If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
INSERM U781, Paris, FranceUniversité Paris V René Descartes, Paris, FranceDepartment of Genetics and Department of Dermatology, Necker Hospital, Paris, France
Lymphoepithelial Kazal-type related inhibitor (LEKTI) is a multidomain proteinase inhibitor whose defective expression causes Netherton syndrome (NS). LEKTI is encoded by SPINK5, which is also a susceptibility gene for atopic disease. In this issue, Fortugno et al. report an elegant and thorough study of the LEKTI proteolytic activation process in which they identify the precise nature of the cleavage sites used and the bioactive fragments generated. They propose a proteolytic activation model in human skin and confirm differential inhibition of kallikrein (KLK) 5, 7, and 14 by the major physiological LEKTI fragments. They show that these bioactive fragments inhibit KLK-mediated proteolysis of desmoglein 1 (DSG1) and suggest a fine-tuned inhibition process controlling target serine proteinase (SP) activity.
Proteases and their inhibitors as essential actors in skin desquamation and inflammation: loss of control in NS
Proteinases and regulation of their activity play important roles in many aspects of skin biology and function, such as epidermal differentiation, barrier formation and shedding, inflammation, immune responses, host defense, wound healing, and hair formation (
). Human tissue kallikreins (KLKs) are secreted SPs with (chymo)trypsin-like specificity that comprise 15 members, several of which are expressed in the skin (
). In particular, KLK5, KLK7, and KLK14 are involved in the desquamation of superficial layers of corneocytes by degrading corneodesmosomal components. They also activate protease-activated receptor-2 (PAR-2), resulting in the production of inflammatory cytokines (
). Protease inhibitors are essential actors in the control of protease activity. In the epidermis, they include LEKTI (or LEKTI-1), LEKTI-2, SPINK6, skin-derived antileukoprotease (SKALP/elafin), secretory leukocyte protease inhibitor, bikunin, hurpin, other SP inhibitors (serpins), and cystatins (
The importance of a tight control of protease activity has been dramatically illustrated in NS, a rare and severe genetic skin disease in which loss of balance between LEKTI, a SP inhibitor, and its target proteinases leads to a profound skin barrier defect with severe inflammatory and allergic manifestations (
). KLK5 degrades DSG1, causing abnormal stratum corneum detachment, leading to a severe skin barrier defect that favors increased allergen penetration in the skin. In parallel, unrestrained KLK5 activity triggers an autonomous inflammatory cascade by activating PAR-2 signaling, resulting in the production of major pro-inflammatory molecules and pro–T helper type 2 (Th2) cytokines such as thymic stromal lymphopoietin (TSLP). This specific KLK5–PAR2–TSLP pathway, in addition to producing thymus and activation-regulated chemokine and macrophage-derived chemokine, IL-1, and tumor necrosis factor-α, contributes to the creation of a pro-inflammatory and pro-allergic Th2 microenvironment that predisposes NS patients to severe skin allergy, asthma, and food allergy (
Elastase 2 is expressed in human and mouse epidermis and impairs skin barrier function in Netherton syndrome through filaggrin and lipid misprocessing.
Infliximab infusions for Netherton syndrome: sustained clinical improvement correlates with a reduction of thymic stromal lymphopoietin levels in the skin.
). LEKTI is encoded by SPINK5, which maps to chromosome 5q33.1 within a gene family cluster including SPINK6 and SPINK9, encoding the single-domain inhibitors SPINK6 and LEKTI-2, respectively (
). SPINK5 consists of 33 exons spread over 61 kb, encoding 15 domains preceded by a peptide signal and interspaced by linker regions. The LEKTI protein primary structure shows that each domain shares a Kazal-type related motif found in SP inhibitors. In fact, only domains 2 and 15 contain the classic Kazal motif defined by the presence of 6-cysteine residues with precise spacing; other domains exhibit only a 4-cysteine pattern. This allows the formation of two or three intramolecular disulfide bonds, resulting in a rigid reactive center loop that mimics the substrate and traps the target protease. Each LEKTI domain, except for D1, D2, and D15, shares an arginine residue at position P1, which predicts activity against trypsin-like SPs (
). Linkers are arginine- and lysine-rich sequences that represent putative cleavage sites for convertin/subtilisin-like enzymes such as furin.
Tartaglia-Polcini et al. have shown that LEKTI is transcribed into three different transcripts that differ through their carboxy-terminal end by alternative processing (
). In addition to the 15-domain isoform (LEKTI-FL), SPINK5 encodes a shorter LEKTI isoform (LEKTI-Sh), containing only the first 13 domains resulting from differential polyadenylation. The third isoform is longer (LEKTI-L), carrying a 30-amino-acid residue insertion between the thirteenth and fourteenth domains. All three SPINK5 alternative transcripts are translated into protein in differentiated keratinocytes, with LEKTI-FL being the most abundant. LEKTI pro-protein is rapidly cleaved intracellularly, generating a number of potentially bioactive fragments (
LEKTI is localized in lamellar granules, separated from KLK5 and KLK7, and is secreted in the extracellular spaces of the superficial stratum granulosum.
LEKTI proteolytic activation cascade generates a wide variety of bioactive fragments
In this issue, Fortugno et al. report the results of a careful and exhaustive study of LEKTI proteolytic processing aimed at identifying the precise nature of the generated LEKTI polypeptides. The investigators combined antibody mapping, N-terminal sequencing, and specific mutagenesis of predicted cleavage sites to identify precisely the cleavage sites used and the corresponding LEKTI fragments generated. Pro-LEKTI proteolytic processing was first studied in normal human epidermis, in differentiated normal human keratinocytes (NHKs), and in human embryonic kidney cells transfected with an expression plasmid containing the cDNA encoding for LEKTI-FL, LEKTI-L, and LEKTI-Sh. The resulting proteolytic patterns showed bands of 65/68, 42, 37, 30, and 23 kDa, comparable to those reported by
. The 68-kDa band seen with LEKTI-L corresponds to the extra 30-amino-acid residues of this isoform in linker region 13. Fragments of 102, 105, and 79 kDa were also detected in human embryonic kidney cells transfected with LEKTI-FL, LEKTI-L, and LEKTI-Sh cDNA, respectively, and in keratinocytes. These fragments correspond to an intermediate form not seen in human epidermis, possibly attributable to rapid processing.
subsequently used in silico analysis to identify potential furin cleavage sites. Numerous potential sites were identified, but four arginine residues with the highest scores were selected for functional studies (Arg355, Arg425, Arg489, and Arg625). Each of these sites was subsequently mutated to inactivate them, and the effect on Pro-LEKTI proteolytic processing was analyzed in human embryonic kidney cells transfected with the mutated cDNA for LEKTI-FL. The 42-kDa LEKTI fragment was used as a prototype to validate this strategy. N-terminal sequencing of the purified 42-kDa band showed that it is generated from cleavage at Arg625 in linker region 9. Site-directed mutagenesis of Arg625 was subsequently used to confirm that the 65- and 68-kDa fragments were generated from cleavage at the same site. These results identified unambiguously the 65- and 68-kDa fragments as being D10–D15 derived from LEKTI-FL and LEKTI-L, respectively, and the 42-kDa band as corresponding to D10–D13 from LEKTI-Sh.
The identity of the other 37-, 30-, and 23-kDa fragments was determined by mutating the other three potential furin cleavage sites with the highest scores (Arg355, Arg425, and Arg489 in linker regions 5, 6, and 7, respectively) in the LEKTI-FL sequence. This strategy allowed the investigators to confirm the use of these cleavage sites for proteolytic processing and assigned the 37-, 30-, and 23-kDa bands to D6–D9, D7–D9, and D8–D9, respectively. As stated by
This study identified the furin-sensitive sites used during pro-LEKTI proteolytic processing, allowing the identification of the LEKTI bioactive fragments generated.
propose the following succession of cleavages during the activation cascade in NHK (Figure 1). The full-length protein (145 kDa) is cleaved into a D6–D15 fragment (102 kDa), which is rapidly processed to generate a D6–D9 fragment (37 kDa), a D10–D15 fragment (65 kDa), and smaller amino-terminal fragments not reported in this study. The D10–D15 fragment shows no evidence of subsequent cleavage. By contrast, the D6–D9 fragment undergoes further proteolysis to release single domains (D6 and then D7), as well as the corresponding three- (D7–D9) and two-domain (D6–D9) fragments. The authors conclude that in normal keratinocytes and human epidermis, pro-LEKTI proteolytic processing results in the generation of a wide variety of LEKTI polypeptides, which comprise at least D6–D9, D7–D9, D8–D9, D6, D7, D10–D15, and D10–D13 fragments, with the latter being generated from LEKTI-Sh, which lacks D14 and D15 (Figure 1).
Figure 1Model of LEKTI proteolytic activation cascade in human keratinocytes and inhibitory properties of physiological fragments. The three LEKTI isoforms—LEKTI-FL, LEKTI-L, and LEKTI-Sh—are represented (see text for their description). The four arginine residues used for cleavage by furin are indicated. The three isoforms are cleaved at Arg 355 to generate D6–D15-FL (102 kDa), D6–D15-L (105 kDa), or D10–D13-Sh (79 kDa) fragments, which are rapidly cleaved at Arg625 to produce fragments D10–D15-FL (65 kDa), D10–D15-L (68 kDa), and D10–D13-Sh (42 kDa), as well as fragment D6–D9 (37 kDa), which is common to the three isoforms. D6–D9 is subsequently cleaved into D6 (7 kDa) and D7–D9 (30 kDa) following cleavage at Arg 425. D7–D9 is further cleaved at Arg 489 to generate D7 (7 kDa) and D8–D9 (23 kDa). Proteolytic processing of fragment D1–D5 was not included in this study and is not yet fully understood. D1 and D5 are represented as single domains because they, like D6, were identified from human blood ultrafiltrates (
). Additional proteolytic fragments derived from D2–D4 are not represented. However, there is evidence that fragments of 20, 15, 13, 11, and 10 kDa are generated from this fragment, the identity of which remains to be established (
). Domains are color coded according to their inhibitory properties found in Fortugno et al. (this issue, 2011) for C-terminal fragments (D6–D9, D10–D13, D10–D15, and derived fragments). For N-terminal fragments (D1–D5), D1–D6 was previously shown to inhibit strongly KLK5 and KLK14 (
), and D1 does not inhibit any of these KLKs. Note that, except for D6–D9, the physiological proteolytic fragments do not match LEKTI fragments previously investigated for inhibition studies. Pink, preferential and strong inhibition of KLK5 and KLK14; blue, preferential but weaker inhibition of KLK7; yellow, unknown or no inhibition of KLK5 or KLK7. LEKTI domains are represented as cylinders; linker regions are shown as curved black lines between domains. The peptide signal sequence is indicated in green at the N-terminal end of the molecules. K, classic Kazal-type domain (D2 and D15).
Fortugno and colleagues mainly investigated the carboxy-terminal region of the protein, from D6 to D15. Previous work on physiological proteolytic processing of the amino-terminal region identified LEKTI fragments of 20, 15, 13, 11, and 10 kDa, recognized by D1–D6 antibodies from human epidermal extract and conditioned medium of cultured NHK (
). The 11- and 10-kDa bands were proposed to correspond to D1, D5, and/or D6 previously isolated from the serum, but the identity of the other bands remains to be determined.
KLK5 and KLK14 as major target SPs of LEKTI physiological fragments
Initial functional studies aimed at assessing LEKTI’s inhibitory capacity were performed with a LEKTI full-length molecule (
). Protease-inhibition assays showed inhibition of the SPs trypsin, plasmin, subtilisin A, cathepsin G, and human neutrophil elastase. The cysteine proteinases papain or cathepsin K, L, or S were not inhibited. However, subsequent studies of rLEKTI fragments reported a different spectrum of inhibition. Specifically, rLEKTI D6-D9 was shown to inhibit trypsin, subtilisin A, KLK5, and KLK7 (
Recombinant production, purification and biochemical characterization of domain 6 of LEKTI: a temporary Kazal-type-related serine proteinase inhibitor.
J Chromatogr B Analyt Technol Biomed Life Sci.2004; 803: 75-81
examined the inhibition profiles of proposed physiological LEKTI fragments against a wide range of human SPs. All fragments tested (D5, D6, D8–D11, and D9–D15) inhibited KLK5, KL7, and KLK14 only. No inhibition was observed with the human SPs KLK8, cathepsin G, tryptase, elastase, plasmin, KLK3, and thrombin. D8–D11 showed the strongest inhibition properties toward KLK5 and KLK14, whereas KLK7 was less efficiently inhibited. D5, D6, and D9–D15 were weaker inhibitors of KLK5, KLK7, and KLK14. From these studies, it appeared that the main target proteases of these LEKTI fragments were KLK5 and KLK14 and, to a lesser extent, KLK7.
tested the inhibition potential of four overlapping rLEKTI multidomain fragments covering the entire molecule against multiple KLKs involved in epidermal desquamation. They found that rLEKTI D1–D6, rLEKTI D6–D9, and rLEKTI D9–D12 were potent inhibitors of KLK5 and KLK14 and that they also inhibited KLK6 and KLK13 to a lesser extent. By contrast, rLEKTI D12–D15 inhibited KLK5 only. This work expanded the spectrum of KLKs inhibited by LEKTI fragments to KLK6 and KLK13, and KLK6 was shown to degrade DSG1.Inhibitors of kallikrein-mediated proteolysis suggest fine-tuned inhibition and a potential for new therapies.
assessed the inhibitory properties of the LEKTI physiological fragments that they identified against KLK5, KLK7, and KLK14. They found D6–D9 and derived multidomain fragments D7–D9 and D8–D9 to be highly efficient inhibitors of KLK5 and KLK14 but weak inhibitors of KLK7. Conversely, they found D10–D15 to be a weak inhibitor of KLK5 and KLK14, but that it inhibited KLK7 more efficiently. Single domains (D6 and D7) showed weaker inhibition than corresponding multidomain fragments.
These results confirm the differential inhibitory properties of LEKTI domains. They also support the notion that KLK5 and KLK14 are major targets of physiological LEKTI fragments, with D6–D9 and its derived multidomain fragments being much more potent KLK5 and KLK14 inhibitors than D10–D15. These results are consistent with previous work that identified D8–D11 as the most effective LEKTI fragment, suggesting that D8–D9 could account for this activity (
also show that KLK7 is inhibited to a lesser extent by D10–D15. However, the fact that LEKTI is degraded by KLK7 suggests that LEKTI is not a major KLK7 inhibitor. In addition, none of the LEKTI domains contains a large hydrophobic Phe, Try, or Tyr residue at position P1, as are found in chymotrypsin-like inhibitors (
). Nevertheless, LEKTI could regulate KLK7 activity through KLK5-mediated KLK7 activation.
From these data, it appears that the inhibitory capabilities of LEKTI fragments and LEKTI domains do not necessarily correlate with the number of domains contained in each fragment. Although unique domains (D6, D7) are usually weaker inhibitors than multidomain fragments, two-domain fragments (D8–D9) can be as potent as four-domain fragments (D6–D9). This suggests that the conformation of multidomain fragments could modify the accessibility of inhibitory reactive loops and/or the binding to putative allosteric sites, thus influencing the inhibitory capacity of the fragment (
What do these results tell us about LEKTI biological function?
These studies identify the nature of physiological LEKTI fragments derived from D6 to D15, crucial information for assessing their biological functions. They confirm that KLK5 and KLK14 are major target proteases of D6–D9 and derived domains, whereas fragments D10–D15 are more potent KLK7 inhibitors. KLK5, KLK14, and KLK7 play a central role in skin desquamation. In fact,
showed that LEKTI fragments derived from D6–D9 had a differential capability to inhibit KLK14- versus KLK5-mediated DSG1 degradation in vitro. Further studies may explore the inhibitory capacities of these physiological fragments toward other epidermal KLKs and SPs.
These results are consistent with the biological cascades described in NS patients and in mouse models of NS (
). In these models, unopposed KLK5 activity initiates an autonomous cascade, leading to stratum corneum detachment and skin barrier defect, inflammation, allergy, filaggrin degradation, and lipid abnormalities. Notably, proteases not inhibited by LEKTI directly, such as elastase 2, also play important roles in NS, as shown by a transgenic mouse model that overexpresses elastase 2 (
Elastase 2 is expressed in human and mouse epidermis and impairs skin barrier function in Netherton syndrome through filaggrin and lipid misprocessing.
The different inhibition properties displayed by LEKTI fragments probably represent an advantage in addressing the diversity of exogenous and/or endogenous target proteinases. Likewise, the large number of LEKTI domains would contribute to quickly release a significant number of bioactive fragments at the site of injury. LEKTI would thus act as a “cluster bomb”—able to rapidly mobilize a significant number of bioactive fragments from a single molecule.
Impact on the development of new treatments for NS and allergy
Understanding the physiological LEKTI fragments and their inhibitory properties has the potential to influence the design of new treatments for NS. In topical protein-replacement therapy, penetration of selected LEKTI fragments would be hampered by their large molecular weight and would require skin delivery systems. Nevertheless, the profound skin barrier defect seen in NS could facilitate diffusion through a disrupted stratum corneum. Alternatively, and perhaps more importantly, these results confirm that KLK5, KLK14, and KLK7 are potential therapeutic targets whose specific inhibition by small lead compounds might block unrestrained protease activity, leading to pharmacological inhibition of the pathogenic cascades.
, spontaneous mutations abolishing or creating furin cleavage sites in linker regions may impact LEKTI processing and proteolytic activation. These mutations could underlie more subtle SP deregulation in the epidermis, which may have a clinical impact and contribute to a genetic predisposition to eczema (
The authors’ work is supported in part by the Agence Nationale de la Recherche (ANR-08-GENO-033) and the Fondation pour la Recherche Médicale (FRM-DAL20051205066).
REFERENCES
Ahmed A.
Kandola P.
Ziada G.
et al.
Purification and partial amino acid sequence of proteins from human epidermal keratinocyte conditioned medium.
Elastase 2 is expressed in human and mouse epidermis and impairs skin barrier function in Netherton syndrome through filaggrin and lipid misprocessing.
Infliximab infusions for Netherton syndrome: sustained clinical improvement correlates with a reduction of thymic stromal lymphopoietin levels in the skin.
LEKTI is localized in lamellar granules, separated from KLK5 and KLK7, and is secreted in the extracellular spaces of the superficial stratum granulosum.
Recombinant production, purification and biochemical characterization of domain 6 of LEKTI: a temporary Kazal-type-related serine proteinase inhibitor.
J Chromatogr B Analyt Technol Biomed Life Sci.2004; 803: 75-81
35069-7&id=gr1.jpg){kind=link}