Atopic dermatitis (AD) is a disease characterized by chronically relapsing-remitting skin inflammation. The etiology of AD is multifactorial, involving gene-environment interaction with strong hereditability (80%) (Larsen, 1993). Recent studies show that epidermal barrier dysfunction is a central feature in the pathogenesis of AD (Cork et al., 2009; Luger et al., 2021). Genetic studies point towards a region that harbours several genes involved in epidermal barrier maintenance (Cookson et al., 2001; Paternoster et al., 2011), known as the “epidermal differentiation complex” (EDC) (Mischke et al., 1996), spanning 1.9 Mbp within chromosome 1q21 and known to be prone to chromosomal rearrangement (Forus et al., 1998; Itoyama et al., 2002; Chen et al., 2003; Wong et al., 2003). The genes of the EDC can be grouped into S100 calcium binding proteins (Eckert et al., 2004; Marenholz et al., 2004), S100 fused-type protein (SFTP) family (Gan et al., 1990; Lee et al., 1993; Krieg et al., 1997; Contzler et al., 2005; Takaishi et al., 2005; Wu et al., 2011; Kypriotou et al., 2012), cornified envelope precursor family (Backendorf and Hohl, 1992), and small proline-rich proteins (Zhao and Elder, 1997; Marshall et al., 2001; Jackson et al., 2005). Among the 63 genes (59 protein coding genes and four pseudogenes) located within the EDC, null mutations in the gene encoding profilaggrin (FLG), an SFTP gene, was shown to be the major risk factor for AD (Morar et al., 2007).

The first suggestion of the involvement of filaggrin in barrier maintenance was reported by Sybert et al. (Sybert et al., 1985) where reduction of its expression was correlated with ichthyosis vulgaris (IV); follow-up studies confirmed reduction or loss of filaggrin expression with epidermal barrier dysfunction (Fleckman et al., 1987; Peña Penabad et al., 1998). A breakthrough came with the discovery that loss of function FLG mutations (R501X and 2282del4) are highly prevalent in the IV (Smith et al., 2006) and AD patients (Palmer et al., 2006); this was replicated on different genetic backgrounds and ethnicities (Marenholz et al., 2006; Ruether et al., 2006; Sandilands et al., 2006; Weidinger et al., 2006; Barker et al., 2007; Morar et al., 2007; Nomura et al., 2007; Enomoto et al., 2008; Rodríguez et al., 2008; Nemoto-Hasebe et al., 2009; Osawa et al., 2010; Cheng et al., 2012; Pigors et al., 2018; Handa et al., 2019; Koseki et al., 2019; Jurakic Toncic et al., 2020; Smieszek et al., 2020), with hundreds of mutations now identified (Karczewski et al., 2020). Importantly, while filaggrin expression is almost entirely restricted to the epidermis, FLG mutations have been also shown to be linked to additional manifestations of atopic march and allergy, including food (Brown et al., 2011) and contact allergies (Novak et al., 2008), asthma (van den Oord and Sheikh, 2009; Rodríguez et al., 2009), allergic rhinitis (van den Oord and Sheikh, 2009) and eosinophilic esophagitis (Sherrill and Blanchard, 2014).

Composed of approximately 4,061 amino acids (aa), profilaggrin is the largest protein of the SFTP family; the protein is structurally complex and composed of 10–12 filaggrin monomer repeats flanked with truncated filaggrin repeats (Gan et al., 1990) and a S100 domain at the N-terminus (Kypriotou et al., 2012). Unlike any other SFTP, profilaggrin contains a “bipartite” nuclear localization signal (Lu et al., 2021) next to the S100 calcium-binding domain, an indicator of its nuclear function (Presland et al., 1992; Aho et al., 2012). Upon expression in keratinocytes, profilaggrin is phosphorylated and stored within keratohyalin granules (KHGs), mainly present in the stratum granulosum layer (Resing et al., 2002), from where it is released by an AKT1-dependent, actin scaffold-driven mechanism that we have recently described (Gutowska-Owsiak et al., 2018). It is speculated that dephosphorylation makes the protein accessible to the pro-protein convertase-mediated cleavage (Resing et al., 1993). The cleaved N-terminal domain translocates into the nucleus, where it is involved in denucleation (Ishida-Yamamoto et al., 1998; Pearton et al., 2002; Yamamoto-Tanaka et al., 2014) and control of epidermal homeostasis (Aho et al., 2012; Naeem et al., 2015); the remaining part is cleaved by SASPase (Matsui et al., 2011) and KLK5 (Sakabe et al., 2013) proteolytic enzymes. Monomeric filaggrin promotes aggregation and collapse of keratin intermediate filaments (IFs), resulting in the formation of more squamous flattened cells (Steinert et al., 1981; Candi et al., 2005). In parallel, further events occur, such as the conversion of arginine (Arg) residues to citrulline (Nachat et al., 2005) and covalent cross-linking of monomers by transglutaminase, which leads to stabilization of the cornified cell envelope (Takahashi et al., 1996). Finally, the crosslinked filaggrin undergoes extensive proteolytic cleavage by caspase-14 (Denecker et al., 2007), calpain-1 (Yamazaki et al., 1997), bleomycin hydrolase (Kamata et al., 2009), elastase-2 (Bonnart et al., 2010), matripase (List et al., 2003), prostatin (Leyvraz et al., 2005) and other proteases, resulting in a pool of hygroscopic aa and derivatives, i.e., urocanic acid (UCA), a derivative of histidine, highly abundant in the protein, and pyrrolidone carboxylic acid (PCA), a glutamine derivative. These constitute the majority of the so-called “natural moisturizing factor” (NMF) (Rawlings and Harding, 2004), contributing to stratum corneum (SC) hydration, as well as acidic pH, with antimicrobial action (Miajlovic et al., 2010). In addition, trans-UCA may protect cells in deeper layers from UVB-mediated mutagenesis by absorbing UVB (Tabachnick, 1957). FLG mutations significantly reduce the amount of NMF in SC compared to the healthy control (Kezic et al., 2011).

The ubiquitin-proteasome system (UPS) is a major proteolytic pathway that removes damaged and unwanted proteins. The selective turnover is initiated by a covalent attachment of a small ubiquitin (Ub) protein, mainly to the internal lysine (Lys) residues, which is mediated by an enzymatic cascade orchestrated by E1, E2, and E3 enzymes (Hershko and Ciechanover, 1992; Komander and Rape, 2012). However, a single ubiquitination event is usually insufficient to target a protein for degradation; several kinds of polyubiquitin chains are formed between the Lys residues of the Ub subunits and these have specific functions (Hershko and Ciechanover, 2003; Husnjak and Dikic, 2012; Oh et al., 2018). The Lys11 and Lys48 linkages, and their combination drive proteasomal degradation (Chau et al., 1989; Jin et al., 2008; Meyer and Rape, 2014), whereas Lys6 and Lys63 linkages mediate the processes of autophagy (Ordureau et al., 2014), endocytosis-exocytosis (Lauwers et al., 2009)^, or lysosomal degradation (Duncan et al., 2006). Interestingly, ubiquitination can also occur on the free amino group of the N-terminus of a protein, as well as on serine (Ser), cysteine (Cys), and threonine (Thr) residues, and can lead to proteasomal turnover of the modified proteins (Kravtsova-Ivantsiv and Ciechanover, 2012; McClellan et al., 2019).

E3 ligases, which transfer Ub onto a substrate protein, mainly recognize substrates through their short linear motif called the primary degron, which may be located at the unstructured ends of the proteins, inducing protein elimination via the N- or C-degron pathways or internally. In addition, proteolytic cleavage may lead to the emergence of a degron motif within the novel N- or C-terminus, subsequently initiating degradation of the cleavage products (Guharoy et al., 2016; Varshavsky, 2019; Guharoy et al., 2022). Moreover, post-translational modifications (PTMs) can also modulate the recognition of the primary degrons by E3 ligases (van Roey et al., 2013; Guharoy et al., 2016; Lucas and Ciulli, 2017; Millar et al., 2019; Varshavsky, 2019; Chen and Kashina, 2021; Guharoy et al., 2022). Following interaction with a degron, E3 mediates ubiquitination in the proximal position(s) (secondary degron), often found close to the intrinsically disordered region (IDR) (tertiary degron), which triggers substrate breakdown by successful activation of the proteasome. Mutations affecting degron components can enhance protein stability and disrupt cellular proteostasis, leading to disease (Tokheim et al., 2021; Guharoy et al., 2022; Kampmeyer et al., 2022). Accordingly, adequately controlled profilaggrin turnover is likely crucial for SC functionality since its reduced expression or instability confers vulnerability and impacts several epidermal functions, leading to pathological conditions (Scott et al., 1982; Resing et al., 1984; Moosbrugger-Martinz et al., 2022). However, despite the central role of profilaggrin and filaggrin in skin barrier function, little is known about the regulation of their degradation, the proteolytic mechanisms involved, and the influence of FLG mutations on their turnover and stability.

In this study, we investigated proteasome inhibition’s effect on the intracellular profilaggrin level and its processing products. We also studied the entire profilaggrin sequence for the presence of degrons and examined possible ubiquitination sites throughout. Furthermore, we analysed the effect of recurrent pathogenic and rare family-specific frameshift and nonsense FLG mutations on the introduction of possible ubiquitination sites, the appearance of novel degrons, and protein stability.

Expression of profilaggrin and its processing into filaggrin monomer units, as well as their further breakdown into the components of the NMF, is essential for the functionality of the epidermal barrier; disturbances at any point of this intricate process are detrimental and lead to pathology. Given their role in the collapse of the keratin-based cytoskeleton and the striking effect on nuclear integrity, it is apparent why the accumulation of free filaggrin monomers into the cytosol of a keratinocyte is toxic to the cell and initiates its programmed death (Dale et al., 1997; Kuechle et al., 2000; Presland et al., 2001). Hence, both the released filaggrin monomers and any free profilaggrin molecules that could be processed must be under rigorous control in the cytosol to prevent premature cell death and allow for keratinocyte differentiation and stratification, critical for the formation of the functional skin barrier. Such control may be executed by different means, including protein sequestration, removal of the excess from the cytosol, and re-direction for turnover by the UPS.

We have previously described two separate mechanisms that control intracellular filaggrin levels during keratinocyte differentiation; the actin-based Akt-1/HspB1-dependent mechanism governing profilaggrin sequestration in KHGs (Gutowska-Owsiak et al., 2018) as well as the small extracellular vesicle (sEV)-mediated removal of excess free-floating profilaggrin/filaggrin from the cytosol (Gutowska-Owsiak, 2022). As for the latter, we also determined that Staphylococcus aureus, a skin pathogen with high prevalence and significant contribution to the pathology in AD patients, enhances filaggrin loading into the sEV cargo, facilitating its removal from the skin. Modeling protein networks indicated that the link between TLR2 signaling, profilaggrin processing, and cargo loading into the sEVS might include proteins involved in ubiquitination, pointing to protein degradation and intracellular trafficking. Indeed, the importance of protein turnover has been shown previously to be critical for controlling cellular protein abundance (Yen et al., 2008; Cambridge et al., 2011), and we envisage that both subcellular localization and propensity for degradation are key for homeostasis within the profilaggrin/filaggrin system. Ubiquitination is important for protein trafficking to diverse cellular localizations, including the compartments of the endocytic system, as well as nuclear localization aiding gene regulation. These pathways are also exploited by pathogenic bacteria and viruses for their successful transmission and immune evasion (Millar et al., 2019; Tokheim et al., 2021; Kampmeyer et al., 2022; Moosbrugger-Martinz et al., 2022).

The importance of the proteasome in profilaggrin turnover and control of its intracellular fate can be speculated already based on the appearance of greatly enlarged KHGs (hypergranulosis) in the skin of patients with an autosomal recessive epidermal abnormality known as keratosis linearis with ichthyosis congenita and sclerosing keratoderma (KLICK) syndrome (Vahlquist et al., 1997)^, (Dahlqvist et al., 2010)^, (Takeichi and Akiyama, 2020). Those patients have a deletion of a single nucleotide at the 5’ untranslated region of the proteasomal maturation protein (POMP), a chaperon that mediates stabilization of the proteasome complex; thus, loss of function leads to insufficiency of the proteasome in differentiating keratinocytes (Dahlqvist et al., 2010; Morice-Picard et al., 2017; Onnis et al., 2018). KLICK patients also demonstrate thickened SC and aberrant filaggrin staining with an antibody directed against filaggrin monomer (Dahlqvist et al., 2010).

In this study we were able to confirm the involvement of the UPS in profilaggrin degradation using a proteasome inhibitor MG132 and deubiquitinases inhibitor PR-619. Treatment with these compounds allowed us to observe the accumulation of profilaggrin and high molecular weight processed products in the lysates, suggesting reduced profilaggrin turnover likely underlying hypergranulosis in the KLICK syndrome. At the same time, the reduction in keratinocyte viability after the overnight treatment could result from the accumulation of the free unsequestered/uncontrolled monomer filaggrin units (Presland et al., 2001). Unfortunately, dead cells where this could be potentially detectable were not evaluated in this study.

Our findings are in a sharp contrast to just published short communication by Briot et al. who failed to observe a difference upon proteasome inhibition and speculated that filaggrin monomer is not degraded via proteasome (Briot et al., 2023). We believe that this discrepancy results from the use of different models. Specifically, we used cells growing in monolayers, where many cells present with low, but detectable profilaggrin expression in relatively undifferentiated cells (Gutowska-Owsiak, 2022). This is where the greatest control over free profilaggrin must be exerted and where it could undergo the UPS-mediated turnover. With increased profilaggrin expression in the 3D epidermal equivalent used by Briot et al., the vast majority of the protein is already sequestered within KGHs, therefore the risk of keratin aggregation by free cytosolic filaggrin is reduced. Importantly, such containment of the protein within KHGs prevents its access to the proteasome, which is not able to sample from those insoluble organelles. The importance of UPS-mediated control over profilaggrin is also supported by the biological pattern of the proteasome expression, which functions primarily in the undifferentiated keratinocytes, corresponding to basal and suprabasal epidermal layers (Zieba et al., 2017); proteasome gets disassembled in the cells at the late stages of differentiation, which results from regulation of POMP expression.

To get a global picture of features important from the perspective of regulation by the UPS, it is important to consider multiple factors, i.e., internal degron motifs, different properties of the N-/C-termini, such as their stability and hydrophobicity, structural features of the protein with regards to the solvent accessibility and IDRs, and residues that may undergo ubiquitination. Here, we determined sequence characteristics by identifying the degron sequences with the DEGRONOPEDIA web server, complemented with the detailed analysis of PTMs, i.e., ubiquitination and phosphorylation with the potential to govern profilaggrin/filaggrin degradation and cellular trafficking, allowing us to propose likely degradation pathways. Protein degradation via the UPS often requires the formation of the Lys-linked polyubiquitin chains, while the nuclear re-directing depends on the Lys monoubiquitination (Trotman et al., 2007). Apart from the Lys residues, the profilaggrin sequence contains multiple Cys, Ser, and Thr residues; such residues have been shown to undergo ubiquitination at these marks. Furthermore, it has been shown that such non-canonical ubiquitination is less thermodynamically stable (McDowell and Philpott, 2013; McClellan et al., 2019; Squair and Virdee, 2022) than the canonical Lys ubiquitination and predominantly drives proteasomal degradation over protein sorting (McClellan et al., 2019). The high enrichment of the canonical ubiquitination sites in the N-terminal domains of profilaggrin and the accumulation of the non-canonical ubiquitination sites within the monomer repeats are interesting observations and suggest potential differences in the degradation pathways between the domains. The N-terminal domain is involved in nuclear signaling events that initiate a positive feedback loop for profilaggrin expression and keratinocyte differentiation. Proteasome activity is considered high in the nucleus, thus, increased predicted stability values could potentially indicate relative resistance of the profilaggrin N-terminal domain.

Furthermore, we found that the entire profilaggrin sequence contains 18 putative primary degron sites. Our screening suggests that the unprocessed profilaggrin protein is likely degraded upon recognition of the Ac/N-degron motif as the second N-terminal residue is Ser and Nα-terminal acetylation (Nt-acetylation) of nascent proteins, whose N-terminus contains Met or a small uncharged residue alanine, Cys, Val, Ser or Thr as these become N-terminal after co-translational removal of Met by Met-aminopeptidase), is one of the most significant and common protein modifications occurring on ∼80% of all human proteins (van Damme et al., 2012; Aksnes et al., 2016). Nt-acetylation can serve as a degron motif for the two E3 ligases: March6/TEB4, a Really Interesting New Gene (RING)-type E3 ligase located in the endoplasmic reticulum membrane, and NOT4, a component of the CCR4-NOT multi-subunit complex, which induces substrate protein degradation via the Ac/N-degron pathway (Varshavsky, 2011; Shemorry et al., 2013; Park et al., 2015; Lee et al., 2016).

The presence of 17 internal degrons in wild-type profilaggrin indicates the possibility of its N-/C-terminus-independent degradation. Among these degrons, the destruction box (DBOX), KEN box, and ABBA motifs are found and substrates containing one or more of these sequences typically undergo polyubiquitylation by the anaphase-promoting complex (APC/C) (Brown et al., 2014). APC/C is a large, multi-subunit E3 ligase that regulates cell cycle progression in eukaryotes. Importantly, it is also involved in proliferation and differentiation regulation in human primary keratinocytes (Quek et al., 2018). Hence, it is probable that APC/C could influence the keratinocyte life cycle and contribute to the disruption in epidermal homeostasis, such as that seen in AD, at least partly through its effect on profilaggrin degradation.

SCFβ-TRCP E3 ligase complex regulates proteasome-dependent degradation of various substrates, including early mitotic inhibitor 1 (EMI1) (Margottin-Goguet et al., 2003), cell cycle homolog 25 (CDC25A) (Busino et al., 2003; Jin et al., 2006) and vascular endothelial growth factor receptor 2 (VEGFR2) (Shaik et al., 2012). Profilaggrin also contains the consensus degron sequence (D(S)Gx{2,3}([ST]) of SCF^β−TRCP where Ser/Thr residues should be phosphorylated for proper motif recognition (x—any aa) (Winston et al., 1999; Guharoy et al., 2016). Interestingly, this motif was shown to be phosphorylated in rat profilaggrin (Resing et al., 1995) but neither the iPTMNet (Huang et al., 2018) nor PhoshoSitePlus (Hornbeck et al., 2015) databases, incorporated into the DEGRONOPEDIA, record any phosphorylation within them.

Profilaggrin has also been proposed as a binding partner of receptors of the cullin E3 ligases: FBXW7 (Xu et al., 2021), DTL (Huttlin et al., 2021), and VHL (Ewing et al., 2007). Present in proliferating cells, FBXW7 (F-Box and WD Repeat Domain Containing 7) is a member of the F-box family of proteins and functions as a substrate recognition element of the SCF E3 ligase. FBXW7 isoforms recognize their substrates through the CDC4 phosphodegron (CPD) motif so that they can be ubiquitinated and targeted for degradation by the proteasome along with the substrate. Substrates of FBXW7 containing CPD variants include MYC, Cyclin E/cyclin-dependent kinase CDK2, JUN, Myeloid cell leukemia-1 (MCL-1), mammalian target of rapamycin (mTOR), and NOTCH1 (Koepp et al., 2001; Wei et al., 2005; Mo et al., 2007; Mao et al., 2008; Inuzuka et al., 2011; King et al., 2013; Yeh et al., 2018). Analysis of the profilaggrin sequence did not reveal the occurrence of CPD motifs, which may suggest a profilaggrin-specific degron/phosphodegron that requires further identification. DTL is another receptor likely to bind to profilaggrin; it associates with CRL4A (cullin 4A-RING ubiquitin ligase) and regulates DNA replication and the cell cycle by regulating proteins such as CDT1, PR-Set7/Set8/KMT5A and p21 (Jin et al., 2006; Abbas and Dutta, 2009; Centore et al., 2010; Shibata et al., 2011; Cui et al., 2019). Since DTL prevents DNA damage after UV irradiation (Ishii et al., 2010), we hypothesize its involvement in skin function, possibly through modulation of profilaggrin or filaggrin monomer levels. In addition, profilaggrin was detected as an interactor of cullin-2 (CUL2)-based E3 ligases (Bennett et al., 2010) and VHL (von Hippel-Lindau). CUL2 is a platform for the Elongin B and Elongin C adaptor protein complex, which interact with various substrate receptors, such as the VHL tumor suppressor protein (Gossage et al., 2014; Cai and Yang, 2016). Since VHL plays a role in skin inflammation, CUL2 and VHL may regulate profilaggrin in psoriasis (Martínez-Torres et al., 2022).

Our analysis indicated that the monomers generated by SASPase cleavage are relatively stable, given their minor decrease in the N-terminal PSI, but a significant increase in the C-terminal PSI and a drop in GHI value; this likely permits their long-lasting functionality required for cross-linking of the proteins within the cornified envelope, and aggregation and collapse of the IF-based cytoskeleton. Intriguingly, the UPS system has been previously implicated in the degradation and turnover of keratins (Yamazaki et al., 2012); increased keratin turnover is observed in a subtype of epidermolysis bullosa simplex (EBS) patients, resulting in shortened lifespan, replicative senescence, and decreased cellular resistance in keratinocytes (Logli et al., 2022).

It is important to note that the UPS-mediated degradation is a pathway separate from the pathway of profilaggrin processing to monomers and subsequently into the compounds of the NMF, which occurs predominantly in the SC containing no living cells. UPS components primarily function in the intracellular environment; however, both the ubiquitinating enzymes and proteasome are also present and active in the extracellular spaces and body fluids (Takada et al., 1997; Sixt and Dahlmann, 2008; Majetschak, 2011; Ben-Nissan et al., 2022), potentially also within the SC. However, the difference in pH could affect the efficiency of the degradation process, given the impact of the acidic environment. All three hydrolytic activities (chymotrypsin-like, trypsin-like, and caspase-like) of the 26S proteasome decreased when the pH was lowered from 7.5 to 7.0 (Ugai et al., 1993; Klinkradt et al., 1997). At the same time, the 20S proteasome from human platelets displayed chymotrypsin-like activity with an optimum pH of 5.0–5.5 (Ostrowska et al., 2009). It can be speculated that similar 20S activity may be maintained within the SC, enabling partial degradation of targeted proteins. Changes in pH may also affect the ubiquitination process, e.g., pH lower than 7.5 destabilized the APC/C complex (Passmore et al., 2005), and we identified potential APC/C degrons in profilaggrin. These observations may suggest that profilaggrin and its derivatives do not undergo intense turnover in the SC.

For this study, we assumed that profilaggrin processing may coincide with the UPS-mediated degradation, at least to some extent. On the other hand, while the majority of profilaggrin processing to monomers can be localized to the SC, i.e., after keratinocyte death, it is unclear how much processing may take place in the live epidermal layers. However, small but detectable amounts of the filaggrin monomer have been observed by Western blot in keratinocytes grown in 2D cultures which were not stratified (Dang et al., 2016), so we could envisage that the process likely begins much earlier, albeit probably at a reduced rate, since the highest expression of the processing enzymes was demonstrated for the top epidermal layers (Miyachi et al., 1986; Pearton et al., 2001; Bernard et al., 2005).

Here we also determined that the common FLG mutations predisposing to AD result in the generation of new products with a different propensity for degradation, supported by changes in stability and ubiquitination-prone residues as well as the emergence of novel degrons. This may suggest that both the filaggrin turnover and organellar distribution in the skin of the patients are affected, and the generated products may undergo altered cellular fates. Indeed, in the case of several mutations, we noted the accumulation of all those pro-degradation motifs. At the same time, while introducing several new ubiquitination-prone residues compared to the wild-type of the same length, one of the most common pathological FLG mutations, 2282del4, does not introduce any novel known degrons. However, it is important to note that it is not clear to what extent profilaggrin could undergo modification at all these potentially ubiquitinated residues; since the protein is known to undergo many different modifications at Lys residues.

FLG mutations resulting in the loss of the C-terminal domain led to practically nonfunctional monomers generated from this allele (Presland et al., 2000; Sandilands et al., 2007); hence, it has been long assumed that the profilaggrin C-terminal domain is critical for processing of the protein. However, it is not clear whether this is of crucial importance, as no experimental data on the C-terminal domain’s involvement in the processing has yet been published. Here, we determined that the C-terminal sequence of the wild-type filaggrin has a stability score in the range that is typical for most human proteins, therefore not allowing us to consider this to be a highly stabilizing part for the entire profilaggrin molecule.

Interestingly, we found the C-terminal PSI scores to be increased for nearly all the mutated products, complicating the picture further. It could be speculated that the increased stability of the mutated product would result in the accumulation of such protein intracellularly, albeit likely not in the form of granules (Sybert et al., 1985). However, there is no evidence in the samples from the patients, and certainly, no evidence indicating increased processing of the protein as this would result in high amounts of NMF being generated and barrier function maintenance. It is possible that in this scenario, the remaining routes (proteasomal degradation, vesicular export) would be enhanced to control the profilaggrin/filaggrin levels in the cytosol. Indeed, we have previously made a very unexpected observation of the increased amount of the profilaggrin/filaggrin protein cargo in the sEVs fractions isolated from the blood of AD patients compared to the healthy controls (Gutowska-Owsiak, 2022), and this would perfectly align with such a scenario. Thus, it is conceivable that keratinocytes could expel the excess profilaggrin/filaggrin-related products to prevent the toxic consequences in the event of a diminished capacity to generate KHGs (Sybert et al., 1985). It would require additional experimental work beyond the scope of this study to determine if the mutated products can indeed be detected in the sEVs found in the blood of the patients.

In summary, we determined that proteasome-mediated profilaggrin degradation is one of the means of controlling intracellular levels of the protein in keratinocytes, and we identified critical components regulating UPS-mediated processing (ubiquitination-prone residues and degrons) within its sequence. Furthermore, we also described how FLG mutations might affect the stability of mutant proteins and potential degradation routes in the cell.