Autophagy Controls the Protein Composition of Hair Shafts

Hair shafts are formed by a special mode of keratinocyte (KC) differentiation (Harland and Plowman, 2018Harland D.P. Plowman J.E. Development of hair fibres.Adv Exp Med Biol. 2018; 1054: 109-154https://doi.org/10.1007/978-981-10-8195-8_10Crossref PubMed Scopus (21) Google Scholar) which is characterized by the accumulation of hair keratins (Krt31-Krt40, Krt81-Krt90) and keratin-associated proteins (KRTAPs), cell death by cornification and the integration of the dead cells into the growing hair shaft (Jones et al., 2018Jones L.A. Harland D.P. Jarrold B.B. Connolly J.E. Davis M.G. The walking dead: sequential nuclear and organelle destruction during hair development.Br J Dermatol. 2018; 178: 1341-1352https://doi.org/10.1111/bjd.16148Crossref Scopus (11) Google Scholar) (Figure 1a). The death of hair KCs is associated with the breakdown of the nucleus and other organelles (Fischer et al., 2011Fischer H. Szabo S. Scherz J. Jaeger K. Rossiter H. Buchberger M. et al.Essential role of the keratinocyte-specific endonuclease DNase1L2 in the removal of nuclear DNA from hair and nails.J Invest Dermatol. 2011; 131: 1208-1215https://doi.org/10.1038/jid.2011.13Abstract Full Text Full Text PDF PubMed Scopus (55) Google Scholar; Jones et al., 2018Jones L.A. Harland D.P. Jarrold B.B. Connolly J.E. Davis M.G. The walking dead: sequential nuclear and organelle destruction during hair development.Br J Dermatol. 2018; 178: 1341-1352https://doi.org/10.1111/bjd.16148Crossref Scopus (11) Google Scholar). Autophagy is a major intracellular degradation process that is controlled by autophagy-related (Atg) genes (Mizushima and Komatsu, 2011Mizushima N. Komatsu M. Autophagy: renovation of cells and tissues.Cell. 2011; 147: 728-741https://doi.org/10.1016/j.cell.2011.10.026Abstract Full Text Full Text PDF PubMed Scopus (4151) Google Scholar). The targeted deletion of Atg7 in Atg7f/f K14-Cre (Atg7ΔKC) mice abolishes autophagy in keratin K14-positive epithelial cells and all cells that are derived from them, including hair KCs (Rossiter et al., 2013Rossiter H. König U. Barresi C. Buchberger M. Ghannadan M. Zhang C.F. et al.Epidermal keratinocytes form a functional skin barrier in the absence of Atg7 dependent autophagy.J Dermatol Sci. 2013; 71: 67-75https://doi.org/10.1016/j.jdermsci.2013.04.015Abstract Full Text Full Text PDF PubMed Scopus (54) Google Scholar). In comparison to fully autophagy-competent Atg7f/f mice, Atg7Δep mice display a thickened stratum corneum (Rossiter et al., 2013Rossiter H. König U. Barresi C. Buchberger M. Ghannadan M. Zhang C.F. et al.Epidermal keratinocytes form a functional skin barrier in the absence of Atg7 dependent autophagy.J Dermatol Sci. 2013; 71: 67-75https://doi.org/10.1016/j.jdermsci.2013.04.015Abstract Full Text Full Text PDF PubMed Scopus (54) Google Scholar), altered lipid composition in the sebum (Rossiter et al., 2018Rossiter H. Stübiger G. Gröger M. König U. Gruber F. Sukseree S. et al.Inactivation of autophagy leads to changes in sebaceous gland morphology and function.Exp Dermatol. 2018; 27: 1142-1151https://doi.org/10.1111/exd.13752Crossref PubMed Scopus (19) Google Scholar) and aberrant elevation of non-cytoskeletal proteins in the cornified nails (Jaeger et al., 2019Jaeger K. Sukseree S. Zhong S. Phinney B.S. Mlitz V. Buchberger M. et al.Cornification of nail keratinocytes requires autophagy for bulk degradation of intracellular proteins while sparing components of the cytoskeleton.Apoptosis. 2019; 24: 62-73https://doi.org/10.1007/s10495-018-1505-4Crossref PubMed Scopus (18) Google Scholar). Here, we tested the hypothesis that autophagy contributes to the cellular remodeling during hair shaft cornification (Figure 1a). Hair was sampled from the backs of 6 Atg7f/f and 7 Atg7Δep mice and subjected to mass spectrometry-based proteome analysis according to an established protocol (Plott et al., 2020Plott T.J. Karim N. Durbin-Johnson B.P. Swift D.P. Scott Youngquist R. Salemi M. et al.Age-related changes in hair shaft protein profiling and genetically variant peptides.Forensic Sci Int Genet. 2020; 47102309https://doi.org/10.1016/j.fsigen.2020.102309Abstract Full Text Full Text PDF PubMed Scopus (8) Google Scholar) (Table S1; Figure S1). Hair keratins were the quantitatively dominant proteins in hair shafts of both genotypes (Figure 1b, c; Table S2), but their share in the total protein was smaller in Atg7Δep hair (Figure 1d). The contents of KRTAPs and histones (Figure 1d) were not significantly different between Atg7f/f and Atg7Δep hair. By contrast, proteins of other functional categories were increased between 2-fold (mitochondrial proteins) and more than 12-fold (chaperonin subunits) in hair of Atg7Δep mice (Figure 1d). While the fold-change of the highly abundant hair keratins was relatively small (0.86 fold), several protein types of lower absolute abundance, such as translation initiation factors, proteasome subunits and chaperonin proteins, were strongly elevated (>4 fold) when autophagy was blocked in KCs (Figure 1d). At the level of individual proteins, 49% (453/819) of the proteins meeting the criteria for quantitation were increased at least 2-fold in Atg7Δep hair (Table S2). Strikingly, the list of the top-20 elevated proteins in Atg7Δep hair (Table 1) comprised only regulators of protein turnover, i.e. factors involved in the synthesis, folding and degradation of proteins. Among them, 6 chaperonin subunits and heat shock protein beta-1 (Hspb1) ensure proper folding of proteins. Two of the top-elevated proteins were tRNA-ligases (aminoacyl-tRNA synthetases), namely cysteine-tRNA ligase (Cars1) and glycine-tRNA ligase (Gars1) (Table 1). These enzymes ligate specific amino acids to the corresponding tRNAs and thereby support protein synthesis at ribosomes. Comparison of the complete set of cytoplasmic tRNA-ligases revealed that all of the 18 tRNA-ligases quantified were elevated in Atg7Δep hair (Figure S2a). Likewise, the abundance levels of translation initiation factors and ribosomal proteins (Table 1; Table S2; Figure S2b) were consistently increased in Atg7-deficient hair, suggesting that the major components of the molecular machinery of protein synthesis accumulated due to blockade of autophagy in hair KCs. Finally, the subunits of the proteasome, which controls the main route of intracellular protein degradation besides autophagy (Rousseau and Bertolotti, 2018), were elevated in Atg7Δep hair, contributing 9 of the 20 top-elevated proteins (Table 1; Figure S2c).Table 1Proteins most strongly increased in Atg7Δep versus Atg7f/f hairRankProteinGeneFCAdjusted P-value1T-complex protein 1 subunit gammaCct322.8<0.0000012T-complex protein 1 subunit alphaTcp120.2<0.0000013T-complex protein 1 subunit zetaCct6a18.70.0060684T-complex protein 1 subunit deltaCct415.50.000025Heat shock protein beta-1Hspb114.1<0.000001626S Proteasome non-ATPase regulatory subunit 3Psmd311.5<0.000001726S Proteasome non-ATPase regulatory subunit 11Psmd1111.3<0.0000018T-complex protein 1 subunit epsilonCct510.80.000004926S Proteasome non-ATPase regulatory subunit 1Psmd110.5<0.00000110Eukaryotic translation initiation factor 2 subunit 3 X-linkedEif2s3x10.20.00000111Cysteine--tRNA ligase cytoplasmicCars110.2<0.0000011226S Proteasome non-ATPase regulatory subunit 12Psmd129.80.00000713Translation initiation factor eIF-2B subunit epsilonEif2b59.70.00239614Proteasome subunit alpha type-7Psma79.4<0.0000011526S Proteasome non-ATPase regulatory subunit 7Psmd79.30.0000116Glycine--tRNA ligaseGars19.3<0.00000117T-complex protein 1 subunit betaCct29.10.00226218Proteasome subunit alpha type-4Psma49.10.00002819Proteasome subunit alpha type-5Psma59.00.00000520Proteasome subunit alpha type-6Psma69.0<0.000001Notes: The 20 proteins with the highest fold change (FC, Atg7Δep : Atg7f/f) and adjusted P<0.01 are shown. Open table in a new tab Notes: The 20 proteins with the highest fold change (FC, Atg7Δep : Atg7f/f) and adjusted P<0.01 are shown. The most abundant protein directly involved in autophagy was gamma-aminobutyric acid receptor-associated protein-like 2 (GABARAPL2), the levels of which increased approximately 6-fold in Atg7Δep hair (Table S2). GABARAPL2 facilitates bulk autophagy of cytosolic proteins (Szalai et al., 2015Szalai P. Hagen L.K. Sætre F. Luhr M. Sponheim M. Øverbye A. et al.Autophagic bulk sequestration of cytosolic cargo is independent of LC3, but requires GABARAPs.Exp Cell Res. 2015; 333: 21-38https://doi.org/10.1016/j.yexcr.2015.02.003Crossref PubMed Scopus (55) Google Scholar) and was previously reported to accumulate upon deletion of Atg7 in the brain (Hui et al., 2019). Interleukin 1 family 10 (IL1F10, also known as IL-38) (Martin et al., 2021Martin P. Goldstein J.D. Mermoud L. Diaz-Barreiro A. Palmer G. IL-1 family antagonists in mouse and human skin inflammation.Front Immunol. 2021; 12652846https://doi.org/10.3389/fimmu.2021.652846Crossref Scopus (25) Google Scholar) was the only member of the IL-1 family detectable in hair shafts, which is remarkable because it is expressed at comparably low levels in other parts of the body (Lachner et al., 2017Lachner J. Mlitz V. Tschachler E. Eckhart L. Epidermal cornification is preceded by the expression of a keratinocyte-specific set of pyroptosis-related genes.Sci Rep. 2017; 717446https://doi.org/10.1038/s41598-017-17782-4Crossref Scopus (55) Google Scholar). IL-38 was significantly elevated in Atg7Δep hair (Table S2), suggesting that autophagy regulates the intracellular turnover or secretion of this anti-inflammatory cytokine in hair KCs. This study reveals that autophagy contributes to the normal maturation of the hair shaft. Although inhibition of autophagy does not result in gross defects of hair (Rossiter et al., 2013Rossiter H. König U. Barresi C. Buchberger M. Ghannadan M. Zhang C.F. et al.Epidermal keratinocytes form a functional skin barrier in the absence of Atg7 dependent autophagy.J Dermatol Sci. 2013; 71: 67-75https://doi.org/10.1016/j.jdermsci.2013.04.015Abstract Full Text Full Text PDF PubMed Scopus (54) Google Scholar, Rossiter et al., 2018Rossiter H. Stübiger G. Gröger M. König U. Gruber F. Sukseree S. et al.Inactivation of autophagy leads to changes in sebaceous gland morphology and function.Exp Dermatol. 2018; 27: 1142-1151https://doi.org/10.1111/exd.13752Crossref PubMed Scopus (19) Google Scholar), it alters the molecular composition of hair shafts. The spectrum of autophagy substrates, deduced from the list of proteins elevated in hair of Atg7Δep relative to Atg7f/f mice, includes functionally diverse cytoplasmic and mitochondrial proteins but excludes the main cytoskeletal proteins of hair shafts, i.e. keratins and KRTAPs. This pattern suggests that a major function of autophagy in hair KCs is to reduce the concentration of non-cytoskeletal proteins in the mature hair shaft. Thus, autophagy-dependent degradation of a broad range of cellular proteins complements the synthesis of few structural proteins, which build the mechanically resilient mature hair shaft (Figure 1c). The finding that regulators of protein homeostasis, i.e. synthesis, folding and proteasomal degradation, are particularly elevated in Atg7Δep hair suggests that they are preferential substrates of autophagy, possibly involving specific receptors (Vargas et al., 2023Vargas J.N.S. Hamasaki M. Kawabata T. Youle R.J. Yoshimori T. The mechanisms and roles of selective autophagy in mammals.Nat Rev Mol Cell Biol. 2023; 24: 167-185https://doi.org/10.1038/s41580-022-00542-2Crossref PubMed Scopus (30) Google Scholar), during normal hair KC differentiation. A similar pattern of proteome changes, especially with regard to the pronounced accumulation of proteasomal and chaperonin proteins, was detected in the nails of Atg7Δep mice (Jaeger et al., 2019Jaeger K. Sukseree S. Zhong S. Phinney B.S. Mlitz V. Buchberger M. et al.Cornification of nail keratinocytes requires autophagy for bulk degradation of intracellular proteins while sparing components of the cytoskeleton.Apoptosis. 2019; 24: 62-73https://doi.org/10.1007/s10495-018-1505-4Crossref PubMed Scopus (18) Google Scholar), indicating that the role of autophagy is conserved in both types of hard skin appendages. The present study demonstrates that mass spectrometry-based proteomics of hair shafts is a non-invasive method suitable for the detection of dysfunctions of autophagy in hair KCs. Previous reports have implicated autophagy in the control of hair growth and in the suppression of hair loss in alopecia areata (Chai et al., 2019Chai M. Jiang M. Vergnes L. Fu X. de Barros S.C. Doan N.B. et al.Stimulation of hair growth by small molecules that activate autophagy.Cell Rep. 2019; 27: 3413-3421https://doi.org/10.1016/j.celrep.2019.05.070Abstract Full Text Full Text PDF PubMed Scopus (51) Google Scholar; Gund and Christiano, 2023Gund R. Christiano A.M. Impaired autophagy promotes hair loss in the C3H/HeJ mouse model of alopecia areata.Autophagy. 2023; 19: 296-305https://doi.org/10.1080/15548627.2022.2074104Crossref Scopus (2) Google Scholar). It remains to be investigated if pathologically relevant dysfunctions of autophagy occur in KCs during hair shaft maturation, which would make them amenable to diagnostic detection by proteomics, or in other cells of the hair follicle. Animal procedures were approved by the Ethics Review Committee for Animal Experimentation of the Medical University of Vienna, Austria (approval number BMWF-66.009/0124-II/10b/2010). No experiments on live animals were performed. The raw data that support the findings of this study are openly available in the MASSive Proteomics repository at http://www.massive.ucsd.edu, reference number MSV000091072 and at http://www.proteomexchange.org, reference number PXD039530. Betz et al., 2015Betz R.C. Petukhova L. Ripke S. Huang H. Menelaou A. Redler S. et al.Genome-wide meta-analysis in alopecia areata resolves HLA associations and reveals two new susceptibility loci.Nat Commun. 2015; 6: 5966https://doi.org/10.1038/ncomms6966Crossref PubMed Scopus (174) Google Scholar. We thank Michelle Salemi and Brett Phinney (University of California Davis Proteomics Core) for expert protein mass spectrometry. This work was supported in part by NIH grant 1S10OD026918-01A1, Austrian Science Fund (FWF): P32777 and Austrian Science Fund (FWF): I 5627-B. Author contributions Conceptualization: SS, NK, FG, ET, RHR, LE; Formal Analysis: SS, NK, RHR, LE; Funding Acquisition: FG, ET, RHR, LE; Investigation: SS, SZ, KJ, HR, IMN, NK, RHR, LE; Methodology: NK, RHR; Resources: FG, ET; Supervision: RHR, LE; Validation: NK, RHR, LE; Visualization: SS, LE; Writing - Original Draft Preparation: SS, NK, RHR, LE; Writing - Review and Editing: SS, NK, KJ, SZ, HR, IMN, FG, ET, RHR, LE. Download .xlsx (.17 MB) Help with xlsx files Tabled 1Table S1. Overview of proteomic analysis of hair shafts from Atg7f/f (wildtype, WT) and Atg7Δep (epithelial knockout, EKO) miceSample# Scans# FeaturesIdentified# Peptides# Sequences# Proteins (including duplicates)# Proteins used for quantitative comparisonMS1MS/MS# Chimera# PSMs# Scans# FeaturesGroupsAllTopWT1623169542205171290281178611350974178216881118933971978814WT2631468035197841241201038910087870071496348112231761839814WT3639768054201581243831098410606908474136538116032761927815WT4629668467201861210469399905374885926513991826431517790WT5636868055207651268751099910595913673966501116932811951817WT6622768947203051276101128410873937875846706114433301891814EKO162667006122113141093158651519813453110199923152144302591819EKO263376822821464138276140331355812073101409248146743132512819EKO36446669702025912824712005116891034188118051138340582327816EKO462986941121364141603145691406712538104699532150744462592819EKO564286751720328135722146061414612695107049690151745482641819EKO664206808521174138171147311417712731105399587148844242581818EKO763866833121230138448142621376512332102779366148543912536818Blank 16239676052002896515308296252221185119253179n.a.Blank 264646606919745905397976626546457058n.a.WT/Atg7f/f mean630668517202861255101080710427892172156352111731841851811EKO/Atg7Δep mean63696837221133137366142961380012309102809342148143732540818Abbreviations: MS, mass spectrometry; PSMs, peptide spectrum matches; n.a., not applicable; #, number.Notes: Protein duplicates are identical proteins (or fragments) with different UniProt accession numbers.Quantitative comparisons were performed for proteins of which at least 2 unique peptides were identified. The values for only one accession number per protein were used. Open table in a new tab Mice The generation Atg7Δep mice by crossing Atg7f/f mice (Komatsu et al. 2005) with Tg(KRT14-cre)1Amc/J mice (Jackson Laboratories, Bar Harbour, ME) and their maintenance were reported previously (Sukseree et al. 2012; Rossiter et al. 2013; Sukseree et al. 2020). Sample preparation for proteomic analysis Mice were sacrificed at an age of 5-6 months. The back was shaved and hair fibers were collected. The hair was stored at room temperature prior to further processing. Samples from each mouse were processed separately according to a protocol modified from a previous study of mouse hair (Rice et al. 2009). Hair samples (4 mg) were rinsed three times in a solution containing 2% sodium dodecanoate and 50 mM ammonium bicarbonate and subsequently stirred in a solution of 2% sodium dodecanoate, 50 mM ammonium bicarbonate and 100 mM dithiothreitol on a magnetic stirrer at room temperature for 6 hours, followed by alkylation with 200 mM iodoacetamide for 45 minutes in the dark. Dodecanoate was extracted from the samples in ethyl acetate at pH 2. After readjusting the pH to 8 with NaOH and ammonium bicarbonate, 40 μg of reductively methylated trypsin (Rice et al. 1977) were added. The samples were stirred on a magnetic stirrer for about 48 hours at room temperature with 40 μg aliquots of reductively methylated trypsin added at 16 h intervals. Mass spectrometry (MS) and protein identification The digested peptide samples from the hair of 6 Atg7f/f and 7 Atg7Δep mice were randomized and analyzed by LC–1MS/MS on a Thermo Scientific Dionex UltiMate 3000 RSLC system using a Thermo PepMap trap for desalting and a PepSep (PepSep, Denmark) ReproSil 8 cm 150 μm I.D. C18 column with 1.5 μm particle size (120 Å pores), heated to 40°C. Separation was performed in a total run time of 60 min with a flow rate of 500 nl/min with mobile phases A: water/0.1% formic acid and B: 80% acetonitrile/0.1% formic acid. Peptides were directly eluted into an Orbitrap Exploris 480 instrument (Thermo Fisher Scientific, Bremen, Germany). Spray voltage was set to 1.8 kV, funnel RF level at 45, and heated capillary temperature at 275°C. The full MS resolution was set to 60000 at m/z 200 and full MS AGC target was 300% with an IT set to Auto. Mass range was set to 350–1500. AGC target value for fragment spectra was set at 200% with a resolution of 15,000 and injection time was set to Standard and Top40. Intensity threshold was kept at 5E3. Isolation width was set at 1.6 m/z, and normalized collision energy was set at 30%. Raw data files from mass spectrometry were searched in PEAKS Studio 10.6 build 20201015 (Bioinformatics Solutions Inc., Waterloo, ON, Canada) against a validated UniProt mouse reference proteome (UniProt-proteome UP000000589_Mouse) containing 63,759 proteins using search parameters reported previously (Plott et al. 2020). Label-free quantitation was performed using the Q-module function of PEAKS Studio (Zhang et al. 2012), which determines protein abundance (abundance units) by measuring the area under the curve in a plot of the extracted ion chromatogram (XIC) intensity over retention time (minutes) for the top 3 peptides of each protein. Statistical analysis The data were analyzed using GraphPad Prism version 8.0.1 for Windows, GraphPad Software, San Diego, CA, USA. For each protein, only one accession number was used to derive quantitative values from mass spectrometry for statistical analysis. The significance of differences (adjusted P-values) was determined by two-tailed t-tests with a correction for multiple comparisons using the Holm-Šídák method with alpha = 0.05. Data are presented in figures as mean +/- standard deviation. Supplementary references Komatsu M, Waguri S, Ueno T, Iwata J, Murata S, Tanida I, et al. Impairment of starvation-induced and constitutive autophagy in Atg7-deficient mice. J Cell Biol. 2005;169:425-434. doi: 10.1083/jcb.200412022. Plott TJ, Karim N, Durbin-Johnson BP, Swift DP, Scott Youngquist R, Salemi M, et al. Age-related changes in hair shaft protein profiling and genetically variant peptides. Forensic Sci Int Genet. 2020;47:102309. doi: 10.1016/j.fsigen.2020.102309. Rice RH, Means GE, Brown WD. Stabilization of bovine trypsin by reductive methylation. Biochim Biophys Acta 1977;492:316-321. doi: 10.1016/0005-2795(77)90082-4. Rice RH, Rocke DM, Tsai HS, Silva KA, Lee YJ, Sundberg JP. Distinguishing mouse strains by proteomic analysis of pelage hair. J Invest Dermatol. 2009;129:2120-2125. doi: 10.1038/jid.2009.52. Rossiter H, König U, Barresi C, Buchberger M, Ghannadan M, Zhang CF, et al. Epidermal keratinocytes form a functional skin barrier in the absence of Atg7 dependent autophagy. J Dermatol Sci. 2013;71:67-75. doi: 10.1016/j.jdermsci.2013.04.015. Sukseree S, Mildner M, Rossiter H, Pammer J, Zhang CF, Watanapokasin R, et al. Autophagy in the thymic epithelium is dispensable for the development of self-tolerance in a novel mouse model. PLoS One 2012;7:e38933. doi: 10.1371/journal.pone.0038933. Sukseree S, Schwarze UY, Gruber R, Gruber F, Quiles Del Rey M, Mancias JD, et al. ATG7 is essential for secretion of iron from ameloblasts and normal growth of murine incisors during aging. Autophagy 2020;16:1851-1857. doi: 10.1080/15548627.2019.1709764. Zhang J, Xin L, Shan B, Chen W, Xie M, Yuen D, et al. PEAKS DB: de novo sequencing assisted database search for sensitive and accurate peptide identification. Mol Cell Proteomics 2012;11:M111.010587. doi: 10.1074/mcp.M111.010587. Figure S2. The abundance of proteins regulating the synthesis and degradation of proteins is elevated in hair shafts formed by autophagy-deficient hair keratinocytes. The abundances of (a) tRNA-ligases, (b) protein components of the ribosome 40S subunit and (c) subunits of proteasomes in hair shafts from the backs of Atg7f/f (wildtype, grey bars) and Atg7Dep (Atg7-epithelial knockout, red bars) mice were determined by mass spectrometry-based proteomics using label-free quantitation. Significant differences (multiple t-tests with correction for multiple comparisons using the Holm-Šídák method) with adjusted P-values <0.05 and <0.01 are marked with one and two asterisks, respectively. Abbreviations: a.u., abundance units; n.s., not significant.View Large Image Figure ViewerDownload Hi-res image Download (PPT)