
      • et al.
      Blocking negative effects of senescence in human skin fibroblasts with a plant extract.
      ). Telomerase-immortalized FBs and corresponding control FBs had been isolated previously from neonatal foreskin in the Pereira-Smith laboratory (
      • Ferenac M.
      • Polancec D.
      • Huzak M.
      • Pereira-Smith O.M.
      • Rubelj I.
      Early-senescing human skin fibroblasts do not demonstrate accelerated telomere shortening.
      ). The skin samples for the cryosections used in this study were approved by the Ethics Committee of the Medical University of Vienna (1149/2016), and written informed consent was obtained from all subjects. The human myeloid leukemia cell line U937 (
      • Sundstrom C.
      • Nilsson K.
      Establishment and characterization of a human histiocytic lymphoma cell line (U-937).
      ) and the human monocytic cell line THP1 were both obtained from ATCC (Manassas, VA). For detailed information, see Supplementary Material.

      Induction of SIPS by H2O2 and doxorubicin

      Middle-aged human dermal FB cells (PD 10–21) were seeded at cell densities of 2,800 to 3,500 cells/cm2 and subsequently treated with H2O2 (216763, Sigma-Aldrich, St. Louis, MO) in sublethal concentrations (60–100 μM) supplemented to their growth media for 1 hour/day. Cells were treated for 8 to 9 days with 2 days of recovery in between as in
      • Chen Q.M.
      • Tu V.C.
      • Catania J.
      • Burton M.
      • Toussaint O.
      • Dilley T.
      Involvement of Rb family proteins, focal adhesion proteins and protein synthesis in senescent morphogenesis induced by hydrogen peroxide.
      . For doxorubicin treatment, cells were seeded at 3,500 cells/cm2 and subsequently treated with 100 nM doxorubicin for 6 days. After a recovery phase of 7 days, cells were used for experiments as in
      • Demidenko Z.N.
      • Blagosklonny M.V.
      Growth stimulation leads to cellular senescence when the cell cycle is blocked.
      . UVB treatment is detailed in the Supplementary Material.

      MALDI-FTICR-MSI

      Tissue sections from OCT-embedded abdominal body lift material biopsies were covered with 1,5-diaminonaphtalene (
      • Holzlechner M.
      • Bonta M.
      • Lohninger H.
      • Limbeck A.
      • Marchetti-Deschmann M.
      Multisensor imaging-from sample preparation to integrated multimodal interpretation of LA-ICPMS and MALDI MS imaging data.
      ). MALDI-FTICR-MSI experiments were performed on a 7T scimaX MRMS (Bruker Daltonik, Bremen, Germany) in the positive ion mode. For detailed information, see Supplementary Material.
      Luminex analysis, transcriptome analysis of U937, lipid isolation and analysis, dextran uptake, qPCR, β-galactosidase staining, and preparation of graphs is provided in the Supplementary Material.

      Data availability statement

      Data related to this article can be found at https://www.ncbi.nlm.nih.gov/gds/?term=GSE93535[Accession], hosted at NCBI Gene Expression Omnibus website (accession number GSE93535) and at https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE157588, hosted at NCBI Gene Expression Omnibus website (accession number GSE157588).

      Conflict of Interest

      JG is a cofounder of Evercyte GmbH, Vienna, Austria; JG has received financial support of the Federal Ministry for Digital and Economic Affairs (BMWFW) of Austria; the National Foundation for Research, Technology, and Development of Austria; CHANEL Parfums et Beauté to the Christian Doppler Laboratory for Biotechnology of Skin Aging; and the Austrian Science Fund PhD Program BioToP–Biomolecular Technology of Proteins (W1224). JG is listed as one of the inventors on a patent application filed. The remaining authors state no conflict of interest.

      Acknowledgments

      The financial support of the Federal Ministry for Digital and Economic Affairs (BMWFW) of Austria; the National Foundation for Research, Technology, and Development of Austria; and CHANEL Parfums et Beauté to the Christian Doppler Laboratory for Biotechnology of Skin Aging and the Christian Doppler Laboratory for Skin Multimodal Imaging of Aging and Senescence is gratefully acknowledged. The support of the Herzfelder'sche Familienstiftung , Austria is gratefully acknowledged. These funding bodies do not issue grant numbers. VP acknowledges support by the Austrian Science Fund PhD Program BioToP–Biomolecular Technology of Proteins (W1224).The research at TU Wien was supported by the TU Wien doctoral program MEIBio (Molecular and Elemental Imaging in Life Sciences), the Federal Ministry Republic of Austria for Education, Science and Research (HRSM 2016). MMD acknowledges the support by the Austrian Bioimaging Initiative for Correlated Multimodal Imaging, and MMD and FG acknowledge the support by European Cooperation in Science and Technology as members of Correlated Multimodal Imaging in Life Sciences (CA17121) and FG as member of EpiLipiNet (CA 19105). LN acknowledges the support by an EMBO Short Term Fellowship and a joint mobility grant by Centre for International Cooperation & Mobility of the Austrian Agency for International Cooperation in Education and Research and the Croatian Science Foundation Project No. HR 16/2020. We would like to acknowledge the help of Markus Jeitler and the Transcriptomics Core Facility of the Medical University of Vienna. Correspondence regarding all aspects of this manuscript should be addressed to either to FG ( [email protected] ) or JG ( [email protected] ).

      Author Contributions

      Conceptualization: FG, JG, MMD, PJD, IR; Data Curation: IL; Formal Analysis: MN, VP, EB, AB, RR, SZ, MMD, LZT; Funding Acquisition: FG, JG, PJD; Investigation: MN, VP, CK, PWS, MCN, SW, LN, SZ; Methodology: IMN, MS, PWS, MCN, LN, AEG; Project Administration: FG, JG, MMD; Resources: IMN, AEG; Supervision: MS, MM, ET; Validation: FG, JG; Visualization: CK, VP; Writing - Original Draft Preparation: FG, IL; Writing - Review and Editing: ET, JG, MMD.

      Supplementary Material

      Figure thumbnail fx2
      Supplementary Figure S1Epilipidomics of secreted PC species of FBs in replicative senescence. HDFs isolated from three different donors underwent a replicative senescence protocol. Secreted OxPC species were quantified with HPLC MS/MS in lipid extracts from the supernatant of dermal FBs at early and end stages of the replicative lifespan. (a) Heatmap representation of the secreted oxidized phospholipid species peak areas normalized to secreted DPPC. Columns represent early and replicative senescent stages of FBs; rows show normalized relative log abundance (left) and fold change (right) of individual OxPC species (color code: normalized relative log abundance from low [orange] to high [green]; fold change: increase [red], decrease [blue]). (b, c) Dot blots showing abundance of the secreted respective species (b) LysoPPC and (c) LysoSPC of early and replicative senescent FBs normalized to DPPC. (d–f) Dot blots showing abundance of the secreted species (d) DPPC, (e) LysoPPC, and (f) LysoSPC normalized to the cell number of the FB, respectively. (n = 3; error bars ± SD). Asterisks represent significant differences (∗P < 0.05; ∗∗P < 0.01) determined by Student’s t-test. DPPC, 1,2-dipalmitoyl-sn-glycero-3-phosphocholine; E, early; FB, fibroblast; fold chg., fold change; HDF, human dermal fibroblast; LysoPPC, 1-palmitoyl-2-lyso-sn-glycero-3-phosphorylcholine; LysoSPC, 1-stearoyl-2-lyso-sn-glycero-3-phosphorylcholine; MS/MS, tandem mass spectrometry; norm., normalized; n.s., not significant; OxPC, oxidized phosphatidylcholine; PC, phosphatidylcholine; RS, replicative senescent.
      Figure thumbnail fx3
      Supplementary Figure S2Comparison of cellular and secreted oxidized PC species of replicative senescent fibroblasts. PC species were quantified with HPLC MS/MS and dot plots show the abundance of the nonoxidized (a–g) PAPC and (b–h) SAPC species, (c, e, i, k) oxidized PC-OOH, (d, j) PC-OH, and (f, l) isoPGJ2-PPC extracted from (a–f) senescent fibroblasts and (g–l) their secreted medium normalized to DPPC × 10,000. (n = 3; error bars indicate SD). Asterisks indicate statistically significant differences (∗P < 0.05; ∗∗P < 0.01) determined by one-way (a–f) ANOVA and (g–l) Student’s t-test. DPPC, 1,2-dipalmitoyl-sn-glycero-3-phosphocholine; E, early; FB, fibroblast; isoPG, isoprostaglandin; MS/MS, tandem mass spectrometry; PAPC, 1-palmitoyl-2-arachidonoyl-sn-glycero-3-phosphorylcholine; PC, phosphatidylcholine; RS, replicative senescent; SAPC, 1-stearoyl-2-arachidonoyl-sn-glycero-3-phosphorylcholine.
      Figure thumbnail fx4
      Supplementary Figure S3Doxorubicin stress–induced premature senescent HDFs show a stop in proliferation and an increase in G2/M phase. Cell cycle and proliferation status was assessed with a BrdU/PI-based flow cytometry assay. Bar graphs show (a) the percentage of cells positive for BrdU incorporation and (b) the fractions allocated to the cell cycle phases (G1, S, G2/M) in doxorubicin-induced premature senescent [SIPS (doxo)], quiescent (Q), and proliferating (Prol) HDFs from two donors. Proliferation and cell cycle analysis was performed with a BrdU/PI-based flow cytometry assay. Whereas 30% to 60% of the proliferating cells stained positive for BrdU incorporation, only 2% to 7% of the doxorubicin-induced senescent or quiescent cells were BrdU-positive. Furthermore, 23% to 33% of the senescent cells were detected to be in G2/M phase, whereas only 6% to 9% of the quiescent and 12% to 14% of the proliferating human dermal FBs belonged to this fraction. doxo, doxorubicin; HDF, human dermal fibroblast; PI, propidium iodide; Prol, proliferating; Q, quiescent; SIPS, stress-induced premature senescence.
      Figure thumbnail fx5
      Supplementary Figure S4Epilipidomics of stress-induced premature senescent HDFs identify differential lipid signatures. HDFs were treated with an H2O2 and doxorubicin protocol for SIPS. (a) Heatmaps representing the relative log abundance and fold change of oxidized phospholipid species of non- and SIPS-treated FBs normalized to DPPC (color code: normalized relative log abundance from low [orange] to high [green]; fold change: increase [red], decrease [blue]). (b–k) Dot blots showing values of secreted (b, g) LysoPPC and (c, h) LysoSPC of SIPS FBs normalized to DPPC and values of secreted (d, i) DPPC, (e, j) LysoPPC, and (f, k) LysoSPC of SIPS FBs normalized to the cell number, respectively (n = 3; error bars indicate ± SD). Asterisks represent significant differences (∗P < 0.05; ∗∗P < 0.01) determined by Student’s t-test. Doxo, doxorubicin; DPPC, 1,2-dipalmitoyl-sn-glycero-3-phosphocholine; FB, fibroblast; fold chg., fold change; H2O2, hydrogen peroxide; HDF, human dermal fibroblast; LysoPPC, 1-palmitoyl-2-lyso-sn-glycero-3-phosphorylcholine; LysoSPC, 1-stearoyl-2-lyso-sn-glycero-3-phosphorylcholine; norm. abund., normalized abundance; NT, nontreated; SIPS, stress-induced premature senescence.
      Figure thumbnail fx6
      Supplementary Figure S5Comparison of cellular and secreted oxidized PC species of stress-induced premature senescent dermal fibroblasts. Nonoxidized and oxidized PC species were analyzed with HPLC MS/MS. Dot plots show the abundance of the (a–g) nonoxidized PAPC and (b–h) SAPC species, (c, e, i, k) oxidized PC-OOH, (d, j) PC-OH, and (f, l) isoPGJ2-PPC extracted from (a–f) stress-induced premature senescent FBs and (g–l) their secreted medium normalized to DPPC × 10,000. (n = 3; error bars indicate SD). Asterisks indicate statistically significant differences (∗P < 0.05; ∗∗P < 0.01) determined by Student’s t-test. Doxo, doxorubicin; DPPC, 1,2-dipalmitoyl-sn-glycero-3-phosphocholine; FB, fibroblast; isoPG, isoprostaglandin; MS/MS, tandem mass spectrometry; n.s., not significant; NT, nontreated; PAPC, 1-palmitoyl-2-arachidonoyl-sn-glycero-3-phosphorylcholine; PC, phosphatidylcholine; rel., relative; SAPC, 1-stearoyl-2-arachidonoyl-sn-glycero-3-phosphorylcholine.
      Figure thumbnail fx7
      Supplementary Figure S6PL detection in naive medium and after 48-hour supernatant collection. Dot blots showing relative amount of secreted (a) DPPC, (d) LysoPPC, and (c) LysoSPC of H2O2-induced senescent fibroblasts and the medium control, normalized to the spike-in control lipid DNPC × 10,000. Error bars indicate ± SD. DNPC, 1,2-dinonanoyl-sn-glycero-3-phosphocholine; DPPC, 1,2-dipalmitoyl-sn-glycero-3-phosphocholine; H2O2, hydrogen peroxide; LysoPPC, 1-palmitoyl-2-lyso-sn-glycero-3-phosphorylcholine; LysoSPC, 1-stearoyl-2-lyso-sn-glycero-3-phosphorylcholine; NT, nontreated; PL, phospholipid; rel., relative.
      Figure thumbnail fx8
      Supplementary Figure S7UVB stress–induced premature senescent human dermal FBs show higher LysoPPC and LysoSPC levels. To induce stress-induced premature senescence, human dermal FBs were treated twice a day with 3,250 J/m2 UVB for 4 days followed by recovery for 11 days. (a, b) Representative cell micrographs of (a) NT or (b) UVB-induced senescent FBs (+UVB) stained with β-galactosidase. Bar = 100 μm. (c, d) Dot blots showing relative amount of (c) LysoPPC and (d) LysoSPC of NT and UVB-induced senescent FBs normalized to intrinsic DPPC. DPPC, 1,2-dipalmitoyl-sn-glycero-3-phosphocholine; FB, fibroblast; LysoPPC, 1-palmitoyl-2-lyso-sn-glycero-3-phosphorylcholine; LysoSPC, 1-stearoyl-2-lyso-sn-glycero-3-phosphorylcholine; NT, nontreated; rel., relative.
      Figure thumbnail fx9
      Supplementary Figure S8Gene regulation by LysoPPC and FSL-1 treatment in U937 monocytes. Cells were treated with LysoPPC (10 μM) or FSL-1 (1 μg/ml) followed by LysoPPC after treatment or left untreated to serve as control. Then, RNA was extracted and global gene expression was assayed with microarray technology (Affymetrix Human Gene 2.1 ST array). (a) Graph shows numbers of upregulated genes in LysoPPC-, FSL-1–, and FSL-1/LysoPPC–treated monocytes. (b) Heatmap showing the activation z-score of the top five pathways of the canonical pathway analysis using IPA software in FSL-1 and FSL-1 followed by LysoPPC exposed cells. IPA, Ingenuity Pathway Analysis; LysoPPC, 1-palmitoyl-2-lyso-sn-glycero-3-phosphorylcholine.
      Figure thumbnail fx10
      Supplementary Figure S9Pretreatment with LysoPPC inhibits AF647-dextran uptake by THP-1 and U937 cells. THP-1 and U937 cells, pretreated with PMA, were treated with either oxPAPC (25 μg/ml), LysoPPC (25 μg/ml), or DPPC (25 μg/ml) or stayed untreated to serve as control and were further incubated with AF647-dextran at 37 °C or 4 °C, respectively. (a) AF647-dextran uptake of THP-1 (left panel) and U937 (right panel) cells was analyzed by FACS. Representative histograms of three similar experiments are shown. Histograms show dextran uptake of control cells at 37 °C and 4 °C, respectively, for (b) THP-1 and (c) U937 cells. DPPC, 1,2-dipalmitoyl-sn-glycero-3-phosphocholine; LysoPPC, 1-palmitoyl-2-lyso-sn-glycero-3-phosphorylcholine; oxPAPC, oxidized 1-palmitoyl-2-arachidonoyl-sn-glycero-3-phosphorylcholine; PMA, phorbol 12-myristate 13-acetate.
      Figure thumbnail fx11
      Supplementary Figure S10MALDI imaging profile spectra (TIC normalized) acquired on a 7T FTMS instrument at different spatial resolution. (a, b) 10 μm. (c, d) 40 μm. Abdominal body lift material from a female 31-year-old (blue) and a female 63-year-old (red) patient are compared. (a, c) Absolute intensities for LysoPPC, LysoSPC, and DPPC over the total investigated area are directly compared using the monoisotopic masses for the respective lipids. (b, d) Profile mass spectra showing all isotopic peaks (marked with an asterisks) for the investigated lipid species. DPPC, 1,2-dipalmitoyl-sn-glycero-3-phosphocholine; LysoPPC, 1-palmitoyl-2-lyso-sn-glycero-3-phosphorylcholine; LysoSPC, 1-stearoyl-2-lyso-sn-glycero-3-phosphorylcholine; MALDI, matrix-assisted laser desorption/ionization.
      Supplementary Table S1Expression of Marker Genes for Cellular Senescence, Lipid Metabolism, and SASP Factor Genes
      Gene SymbolLog2 Fold Change (SIPS/Q)Q-Value (Adjusted for FDR)
      Cellular senescence markers
      CDKN1A1.09757<0.001
      LMNB1−1.88363<0.001
      SASP factors
      CCL21.98317<0.001
      CCL261.077770.0273215
      CCL51.628280.00470281
      CTSB0.693733<0.001
      CXCL51.767950.0212036
      CXCL81.43636<0.001
      ICAM30.7270870.00332154
      IGFBP22.69489<0.001
      IGFBP41.39149<0.001
      IGFBP60.3977410.0327422
      IL111.60667<0.001
      IL6R0.9768210.013262
      IL6ST0.4072450.0264479
      INHBA1.17205<0.001
      KITLG0.598603<0.001
      LEP1.432640.0149162
      LIF0.951335<0.001
      MMP10.9934520.0159062
      MMP30.721386<0.001
      MMP101.788590.00332154
      MMP122.696510.0037834
      MMP140.4797840.00979227
      PLAT1.0844<0.001
      PLAU0.8097470.0159062
      PLAUR0.5465410.0037834
      SERPINB23.47154<0.001
      TIMP10.4681020.012933
      TNFRSF10C1.84891<0.001
      Lipid metabolism
      ACSL10.4969310.0146057
      ACSL40.5762370.00183005
      ACSL52.02924<0.001
      LPCAT11.2984<0.001
      MBOAT71.15925<0.001
      PLA2G150.988566<0.001
      PLA2G4A1.24725<0.001
      Abbreviations: FDR, false discovery rate; LysoPC, lysophosphatidylcholine; RNA-seq, RNA sequencing; SASP, senescence-associated secretory phenotype; SIPS, stress-induced premature senescence.
      Reanalysis of previously published RNA-seq data (
      • Lämmermann I.
      • Terlecki-Zaniewicz L.
      • Weinmüllner R.
      • Schosserer M.
      • Dellago H.
      • de Matos Branco A.D.
      • et al.
      Blocking negative effects of senescence in human skin fibroblasts with a plant extract.
      ) from our hydrogen peroxide–based SIPS protocol revealed that, in addition to previously reported genes associated with cellular senescence and SASP, several genes of the Lands cycle were upregulated in senescent cells. This indicates an increased turnover of phospholipids and enzymatic hydrolysis as a dominant source for the generation of LysoPCs.
      Next-generation sequencing and data analysis: Library preparation and sequencing were performed on an Illumina HighSeq 2000 Platform (GATC Biotech AG, Konstanz, Germany). All analysis steps were done according to the Tuxedo Suite Pipeline (
      • Trapnell C.
      • Roberts A.
      • Goff L.
      • Pertea G.
      • Kim D.
      • Kelley D.R.
      • et al.
      Differential gene and transcript expression analysis of RNA-seq experiments with TopHat and Cufflinks.
      ). Briefly, Illumina Casava 1.8.2 software was used for base calling. RNA-seq reads were aligned to hg19 genome assembly using TOPHAT Version 2.0.13 with default parameters. Transcripts were assembled in Cufflinks Version 2.1.1 and differentially expressed genes were predicted by Cuffdiff.
      Supplementary Table S2Database Search for Accurately Measured Ion Signals Correlating to Lipid Species Identified in Bulk Samples
      m/zMass Error (ppm)± Uncertainty in m/z
      496.339810.10.00005
      524.370960.10.000052
      734.569070.10.000073
      LMSD Results for m/z 496.33981
      Input MassMatched MassDeltaNameFormulaIon
      496.33981496.339700.000101-(2-methoxy-6Z-octadecenyl)-sn-glycero-3-phosphoethanolamineC24H51NO7P[M+H]+
      496.33981496.339700.00010PC(0:0/16:0)C24H51NO7P[M+H]+
      496.33981496.339700.00010PC(16:0/0:0)C24H51NO7P[M+H]+
      496.33981496.339700.00010PC(16:0/0:0)[rac]C24H51NO7P[M+H]+
      496.33981496.339700.00010PC(O-14:0/2:0)C24H51NO7P[M+H]+
      496.33981496.339700.00010PE(19:0/0:0)C24H51NO7P[M+H]+
      496.33981496.339700.00010(7Z,10Z,13Z,16Z,19Z)-docosapentaenoylcarnitineC29H47NO4Na[M+Na]+
      496.33981496.339700.00010Clupanodonyl carnitineC29H47NO4Na[M+Na]+
      496.33981496.339700.00010Docosa-4,7,10,13,16-pentaenoyl carnitineC29H47NO4Na[M+Na]+
      LMSD Results for m/z 524.37096
      Input MassMatched MassDeltaNameFormulaIon
      524.37096524.371100.00010PC(0:0/18:0)C26H55NO7P[M+H]+
      524.37096524.371100.00010PC(18:0/0:0)C26H55NO7P[M+H]+
      524.37096524.371100.00010PC(O-16:0/2:0)C26H55NO7P[M+H]+
      524.37096524.371100.00010PE(21:0/0:0)C26H55NO7P[M+H]+
      LMSD Results for m/z 734.56907
      Input MassMatched MassDeltaNameFormulaIon
      734.56907734.569400.00040PC(10:0/22:0)C40H81NO8P[M+H]+
      734.56907734.569400.00040PC(11:0/21:0)C40H81NO8P[M+H]+
      734.56907734.569400.00040PC(12:0/20:0)C40H81NO8P[M+H]+
      734.56907734.569400.00040PC(13:0/19:0)C40H81NO8P[M+H]+
      734.56907734.569400.00040PC(14:0/18:0)C40H81NO8P[M+H]+
      734.56907734.569400.00040PC(15:0/17:0)C40H81NO8P[M+H]+
      734.56907734.569400.00040PC(16:0/16:0)C40H81NO8P[M+H]+
      734.56907734.569400.00040PC(17:0/15:0)C40H81NO8P[M+H]+
      734.56907734.569400.00040PC(18:0/14:0)C40H81NO8P[M+H]+
      734.56907734.569400.00040PC(19:0/13:0)C40H81NO8P[M+H]+
      734.56907734.569400.00040PC(20:0/12:0)C40H81NO8P[M+H]+
      734.56907734.569400.00040PC(21:0/11:0)C40H81NO8P[M+H]+
      734.56907734.569400.00040PC(22:0/10:0)C40H81NO8P[M+H]+
      734.56907734.569400.00040PC(9:0/23:0)C40H81NO8P[M+H]+
      734.56907734.569400.00040PE(13:0/22:0)C40H81NO8P[M+H]+
      734.56907734.569400.00040PE(14:0/21:0)C40H81NO8P[M+H]+
      734.56907734.569400.00040PE(15:0/20:0)C40H81NO8P[M+H]+
      734.56907734.569400.00040PE(16:0/19:0)C40H81NO8P[M+H]+
      734.56907734.569400.00040PE(17:0/18:0)C40H81NO8P[M+H]+
      734.56907734.569400.00040PE(18:0(10(R)Me)/16:0)C40H81NO8P[M+H]+
      734.56907734.569400.00040PE(18:0/17:0)C40H81NO8P[M+H]+
      734.56907734.569400.00040PE(19:0/16:0)C40H81NO8P[M+H]+
      734.56907734.569400.00040PE(20:0/15:0)C40H81NO8P[M+H]+
      734.56907734.569400.00040PE(21:0/14:0)C40H81NO8P[M+H]+
      734.56907734.569400.00040PE(22:0/13:0)C40H81NO8P[M+H]+
      Abbreviations: LMSD, LIPID MAPS structure database; RMS, root mean square.
      Experimental data note: calibration RMS error = 0.067 ppm.
      Experimental data are listed with their respective experimental accuracy. Database search results for the three most interesting m/z values are listing isobaric structures and lipid species that have to be considered under the given experimental conditions (positive ion, singly charged, proton or salt adduct ions).
      Supplementary Table S3Ion Intensity Levels after TIC Normalization over the Investigated Tissue Samples
      Experiment at 10 μm spatial resolution
      Analytem/zAgeAbsolute Intensity (TIC Norm)Number of SpectraArea (mm2)Int/SpectrumInt/Area
      LysoPPC496.3398163y462,2724,9540.495493.31933,128.78
      31y277,3508,7550.875531.68316,790.41
      LysoSPC524.3709663y116,1144,9540.495423.44234,384.34
      31y105,8898,7550.875512.09120,946.89
      DPPC734.5690763y388,1354,9540.495478.35783,478.00
      31y580,1218,7550.875566.26662,616.79
      Experiment at 40 μm spatial resolution
      Analytem/zAgeAbsolute Intensity (TIC Norm)Number of SpectraArea (mm2)Int/SpectrumInt/Area
      LysoPPC496.3398163y1,077,2905,4038.6448199.39124,617.11
      31y536,7004,5017.2016119.2474,525.11
      LysoSPC524.3709663y538,3165,4038.644899.6362,270.50
      31y323,3874,5017.201671.8544,904.88
      DPPC734.5690763y884,1915,4038.6448163.65102,280.10
      31y1,663,7404,5017.2016369.64231,023.66
      Abbreviations: DPPC, 1,2-dipalmitoyl-sn-glycero-3-phosphocholine; Int, intensity; LysoPPC, 1-palmitoyl-2-lyso-sn-glycero-3-phosphorylcholine; LysoSPC, 1-stearoyl-2-lyso-sn-glycero-3-phosphorylcholine; Norm, normalization; TIC, total ion current.
      For each experiment, the absolute ion intensity for the respective lipid species is given together with the number of spectra collected over the investigated area.

      Materials and Methods

      Cell culture and tissue

      Human dermal fibroblasts (HDFs) isolated from adult female donors were obtained from Evercyte (Vienna, Austria), and site-matched papillary and reticular HDFs were isolated from the dermis of healthy donors by the Department of Dermatology of the Leiden University Medical Center (Leiden, The Netherlands) according to article 467 of the Dutch Law on Medical Treatment Agreement and the Code for Proper Use of Human Tissue of the Dutch Federation of Biomedical Scientific Societies as described by
      • Janson D.
      • Saintigny G.
      • Mahé C.
      • El Ghalbzouri A.
      Papillary fibroblasts differentiate into reticular fibroblasts after prolonged in vitro culture.
      . Their identity was confirmed by measurement of the expression levels of three papillary and three reticular mRNA markers as described previously (
      • Lämmermann I.
      • Terlecki-Zaniewicz L.
      • Weinmüllner R.
      • Schosserer M.
      • Dellago H.
      • de Matos Branco A.D.
      • et al.
      Blocking negative effects of senescence in human skin fibroblasts with a plant extract.
      ). HCA2 (MJ90 telomerase-immortalized fibroblasts) and corresponding control fibroblasts had been isolated previously from neonatal foreskin in the Pereira-Smith laboratory (
      • Ferenac M.
      • Polancec D.
      • Huzak M.
      • Pereira-Smith O.M.
      • Rubelj I.
      Early-senescing human skin fibroblasts do not demonstrate accelerated telomere shortening.
      ). All fibroblasts were maintained in DMEM/Ham's F-12 (1:1 mixture) (F4815, Biochrome, Berlin, Germany) supplemented with 10% fetal calf serum (F7524, Sigma-Aldrich, St. Louis, MO) and 4 mM L-glutamine (G7513, Sigma-Aldrich) at 37 °C at 5% CO2. The skin samples (surplus material from cosmetic surgery) for the cryosections used in this study were approved by the Ethics Committee of the Medical University of Vienna (1149/2016), and written informed consent was obtained from all subjects. The human myeloid leukemia cell line U937 (
      • Sundstrom C.
      • Nilsson K.
      Establishment and characterization of a human histiocytic lymphoma cell line (U-937).
      ) and the human monocytic cell line THP1, both obtained from ATCC (Manassas, VA), were cultured in RPMI 1640 (FG1215, Biochrome) supplemented with 10% fetal calf serum and 4 mM L-glutamine at 37 °C at 5% CO2.

      Induction of stress-induced premature senescence by hydrogen peroxide and doxorubicin

      Middle-aged HDF cells (population doubling 10–21) were seeded at cell densities of 2,800 to 3,500 cells/cm2 and subsequently treated with hydrogen peroxide (216763, Sigma-Aldrich) in sublethal concentrations (60–100 μM) supplemented to their growth media for 1 hour/day. Cells were treated for 8 to 9 days with 2 days of recovery in between as in
      • Chen Q.M.
      • Tu V.C.
      • Catania J.
      • Burton M.
      • Toussaint O.
      • Dilley T.
      Involvement of Rb family proteins, focal adhesion proteins and protein synthesis in senescent morphogenesis induced by hydrogen peroxide.
      and an 11-day recovery period before sampling.
      For doxorubicin treatment, cells were seeded at 3,500 cells/cm2 and subsequently treated with 100 nM doxorubicin for 6 days. After a recovery phase of 7 days, cells were used for experiments as in
      • Demidenko Z.N.
      • Blagosklonny M.V.
      Growth stimulation leads to cellular senescence when the cell cycle is blocked.
      . For UVB treatment, HDFs were seeded at 20,000 cells/cm2. UVB treatment was carried out as described in
      • Greussing R.
      • Hackl M.
      • Charoentong P.
      • Pauck A.
      • Monteforte R.
      • Cavinato M.
      • et al.
      Identification of microRNA-mRNA functional interactions in UVB-induced senescence of human diploid fibroblasts.
      . For irradiation, cells were washed twice with Hank’s Balanced Salt Solution (Sigma, Steinheim, Germany) and covered by a thin layer of this same buffer. The irradiation was carried using a UVB Narrowband Phototherapy Lamp (Philips TL20W/01 lamp, UV spectrum 290–315 nm; Philips, The Netherlands). Irradiation time was calculated with the formula: time (s) = (mJ/cm2)/(mW/cm2), and the power per area was determined by a UVX Radiometer (Thermo Fisher Scientific, Waltham, MA). The output of the lamp without any filter was determined as 14.2 ± 0.5 W/m2. For experiments, the lamp was preheated for 15 minutes. Under these conditions, cells were irradiated twice a day with a dose of 3,250 mJ/cm2 for 4 consecutive days, where irradiation time was 302 ± 10 seconds.

      Real-time qPCR p21 fibroblast

      HDF cells were lysed in TRI reagent (T9424, Sigma) and cellular RNA was subsequently isolated using chloroform extraction and isopropanol precipitation following the manufacturer’s instructions. As determined by Nanodrop spectrometer (ND-1000), 500 ng of total RNA were reversely transcribed into cDNA using the High-Capacity cDNA Reverse Transcription Kit (4368814, Applied Biosystems, Foster City, CA) according to the manufacturer’s instructions. Gene expression levels were quantified using the 5x HOT FIREPol EvaGreen qPCR Mix Plus with ROX (SB_08-24-GP, Solis BioDyne, Tartu, Estonia) with a Rotor-Gene Q cycler (Qiagen, Hilden, Germany) and normalized to GAPDH levels.
      The following primers were used: GAPDH (sense), CGACCACTTTGTCAAGCTC; GAPDH (antisense), TGTGAGGAGGGGAGATTCA; p21 (sense), GGCGGCAGACCAGCATGACAGAT; p21 (antisense), GCAGGGGGCGGCCAGGGTA.

      Senescence-associated β-galactosidase staining

      Senescence-associated β-galactosidase staining was performed according to
      • Dimri G.P.
      • Lee X.
      • Basile G.
      • Acosta M.
      • Scott G.
      • Roskelley C.
      • et al.
      A biomarker that identifies senescent human cells in culture and in aging skin in vivo.
      . Pictures were taken at 10 random evenly distributed positions per well at 100-fold magnification. Evaluation was done in a blinded fashion to ensure unbiased results.

      Lipid isolation and analysis

      Analysis of purified phospholipids was performed at FTC-Forensic Toxicological Laboratory, Vienna, Austria as recently described by us (
      • Gruber F.
      • Bicker W.
      • Oskolkova O.V.
      • Tschachler E.
      • Bochkov V.N.
      A simplified procedure for semi-targeted lipidomic analysis of oxidized phosphatidylcholines induced by UVA irradiation.
      ). In brief, fibroblasts were washed with PBS containing diethylenetriamine-pentaacetic acid (0.5 mM) and scraped with methanol/acetic acid (3%)/BHT (0.01%). A total of 10 ng of internal standard 1,2-dinanoyl-sn-glycero-3-phosphocholine (Avanti Polar Lipids, Alabaster, AL) was added to each sample. Organic phases were purified by the hexane liquid–liquid extraction procedure and stored at −20 °C until mass spectrometry analysis. For analysis, purified lipid samples reconstituted in 85% aqueous methanol containing 5 mM ammonium formate and 0.1% formic acid were injected onto a core–shell type C18 column (Kinetex 2.6 μm, 50 mm, 3.0 mm ID; Phenomenex, Torrance, CA) using a 1200 series HPLC system (Agilent Technologies, Santa Clara, CA), which was coupled to a 4000 QTrap triple quadrupole linear ion trap hybrid mass spectrometer system equipped with a Turbo V electrospray ion source (Applied Biosystems). Detection was carried out in positive ion mode by selected reaction monitoring of 99 tandem mass spectrometry transitions using a phosphatidylcholine-specific product ion (m/z 184). Data acquisition and instrument control were performed with Analyst software, version 1.6 (Applied Biosystems) and individual values were normalized to the intrinsic 1,2-dipalmitoyl-sn-glycero-3-phosphocholine (DPPC). The method allows relative quantification of lipid species normalized to DPPC within one experiment but not comparison of the relative quantification values between different experiments. The readout for the abundant lysophosphatidylcholine species is presented in the diagrams as LysoPC/DPPC, whereas the values for the less abundant species have been multiplied by 10,000 for better visualization. Heatmaps generated in Excel were used to visualize the normalized relative logarithmic abundance of the analyzed phospholipids (minimum: 3; maximum: 16.5) and the fold change (minimum: 0.25; maximum: 5) relative to the control. For isolation of lipids from cell culture supernatants, material from 1.5 ml fetal calf serum–free DMEM/Ham’s F12 medium in which the complete secretome had been collected for 48 hours. Total lipids were isolated by Folch’s (
      • Folch J.
      • Lees M.
      • Sloane Stanley G.H.
      A simple method for the isolation and purification of total lipides from animal tissues.
      ) method (in a 9-fold volume of chloroform:methanol 2:1) and then subjected to the liquid–liquid extraction as described.
      The statistical significance of differences between the groups was determined with ordinary one-way ANOVA with Tukey’s multiple comparison for the replicative senescent fibroblasts, including hTERT cells. For the analysis of the stress-induced premature senescence–treated papillary and reticular fibroblasts, two-tailed Student’s t-test was applied.

      Matrix-assisted laser desorption/ionization Fourier transform ion cyclotron resonance mass spectrometry imaging

      Tissue sections (10 μm thickness) of OCT-embedded abdominal body lift material biopsies were mounted after cryosectioning onto a indium tin-oxide–covered microscopic glass slide (Sigma-Aldrich) and covered with 0.26 mg 1,5-diaminonaphtalene/cm2 using a home-built sublimation device (TU Wien, Institute of Chemical Technologies and Analytics) (
      • Holzlechner M.
      • Bonta M.
      • Lohninger H.
      • Limbeck A.
      • Marchetti-Deschmann M.
      Multisensor imaging-from sample preparation to integrated multimodal interpretation of LA-ICPMS and MALDI MS Imaging Data.
      ). To ensure comparability, tissue sections from young and old donors were mounted on the same slide and measured in one mass spectrometry run to receive identical experimental conditions. After matrix application, matrix-assisted laser desorption/ionization Fourier transform ion cyclotron resonance mass spectrometry imaging experiments were immediately performed on a 7T scimaX MRMS (Bruker Daltonik, Bremen, Germany) equipped with a dual electrospray ionization and matrix-assisted laser desorption/ionization ion source and the Smatbeam laser set to small focus (90%). Experiments were carried out in the positive ion mode. The m/z range acquired was 107.5 to 2,000 with a transient length of 1,048.6 ms at 4 megaword for a broad band resolving power of 160,000 at 838 m/z. Single-scan spectra consisting of 50 accumulated laser shots were acquired at a rate of 1 kHz either at 10 μm or 40 μm of spatial resolution. Instrument calibration was performed with sodium trifluoroacetate (98% from Sigma-Aldrich) before matrix-assisted laser desorption/ionization imaging experiments (root mean square error, 0.067 ppm), followed by lock mass online calibration using the 1,5-diaminonaphtalene ion at m/z 317.176073 before saving. Raw data were post-processed using SCiLS 2020b Pro (Bruker Daltonik). Data were not denoised but normalized (total ion current normalization) before further processing. Lipid identification is based on accurate mass and database search using the structure database of LipidMaps (search parameters: protonation, singly charged ions, positive ion mode, mass accuracy ± 0.001).

      Proliferation and cell cycle analysis

      For the cell proliferation assay, cells were incubated with 10 μM BrdU under optimal growth conditions for 24 hours. After trypsinization and centrifugation at 170g for 10 minutes, the cells were fixated with 70% ethanol. Harvested cells were incubated with 2M HCl containing 1% Triton, pelleted by centrifugation (170g/10 min), and resuspended in 0.1M Na-Borat pH 8.5 for a maximum of 24 hours at 4 °C before continuing with the staining procedure. Monoclonal mouse anti-BrdU (BD Biosciences, San Jose, CA, 347580) (1:50) and anti-mouse FITC-conjugated antibody (Sigma, F-8264) (1:100) were added in Tris-buffered saline with centrifugation steps in between to wash cells. Finally, propidium iodide in PBS (1:1,000) was added and the samples analyzed by flow cytometry.

      Luminex analysis

      Young cells (> population doubling 12) from three different donors were treated with vehicle (basal medium) alone or 6.25 μg/ml, 12 μg/ml, 25 μg/m, and 50 μg/ml 1-palmitoyl-2-hydroxy-sn-glycero-3-phosphocholine, 1-stearoyl-2-hydroxy-sn-glycero-3-phosphocholine, or DPPC, obtained from Avanti Polar Lipids, for 4 hours. Then the medium was changed to fetal calf serum–free medium, and the cells were left to secrete for another 4 hours. The supernatants (quadruplicates of each donor) were harvested and Luminex analysis was performed for IL-6 and IL-8 independently. Outliers were taken out according to Grubbs’s outlier test. Two-way ANOVA and Dunnett’s test comparing all conditions to vehicle using raw values (all replicates) was performed. Plotted are relative values to vehicle for each donor.

      Transcriptome analysis of U937

      The human myeloid leukemia cell line U937 (1.5 × 106 cells/ml) was pretreated for 1 hour with 10 μM 1-palmitoyl-2-lyso-sn-glycero-3-phosphorylcholine (Avanti Polar Lipids) before supplementing 1 μg/ml FSL-1 (TLRL-FSL; Invivogen, San Diego, CA). After a total treatment time of 3 hours, cells were lysed in TriFast Reagent (VWR Peqlab, Radnor, PA) and total RNA was isolated using chloroform extraction and isopropanol precipitation according to the manufacturer’s instructions. The RNA cleanup and concentration were performed using the RNeasy MinElute Cleanup Kit (Qiagen) according to the manufacturer’s instructions. A total of 200 ng of each sample were used for gene expression analysis with human Affymetrix (Santa Clara, CA) 2.0 Gene Arrays. Hybridization and scanning were performed according to manufacturer's protocol (http://www.affymetrix.com), and robust multiarray average signal extraction and normalization were performed using custom chip description file. The experiment was performed on quadruplicate biological samples. The full microarray data was uploaded to the Gene Expression Omnibus. After exclusion of all Gene IDs with a robust multiarray average value of less than 50 in all conditions, log2 transformation was applied. Differential expression between the conditions was tested using moderated t-tests as described (R version 3.2.2/Bioconductor software package Limma). Genes that were regulated more than 1.5-fold with a P-value < 0.05 were analyzed using the software Qiagen’s Ingenuity Pathway Analysis (Qiagen, www.qiagen.com/ingenuity). The software was used to forecast which signaling pathways were likely to be activated based on literature evidence. Heatmaps were generated using the Ingenuity Pathway Analysis software package.

      Dextran uptake

      U937 and THP-1 cells were treated with 100 nM phorbol 12-myristate 13-acetate (P1585, Merk, Vienna, Austria) at a cell density of 0.5 × 106 cells/ml for 72 hours. Before treatment, cells were washed twice with PBS and were pretreated for 1 hour with 1-palmitoyl-2-lyso-sn-glycero-3-phosphorylcholine, DPPC, and oxidized 1-palmitoyl-2-arachidonoyl-sn-glycero-3-phosphorylcholine dissolved in serum-free RPMI medium at a concentration of 25 μg/ml. After washing twice with PBS, cells were incubated with dextran (AF647 10,000 MW [RefD22914, Life Technologies, Carlsbad, CA]) with an end concentration of 30 μg/ml for 3 hours at 37 °C or 4 °C to serve as control. Cells were immediately analyzed at the FACS Calibur (BD Bioscience). Data were evaluated using FlowJo software (Tree Star, Ashland, OR). The experiment was performed on triplicate biological samples. Statistical significance of differences between the groups was determined by ordinary one-way ANOVA with Tukey’s multiple comparison.

      Graphs

      Dot plots were generated with GraphPad Prism software version 8 (GraphPad Software Inc, San Diego, CA). Error bars represent overall mean ± SD. Significant differences are indicated by asterisks (∗P < 0.05; ∗∗P < 0.01; NS, no significant differences).

      References

        • Abe T.
        • Shimamura M.
        • Jackman K.
        • Kurinami H.
        • Anrather J.
        • Zhou P.
        • et al.
        Key role of CD36 in toll-like receptor 2 signaling in cerebral ischemia.
        Stroke. 2010; 41: 898-904
        • Ademowo O.S.
        • Dias H.K.I.
        • Milic I.
        • Devitt A.
        • Moran R.
        • Mulcahy R.
        • et al.
        Phospholipid oxidation and carotenoid supplementation in Alzheimer's disease patients.
        Free Radic Biol Med. 2017; 108: 77-85
        • Antonelli A.
        • Rotondi M.
        • Fallahi P.
        • Ferrari S.M.
        • Paolicchi A.
        • Romagnani P.
        • et al.
        Increase of CXC chemokine CXCL10 and CC chemokine CCL2 serum levels in normal ageing.
        Cytokine. 2006; 34: 32-38
        • Baker D.J.
        • Childs B.G.
        • Durik M.
        • Wijers M.E.
        • Sieben C.J.
        • Zhong J.
        • et al.
        Naturally occurring p16(Ink4a)-positive cells shorten healthy lifespan.
        Nature. 2016; 530: 184-189
        • Basisty N.
        • Kale A.
        • Jeon O.H.
        • Kuehnemann C.
        • Payne T.
        • Rao C.
        • et al.
        A proteomic atlas of senescence-associated secretomes for aging biomarker development.
        PLoS Biol. 2020; 18e3000599
        • Beer L.
        • Zimmermann M.
        • Mitterbauer A.
        • Ellinger A.
        • Gruber F.
        • Narzt M.S.
        • et al.
        Analysis of the secretome of apoptotic peripheral blood mononuclear cells: impact of released proteins and exosomes for tissue regeneration.
        Sci Rep. 2015; 5: 16662
        • Binder C.J.
        • Papac-Milicevic N.
        • Witztum J.L.
        Innate sensing of oxidation-specific epitopes in health and disease.
        Nat Rev Immunol. 2016; 16: 485-497
        • Bochkov V.
        • Gesslbauer B.
        • Mauerhofer C.
        • Philippova M.
        • Erne P.
        • Oskolkova O.V.
        Pleiotropic effects of oxidized phospholipids.
        Free Radic Biol Med. 2017; 111: 6-24
        • Bochkov V.N.
        • Kadl A.
        • Huber J.
        • Gruber F.
        • Binder B.R.
        • Leitinger N.
        Protective role of phospholipid oxidation products in endotoxin-induced tissue damage.
        Nature. 2002; 419: 77-81
        • Briganti S.
        • Flori E.
        • Mastrofrancesco A.
        • Kovacs D.
        • Camera E.
        • Ludovici M.
        • et al.
        Azelaic acid reduced senescence-like phenotype in photo-irradiated human dermal fibroblasts: possible implication of PPARγ.
        Exp Dermatol. 2013; 22: 41-47
        • Burton D.G.
        • Krizhanovsky V.
        Physiological and pathological consequences of cellular senescence.
        Cell Mol Life Sci. 2014; 71: 4373-4386
        • Catalano A.
        • Rodilossi S.
        • Caprari P.
        • Coppola V.
        • Procopio A.
        5-Lipoxygenase regulates senescence-like growth arrest by promoting ROS-dependent p53 activation.
        EMBO J. 2005; 24: 170-179
        • Cavinato M.
        • Koziel R.
        • Romani N.
        • Weinmüllner R.
        • Jenewein B.
        • Hermann M.
        • et al.
        UVB-induced senescence of human dermal fibroblasts involves impairment of proteasome and enhanced autophagic activity.
        J Gerontol A Biol Sci Med Sci. 2017; 72: 632-639
        • Chen Q.M.
        • Tu V.C.
        • Catania J.
        • Burton M.
        • Toussaint O.
        • Dilley T.
        Involvement of Rb family proteins, focal adhesion proteins and protein synthesis in senescent morphogenesis induced by hydrogen peroxide.
        J Cell Sci. 2000; 113: 4087-4097
        • Childs B.G.
        • Baker D.J.
        • Kirkland J.L.
        • Campisi J.
        • van Deursen J.M.
        Senescence and apoptosis: dueling or complementary cell fates?.
        EMBO Rep. 2014; 15: 1139-1153
        • Childs B.G.
        • Baker D.J.
        • Wijshake T.
        • Conover C.A.
        • Campisi J.
        • van Deursen J.M.
        Senescent intimal foam cells are deleterious at all stages of atherosclerosis.
        Science. 2016; 354: 472-477
        • Chong M.
        • Yin T.
        • Chen R.
        • Xiang H.
        • Yuan L.
        • Ding Y.
        • et al.
        CD36 initiates the secretory phenotype during the establishment of cellular senescence.
        EMBO Rep. 2018; 19: e45274
        • Coppé J.P.
        • Patil C.K.
        • Rodier F.
        • Sun Y.
        • Muñoz D.P.
        • Goldstein J.
        • et al.
        Senescence-associated secretory phenotypes reveal cell-nonautonomous functions of oncogenic RAS and the p53 tumor suppressor.
        PLoS Biol. 2008; 6: 2853-2868
        • Demaria M.
        • Desprez P.Y.
        • Campisi J.
        • Velarde M.C.
        Cell autonomous and non-autonomous effects of senescent cells in the skin.
        J Invest Dermatol. 2015; 135: 1722-1726
        • Demaria M.
        • Ohtani N.
        • Youssef S.A.
        • Rodier F.
        • Toussaint W.
        • Mitchell J.R.
        • et al.
        An essential role for senescent cells in optimal wound healing through secretion of PDGF-AA.
        Dev Cell. 2014; 31: 722-733
        • Demidenko Z.N.
        • Blagosklonny M.V.
        Growth stimulation leads to cellular senescence when the cell cycle is blocked.
        Cell Cycle. 2008; 7: 3355-3361
        • Dimri G.P.
        • Lee X.
        • Basile G.
        • Acosta M.
        • Scott G.
        • Roskelley C.
        • et al.
        A biomarker that identifies senescent human cells in culture and in aging skin in vivo.
        Proc Natl Acad Sci USA. 1995; 92: 9363-9367
        • Egashira M.
        • Hirota Y.
        • Shimizu-Hirota R.
        • Saito-Fujita T.
        • Haraguchi H.
        • Matsumoto L.
        • et al.
        F4/80+ macrophages contribute to clearance of senescent cells in the mouse postpartum uterus.
        Endocrinology. 2017; 158: 2344-2353
        • Ferenac M.
        • Polancec D.
        • Huzak M.
        • Pereira-Smith O.M.
        • Rubelj I.
        Early-senescing human skin fibroblasts do not demonstrate accelerated telomere shortening.
        J Gerontol A Biol Sci Med Sci. 2005; 60: 820-829
        • Frescas D.
        • Roux C.M.
        • Aygun-Sunar S.
        • Gleiberman A.S.
        • Krasnov P.
        • Kurnasov O.V.
        • et al.
        Senescent cells expose and secrete an oxidized form of membrane-bound vimentin as revealed by a natural polyreactive antibody.
        Proc Natl Acad Sci USA. 2017; 114: E1668-E1677
        • Gorbunova V.
        • Seluanov A.
        • Pereira-Smith O.M.
        Expression of human telomerase (hTERT) does not prevent stress-induced senescence in normal human fibroblasts but protects the cells from stress-induced apoptosis and necrosis.
        J Biol Chem. 2002; 277: 38540-38549
        • Greenberg M.E.
        • Li X.M.
        • Gugiu B.G.
        • Gu X.
        • Qin J.
        • Salomon R.G.
        • et al.
        The lipid whisker model of the structure of oxidized cell membranes.
        J Biol Chem. 2008; 283: 2385-2396
        • Gruber F.
        • Bicker W.
        • Oskolkova O.V.
        • Tschachler E.
        • Bochkov V.N.
        A simplified procedure for semi-targeted lipidomic analysis of oxidized phosphatidylcholines induced by UVA irradiation.
        J Lipid Res. 2012; 53: 1232-1242
        • Gruber F.
        • Kremslehner C.
        • Narzt M.S.
        The impact of recent advances in lipidomics and redox lipidomics on dermatological research.
        Free Radic Biol Med. 2019; 144: 256-265
        • Hoebe K.
        • Georgel P.
        • Rutschmann S.
        • Du X.
        • Mudd S.
        • Crozat K.
        • et al.
        CD36 is a sensor of diacylglycerides.
        Nature. 2005; 433: 523-527
        • Holzlechner M.
        • Bonta M.
        • Lohninger H.
        • Limbeck A.
        • Marchetti-Deschmann M.
        Multisensor imaging-from sample preparation to integrated multimodal interpretation of LA-ICPMS and MALDI MS imaging data.
        Anal Chem. 2018; 90: 8831-8837
        • Holzlechner M.
        • Strasser K.
        • Zareva E.
        • Steinhäuser L.
        • Birnleitner H.
        • Beer A.
        • et al.
        In situ characterization of tissue-resident immune cells by MALDI mass spectrometry imaging.
        J Proteome Res. 2017; 16: 65-76
        • Hrelia S.
        • Fiorentini D.
        • Maraldi T.
        • Angeloni C.
        • Bordoni A.
        • Biagi P.L.
        • et al.
        Doxorubicin induces early lipid peroxidation associated with changes in glucose transport in cultured cardiomyocytes.
        Biochim Biophys Acta. 2002; 1567: 150-156
        • James E.L.
        • Lane J.A.
        • Michalek R.D.
        • Karoly E.D.
        • Parkinson E.K.
        Replicatively senescent human fibroblasts reveal a distinct intracellular metabolic profile with alterations in NAD+ and nicotinamide metabolism.
        Sci Rep. 2016; 6: 38489
        • Janson D.
        • Saintigny G.
        • Mahé C.
        • El Ghalbzouri A.
        Papillary fibroblasts differentiate into reticular fibroblasts after prolonged in vitro culture.
        Exp Dermatol. 2013; 22: 48-53
        • Jimenez-Dalmaroni M.J.
        • Xiao N.
        • Corper A.L.
        • Verdino P.
        • Ainge G.D.
        • Larsen D.S.
        • et al.
        Soluble CD36 ectodomain binds negatively charged diacylglycerol ligands and acts as a co-receptor for TLR2.
        PLoS One. 2009; 4: e7411
        • Jin H.
        • Zhang Y.
        • Liu D.
        • Wang S.S.
        • Ding Q.
        • Rastogi P.
        • et al.
        Innate immune signaling contributes to tubular cell senescence in the Glis2 knockout mouse model of nephronophthisis.
        Am J Pathol. 2020; 190: 176-189
        • Jørgensen P.
        • Milkovic L.
        • Zarkovic N.
        • Waeg G.
        • Rattan S.I.
        Lipid peroxidation-derived 4-hydroxynonenal-modified proteins accumulate in human facial skin fibroblasts during ageing in vitro.
        Biogerontology. 2014; 15: 105-110
        • Kadl A.
        • Meher A.K.
        • Sharma P.R.
        • Lee M.Y.
        • Doran A.C.
        • Johnstone S.R.
        • et al.
        Identification of a novel macrophage phenotype that develops in response to atherogenic phospholipids via Nrf2.
        Circ Res. 2010; 107: 737-746
        • Kale A.
        • Sharma A.
        • Stolzing A.
        • Desprez P.Y.
        • Campisi J.
        Role of immune cells in the removal of deleterious senescent cells.
        Immun Ageing. 2020; 17: 16
        • Karin O.
        • Agrawal A.
        • Porat Z.
        • Krizhanovsky V.
        • Alon U.
        Senescent cell turnover slows with age providing an explanation for the Gompertz law.
        Nat Commun. 2019; 10: 5495
        • Karuppagounder V.
        • Giridharan V.V.
        • Arumugam S.
        • Sreedhar R.
        • Palaniyandi S.S.
        • Krishnamurthy P.
        • et al.
        Modulation of macrophage polarization and HMGB1-TLR2/TLR4 cascade plays a crucial role for cardiac remodeling in senescence-accelerated prone mice.
        PLoS One. 2016; 11e0152922
        • Katsuumi G.
        • Shimizu I.
        • Yoshida Y.
        • Minamino T.
        Vascular senescence in cardiovascular and metabolic diseases.
        Front Cardiovasc Med. 2018; 5: 18
        • Kendall A.C.
        • Koszyczarek M.M.
        • Jones E.A.
        • Hart P.J.
        • Towers M.
        • Griffiths C.E.M.
        • et al.
        Lipidomics for translational skin research: a primer for the uninitiated.
        Exp Dermatol. 2018; 27: 721-728
        • Korosec A.
        • Frech S.
        • Gesslbauer B.
        • Vierhapper M.
        • Radtke C.
        • Petzelbauer P.
        • et al.
        Lineage identity and location within the dermis determine the function of papillary and reticular fibroblasts in human skin.
        J Invest Dermatol. 2019; 139: 342-351
        • Lämmermann I.
        • Terlecki-Zaniewicz L.
        • Weinmüllner R.
        • Schosserer M.
        • Dellago H.
        • de Matos Branco A.D.
        • et al.
        Blocking negative effects of senescence in human skin fibroblasts with a plant extract.
        NPJ Aging Mech Dis. 2018; 4: 4
        • Li X.M.
        • Salomon R.G.
        • Qin J.
        • Hazen S.L.
        Conformation of an endogenous ligand in a membrane bilayer for the macrophage scavenger receptor CD36.
        Biochemistry. 2007; 46: 5009-5017
        • Lizardo D.Y.
        • Lin Y.L.
        • Gokcumen O.
        • Atilla-Gokcumen G.E.
        Regulation of lipids is central to replicative senescence.
        Mol Biosyst. 2017; 13: 498-509
        • Matt U.
        • Sharif O.
        • Martins R.
        • Knapp S.
        Accumulating evidence for a role of oxidized phospholipids in infectious diseases.
        Cell Mol Life Sci. 2015; 72: 1059-1071
        • Mauerhofer C.
        • Philippova M.
        • Oskolkova O.V.
        • Bochkov V.N.
        Hormetic and anti-inflammatory properties of oxidized phospholipids.
        Mol Aspects Med. 2016; 49: 78-90
        • Medina C.B.
        • Ravichandran K.S.
        Do not let death do us part: 'find-me' signals in communication between dying cells and the phagocytes.
        Cell Death Differ. 2016; 23: 979-989
        • Miller Y.I.
        • Choi S.H.
        • Wiesner P.
        • Fang L.
        • Harkewicz R.
        • Hartvigsen K.
        • et al.
        Oxidation-specific epitopes are danger-associated molecular patterns recognized by pattern recognition receptors of innate immunity.
        Circ Res. 2011; 108: 235-248
        • Minamino T.
        • Miyauchi H.
        • Yoshida T.
        • Ishida Y.
        • Yoshida H.
        • Komuro I.
        Endothelial cell senescence in human atherosclerosis: role of telomere in endothelial dysfunction.
        Circulation. 2002; 105: 1541-1544
        • Mine S.
        • Fortunel N.O.
        • Pageon H.
        • Asselineau D.
        Aging alters functionally human dermal papillary fibroblasts but not reticular fibroblasts: a new view of skin morphogenesis and aging.
        PLoS One. 2008; 3: e4066
        • Muñoz-Espín D.
        • Serrano M.
        Cellular senescence: from physiology to pathology.
        Nat Rev Mol Cell Biol. 2014; 15: 482-496
        • Nelson G.
        • Kucheryavenko O.
        • Wordsworth J.
        • von Zglinicki T.
        The senescent bystander effect is caused by ROS-activated NF-κB signalling.
        Mech Ageing Dev. 2018; 170: 30-36
        • Ni C.
        • Narzt M.S.
        • Nagelreiter I.M.
        • Zhang C.F.
        • Larue L.
        • Rossiter H.
        • et al.
        Autophagy deficient melanocytes display a senescence associated secretory phenotype that includes oxidized lipid mediators.
        Int J Biochem Cell Biol. 2016; 81: 375-382
        • Ni Z.
        • Goracci L.
        • Cruciani G.
        • Fedorova M.
        Computational solutions in redox lipidomics - current strategies and future perspectives.
        Free Radic Biol Med. 2019; 144: 110-123
        • Nicolaou A.
        • Pilkington S.M.
        • Rhodes L.E.
        Ultraviolet-radiation induced skin inflammation: dissecting the role of bioactive lipids.
        Chem Phys Lipids. 2011; 164: 535-543
        • Niki E.
        Lipid oxidation in the skin.
        Free Radic Res. 2015; 49: 827-834
        • O'Donnell V.B.
        • Aldrovandi M.
        • Murphy R.C.
        • Krönke G.
        Enzymatically oxidized phospholipids assume center stage as essential regulators of innate immunity and cell death.
        Sci Signal. 2019; 12eaau2293
        • Perrott K.M.
        • Wiley C.D.
        • Desprez P.Y.
        • Campisi J.
        Apigenin suppresses the senescence-associated secretory phenotype and paracrine effects on breast cancer cells.
        Geroscience. 2017; 39: 161-173
        • Quinn M.T.
        • Parthasarathy S.
        • Steinberg D.
        Lysophosphatidylcholine: a chemotactic factor for human monocytes and its potential role in atherogenesis.
        Proc Natl Acad Sci USA. 1988; 85: 2805-2809
        • Ressler S.
        • Bartkova J.
        • Niederegger H.
        • Bartek J.
        • Scharffetter-Kochanek K.
        • Jansen-Dürr P.
        • et al.
        p16INK4A is a robust in vivo biomarker of cellular aging in human skin.
        Aging Cell. 2006; 5: 379-389
        • Ritschka B.
        • Storer M.
        • Mas A.
        • Heinzmann F.
        • Ortells M.C.
        • Morton J.P.
        • et al.
        The senescence-associated secretory phenotype induces cellular plasticity and tissue regeneration.
        Genes Dev. 2017; 31: 172-183
        • Sagiv A.
        • Krizhanovsky V.
        Immunosurveillance of senescent cells: the bright side of the senescence program.
        Biogerontology. 2013; 14: 617-628
        • Salama R.
        • Sadaie M.
        • Hoare M.
        • Narita M.
        Cellular senescence and its effector programs.
        Genes Dev. 2014; 28: 99-114
        • Semba R.D.
        • Zhang P.
        • Adelnia F.
        • Sun K.
        • Gonzalez-Freire M.
        • Salem Jr., N.
        • et al.
        Low plasma lysophosphatidylcholines are associated with impaired mitochondrial oxidative capacity in adults in the Baltimore Longitudinal Study of Aging.
        Aging Cell. 2019; 18: e12915
        • Shimizu R.
        • Kanno K.
        • Sugiyama A.
        • Ohata H.
        • Araki A.
        • Kishikawa N.
        • et al.
        Cholangiocyte senescence caused by lysophosphatidylcholine as a potential implication in carcinogenesis.
        J Hepatobiliary Pancreat Sci. 2015; 22: 675-682
        • Song X.
        • Narzt M.S.
        • Nagelreiter I.M.
        • Hohensinner P.
        • Terlecki-Zaniewicz L.
        • Tschachler E.
        • et al.
        Autophagy deficient keratinocytes display increased DNA damage, senescence and aberrant lipid composition after oxidative stress in vitro and in vivo.
        Redox Biol. 2017; 11: 219-230
        • Sundstrom C.
        • Nilsson K.
        Establishment and characterization of a human histiocytic lymphoma cell line (U-937).
        Int J Cancer. 1976; 17: 565-577
        • Tavasoli M.
        • Lahire S.
        • Reid T.
        • Brodovsky M.
        • McMaster C.R.
        Genetic diseases of the Kennedy pathway for phospholipid synthesis [e-pub ahead of print].
        J Biol Chem. 2020; (accessed 14 November 2020)https://doi.org/10.1074/jbc.REV120.013529
        • Terlecki-Zaniewicz L.
        • Lämmermann I.
        • Latreille J.
        • Bobbili M.R.
        • Pils V.
        • Schosserer M.
        • et al.
        Small extracellular vesicles and their miRNA cargo are anti-apoptotic members of the senescence-associated secretory phenotype.
        Aging (Albany NY). 2018; 10: 1103-1132
        • Terlecki-Zaniewicz L.
        • Pils V.
        • Bobbili M.R.
        • Lämmermann I.
        • Perrotta I.
        • Grillenberger T.
        • et al.
        Extracellular vesicles in human skin: cross-talk from senescent fibroblasts to keratinocytes by miRNAs.
        J Invest Dermatol. 2019; 139: 2425-2436.e5
        • Thompson P.J.
        • Shah A.
        • Ntranos V.
        • Van Gool F.
        • Atkinson M.
        • Bhushan A.
        Targeted elimination of senescent beta cells prevents type 1 diabetes.
        Cell Metab. 2019; 29: 1045-1060.e10
        • Trayssac M.
        • Hannun Y.A.
        • Obeid L.M.
        Role of sphingolipids in senescence: implication in aging and age-related diseases.
        J Clin Invest. 2018; 128: 2702-2712
        • Tseng H.C.
        • Lin C.C.
        • Hsiao L.D.
        • Yang C.M.
        Lysophosphatidylcholine-induced mitochondrial fission contributes to collagen production in human cardiac fibroblasts.
        J Lipid Res. 2019; 60: 1573-1589
        • Tsuchiya S.
        • Kobayashi Y.
        • Goto Y.
        • Okumura H.
        • Nakae S.
        • Konno T.
        • et al.
        Induction of maturation in cultured human monocytic leukemia cells by a phorbol diester.
        Cancer Res. 1982; 42: 1530-1536
        • Waaijer M.E.C.
        • Gunn D.A.
        • van Heemst D.
        • Slagboom P.E.
        • Sedivy J.M.
        • Dirks R.W.
        • et al.
        Do senescence markers correlate in vitro and in situ within individual human donors?.
        Aging (Albany NY). 2018; 10: 278-289
        • Waldera Lupa D.M.
        • Kalfalah F.
        • Safferling K.
        • Boukamp P.
        • Poschmann G.
        • Volpi E.
        • et al.
        Characterization of skin aging-associated secreted proteins (SAASP) produced by dermal fibroblasts isolated from intrinsically aged human skin.
        J Invest Dermatol. 2015; 135: 1954-1968
        • Watson A.D.
        • Leitinger N.
        • Navab M.
        • Faull K.F.
        • Hörkkö S.
        • Witztum J.L.
        • et al.
        Structural identification by mass spectrometry of oxidized phospholipids in minimally oxidized low density lipoprotein that induce monocyte/endothelial interactions and evidence for their presence in vivo.
        J Biol Chem. 1997; 272: 13597-13607
        • Weinmüllner R.
        • Zbiral B.
        • Becirovic A.
        • Stelzer E.M.
        • Nagelreiter F.
        • Schosserer M.
        • et al.
        Organotypic human skin culture models constructed with senescent fibroblasts show hallmarks of skin aging.
        NPJ Aging Mech Dis. 2020; 6: 4
        • Wennberg A.M.V.
        • Schafer M.J.
        • LeBrasseur N.K.
        • Savica R.
        • Bui H.H.
        • Hagen C.E.
        • et al.
        Plasma sphingolipids are associated with gait parameters in the Mayo Clinic study of aging.
        J Gerontol A Biol Sci Med Sci. 2018; 73: 960-965
        • Wiley C.D.
        • Brumwell A.N.
        • Davis S.S.
        • Jackson J.R.
        • Valdovinos A.
        • Calhoun C.
        • et al.
        Secretion of leukotrienes by senescent lung fibroblasts promotes pulmonary fibrosis.
        JCI Insight. 2019; 4e130056
        • Witztum J.L.
        CEP is an important and ubiquitous oxidation specific epitope recognized by innate pattern recognition receptors.
        Circ Res. 2015; 117: 305-308
        • Young A.R.
        • Narita M.
        SASP reflects senescence.
        EMBO Rep. 2009; 10: 228-230
        • Zdanov S.
        • Bernard D.
        • Debacq-Chainiaux F.
        • Martien S.
        • Gosselin K.
        • Vercamer C.
        • et al.
        Normal or stress-induced fibroblast senescence involves COX-2 activity.
        Exp Cell Res. 2007; 313: 3046-3056
        • Zhang C.F.
        • Gruber F.
        • Ni C.
        • Mildner M.
        • Koenig U.
        • Karner S.
        • et al.
        Suppression of autophagy dysregulates the antioxidant response and causes premature senescence of melanocytes.
        J Invest Dermatol. 2015; 135: 1348-1357
        • Zhao Y.
        • Zhang C.F.
        • Rossiter H.
        • Eckhart L.
        • König U.
        • Karner S.
        • et al.
        Autophagy is induced by UVA and promotes removal of oxidized phospholipids and protein aggregates in epidermal keratinocytes.
        J Invest Dermatol. 2013; 133: 1629-1637