Increased basal transepidermal water loss leads to elevation of some but not all stratum corneum serine proteases


Rainer Voegeli, DSM Nutritional Products Ltd, Branch Pentapharm, Engelgasse 109, 4002 Basel, Switzerland. Tel.: +41 61 706 4848; fax: +41 61 706 4800; e-mail:


There are indications of elevation of some inflammatory serine proteases in barrier damaged skin (e.g. plasmin and urokinase). Moreover, many other serine protease activities are present such as desquamatory enzymes as well as a newly detected tryptase-like serine protease. However, the activities of these proteases have never been correlated with stratum corneum (SC) barrier function. The activity of extractable key serine proteases (SC trypsin-like kallikreins, SC chymotrypsin-like kallikreins, SC tryptase-like serine protease, urokinase and plasmin) was measured from the outermost layers of SC obtained from facial tape strippings in clinically normal subjects. The protein content of the tape strippings was quantified by absorption measurements with the novel infrared densitometer SquameScanTM 850A and the protease activities by the use of fluorogenic peptide substrates. SC barrier function, SC hydration and skin surface pH were measured using AquaFluxTM, NOVA dermal phase meter and Skin-pH-Meter®, respectively. As expected, SC hydration was reduced with increased transepidermal water loss (TEWL) values indicative of barrier impairment. Surprisingly, SC chymotrypsin-like activity showed no correlation with hydration or TEWL, whereas all other enzymes positively correlated with impaired barrier function and some were statistically significant: SC trypsin-like kallikreins (R= 0.66, < 0.01), SC tryptase-like enzyme (R= 0.95, < 0.001), plasmin (R= 0.86, < 0.001) and urokinase (R= 0.50, < 0.05). All enzymes except urokinase also negatively correlated with SC hydration. Elevated levels of SC serine proteases have been associated with some dermatological disorders, such as atopic dermatitis, psoriasis and rosacea but these results indicate that these enzymes are also elevated with milder forms of barrier disruption, which is not clinically evident as irritated skin. As these proteases are elevated in the SC, they will also be elevated in the epidermis where they can be involved in neurogenic inflammation and epidermal barrier impairment via activation of the protease-activated receptors. These results highlight the need for using serine protease inhibitors especially for urokinase and plasmin, SC tryptase-like serine protease and possibly SC trypsin-like kallikreins even in milder forms of barrier damage.


Il existe des indices sur l’augmentation de certaines protéases à sérine inflammatoires dans la barrière cutanée endommagée (plasmine et urokinase par exemple). De plus, de nombreuses autres activités de protéases à sérine sont présentes comme les enzymes desquamantes ou la protéase à sérine tryptase-like nouvellement détectée. Cependant, les activités de ces protéases n’ont jamais été corrélées à la fonction barrière du stratum corneum (SC). L’activité de protéases à sérine clées (kallikréine trypsine-like du SC, kallikréine chymotrypsine-like du SC, protéase à sérine tryptase-like du SC, urokinase et plasmine) a été mesurée dans les premières couches du SC obtenues par stripping du visage de sujets cliniquement normaux. La teneur en protéines des strippings a été quantifiée par des mesures d’absorption avec le nouveau densitomètre IR SquameScanTM 850A et les activités protéases ont été mesurées en utilisant des substrats peptidiques fluorogéniques. La fonction barrière du SC, l’hydratation du SC et le pH de la surface de la peau ont été mesurées en utilisant respectivement l’AquaFluxTM, le « Dermal phase Meter » Nova et le Skin-pH-meter®.Comme attendu, l’hydratation du SC est réduite avec l’augmentation des valeurs de perte en eau transépidermique (TEWL) indicatrice d’une détérioration de la barrière. De façon surprenante, l’activité chymotrypsine-like du SC ne montre aucune corrélation avec l’hydratation ou TEWL, alors que les autres enzymes corrèlent positivement avec la fonction barrière détériorée et que certaines d’entre elles sont statistiquement significatives : kallikréine trypsine-like du SC (R= 0.66, < 0.01), enzymes tryptase-like du SC (R= 0.95, < 0.001), plasmine (R= 0.86, < 0.001) et urokinase (R= 0.5, < 0.05). Toutes les enzymes, à l’exception du l’urokinase, corrèlent négativement avec l’hydratation du SC. Des valeurs élevées en protéases à sérine du SC ont été associées à des désordres dermatologiques, comme la dermatite atopique, le psoriasis et la rosacée, mais ces résultats indiquent que ces enzymes sont aussi élevées que dans des formes douces de détérioration de la barrière cliniquement non évidentes, comme une peau irritée. Comme ces teneurs en protéases sont élevées dans le SC, elles le seront aussi dans l’épiderme où elles peuvent être impliquées dans une inflammation neurogénique et dans une dégradation de la barrière épidermique via une activation des récepteurs activés par protéases. Ces résultats mettent en évidence un besoin d’utiliser des inhibiteurs de protéases à sérine, spécialement de l’urokinase et de la plasmine, de la protéase à sérine trypsine-like du SC et certainement des kallikréines trypsine-like du SC, même dans les cas les plus légers d’endommagement de la barrière.


Several serine proteases have been shown to be present in the epidermis and especially in the stratum corneum (SC). These enzymes have a wide spectrum of specificities and functions and play important roles in numerous physiological and pathological reactions in cells and tissues [1]. In skin, they are involved in epidermal proliferation, differentiation, lipid barrier homeostasis and tissue remodelling. Most importantly, kallikreins (KLKs) together with other enzymes are involved in the proteolysis of corneodesmosomes, a crucial event prior to desquamation [2]. Impaired corneodesmolysis is known to occur in winter and soap-induced xerosis [3, 4]. In this respect, reduced activity of KLK7 (SC chymotryptic enzyme) [5] and KLK5 (SC tryptic enzyme or SCTE) [6] has been observed in the outer layers of the SC in dry skin. However, an increase in their activities in total SC has been reported after inflammatory challenges to skin for instance following ultraviolet radiation or treatment with surfactants [7, 8]. Moreover, an increased epidermal expression of KLK7 has been reported in psoriasis and atopic dermatitis, two major chronic inflammatory diseases [9].

Komatsu et al. [9, 10] identified several KLKs in SC extracts immunologically (KLK5, KLK6, KLK7, KLK8, KLK10, KLK11, KLK13 and KLK14). KLK5, KLK8, KLK10, KLK11 and KLK13 exert SC trypsin-like activity. Brattsand et al. [11] found that at most 50% of the total activity in SC extracts was because of KLK5, whereas a major part of the remaining activity was associated with KLK14. KLK14 is a dual specificity enzyme with both trypsin- and chymotrypsin-like activity [12]. However, its chymotrypsin-like activity is low. KLK7 is the major chymotrypsin-like KLK in the SC [10] and almost all chymotrypsin-like activity in the SC can be ascribed to KLK7 [11].

Elevation of skin surface pH has been shown to induce increased activity of serine proteases. This can lead to barrier perturbation because of the degradation of lipid processing enzymes, uncontrolled sustained corneodesmolysis and reduced SC integrity and cohesion [13]. Elevated skin surface pH is reported to cause this activity in aged skin [14]. Although reduced cohesion in aged skin has been demonstrated using cohesography [15], desquamation is actually reduced and not enhanced, which is consistent with the reduced levels of KLK5 reported in ageing skin [16].

Many enzymes other than KLKs are present in the SC that may or may not be involved in the desquamatory process. Some of these enzymes, especially serine proteases, are key markers for the underlying and sometimes non-observable skin inflammation. In this respect, elevated activity of the plasminogen/plasmin system is known to occur in tape-stripped and sodium lauryl sulfate-treated skin [17]. The urokinase-type plasminogen activator has been reported to be activated following barrier damage [18]. Urokinase activity was observed in tape strippings from the cheeks of subjects with dry skin, which was elevated with increased transepidermal water loss (TEWL) levels [19]. However, this was a relatively insensitive assay approach. More recently, increased levels of several serine proteases have been identified on facial SC compared with volar forearm SC [20]. SC chymotrypsin-like activity was shown to be activated from the inner layers to the outer layers of the SC on the face but SC trypsin-like activity was abnormally activated in the inner layers of facial SC compared with the corresponding forearm samples. Equally, high levels of urokinase and plasmin activities were found as expected but a potential SC tryptase-like serine protease was also detected, which was equally elevated in facial samples. In the same subjects examined in the winter and summer months of the year, reduced SC trypsin-like activity and plasmin were observed in winter. The precise correlation of all these enzymes with TEWL, SC hydration and skin surface pH has not been conducted before. Therefore, the objective of this study was to compare the activities of key desquamatory enzymes, SC trypsin-like KLKs and SC chymotrypsin-like KLKs, together with inflammatory serine proteases, plasmin, urokinase and a novel SC tryptase-like serine protease with TEWL, SC hydration and skin surface pH on samples taken from the face in the winter months of the year. This analysis was only conducted with serine proteases extracted from SC samples. Therefore the data does not necessarily correspond to in situ activities. Nevertheless, because of the known activation of SC chymotrypsin-like activity towards the surface layers of the SC, nine tape-strippings were taken and evaluated.

Materials and methods

Healthy Caucasian subjects (skin type II and III) participated in the studies (= 10, mean age 37.5 ± 8.7 years). All subjects came from Pentapharm (Basel, Switzerland) and the studies were approved by the Pentapharm Safety and Regulatory Group. All studies complied with the Guidelines for Medical Experiments in non-patient human volunteers published by the Association of the British Pharmaceutical Industry (ABPI) and the World Medical Association’s Declaration of Helsinki (2000) concerning biomedical research involving human subjects. Written informed consent was given.

Before conducting the sequential nine-fold tape-stripping (D-Squame®; CuDerm Corporation, Dallas, TX, U.S.A.) on the cheek, SC hydration was measured using a NOVA DPM 9003 (NOVA Technology, Portsmouth, NH, U.S.A.), pH with a Skin-pH-Meter® PH 905 (Courage & Khazaka Electronic, Cologne, Germany) and TEWL using an AquaFlux AF103 (Biox Systems, London, U.K.). SC hydration and skin surface pH measurements were expressed as the mean value of three recordings, TEWL was measured once. Subjects were advised not to apply any cosmetics for at least 12 h before the SC was sampled. As this study was a range finding study examining the relationship between baseline TEWL and extractable SC protease activity, this washout period was felt to be sufficient. Skin physiology in normal life conditions in Western countries usually means regular use of cosmetic products. A wash-out phase of a few days instead of 12 h would not reflect the natural situation for “modern” skin anymore and would have created an artificial baseline. We therefore chose to request that panelists only stop using any products they were using for 12 h. Moreover, any elevations of TEWL and thereby enzyme levels are more likely to be environmentally induced (UV irradiation etc). First, 15 min before the tape-stripping procedure, the skin was cleaned by gentle swabbing with a cotton pad soaked with distilled water of ambient temperature and allowed to dry. Subjects were acclimated for 20 min in an environmental room under standard conditions (21 ± 1°C and 50 ± 10% relative humidity) prior to any measurements. The skin sites were marked with a surgical marker to ensure that the measurement probes and the tapes were consistently applied to the same area. Standard D-Squame® discs with a diameter of 2.2 cm and an area of 3.8 cm2 were placed on the skin under 225 g cm−2 of pressure with a pressure device (CuDerm Corporation, Dallas, TX, U.S.A.) for 5 s. To minimize variations, the procedure was conducted by the same technician. The interval between the strippings was 20 ± 5 s.

The protein content of the tape strippings was quantified by absorption measurements at 850 nm with the novel infrared densitometer SquameScanTM 850A (Heiland Electronic, Wetzlar, Germany) [21]. SquameScanTM 850A is especially designed for the application of standard D-Squame® discs. For protein quantification, the following equation was used [20]:


Immediately after absorption measurement, each tape stripping was transferred into a 1.5-mL Eppendorf tube and extracted for 15 min at 25°C and shaken at 1000 rpm in 750 μL of a buffer composed of 0.1 M Tris–HCl and 0.5% Triton X-100 at pH 8.0. The extracts of tape strippings (1–9) were pooled and to each 250 μL of the combined extracts, 1.25 μL of 5 mM fluorogenic peptide substrates (Table I) dissolved in dimethylsulphoxide was added (final substrate concentration, 25 μM). The solutions were mixed at 37°C and 1000 rpm. The reaction was stopped after 2 h by adding 100 μL of 1% acetic acid to 100 μL of reaction mixture. The released aminomethyl coumarin (AMC) was quantified by reversed-phase high performance liquid chromatography with gradient elution (80% water/20% acetonitrile/0.07% trifluoroacetic acid to 50% water/50% acetonitrile/0.07% trifluoroacetic acid). The column used was Symmetry C18, 3.5 μm, 4.6 mm × 75 mm (Waters, Milford, MA, U.S.A.). The flow rate was 1 mL min−1, the injection volume 5 μL and the retention time of AMC 3.5 min. The wavelength for emission was 442 nm and for excitation 354 nm.

Table I.   Fluorogenic peptide substrates used for serine protease assays
Serine proteaseSubstrateSource
Trypsin-like kallikreinsBoc-Phe-Ser-Arg-AMCI-1400 (Bachem)
Chymotrypsin-like kallikreinsMeOSuc-Arg-Pro-Tyr-AMCPefafluor 180-23 (Pentapharm)
PlasminMeOSuc-Ala-Phe-Lys-AMCI-1275 (Bachem)
UrokinaseBz-β-Ala-Gly-Arg-AMCPefafluor uPA (Pentapharm)
Tryptase-like serine proteaseTos-Gly-Pro-Lys-AMCI-1370 (Bachem)

Because of the potential reactivity of the KLK5 substrate with other KLKs, we refer to this activity as trypsin-like KLK activity. All data were collected in Microsoft® Excel 2002 before transferring to Graphpad prism (5.00 for Windows; GraphPad Software, San Diego, CA, U.S.A.). Data were checked for normality using the D’Agostino & Pearson omnibus normality test and then a two-tailed Pearson correlation and linear regression was performed to calculate the coefficient of determination (R2). Statistical significance was set at 95%.


Characterization and bioengineering data of subjects are shown in Table II. As expected, reduced facial SC hydration correlated with increased TEWL (R= 0.64, < 0.01; Fig. 1). However, skin surface pH did not correlate with either of these two measurements (data not shown). All SC-derived proteases, except the SC chymotrypsin-like KLKs, positively correlated with increasing levels of TEWL and negatively correlated with SC hydration but not with skin surface pH.

Table II.   Characterization and bio-engineering data of subjects
Subject #GenderAge (years)TEWL (g m−2 h−1)Hydration (DPM units)pH
  1. F, female; M, male; TEWL, transepidermal water loss.

Figure 1.

 Correlation of skin stratum corneum hydration (NOVA DPM 9003) and transepidermal water loss (TEWL) (AquaFlux AF103).

Correlating with TEWL (Fig. 2), plasmin activity had a coefficient of determination of R= 0.86 (< 0.001), urokinase activity of R= 0.5 (P < 0.05), SC tryptase-like activity of R= 0.95 (P < 0.001) and SC trypsin-like KLKs (KLK5) of R= 0.66 (P < 0.01). SC chymotrypsin-like KLKs (KLK7) presented no correlation with TEWL.

Figure 2.

 (a) Correlation of kallikrein (KLK) activities and transepidermal water loss (TEWL): SC trypsin-like KLKs (•), SC chymotrypsin-like KLKs (bsl00001) (b) correlation of inflammatory serine protease activities and TEWL: plasmin (•) and urokinase (bsl00001) and (c) correlation of SC tryptase-like activity and TEWL. SC, stratum corneum.

Conversely correlating with SC hydration (Fig. 3), plasmin activity had a coefficient of determination of R= 0.47 (P < 0.05) and SC tryptase-like activity of R= 0.67 (P < 0.005). Urokinase activity did not correlate with SC hydration (R2 = 0.14, n.s.). SC trypsin-like KLKs (KLK5) but not SC chymotrypsin-like KLKs (KLK7) correlated with SC hydration (R= 0.47, P < 0.05 and R= 0.09, n.s. respectively).

Figure 3.

 (a) Correlation of kallikrein activities and SC hydration: SC trypsin-like kallikreins (•), SC chymotrypsin-like kallikreins (bsl00001) (b) correlation of inflammatory serine protease activities and SC hydration: plasmin (•) and urokinase (bsl00001) and (c) correlation of SC tryptase-like activity and SC hydration. SC, stratum corneum.


Reduced activities of KLK5 and KLK7 have been reported in extracted SC samples from the subjects with dry skin [5, 6]. This is presumably because of the effects of environmental dehydration and the effects of cleansing on enzyme denaturation or their leaching from the SC. However, increased activities of KLK5 and KLK7 have also been reported in SLS-induced barrier damage and in sunburn peelings [7, 8]. When considering the non-desquamatory serine proteases, Kitamura et al. [22] demonstrated that plasminogen that was only located at the basal layer in normal subjects was expressed in all epidermal cell layers in dry skin. However, Kawai et al. [19] reported that urokinase was also present in the SC especially in experimentally induced dry skin on back skin of individuals. They further demonstrated that increased extractable urokinase activity was present in SC samples from the cheek in subjects with visibly dry skin and subjects with elevated TEWL levels. If subjects had normal appearing skin and a TEWL of less than approximately 16.4 g m−2 h−1, then no activity was found. Moreover, abnormally high proteolytic activity had been observed in several barrier compromised skin conditions such as atopic dermatitis and psoriasis. Komatsu et al. [23] found elevated SC plasmin activity in subjects with atopic dermatitis. Although no increase in extractable chymotrypsin-like and trypsin-like activities was found, increased mass levels of KLK5, KLK6, KLK7, KLK8, KLK10, KLK11 and KLK14 were observed. Similarly, the same group reported increased plasmin activity in SC samples from subjects with psoriasis [24] even in non-lesional skin compared with healthy skin. In this case, however, SC trypsin-like activity was increased as was the mass level of the KLKs reported above. Nevertheless, recently Simon et al. [25] reported that the mass levels of KLK7 in SC samples were similar in lesional vs. non-lesional psoriatic skin. These results would suggest that we will find increased levels of serine protease activities in SC samples from subjects with dry skin.

In our study, we examined the levels of extractable SC-derived serine proteases (chymotrypsin-like KLKs, trypsin-like KLKs, plasmin, urokinase and SC tryptase-like serine protease) and attempted to correlate their activities with TEWL, SC hydration and skin surface pH. We found no dependences between skin surface pH and serine protease activities. SC tryptase-like serine protease and plasmin activities highly significantly correlated with TEWL values and although SC trypsin-like KLKs and urokinase had a lower association, they were still statistically significant. SC chymotrypsin-like KLK activity had no relationship. Similar, but less distinct negative correlations were found with SC hydration. However, urokinase and again chymotrypsin-like KLKs did not show a relationship.

Use of serine protease inhibitors and especially trypsin-like protease inhibitors have been shown to accelerate barrier recovery in experimentally damaged skin barrier models and in subjects with dry skin [17, 22]. We have now found evidence of not only elevated extractable plasmin and urokinase activities that might be contributing to barrier compromised states but also extractable SC trypsin-like and SC tryptase-like activities are likely to be involved. Surprisingly, there was no elevation of SC chymotrypsin-like activity with increasing TEWL. This would be in keeping with the studies of Komatsu et al. [23, 24] and Simon et al. [25] in subjects with atopic dermatitis and psoriasis as well as the known lack of effect of chymotrypsin-like inhibitors improving barrier function and dry skin [17, 22].

The excess serine proteases probably have an effect on the epidermal differentiation process leading to barrier impairment (we are using the SC as a mirror of underlying epidermal activity), although they could also be having an effect on SC corneodesmolysis and/or degradation of the lipid processing enzymes especially when the skin surface pH is elevated [13, 26, 27]. Skin surface pH, however, was not elevated in this study and as a result, it is anticipated that the elevated expression of these enzymes is influencing epidermal differentiation in some way. Differences in SC pH in the deeper layers of the SC in situ may be influencing protease activity. We are, however, measuring the maximal activity of the extractable proteases.

Nevertheless, there is growing evidence that the protease-activated receptor 2 (PAR-2) plays a key role in epidermal permeability barrier homeostasis. Tape-stripping of the SC leads to an increase of PAR-2 activation and serine protease inhibitors have been shown to accelerate barrier recovery [17, 28]. Stefansson et al. [29] found that some but not all KLKs can activate PAR-2. KLK5 and KLK14, but neither KLK7 nor KLK8 induced keratinocyte PAR-2 signalling. Similarly, Oikonomopoulou et al. [30, 31] found that KLK5 and KLK6 yielded PAR peptide cleavage patterns consistent with PAR-2 receptor activation. Thus, like trypsin, tryptase [32, 33] and plasmin [34] KLKs such as KLK5 (SC trypsin-like activity) are important activators of PAR-2. As urokinase is the activator of plasminogen, this enzyme also clearly needs consideration in this respect.

We believe that the extractable SC protease activities are elevated because of the increased TEWL. Hellemans et al. [35] reported increased surface SCTE (KLK5, SC trypsin-like activity) 3 weeks after damaging the barrier by tape-stripping i.e. elevating TEWL. Collectively, all of these findings suggest that an improvement in skin condition is possible using protease inhibitors. In this respect, Kitamura [36] using trans-aminocyclohexane carboxylic acid in placebo-controlled studies demonstrated an improvement in various symptoms of dry skin such as a faster improvement in skin texture, barrier function and skin hydration. This author, however, only considered elevated plasmin and urokinase activity. In conclusion, these results provide further evidence for the need of using serine protease inhibitors to improve skin barrier function. Moreover, they suggest the need for not only plasmin/urokinase inhibitors but also inhibitors for SC tryptase-like serine protease.


This study was fully funded by DSM Nutritional Products Ltd Branch Pentapharm, Basel.