MATERIALS AND METHODS

Reagents, solvents, media

Recombinant interleukin (rIL)-2 and [³H]-thymidine (2 Ci per mmol) were obtained as described (^{Hales & Camp 1998}). Standard culture medium (RH10) was RPMI 1640 supplemented with 10% human AB serum and 3 mM glutamine but without antibiotics (^{Hales & Camp 1998}). Earle's balanced salts solution (EBSS) was from Life Technologies (Paisley, U.K.); high performance liquid chromatography (HPLC) grade solvents from Fisher (Loughborough, U.K.); phenylmethylsulfonyl fluoride (PMSF) from ICN (Thame, U.K.); and trifluoroacetic acid (TFA), ethylenediaminetetraacetic acid (EDTA), leupeptin, and pepstatin A from Sigma (Poole, U.K.). Novex MultiMark molecular weight standards for electrophoresis were from R&D Systems (Abingdon, U.K.).

Sample extraction, protease inhibition, and proteinase K digestion

SC samples were recovered as electric razor shavings (ERS) and heel scrapings from two male volunteers with normal skin, and were acetone washed, sterilised by [⁶⁰Co] irradiation, extracted by shaking in 0.146 M NaCl at 4°C for 48 h, and stored at –80°C as described (^{Hales & Camp 1998}). ERS were also subjected to glass-on-glass homogenisation (90 mg ERS per ml 0.146 M NaCl with or without 10 mM EDTA, 40 M leupeptin, 1 mM PMSF, and 10 M pepstatin A) for 5 min at 4°C. Extracts with or without protease inhibitors were microcentrifuged, filtered (0.45 M), and 1 ml aliquots subjected immediately to reversed phase (RP)-HPLC and bioassay of 1.5 ml fractions as described below.

Full thickness normal skin was obtained from patients undergoing breast reduction surgery, trimmed of subcutaneous fat, and pieces of approximately 4 cm 8 cm immersed in distilled, deionised water at 60°C for 1 min. Epidermis was gently stripped from underlying dermis with forceps and a scalpel, frozen on CO₂ ice, and powdered by CO₂ ice-cooled mortar and pestle. The powder was washed twice in excess acetone, centrifuged, and the pellet vacuum desiccated. The residue was sterilized by [⁶⁰Co] (at least 10 kGy), resuspended in PBS (1 g per 20 ml), agitated for 48 h at 4°C, centrifuged, the supernatant filtered (0.45 m), and stored at –80°C.

Proteinase K digestions were carried out as previously described (^{Beckman et al. 1994}). Briefly, ERS extract was purified by RP-HPLC, and biologically active fractions of intermediate hydrophobicity (see Results) pooled, assayed for total protein (Micro BCA kit, Pierce and Warriner, Chester, U.K.), and evaporated. Aliquots (100 g) were incubated with 10 g proteinase K (specific activity 20 U per ml; Boehringer Mannheim, Lewes, U.K.) in 0.05 M NH₄ HCO₃ and 1 mM CaCl₂ (final volume 220 l) for 2 h at 37°C. Following heat inactivation for 5 min at 65°C, a further 10 g proteinase K was added, the incubation continued for 15 h at 37°C and then frozen at –20°C. Control incubations, which contained no proteinase K, were carried out in parallel. Subsequently thawed samples were serially diluted in RH10 and subjected to T cell proliferation assays with both PBMC alone and an ERS-responsive T cell line with irradiated, autologous PBMC as APC (see below). Thawed samples were also purified by RP-HPLC and appropriate fractions evaporated and bioassayed.

HPLC

Solvent was delivered by a Waters 626 non-metallic system. RP-HPLC columns included a 4.6 100 mm Brownlee Aquapore RP-300 C8 analytical column, a 10 100 mm Brownlee Aquapore Prep-10 C8 preparative column (Applied Biosystems), a 4.6 250 mm Atlantis 5 C18 300Å analytical column (Phenomenex), and a 2 250 mm Jupiter 5 C18 300Å microbore column (Phenomenex), as described under Results. Chromatofocusing was carried out on a 7.5 mm 75 mm Bio-Gel DEAE-5-PW anion exchange column eluted at 1 ml per min with a mixture of three ammonium acetate-containing buffers: buffer A (20 mM ammonium acetate adjusted to pH 10 with 0.88 specific gravity ammonia), buffer B (20 mM ammonium acetate adjusted to pH 8 with ammonia), and buffer C (20 mM ammonium acetate adjusted to pH 4 with trifluoroacetic acid). A linear gradient of 100% buffer A to 100% buffer B was delivered over 10 min, followed by a linear gradient to 90% buffer B/10% buffer C at 25 min, to 85% buffer B/15% buffer C at 35 min, to 80% buffer B/20% buffer C at 40 min, and finally to 100% buffer C at 45 min. This system enables delivery of approximately linear effluent pH gradient from 10 to 4 over 60 min. Furthermore, the volatility of the solvent components allows lyophilization of effluent fractions and bioassay of residues without the need for dialysis. HPLC fractions were collected and freeze-dried in a centrifugal evaporator. Residues were tested in proliferation assays with PBMC or T cell lines.

Gel electrophoresis, nitrocellulose immunoblotting, and amino acid sequencing

In preliminary experiments 8 mm 8 mm pieces of nitrocellulose membrane (0.45 m, pore-sized Hybond-ECL, Amersham, U.K.) were soaked in an excess of aqueous ERS extract for 60 min at room temperature. Both undenatured extract and that denatured by heating at 99°C in 5% vol/vol 2-mercaptoethanol, 10% wt/vol SDS for 5 min before being allowed to cool for 85 min at room temperature, were used. The nitrocellulose pieces were then washed successively in an excess of 20% methanol in Earle's balanced salts solution (EBSS), then 20% ethanol in EBSS, then finally two washes in EBSS alone. Each wash was continued for 10 min at room temperature with agitation. The nitrocellulose pieces were cut into four 4 mm 4 mm portions, and each portion placed on the base of a flat-bottomed microtitre plate well. An ERS-reactive T cell line (2 10⁵ per ml) and irradiated PBMC (2 10⁶ per ml) were added to each well (total volume 0.2 ml RH10 medium) over the nitrocellulose pieces and incubated for 3 d as in standard proliferation assays. Cultures were pulsed with [³H]thymidine, cells harvested and thymidine incorporation determined as described below. Thereafter RP-HPLC-purified fractions were denatured and subjected to Tris-Tricine-SDS-PAGE (^{Schagger & von Jagow 1987}) on 16 cm 1 mm vertical slab gels containing 16% acrylamide, 0.5% bisacrylamide, and 1.3% glycerol. Separated polypeptides were electroblotted onto Hybond-ECL nitrocellulose membranes that were washed as described above, dried, and lanes cut into 10 mm 2 mm strips. The strips were divided and the two halves placed in a microtitre well with the gel-facing surface upwards, irradiated PBMC (10⁶ per ml) and an autologous ERS-responsive T cell line (2 10⁵ per ml) added, and proliferation determined in 3-d assays as described below. For amino acid sequencing, parallel lanes were simultaneously electroblotted from Tris-Tricine gels onto adjacent nitrocellulose and Fluorotrans 0.2 m polyvinylidine difluoride (PVDF) membranes (Flowgen, Lichfield, U.K.). Typically, electroblotting was performed in buffer containing 48 mM Tris, 39 mM glycine, 10% methanol, and 0.03% SDS, for 2 h at 250 mAmp in a cold room. PVDF membranes were stained with Coomassie Blue, destained, washed several times in water, dried, and when possible protein bands co-migrating with biologic activity on nitrocellulose strips were identified. These were excised and subjected to N-terminal amino acid sequencing by Edman degradation in an ABI 476 liquid phase protein sequencer (Applied Biosystems). PVDF membranes, which were required for sequencing, could not readily be used in T cell proliferation assays in place of nitrocellulose as their hydrophobic nature caused the small strips to float on the surface of culture media.

T cell proliferation assays

PBMC (5 10⁵ per ml) and ERS-responsive T cell lines (2 10⁵ per ml) were prepared from venous blood of healthy volunteer JMH, and used in 3- and 5 d proliferation assays respectively (total volume 0.2 ml) as described previously (^{Hales & Camp 1998}). Autologous, irradiated (25 Gy) PBMC (0.5–2 10⁶ per ml) were used as antigen presenting cells (APC) in assays with T cell lines. Cultures were pulsed with [³H]thymidine (1 Ci) for the last 6 h, cells harvested, counted by liquid scintillation, and results expressed as counts per min (cpm) or stimulation indices (SI; cpm in presence of test material ÷ cpm from control incubations).

RESULTS

Biologic activity in ERS extracts is stable to denaturation and induces T cell proliferation following adsorption to nitrocellulose

Native ERS extract and that denatured by boiling with 2-mercaptoethanol and SDS were adsorbed onto nitrocellulose membranes which were thoroughly washed as described under Materials and Methods. Pieces of membrane were placed in microtitre wells with an ERS-responsive T cell line and irradiated autologous PBMC, in standard 3-d proliferation assays. Substantial activation of the T cell line by adsorbed, native ERS extract was demonstrated (12,132 cpm, SI 111, mean of quadruplicate assays), and this increased following denaturation (26,760 cpm, SI 205, mean of quadruplicate assays).

Purification of SC and epidermal extracts by RP-HPLC demonstrates biologic activity of intermediate hydrophobicity, which is unaffected by protease inhibitors

RP-HPLC of ERS extract was done on a 4.6 100 mm Brownlee Aquapore analytical column eluted with 0.1% aqueous TFA for 5 min at 1 ml per min, followed by a linear gradient to 0.1% TFA in acetonitrile at 35 min. Assay of fractions for PBMC proliferation revealed activity in adjacent fractions eluting at approximately 33–55% acetonitrile (16–22 min) and coinciding with several UV absorbing peaks. The same profile of biologic activity was seen if redissolved residues from HPLC fractions were placed in proliferation assays for a full 5 d, or if PBMC were pulsed with redissolved residue for 24 h, washed, then re-incubated for a further 4 d (Figure 1a–c). The latter procedure was adopted in view of the evidence that SC extracts contain T cell inhibitory material, the effects of which may be overcome by pulsing cultures with extract for 24 h (^{Hales & Camp 1998}). Typically, total protein in a 1 ml ERS sample was 600 g, whereas total protein in a corresponding 16–22 min RP-HPLC fraction was 30 g (Micro BCA kit), indicating approximately 20-fold purification.

Figure 1.

Biologic activity in ERS and heel SC extracts has discrete mobility in RP-HPLC. (A, B) UV absorbance profile and biologic activity in fractions following RP-HPLC of a 300 l ERS sample. Evaporated 1 min fractions were redissolved in 300 l RH10 medium and PBMC proliferation determined in 5 d assays at 30% () and 10% () final concentration. (C) Biologic activity in fractions after RP-HPLC of a further 300 l ERS sample. Fraction residues were redissolved in 300 l RH10 and incubated for 24 h with PBMC at 75% final concentration. PBMC were then washed, resuspended in medium, and thymidine incorporation determined after incubation for a further 4 d. (D, E) UV absorbance profile and biologic activity in fractions following RP-HPLC of a 300 l heel SC sample. Evaporated 1 min fractions were redissolved in 300 l RH10 and PBMC proliferation determined in 5 d assays at 30% () and 10% () final concentration. Mean cpm from triplicate or quadruplicate assays are shown (B, C, E).

Full figure and legend (16K)

When heel SC extract was purified by the same RP-HPLC system and fractions tested in 5-d PBMC proliferation assays, biologic activity coinciding with several UV absorbing peaks was again seen in fractions eluting at about 50% acetonitrile (Figure 1d,e). Analysis of an aqueous extract of heat separated, full thickness epidermis from normal skin in the same RP-HPLC system, also showed PBMC stimulating material with a similar retention time (Figure 2).

Figure 2.

Extract of whole epidermis also contains biologic activity of intermediate hydrophobicity on RP-HPLC. UV absorbance profile (A) and biologic activity in fractions (B) following RP-HPLC of a 1.1 ml sample. Evaporated 1.5 min fractions were redissolved in 440 l RH10 medium and PBMC proliferation determined in 5 d assays at 50% final concentration. Mean cpm from quadruplicate assays are shown.

Full figure and legend (9K)

RP-HPLC (Brownlee analytical column) of ERS extracts obtained in the presence and absence of protease inhibitors as described under Materials and Methods, and assay of fractions for PBMC proliferation, showed the same profiles of biologic activity (Figure 3), indicating that the T cell stimulatory material in SC extracts is not generated during extraction by the action of solubilized proteases. HPLC purification was introduced as the protease inhibitors interfered with PBMC proliferation if extracts were assayed directly (data not shown). The broader peaks of biologic activity in the HPLC of Figure 3 compared with Figure 1 are due to loading of larger amounts of protein.

Figure 3.

Protease inhibitors present during ERS extraction do not affect the level of PBMC activation. Supernatants (1 ml) obtained in the absence (A) and presence (B) of protease inhibitors were subjected to RP-HPLC. Evaporated 1.5 min fractions were redissolved in 320 l RH10 medium and PBMC proliferation determined in 5 d assays at 8% final concentration. Mean cpm from duplicate assays are shown.

Full figure and legend (12K)

RP-HPLC and SDS-PAGE with nitrocellulose T cell immunoblotting of ERS extract show 5–18 kDa biologic activity

ERS extract was purified by reversed phase HPLC as above, residues from biologically active (16–22 min) fractions were denatured by boiling with 2-mercaptoethanol and SDS, and subjected to electrophoresis and immunoblotting as described under Materials and Methods. In preliminary experiments with conventional SDS-PAGE minigels, T cell immunoblots consistently showed migration of the major active material to the bottom of the gels (maximum cpm 52,590-84, 100; SI 177-353; n = 2; data not shown), indicating relatively low molecular weight. Tris-Tricine gels (16 cm) were therefore used subsequently. The 16–22 min residue following RP-HPLC of 1 ml ERS extract was halved and subjected to Tris-Tricine SDS-PAGE in parallel lanes. Following parallel electroblotting onto nitrocellulose and PVDF membranes as described under Materials and Methods, proliferation assays incorporating 2 mm nitrocellulose strips, an ERS-responsive T cell line, and irradiated autologous PBMC revealed biologic activity within the range of 5–18 kDa (Figure 4b). The major biologically active peak (Mr value approximately 13 kDa, maximum SI 102) co-migrated with a prominent Coomassie Blue-stained band on the PVDF paper (Figure 4a). This band was excised and sequenced, giving a 19-residue N-terminal sequence (ILE-PRO-GLY-GLYLEU-SER-GLU-ALA-LYS-PRO-ALA-THR-PRO-GLU-ILE-GLN-GLU-ILE-VAL) completely homologous with that of human epidermal cystatin A (^{Takeda et al. 1989}). This is an 11 kDa, pI 5 cysteine protease inhibitor that was previously reported to be identical to cystatin A from liver, spleen, and leukocytes except for the lack of an N-terminal methionine residue (^{Takeda et al. 1989}), as found in the present work. Commercially available human placental cystatin A (Calbiochem, Nottingham, U.K.; N-terminal sequence MET – ILE – PRO – GLY – GLY – LEU – SER – GLU ALA-LYS-PRO-ALA-THR-PRO-GLU-ILE-GLN) was tested (1, 2, 4, 8, 16, 32 and 84 g per ml) in proliferation assays containing an ERS-reactive T cell line and irradiated PBMC. No activity significantly greater than control RH10 medium alone was obtained (cpm <200). It was concluded that epidermal cystatin A was likely to be a contaminant co-migrating with the major biologically active peak in Figure 4b. More rigorous sequential HPLC purifications were therefore undertaken prior to further SDS-PAGE and T cell immunoblotting.

Figure 4.

Tris-Tricine SDS-PAGE of RP-HPLC-purified ERS extract. Following HPLC of 1 ml ERS extract, the residue from a 16–22 min fraction was denatured, halved, and subjected to SDS-PAGE in separate lanes. One lane was electroblotted onto a PVDF membrane and stained with Coomassie Blue (A). The major band, migrating at Mr approximately 13 kDa, was excised and sequenced. The second lane was electroblotted onto a nitrocellulose membrane, 2 mm strips were cut and assayed for stimulatory activity (B).Migration of molecular weight markers (Da) is shown at the top of the figure.

Full figure and legend (12K)

Bulk purification of ERS extract by successive RP-HPLC, chromatofocusing, and preparative SDS-PAGE shows several potent T-cell stimulating components

In preliminary experiments, 5 ml ERS extract was subjected to RP-HPLC on the 4.6 mm 10 mm Brownlee analytical column and biologically active 16–22 min fractions from repeated runs pooled and evaporated. The residue was re-purified by chromatofocusing on the Bio-Gel DEAE anion exchange column as described under Materials and Methods. A biologically active peak eluting early (effluent pH 10) and two later eluting biologically active peaks (effluent pH 6 and 4.5 respectively) were individually repurified by RP-HPLC on a Phenomenex 4.6 250 mm column using shallow acetonitrile gradients. UV absorbing peaks eluting at 35–40% acetonitrile were collected and found to induce potent activation of an ERS-reactive T cell line in the presence of autologous, irradiated PBMC (SI values >200; data not shown). Finally, the three components were subjected to denaturing Tris-Tricine SDS-PAGE and nitrocellulose T cell immunoblotting, as described above. The components with apparent isoelectric points of 10, 6, and 4.5 showed Mr values of 5, 13.5, and 18 kDa respectively, and their biologic potency was indicated by SI values of 94–532 (Figure 5a–c). Following these purifications, insufficient material was left for amino acid sequencing.

Figure 5.

Tris-Tricine SDS-PAGE and nitrocellulose T cell immunoblotting of three ERS components. The components were purified initially by two RP-HPLC steps and chromatofocusing (apparent pI values approximately 10, 6, and 4.5), and show Mr values of approximately 5 (A), 13.5 (B), and 18 (C) kDa respectively. SI values refer to the most active fraction in each profile. Migration of molecular weight markers (kDa) is shown at the top of the figure.

Full figure and legend (11K)

We therefore elected to bulk purify the pI 10, 5 kDa component that gave the SI value of >500 following T cell immunoblotting (Figure 5a) and appeared to be the most potent component in other experiments (see Figure 10 below). Initial bulk RP-HPLC purification of ERS extract was carried out on the 10 mm 100 mm Brownlee Prep-10 column eluted with 0.1% TFA in water at 3 ml per min. Three 3.5 ml aliquots of ERS extract were injected successively prior to initiation of a linear solvent gradient to 0.1% TFA in acetonitrile over 30 min. Similar UV absorbance profiles to that shown in Figure 1a were obtained (data not shown) and a group of fractions containing biologic activity, equivalent to those shown in Figure 1b, were collected, pooled, and evaporated. A total of 65 ml ERS extract was subjected to this preparative RP-HPLC system, the evaporated residue redissolved in 3 ml 20 mM ammonium acetate adjusted to pH 10 with ammonia, filtered (0.45 m) and immediately subjected to a single chromatofocusing run on the Bio-Gel anion exchange column as described under Materials and Methods. Assay of a portion of each evaporated 1.5 ml fraction for activation of an ERS-responsive T cell line in the presence of autologous irradiated PBMC as APC, showed five sets of biologically active fractions (peaks 1–5 in Figure 6). Separate experiments showed that the biologic profiles obtained with autologous and allogeneic ERS extract were similar (data not shown). The residue from the fraction eluting at 20 min, immediately after peak 2 (arrow, Figure 6b) was assayed after serial dilution, which unmasked potent T cell activation at 80-fold dilution (Figure 6b, inset). The steep nature of the dose–response curve generated by this fraction, with near background proliferation at 400-fold dilution but maximal stimulation following a single further dilution to 80-fold, illustrates the care required to overcome inhibitory activity in these samples, as was found previously when assaying unpurified material (^{Hales & Camp 1998}). In contrast, serial dilution and assay of the residue in the 47 min fraction eluting immediately after peak 4 (Figure 6b) did not unmask any T cell stimulation (data not shown).

Figure 10.

Time dependence of the proliferative responses of unprimed PBMC to residues from fractions equivalent to peaks 1–5 in Figure 6b. Appropriate sample dilutions were established in preliminary experiments, and thymidine incorporation determined after 3, 5, and 7 d. The figures at the top of each column represent SI values. *82,108 cpm.

Full figure and legend (15K)

Figure 6.

Chromatofocusing of RP-HPLC-purified ERS extract shows five groups of biologically active fractions. (A) UV absorbance profile and pH of effluent fractions. (B) One-tenth of each 1.5 ml fraction was evaporated, residues redissolved in 320 l RH10 medium and tested for stimulation of an ERS-responsive T cell line in the presence of irradiated PBMC, at 50% final dilution. Each column shows the mean of triplicate assays. Five sets of biologically active fractions, designated 1, 2, 3, 4, and 5, are shown. Inset: Proliferation of the ERS line in response to the residue from the fraction eluting at 20 min (vertical arrow). The residue was redissolved in 350 l RH10 and assayed in triplicate at the dilutions indicated.

Full figure and legend (12K)

Peak 1 (Figure 6b) was repurified on the Phenomenex analytical RP-HPLC column eluted at 1 ml per min with 0.1% TFA in water for 5 min, followed by a linear gradient to 30% acetonitrile at 65 min, then to 60% acetonitrile at 95 min. The UV absorbance profile and biologic activity in collected, evaporated fractions (tested with an ERS-responsive T cell line and autologous, irradiated PBMC as APC) is shown in Figure 7. Substantial T cell stimulating activity was seen in association with minor UV absorbing peaks at 70 min. The remaining residue from the single, major biologically active peak was repurified by microbore RP-HPLC on a Phenomenex 2 250 mm column. Elution was at 0.2 ml per min with 0.1% TFA in water for 5 min, followed by a linear gradient to 30% acetonitrile at 80 min, then to 50% acetonitrile at 120 min. A group of three UV absorbing peaks eluted close together at 85 min, and substantial T cell stimulatory activity was associated with the tail of the UV absorbing peaks (Figure 8). The residue from the single biologically active fraction was finally repurified on the microbore RP-HPLC column eluted at 0.2 ml per min with 0.1% TFA in water for 5 min, followed by an extremely shallow linear gradient to 22% acetonitrile at 60 min, then to 40% acetonitrile at 7 h. Major T cell stimulatory activity eluted at 145–150 min, but was unexpectedly associated with a tail of minor UV absorbing material containing minimal detectable protein (Figure 9) that was insufficient for conventional amino acid sequencing.

Figure 7.

Analytical RP-HPLC of the residue from peak 1, Figure 6b. (A) UV absorbance profile. (B) Biologic activity in fractions. HPLC conditions are described under Results. One-eighth of each fraction was evaporated, redissolved in 320 l RH10, and tested at 50% final concentration for stimulation of an ERS-responsive T cell line in the presence of irradiated PBMC. Mean results of triplicate assays are shown.

Full figure and legend (5K)

Figure 8.

Microbore RP-HPLC of the residue from the main biologically active fraction in Figure 7b. (A) UV absorbance profile. (B) Biologic activity in fractions. HPLC conditions are described under Results. One-sixth of each fraction was evaporated, redissolved in 320 l RH10, and tested at 50% final concentration for stimulation of an ERS-responsive T cell line in the presence of irradiated PBMC. Mean results of triplicate assays are shown.

Full figure and legend (5K)

Figure 9.

Microbore RP-HPLC of the residue from the biologically active fraction in Figure 8. (A) UV absorbance profile. (B) Biologic activity in fractions. HPLC conditions are described under Results. One-tenth of each fraction was evaporated, redissolved in 320 l RH10, and tested at 50% final concentration for stimulation of an ERS-responsive T cell line in the presence of irradiated PBMC. Mean results of triplicate assays are shown.

Full figure and legend (6K)

Five partially-purified, SC-derived components elicit maximal PBMC proliferation at 5–7 d

A further aliquot of ERS extract was purified by Prep-10 RP-HPLC as described above. Appropriate biologically active fractions were pooled, evaporated, and repurified by chromatofocusing on the Bio-Gel anion exchange column as above. UV absorbance profiles and biologic activity were very similar to those in Figure 6(data not shown). The biologically active fractions equivalent to peaks 1–5 in Figure 6b were evaporated and individually tested for their ability to stimulate proliferation of PBMC (5 10⁵ per ml) in 3-, 5-, and 7-d assays. Appropriate final dilutions of the individual residues were determined in preliminary experiments. Material from each peak stimulated PBMC proliferation, the maximal responses occurring at days 5 or 7, with relatively little activity at day 3. The material in peak 1 appeared to be the most potent (Figure 10).

T cell stimulating activity in ERS extract is proteinase K sensitive

T cell proliferation elicited by RP-HPLC (Brownlee Prep-10)-purified ERS extract was largely abolished by incubation with proteinase K, following assays with both PBMC (Figure 11a) and with an ERS-responsive T cell line and autologous, irradiated PBMC as APC (Figure 11b). The samples that contained the added proteinase K caused non-specific inhibition of PBMC proliferation when present in amounts greater than found in the dilutions represented in Figure 11a,b. This was confirmed by separate experiments in which a wide range of concentrations of proteinase K were added to ERS samples immediately before analysis in PBMC proliferation assays (data not shown). Therefore, to confirm that abolition of biologic activity by proteinase K was not due to a non-specific inhibitory effect of the protease in the T cell proliferation assays, both control samples and those incubated with proteinase K were re-purified by RP-HPLC prior to bioassay of appropriate fractions. Biologic activity was seen in HPLC fractions of samples from control incubations (Figure 11c,d) but was again barely detectable in samples incubated with proteinase K (Figure 11e,f).

Figure 11.

Abolition of T cell proliferative responses to RP-HPLC-purified ERS extract following digestion with proteinase K. ERS samples were purified by preparative RP-HPLC (Brownlee Prep-10 column) and a fraction of intermediate hydrophobicity, representing the biologic activity illustrated in Figure 1b, used for digestions and control incubations as described in Materials and Methods. (A) Proliferation of unprimed PBMC and (B) proliferation of an ERS-reactive T cell line in the presence of irradiated, autologous PBMC as APC, in response to samples from control incubations () and those containing proteinase K (). (C, E) UV absorbance profiles and (D, F) biologic activity in fractions following RP-HPLC of control incubations (C, D) and those containing proteinase K (E, F). RP-HPLC was done on a 4.6 100 mm Brownlee column eluted as described in the second paragraph of Results. The proteinase K peak, shown in (E), eluted about 1 min later than the main ERS-derived peak seen with samples from control incubations (C). Fractions (1 min) were evaporated, redissolved in 320 l RH10 medium, and assayed at 50% final concentration for proliferative responses of an ERS-reactive T cell line in the presence of autologous, irradiated PBMC. Mean results of triplicate bioassays are shown.

Full figure and legend (19K)

DISCUSSION

Previously unrecognised T cell stimulatory materials with similar retention times on RP-HPLC have been recovered from both ERS and heel SC and from whole epidermal samples from normal skin. Although recent work, in which T cell clones generated in response to heel SC extracts were shown to respond to ERS extract, suggests that the antigenic material from the two sources is at least in part similar (^{Hales & Camp 1998}), detailed chromatographic and electrophoretic characterization has to date only been carried out with ERS samples. Proteinase K digestion abolished the activity, confirming its polypeptide nature, but the activity was enhanced by denaturation. The latter finding is consistent with the presence of polypeptide antigen and is likely to be the result of enhanced antigen processing, which has been shown in several studies to be dependent on disulfide reduction (reviewed in^{Watts 1997}). The stability of the biologic activity to denaturation also indicated the possibility of including denaturing SDS-PAGE and nitrocellulose T cell immunoblotting (^{Lamb et al. 1988}) in the purification strategy, and this was subsequently used to demonstrate the limited molecular weight range (5–18 kDa) of the active materials.

SDS-PAGE, T cell immunoblotting, and amino acid sequencing of partly purified ERS extract showed co-migration of epidermal cystatin A (^{Takeda et al. 1989}) with the major T cell stimulating activity. Authentic cystatin A, however, did not activate an ERS-responsive T cell line, suggesting that this polypeptide was a contaminant. The ability to obtain an N-terminal sequence after purification of less than 1 ml ERS extract points to the abundance of this cysteine protease inhibitor in human epidermis. Although epidermal cystatin A appears not to be autoantigenic, its abundance suggests that it may play an important role in epidermal biology, e.g., in inhibiting IL-1 converting enzyme (^{Miller et al. 1993}) or the cysteine protease activity of the house dust mite antigen Der p 1 (^{Hewitt et al. 1995;Schulz et al. 1995,1997,1998}). Preliminary studies have shown that cystatin A does not inhibit caspase 3-like activity involved in apoptosis (S Chow and J Hales, unpublished).

Normal serum antibodies that bind to components in aqueous extracts of human SC and in SC of frozen sections of human skin, have been known for many years (^{Binder et al. 1980;Dabski et al. 1985}), but the epitopes involved and their pathogenic relevance are unclear. Although serum antibodies against keratin intermediate filaments have been described (^{Hintner et al. 1985}), evidence indicates that SC antibodies are not directed against keratin intermediate filaments (^{Qutaishat et al. 1990}). In addition, the similarity between the SC antigens recognized by serum antibodies and by T cells is not known.

The time course of PBMC proliferation in response to five partly purified SC components showed maximal stimulation at 5–7 d with relatively low or minimal responses at 3 d. These properties suggest the presence of antigen as opposed to superantigen, which is capable of inducing potent proliferation of large populations of PBMC in short-term assays (^{Choi et al. 1989;Kotzin et al. 1993}). Previous work has shown that ERS and heel SC extracts were not capable of activating a polyclonal PPD-reactive T cell line (^{Hales & Camp 1998}). In addition, the activity in ERS extract is enhanced by denaturation and one of the most potent components has a molecular weight of only 5 kDa. Although these findings provide additional circumstantial evidence for the presence of antigen, further work preferably with homogeneously pure material will be required to exclude superantigenic activity.

The cellular source and pathogenic relevance of the active materials have not yet been established, and the possibility that at least some are derived from commensal skin microbes cannot be excluded. This possibility does not diminish the potential pathogenic importance of the activity, as responses to microflora have been strongly implicated in skin diseases including psoriasis (^{Valdimarsson et al. 1995}) and atopic dermatitis (^{Rokugo et al. 1990;McFadden et al. 1993}). The ability of extracts of whole epidermis to induce T cell responses (Figure 2), however, suggests the viable epidermis as a source of at least part of the active materials. Whatever the source, the most promising disease association appears to be with atopic dermatitis, in which positive epicutaneous patch tests to "human dander" have been clearly demonstrated (^{Simon 1949;Uehara & Ofuji 1969,1976;Young et al. 1985}). Further in vitro and in vivo studies with pure components are required, the limited chromatographic and electrophoretic diversity of the active materials, as shown in this study, indicating that this should be feasible. Structural identification will be essential, but T cells may respond to concentrations of peptide/MHC complexes that are lower than can be detected by conventional biochemical means (^{Shastri 1996;Watts 1997}). This is exemplified by our recovery of substantial T cell stimulating activity but minimal detectable protein in a microbore RP-HPLC fraction following rigorous bulk purification of the 5 kDa component (Figure 9), and by the fact that, to our knowledge, there has been only one previous report of the successful identification of a CD4⁺ T cell antigen following conventional biochemical purification of a human tissue extract (^{Van Noort et al. 1995}). In the latter work, homogeneously pure material was apparently obtained from multiple sclerosis brain tissue following a single RP-HPLC step. The material induced SI of <8 with primed T cell lines, and a sequence for B-crystallin was obtained following trypsinization, indicating the presence of relatively abundant polypeptide of modest antigenic potency.

Initially, our finding of potent T cell stimulation but minimal detectable protein in the same microbore RP-HPLC fraction (Figure 9) led us to suspect the presence of CD1-associated, non-peptide, lipoglycan-like antigen, perhaps microbially derived (^{Beckman et al. 1994;Sieling et al. 1995;Sugita et al. 1996}). This was disproved, however, by the effects of proteinase K, the inhibitory effects of HLA-DR monoclonal antibody, and the fact that ERS and heel SC responsive T cell lines are 88–93% CD4⁺ (^{Hales & Camp 1998}), as CD1-associated glycolipids are recognized by CD4^– human T cells (^{Porcelli et al. 1992;Gong et al. 1998}).

The relevance of antigens that are present in normal SC and activate the PBMC not only of psoriasis but atopic dermatitis patients, but also of normal donors (^{Hales & Camp 1998}), may be questioned. It should be noted, however, that the previous assays showing the response of atopic, psoriatic, and normal PBMC to SC extract, were not intended to be precisely quantitative, and differences between the subject groups may have been missed. Precise differences between the groups may only be demonstrable once individual, homogeneously pure components are tested. Furthermore, the status of SC antigens may be reflected in findings with several well-recognized antigens, including those in house dust mite species, which are reported in some studies to induce similar in vitro proliferative responses with peripheral blood T cells from both normal and atopic subjects (^{Cavaillon et al. 1988;van Neerven et al. 1996}), but only to give positive skin tests in the atopics. This apparent anomaly may be explained in a number of ways, including the potency of antigen presentation by peripheral blood dendritic cells used in assays incorporating PBMC as APC (^{Banchereau & Steinman 1998}), and the fact that antigen-specific T cells in patients with active psoriasis and atopic dermatitis may have migrated to the skin (^{van der Heijden et al. 1991}), leading to their relative depletion in peripheral blood. In addition, in vivo T cell responses to antigen in atopics may be enhanced greatly by the presence of antigen-specific IgE (^{Mudde et al. 1990;van der Heijden et al. 1993}), which is usually not present in in vitro assays incorporating human AB serum, as in the present experiments. Finally, SC antigens in the skin of normal subjects may not be presented efficiently by Langerhans cells. Thus, immature Langerhans cells in the normal epidermis, although well equipped to capture and process antigen, are found in many experiments only to be weak T cell stimulators (^{Banchereau & Steinman 1998}). In contrast, pathologic states characterized by inflammation and release of cytokines including IL-4 and granulocyte-macrophage colony-stimulating factor, may lead to Langerhans cell maturation into cells that express increased levels of major histocompatibility complex proteins and accessory molecules, and are potent T cell stimulators (^{Banchereau & Steinman 1998}). This concept is supported by work showing that fresh epidermal cell suspensions from normal human skin are poor at alloantigen presentation, whereas epidermal cell suspensions prepared from psoriatic lesions demonstrate potent alloantigen presenting capacity (^{Baadsgaard et al. 1989}).

Recently,^{Valenta et al. (1998)} have reported the presence in skin and several other tissues of an autoallergen, Hom s 1, recognized by serum IgE from atopic dermatitis patients. The relationship between Hom s 1 and the T cell antigens defined in this work, is unclear. Hom s 1 reacts with IgE and there is no available evidence to suggest that it stimulates T cells. Furthermore, its apparent molecular weight (55 kDa) is much greater than that of the major T cell antigens reported here (5–18 kDa; see Figure 4 and Figure 5).

Structural identification of the T cell stimulating components in human SC and epidermis should open important avenues in research into the pathogenesis of T cell mediated skin diseases and may reveal targets for immunotherapy. Whether HPLC purification of a very large bulk of SC extract will generate sufficient amounts of pure polypeptide to allow amino acid sequencing by tandem mass spectrometry, which is more sensitive than conventional sequencing (^{Chicz et al. 1992,1993;Pappin et al. 1996}), is uncertain. An expression cloning approach, which has been used for the identification of a range of T cell peptide antigens (^{Shastri 1996}), may be more successful and is currently being pursued.

References

1. Baadsgaard, O, Gupta, AK, Taylor, RS, Ellis, CN, Voorhees, JJ, Cooper, KD: Psoriatic epidermal cells demonstrate increased numbers and function of non-Langerhans antigen-presenting cells. J Invest Dermatol 1989 92:190–195,  | Article | PubMed | ISI | ChemPort |
2. Banchereau, J, Steinman, RM: Dendritic cells and the control of immunity. Nature 1998 392:245–252,  | Article | PubMed | ISI | ChemPort |
3. Beckman, EM, Porcelli, SA, Morita, CT, Behar, SM, Furlong, ST, Brenner, MB: Recognition of a lipid antigen by CD1-restrict + T cells. Nature 1994 372:691–694,  | Article | PubMed | ChemPort |
4. Binder, WL, Beutner, EH, Jablonska, S: Studies in immunodermatology. VIII. Immunofluorescence studies of stratum corneum antibodies. Int Arch All Appl Immunol 1980 61:407–416,  | ISI | ChemPort |
5. Bos, JD, Hulsebosch, HJ, Krieg, SR: Immunocompetent cells in psoriasis: in situ immunophenotyping by monoclonal antibodies. Arch Dermatol Res 1983 275:181–189,  | Article | PubMed | ISI | ChemPort |
6. Cavaillon, J-M, Fitting, C, Guinnepain, M-T, Rassemont, R, David, B: Lymphocyte proliferative responses to the purified Dermatophagoides farinae major allergen in untreated and hyposensitized atopic patients. Allergy 1988 43:146–151,  | PubMed | ISI | ChemPort |
7. Chicz, RM, Urban, RG, Lane, WS, Gorga, JC, Stern, LJ, Vignali, DAA, Strominger, JL: Predominant naturally processed peptides bound to HLA-DR1 are derived from MHC-related molecules and are heterogeneous in size. Nature 1992 358:764–768,  | Article | PubMed | ISI | ChemPort |
8. Chicz, RM, Urban, RG, Gorga, JC, Vignali, DAA, Lane, WS, Strominger, JL: Specificity and promiscuity among naturally processed peptides bound to HLA-DR alleles. J Exp Med 1993 178:27–47,  | Article | PubMed | ISI | ChemPort |
9. Choi, Y, Kotzin, B, Herron, L, Callahan, J, Marrack, P, Kappler, J: Interaction of Staphylococcus aureus toxin "superantigens" with human T cells. Proc Natl Acad Sci 1989 86:8941–8949,  | PubMed | ChemPort |
10. Dabski, K, Glinski, W, Beutner, EH, Jablonska, S: In vitro effects of enzymes of polymorphonuclear leukocytes on the antigenicity of stratum corneum. Int Arch All Appl Immunol 1985 77:287–291,  | ISI | ChemPort |
11. Gong, J, Stenger, S, Zack, JA, et al: Isolation of mycobacterium-reactive CD1-restricted T cells from patients with human immunodeficiency virus infection. J Clin Invest 1998 101:383–389,  | PubMed | ISI | ChemPort |
12. Hales, JM, Camp, RDR: Potent T cell stimulatory material with antigenic properties in stratum corneum of normal human skin. J Invest Dermatol 1998 110:725–729,  | Article | PubMed | ISI | ChemPort |
13. Hewitt, CRA, Brown, AP, Hart, BJ, Pritchard, DI: A major house dust mite allergen disrupts the immunoglobulin E network by selectively cleaving CD23: innate protection by antiproteases. J Exp Med 1995 182:1537–1544,  | Article | PubMed | ISI | ChemPort |
14. Hintner, H, Neises, GR, Lawley, TJ: Immunologic properties of enzymatically degraded human keratin intermediate filaments. J Invest Dermatol 1985 84:108–113,  | Article | PubMed | ISI | ChemPort |
15. Horricks, C, Holder, JE, Berth-Jones, J, Camp, RDR: Antigen-independent expansion of T-cells from psoriatic skin lesions: phenotypic characterisation and antigen reactivity. Br J Dermatol 1997 137:331–338,  | Article | PubMed | ISI | ChemPort |
16. Kotzin, BL, Leung, DYM, Kappler, J, Marrack, P: Superantigens and their potential role in human disease. Adv Immunol 1993 54:99–106,  | PubMed | ISI | ChemPort |
17. Lamb, JR, O'hehir, RE, Young, DB: The use of nitrocellulose immunoblots for the analysis of antigen recognition by T lymphocytes. J Immunol Methods 1988 110:1–10,  | Article | PubMed | ISI | ChemPort |
18. Lever, R, Turbitt, M, Sanderson, A, MacKie, R: Immunophenotyping of the cutaneous infiltrate and of the mononuclear cells in peripheral blood in patients with atopic dermatitis. J Invest Dermatol 1987 89:4–7,  | PubMed | ISI | ChemPort |
19. McFadden, J, Noble, WC, Camp, RDR: Superantigenic exotoxin-secreting potential of staphylococci isolated from atopic eczematous skin. Br J Dermatol 1993 128:631–632,  | PubMed | ISI | ChemPort |
20. Miller, DK, Calaycay, JR, Chapman, KT, Howard, AD, Kostura, MJ, Molineaux, SM, Thornberry, NA: The IL-1 converting enzyme as a therapeutic target. Ann N Y Acad Sci 1993 696:133–148,  | PubMed | ISI | ChemPort |
21. Mudde, GC, van Reijsen, FC, Boland, GJ, de Gast, GC, Bruijnzeel, PLB, Bruijnzeel-Koomen, CAFM: Allergen presentation by epidermal Langerhans cells from patients with atopic dermatitis is mediated by IgE. Immunology 1990 69:335–341,  | PubMed | ISI | ChemPort |
22. Pappin, DJC, Rahman, D, Hansen, HF, Bartlet-Jones, M, Jeffery, W, Bleasby, AJ: Chemistry, mass spectrometry and peptide-mass databases: evolution of methods for the rapid identification and mapping of cellular proteins. In: Burlingame AL, Carr SA (eds). Mass Spectrometry in the Biological Sciences 1996 Totowa: Humana, pp. 135–150,
23. Porcelli, S, Morita, CT, Brenner, MB: CD1b restricts the response of human CD4–8– T lymphocytes to a microbial antigen. Nature 1992 360:593–597,  | Article | PubMed | ISI | ChemPort |
24. Qutaishat, SS, Kumar, V, Beutner, EH, Jablonska, S: Stratum corneum antibodies detected by hemagglutination are not directed against keratin intermediate filaments. Arch Dermatol Res 1990 282:89–92,  | Article | PubMed | ISI | ChemPort |
25. Rokugo, M, Tagami, H, Usuba, Y, Tomita, Y: Contact sensitivity to Pityrosporum ovale in patients with atopic dermatitis. Arch Dermatol 1990 126:627–632,  | Article | PubMed | ISI | ChemPort |
26. Ruzicka, T, Bieber, T, Schopf, E, et al, for the European Tacrolimus Multicenter Atopic Dermatitis Study Group: A short-term trial of tacrolimus (FK506) ointment for atopic dermatitis. N Engl J Med 1997 337:816–821,  | Article | PubMed | ISI | ChemPort |
27. Schagger, H, von Jagow, G: Tricine-sodium dodecyl sulphate-polyacrylamide gel electrophoresis for the separation of proteins in the range from 1 to 100 kDa. Anal Biochem 1987 166:368–379,  | Article | PubMed | ISI | ChemPort |
28. Schulz, O, Laing, P, Sewell, HF, Shakib, F: Der p. 1, a major allergen of the house dust mite, proteolytically cleaves the low-affinity receptor for human IgE. Eur J Immunol 1995 25:3191–3194,  | PubMed | ISI | ChemPort |
29. Schulz, O, Sutton, BJ, Beavil, RL, et al: Cleavage of the low affinity receptor for human IgE (CD23) by a mite cysteine protease: nature of the cleaved fragment in relation to the structure and function of CD23. Eur J Immunol 1997 27:584–588,  | PubMed | ISI | ChemPort |
30. Schulz, O, Sewell, HF, Shakib, FL: Proteolytic cleavage of CD25, the subunit of the human T cell interleukin 2 receptor, by Der, p. 1, a major mite allergen with cysteine protease activity. J Exp Med 1998 187:271–275,  | Article | PubMed | ISI | ChemPort |
31. Shastri, N: Needles in haystacks: identifying specific peptide antigens for T cells. Current Opinion Immunol 1996 8:271–277,  | ISI | ChemPort |
32. Sieling, PA, Chatterjee, D, Porcelli, SA, et al: CD1-restricted T cell recognition of microbial lipoglycan antigens. Science 1995 269:227–230,  | PubMed | ISI | ChemPort |
33. Simon, FA: Allergy to human dander in infantile eczema. Prog Allergy 1949 2:246–264,  | ISI |
34. Sowden, JM, Berth-Jones, J, Ross, JS, et al: Double-blind, controlled, cross-over study of cyclosporin in adults with severe refractory atopic dermatitis. Lancet 1991 338:137–140,  | Article | PubMed | ISI | ChemPort |
35. Sugita, M, Jackman, RM, van Donselaar, E, et al: Cytoplasmic tail-dependent localization of CD1b antigen-presenting molecules to MIICs. Science 1996 273:349–352,  | PubMed | ISI | ChemPort |
36. Takeda, A, Kaji, H, Nakaya, K, Nakamura, Y, Samejima, T: Comparative studies on the primary structure of human cystatin As from epidermis, liver, spleen, and leukocytes. J Biochem 1989 105:986–991,  | PubMed | ISI | ChemPort |
37. Uehara, M, Ofuji, S: Delayed skin reaction to human dander in atopic dermatitis. Acta Derm Venereol (Stockh) 1969 49:294–298,  | ChemPort |
38. Uehara, M, Ofuji, S: Patch test reactions to human dander in atopic dermatitis. Arch Dermatol 1976 112:951–954,  | Article | PubMed | ISI | ChemPort |
39. Valdimarsson, H, Baker, BS, Jonsdottir, I, Powles, A, Fry, L: Psoriasis: a T-cell-mediated autoimmune disease induced by streptococcal superantigens. Immunol Today 1995 16:145–149,  | Article | PubMed | ISI | ChemPort |
40. Valenta, R, Natter, S, Seiberler, S, et al: Molecular characterization of an autoallergen, Hom s 1, identified by serum IgE from atopic dermatitis patients. J Invest Dermatol 1998 111:1178–1183,  | Article | PubMed | ISI | ChemPort |
41. van der Heijden, FL, Wierenga, EA, Bos, JD, Kapsenberg, ML: High frequency of IL-4-producing CD4+ allergen specific T lymphocytes in atopic dermatitis lesional skin. J Invest Dermatol 1991 97:389–394,  | Article | PubMed | ChemPort |
42. van der Heijden, FL, van Neerven, RJ, van Katwijk, M, Bos, JD, Kapsenberg, ML: Serum-IgE-facilitated allergen presentation in atopic disease. J Immunol 1993 150:3643–3650,  | PubMed | ChemPort |
43. Van Joost, T, Bos, JD, Heule, F, Meinardi, MMHM: Low-dose cyclosporin A in severe psoriasis. A double-blind study. Br J Dermatol 1988 118:183–190,  | PubMed | ChemPort |
44. van Neerven, RJJ, Ebner, C, Yssel, H, Kapsenberg, ML, Lamb, JR: T-cell responses to allergens: epitope-specificity and clinical relevance. Immunol Today 1996 17:526–532,  | Article | PubMed | ChemPort |
45. Van Noort, JM, Van Sechel, AC, Bajramovic, JJ, El Ouagmiri, M, Polman, CH, Lassman, H, Ravid, R: The small heat-shock protein B-crystallin as candidate autoantigen in multiple sclerosis. Nature 1995 375:798–801,  | Article | PubMed | ChemPort |
46. Watts, C: Capture and processing of exogenous antigens for presentation on MHC molecules. Annu Rev Immunol 1997 15:821–850,  | Article | PubMed | ISI | ChemPort |
47. Wierenga, EA, Snoek, M, de Groot, C, Chretien, I, Bos, JD, Jansen, HM, Kapsenberg, ML: Evidence for compartmentalization of functional subsets of CD4+ T lymphocytes in atopic dermatitis. J Immunol 1990 144:4651–4656,  | PubMed | ISI | ChemPort |
48. Young, E, Bruijnzeel-Koomen, C, Berrens, L: Delayed type hypersensitivity in atopic dermatitis. Acta Derm Venereol (Stockh) 1985 Suppl. 114: 77–81,  | ChemPort |

Acknowledgments

We are indebted to Dr. Kathryn Lilley, Protein and Nucleic Acid Chemistry Laboratory, University of Leicester, for carrying out amino acid sequencing. This work was supported by grants from Novartis Pharmaceuticals UK and the Leicester Dermatology Research Foundation.