British Journal of Dermatology {1991 \ 124, 236-241.
ADONIS 000709639100057S
Tissue kallikrein and kininogen in human sweat glands and psoriatic skin M.T.POBLETE. N.J.REYNOLDS,* CD.FIGtJEROA, J.L.BURTON,* W.MULLER-ESTERLt AND K.D.BHOOLA Departments of I'harmacolofiy and *Derma(o/ogiy. Uniwrsitfi of Bristol U.K. \Departmenl of Pathobioiitemistrn. Institute of Physioiogival Chemistry. Umversity of Mainz. Germany Accepted for publication 28 August 1990
Summary
The cellular localization of immunoreactive tissue kallikrcin and kininogen was studied in normal and psoriatic human skin. Immunoreactivity to both enzyme and suhstrate was observed in secretory granules of the dark cells in the secretory fundus (acinus) of the sweat glands. Double immunostaining revealed a segmental distribution of the two antigens. Each acinar section contained either tissue kallikrein or kininogen. However, there appeared to be a junctional zone in which both were present, but in separate dark cells. Immunoreactivity for both antigens was also observed in close apposition to the luminal microvilli of the duct cells. No specific immunostaining was seen in sebaceous glands, hair follicles, keratinocytes and other cells of the secretory unit such as myoepithelial or clear cells. In psoriatic skin there were in addition many neutrophiis immunoreactive for tissue kallikrein in the epidermis and psoriatic scales. Mitogenic action of kinins may account to some extent for the characteristic accelerated turnover of epidermal cells in psoriasis and locally applied kinin antagonists may prove of value in the treatment of this disease.
The presence of a kinin-forming enzyme in normal human eccrine sweat was first reported in 1958.' Later, this kininogenase enzyme in sweat^ was shown to closely resemble a family of proteases known as kallikreins.' However, these findings were questioned by Frewin et al..'^ because of their inability to detect kinins or kininogenase activity in normal human sweat. Further differences emerged when a kininogenase. sensitive to inhibition by soya bean trypsin inhibitor but immunologically distinct from plasma or tissue kallikrein. was purified from extracts of human skin."' Subsequently, using a specific radioimmunoassay. Mann et al. ^ reported the presence of a tissue kallikrein in human sweat, and this was later partially purified by Hibino et al.' More recently, variations in the amount of immunoreactive tissue kallikrein have been reported in sweat collected from different regions of the body, the highest levels being found in samples obtained from the trunk and forehead.^
substrate, kininogen, in human skin biopsies obtained from both normal volunteers and patients with psoriasis.
Methods Tissue samples
Skin biopsies (4-mm punch) were taken under local anaesthetic (2% lignocaine with 1:8().0(K) adrenaline) from the upper arm of six normal volunteers and from the lesional skin of six patients with psoriasis. They were fixed in formal-saline (10% v/v) at room temperature for 24-48 h. The tissue blocks were dehydrated in a graded series of ethanol and embedded in paraffin wax. Sections 4-5 fim thick were mounted on glass slides coated with polylysine (Sigma Chemicals. Poole. Dorset. U.K.).
Antisera
The role of kinins in inflammatory skin disease is poorly understood, but extracts prepared from the epidermal scales ofpsoriatic patients have been shown to contain a tissue kallikrein-like enzyme which forms a kinin from bovine L-kininogen.'' We have now examined the cellular localization of tissue kallikrein and its
Tissue kaUikrcin. Immunocytochcmistry was performed with three polyclonal antibodies raised against human urinary kallikrein: two rabbit antisera'"" and a goat antiserum (Protogen AG. Switzerland).
Correspondence: Dr K.D.Bhooia. Department of Pharmacology. University of Bristol. Bristol BSS lTD. U,K,
Kininogen. A polyclonal antibody raised against the whole L-kininogen molecule'- was used in this study.
236
TISSUE KALLIKRHIN AND KININOGEN IN SWEAT GLANDS
Immunostaining Two different immunostaining procedures were carried out: the peroxidase anti-peroxidase method (PAP) as described previously'' and the immunogold silver staining (IGSS) method. FAP method. Skin tissue sections were dewaxed with xylene and rehydrated through graded series of ethanol. The sections were treated with ahsolute methanol and 3% (v/v) hydrogen peroxide to hlock endogenous pseudoperoxidase activity'"^ and then rinsed in Tris-HCl (0-05 mol/I)-sodium chloride (()-15 mol/1) buffer (pH 7-8). The samples were then incubated with anti-tissue kallikrein (1; 500-1: 5()()()) or anti-kininogen {1:4001:1000) antihody for IK h. When a goat primary antibody was used, the second step included incubation with rabbit anti-goat IgG serum (1:1000) (Sigma) for 50 min. Swine anti-rabbit IgG and PAP complexes (Dakopatts. Denmark, Copenhagen) were used at dilutions of 1:40 and 1:100. respectively. All incubations were performed at 22°C in a water hath that was used as a humidifier. Peroxidase activity was demonstrated with 0-1% 5.5 diaminobeny.idine tetrahydrochloride in the presence of ()()5% hydrogen peroxide. Finally, the sections were counterstained with haematoxylin. IGSS nietlwd. Ihe tissue sections were dewaxed. rehydrated as previously described and rinsed twice for 5 min each in PBS (0 ()1 mol/1 pH 7 2 ) . They were then immunostained by incubation with: (i| S% non-immune goat serum for 20 min: (ii) anti-tissue kallikrein serum (1: SOOO) (Protogen A.G.. Switzerland) at 22X for 18 h: (iii) rabbit anti-goat IgG (1:1000) (Sigma) for 30 min: (iv) gold-labelled goat anti-rabbit IgG (Janssen Biotech. Belgium) (1 :40) for 1 h; (v) silver enhancement for 1020 min. The silver enhancement reagent (Janssen) was prepared by mixing equal amounts of enhancer and initiator solutions. Finally, the slides were washed thoroughly with distilled water, counterstained with 1 % neutral red and then mounted in DPX. Douhh' immiinostaming. A double immunostaining method'^ was used to demonstrate tissue kallikrein and kininogen in the same sections, using two different chromogens. Tissue sections were immunostained for tissue kallikrein as described previously to develop the first colour (brown). Before the second immunostaining procedure, an elution step was performed. Sections were washed for 1 h with several changes of glycine-HCl (0 2 mol/l)-sodium chloride (0-5 mol/1. pH 2 3) and 10%
237
dimethylformamide (Sigma) in Tris-HCl buffer (0-05 mol/l. pH 7-8).''' The sections were then incubated with rabbit anti-kininogen serum (1:200). anti-rabbit IgG (1:40 v/v) and PAP (1:100 v/v). Peroxidase activity was developed with 4-chloro~l-naphthol (Sigma). Finally, the sections were covered with a glass coverslip using a solution of glycerol and Tris-HCl-saline (1:1 v/v). Controls. The specificity of the immunostaining was checked by prior absorption of the tissue kallikrein antisera with purified human urinary kallikrein (50 //g/ ml). The antiserum against i,-kininogen was absorbed with purified human i,-kininogen (20-50 /ig/ml). Other controls included omission of the specific antiserum and replacement by non-immune serum.
Results Normal skin Tissue kaUikreiu. A granular pattern of immunostaining was observed in the acinar dark cells arranged around the lumen of the secretory unit of sweat glands (Fig. 1). Other components of the secretory unit such as the clear cells and myoepithelial cells showed no immunostaining. Staining was also seen as an intense line along the luminal microvilli of the duct cells up to its opening on the surface of the skin (Fig. 2). There was no evidence of specific staining in the sehaceous glands, hair follicles, keratinocytes. dermal vessels or in the interstitial tissue space. Both PAP and IGSS methods gave identical results. Kininoijen. An intense immunoreactivity was observed in the dark cells of the acinus which was different from those containing tissue kallikrein (Fig. i). As with tissue kallikrein. no specific immunoreactivity to kininogen was seen in other cells of the acinus or in the interstitial tissue space but staining was observed in the capillaries. Deep staining for kininogen was observed in apposition to the microvilli of the luminal duct cells. Double immunostaining. The use of the double-immunostaining method allowed the visualisation of both antigens in the same section. Serial sections showed that there was a segmental distribution of the two antigens. Individual sections of the secretory fundus contained either one or other of the antigens, but there appeared to be a junctional region of the fundus in which both antigens were present in separate dark cells (Fig. 4).
2 38
M.T.POBLETE d a/.
cc
TISSUE KALLIKRHIN AND KININOGEN IN SWEAT GLANDS
Psoriatic skin
Tissue kallikrein. In addition to the immunoreactivity present in the dark cells of the secretory fundus. a clear granular pattern of immunostaining was observed in the cytoplasm of neutrophils that had migrated into the epidermis and had accumulated in the psoriatic scales (I'iji. S and dl. In the epidermis the neutrophils orientated either individually or in clusters (Fig. 6).
239
of the enzyme and substrate into the acinar lumen probably occurs by exocytosis. The structural organisa-
EPIDERMtS
Kininogeii. A distrihution similar to that described for the normal skin was observed in the hiopsies from psoriatic patients.
Discussion Following the description of a bradykinin-forming enzyme in human sweat by Fox and Hilton.' several investigators proposed a role for kinins in sweat gland function. However, the precise segment of the gland and the cells that contain the enzyme(s) and the substrate {kininogen) necessary for the formation of kinins were not clearly delineated. Using immunocytochemical techniques we have now localized tissue kallikrein and its substrate, kininogen, in the dark cells of the secretory fundus of sweat glands. Furthermore, using doubleimmunostaining procedures it was found that tissue kallikrein and kininogen were located in distinct regions of the fundus. In the junctional zone the enzyme and substrate were each seen in separate dark cells, which were intermingled (Fig. 7). A similar anatomical relationship has been found in the human kidney.'' (lur observation that the secretory granules containing tissue kallikrein or kininogen are orientated to the apical end of the cell suggests that the primary secretion
KININOGEN KALLIKREIN
SECRETORY FUNDUS OR ACINUS
[•'igurc 7. i)i;igrummulic rcprcscntiilion of the striicturiil and (.clluliir iippciiriincf i)f Ihc luiniiin swi'iit yhuid. BM. biisiil TncTuhnine; M. myoepitholial cell: CC. dear cell: DC, dark cell; T), tight junction; CA, cannalicutus.
Figure 1. Tissuf kallikrein, normal skin. Intense itnmunostaining in the dark cells (DC) of the secretory fundus. Clear cells {CO and myoepithelial cells (M) remain unstained. KISS method x 2 JOl). Figure 2. Tissue kallikrein, normal skin. Arrows indicate immunoreactivity visualised along the luminal microvilli of the duct cells. PAP method X 2000. Figure J. Kininogen. normal skin. Immunoreactive kininogen is contained in dark celts (DC) of the secretory fundus. Arrows point to unstained DC which showed immunoreactivity for tissue kallikrein in the next section. Immunoreactive kininogen is also observed in the lumen of a capillary (C). CC. clear cells: M, myoepithelia! cell. PAl' method x 1000, Figure 4. Tissue kallikrein-kininogen double immunostaining. normal skin, [unctional region of the secretory fundus in which the dark cells containing tissue katlikrein (TK. brown) und kininogen (KNC), purple) intermingle. PAP method x 12()fl. Figure S. Tissue kallikrein, psoriatic skin. Dense neutrophil infiltrates (arrows) invade the epidermis (R) and psoriatic scales (S). Most of the neulrophils contain immunoreactive tissue kallikrein. PAP method x 500. I'igure fi. Tissue kallikrein. psoriatic skin. Immunoreactive tissue kallikrein is localized in the cytoplasmic granules of neutrophils (arrows). PAP method X25OO.
240
M.T.POBLETEff al.
tion of the dark cells and intercellular tight junctions"' is such that during active sweat formation hackward transport of macromolecules'" into the interstitial tissue space is unlikely"'. In our study we found no immunoreactive tissue kallikrein or kininogen beyond the outer surface of the basal lamina of the secretory epithelium, During prolonged stimulation dark cells that cannot sustain secretion are discarded from the secretory epithelium.'" Disintegration iff dark cells causes large intercellular gaps to appear, thereby exposing the myoepithelial cells of the secretory coil to the contractile action of kinins. in addition to the neuronaliy released acetylcholine. In the recovery phase ibilowing intense sweating, the dead cells are replaced by direct division of remaining epithelial cells. The known mitogenic property of kinins''' -' may induce mitotic activity in the secretory epithelium after recovery from sweating. The iontophoretic application of hradykinin produces an abrupt fall of sweat osmolalityr- Moreover. kalHdin (lys-bradykinin) increases sodium absorption across cultured sweat epithelium in a manner that seems to be independent of the generation of prostaglandins.-' However, these effects were only observed when the peptide was added to the basolateral and not to the apical face of the monoiayers.-' During thermal stimulation of sweat glands, small clear vesicles appear on the luminal surface that constitute a resorptive mechanism for the movement t)f molecules to theserosai side.-"* In addition, the intercellular spaces between the luminal cells of the duct widen during sweating. If. therefore, kinins are formed in the duct lumen, permeation to the serosal side of the iuminal cells could modulate the excretion of sodium and water by the gland, assuming they survive the action of kininases. A kininase with properties similar to those of angiotensln I converting enzyme (kininase} has been recently partially purified from human sweat by Hibino ct ai' Sideness-prelerence (hasoiateral versus apical) for hradykinin and kallidln in ion transport has also been described using isolated cortical collecting duct.s from rabbit kidneys. At this level kinins inhibit sodium absorption and arginine vasopressin-mediated water resfirption.^"'-''' Furthermore, the formed kinins could enhance the contractile movements of the microvilli in order to increase propulsion of sweat to the epidermis. One of the characteristic histopathological features of psoriasis is the presence of a dense Infiltrate of polymorphonuclear leucocytes, that may result in microabscess formation within the stratum corneum.-' I'soriatic scales contain leucotactic substances that prohahly correspond to complement-associated peptides generated hy the action of proteolytic enzymes.•'"-'* We have shown that circulating neutrophils which infiltrate the
epidermis of these patients contain immunoreactive tissue kallikrein.'" In addition, neutrophils possess receptors for H-kininogen and contain immunoreactive I.- and H-kininogens." '- Kininogen is increased in several inflammatory diseases including rheumatoid arthritis'' and psoriasis. '^ As kininogens are synthesized hy hepatocytes in the liver, cytokines or similar mediators produced locally in inflamed tissue may be responsible for switching on the hepatic synthesis and secretion of the kininogen into the circulation. Kinins are believed to act as mitogens. stimulating DNA synthesis and cell division.'''"'' Intradermal occlusions of the sweat gland duct unit'^ "' together with the migration of neutrophils into pustules and microabscesses may result in an increaf^ed formation of the peptide. The higher kinin levels could in part account for the increased mitotic activity seen in the psoriatic epidermis. It is therefore possible that kinins contribute to the pathogenesis of psoriasis, and locally applied kinin antagonists may prove of therapeutic value.
A c k n o wied gm en ts We thank the Psoriasis Association for flnancial support. The study was approved by the Bristol Royal Infirmary Ethical Ct)mmittee,
References 1 Fox RH, Hilton SM. Bradykinin formation in human skin as a factor in heat vasodilatation. / Phiisiol 195K: 142: 219-J2. 2 Fraki JK, lansen Cl\ HopsLi-Havu VK, Human sweat kallikrein, Acta Demi Vviu-reol {Storkh) 1970: 50: ?21-f). i Schachter M. Kullikreins (kiiijnoti;cnases)—a group of scrine proteases with bioregulalory actions, Phiirmucol Rev 1980: 11: 1-17. 4 Prcwiii ilB, McConnell D|, Downey | A. Is a kininogenase necessary for human sweating? iMicel ] 9 7 i : ii: 744, 5 Toki N. Yamura T, Kinin-forming enzyme in human skin: The purification and characlerization of a kinin furming ennyme I tnvfsl I\'rmatol 197'*: 7J: 297-J(12, b Mann K, Lipp B. tJrunst | el al. Determination of kallikrein by radioimmunoasskiy in human body lluids, A(ii-nts Actions 19S(); 10: J29-i4, 7 Hibino T, Takemura T, Sato K, Demonstration of glandular kallikrein and angiotensin converting ennyme ikininase-II) in human eccrine sweat. / invest Derinato} 1988; 90: 569, 8 Maylield RK. Sens VA. |affa A. Margolius S, Studies of sweat kallikrein in normal human subjects. Adv Exp Med Hiol iKiiiim V> 9 Hibino T, Isaki S, Kimura H el
ADONIS 000709639100057S
Tissue kallikrein and kininogen in human sweat glands and psoriatic skin M.T.POBLETE. N.J.REYNOLDS,* CD.FIGtJEROA, J.L.BURTON,* W.MULLER-ESTERLt AND K.D.BHOOLA Departments of I'harmacolofiy and *Derma(o/ogiy. Uniwrsitfi of Bristol U.K. \Departmenl of Pathobioiitemistrn. Institute of Physioiogival Chemistry. Umversity of Mainz. Germany Accepted for publication 28 August 1990
Summary
The cellular localization of immunoreactive tissue kallikrcin and kininogen was studied in normal and psoriatic human skin. Immunoreactivity to both enzyme and suhstrate was observed in secretory granules of the dark cells in the secretory fundus (acinus) of the sweat glands. Double immunostaining revealed a segmental distribution of the two antigens. Each acinar section contained either tissue kallikrein or kininogen. However, there appeared to be a junctional zone in which both were present, but in separate dark cells. Immunoreactivity for both antigens was also observed in close apposition to the luminal microvilli of the duct cells. No specific immunostaining was seen in sebaceous glands, hair follicles, keratinocytes and other cells of the secretory unit such as myoepithelial or clear cells. In psoriatic skin there were in addition many neutrophiis immunoreactive for tissue kallikrein in the epidermis and psoriatic scales. Mitogenic action of kinins may account to some extent for the characteristic accelerated turnover of epidermal cells in psoriasis and locally applied kinin antagonists may prove of value in the treatment of this disease.
The presence of a kinin-forming enzyme in normal human eccrine sweat was first reported in 1958.' Later, this kininogenase enzyme in sweat^ was shown to closely resemble a family of proteases known as kallikreins.' However, these findings were questioned by Frewin et al..'^ because of their inability to detect kinins or kininogenase activity in normal human sweat. Further differences emerged when a kininogenase. sensitive to inhibition by soya bean trypsin inhibitor but immunologically distinct from plasma or tissue kallikrein. was purified from extracts of human skin."' Subsequently, using a specific radioimmunoassay. Mann et al. ^ reported the presence of a tissue kallikrein in human sweat, and this was later partially purified by Hibino et al.' More recently, variations in the amount of immunoreactive tissue kallikrein have been reported in sweat collected from different regions of the body, the highest levels being found in samples obtained from the trunk and forehead.^
substrate, kininogen, in human skin biopsies obtained from both normal volunteers and patients with psoriasis.
Methods Tissue samples
Skin biopsies (4-mm punch) were taken under local anaesthetic (2% lignocaine with 1:8().0(K) adrenaline) from the upper arm of six normal volunteers and from the lesional skin of six patients with psoriasis. They were fixed in formal-saline (10% v/v) at room temperature for 24-48 h. The tissue blocks were dehydrated in a graded series of ethanol and embedded in paraffin wax. Sections 4-5 fim thick were mounted on glass slides coated with polylysine (Sigma Chemicals. Poole. Dorset. U.K.).
Antisera
The role of kinins in inflammatory skin disease is poorly understood, but extracts prepared from the epidermal scales ofpsoriatic patients have been shown to contain a tissue kallikrein-like enzyme which forms a kinin from bovine L-kininogen.'' We have now examined the cellular localization of tissue kallikrein and its
Tissue kaUikrcin. Immunocytochcmistry was performed with three polyclonal antibodies raised against human urinary kallikrein: two rabbit antisera'"" and a goat antiserum (Protogen AG. Switzerland).
Correspondence: Dr K.D.Bhooia. Department of Pharmacology. University of Bristol. Bristol BSS lTD. U,K,
Kininogen. A polyclonal antibody raised against the whole L-kininogen molecule'- was used in this study.
236
TISSUE KALLIKRHIN AND KININOGEN IN SWEAT GLANDS
Immunostaining Two different immunostaining procedures were carried out: the peroxidase anti-peroxidase method (PAP) as described previously'' and the immunogold silver staining (IGSS) method. FAP method. Skin tissue sections were dewaxed with xylene and rehydrated through graded series of ethanol. The sections were treated with ahsolute methanol and 3% (v/v) hydrogen peroxide to hlock endogenous pseudoperoxidase activity'"^ and then rinsed in Tris-HCl (0-05 mol/I)-sodium chloride (()-15 mol/1) buffer (pH 7-8). The samples were then incubated with anti-tissue kallikrein (1; 500-1: 5()()()) or anti-kininogen {1:4001:1000) antihody for IK h. When a goat primary antibody was used, the second step included incubation with rabbit anti-goat IgG serum (1:1000) (Sigma) for 50 min. Swine anti-rabbit IgG and PAP complexes (Dakopatts. Denmark, Copenhagen) were used at dilutions of 1:40 and 1:100. respectively. All incubations were performed at 22°C in a water hath that was used as a humidifier. Peroxidase activity was demonstrated with 0-1% 5.5 diaminobeny.idine tetrahydrochloride in the presence of ()()5% hydrogen peroxide. Finally, the sections were counterstained with haematoxylin. IGSS nietlwd. Ihe tissue sections were dewaxed. rehydrated as previously described and rinsed twice for 5 min each in PBS (0 ()1 mol/1 pH 7 2 ) . They were then immunostained by incubation with: (i| S% non-immune goat serum for 20 min: (ii) anti-tissue kallikrein serum (1: SOOO) (Protogen A.G.. Switzerland) at 22X for 18 h: (iii) rabbit anti-goat IgG (1:1000) (Sigma) for 30 min: (iv) gold-labelled goat anti-rabbit IgG (Janssen Biotech. Belgium) (1 :40) for 1 h; (v) silver enhancement for 1020 min. The silver enhancement reagent (Janssen) was prepared by mixing equal amounts of enhancer and initiator solutions. Finally, the slides were washed thoroughly with distilled water, counterstained with 1 % neutral red and then mounted in DPX. Douhh' immiinostaming. A double immunostaining method'^ was used to demonstrate tissue kallikrein and kininogen in the same sections, using two different chromogens. Tissue sections were immunostained for tissue kallikrein as described previously to develop the first colour (brown). Before the second immunostaining procedure, an elution step was performed. Sections were washed for 1 h with several changes of glycine-HCl (0 2 mol/l)-sodium chloride (0-5 mol/1. pH 2 3) and 10%
237
dimethylformamide (Sigma) in Tris-HCl buffer (0-05 mol/l. pH 7-8).''' The sections were then incubated with rabbit anti-kininogen serum (1:200). anti-rabbit IgG (1:40 v/v) and PAP (1:100 v/v). Peroxidase activity was developed with 4-chloro~l-naphthol (Sigma). Finally, the sections were covered with a glass coverslip using a solution of glycerol and Tris-HCl-saline (1:1 v/v). Controls. The specificity of the immunostaining was checked by prior absorption of the tissue kallikrein antisera with purified human urinary kallikrein (50 //g/ ml). The antiserum against i,-kininogen was absorbed with purified human i,-kininogen (20-50 /ig/ml). Other controls included omission of the specific antiserum and replacement by non-immune serum.
Results Normal skin Tissue kaUikreiu. A granular pattern of immunostaining was observed in the acinar dark cells arranged around the lumen of the secretory unit of sweat glands (Fig. 1). Other components of the secretory unit such as the clear cells and myoepithelial cells showed no immunostaining. Staining was also seen as an intense line along the luminal microvilli of the duct cells up to its opening on the surface of the skin (Fig. 2). There was no evidence of specific staining in the sehaceous glands, hair follicles, keratinocytes. dermal vessels or in the interstitial tissue space. Both PAP and IGSS methods gave identical results. Kininoijen. An intense immunoreactivity was observed in the dark cells of the acinus which was different from those containing tissue kallikrein (Fig. i). As with tissue kallikrein. no specific immunoreactivity to kininogen was seen in other cells of the acinus or in the interstitial tissue space but staining was observed in the capillaries. Deep staining for kininogen was observed in apposition to the microvilli of the luminal duct cells. Double immunostaining. The use of the double-immunostaining method allowed the visualisation of both antigens in the same section. Serial sections showed that there was a segmental distribution of the two antigens. Individual sections of the secretory fundus contained either one or other of the antigens, but there appeared to be a junctional region of the fundus in which both antigens were present in separate dark cells (Fig. 4).
2 38
M.T.POBLETE d a/.
cc
TISSUE KALLIKRHIN AND KININOGEN IN SWEAT GLANDS
Psoriatic skin
Tissue kallikrein. In addition to the immunoreactivity present in the dark cells of the secretory fundus. a clear granular pattern of immunostaining was observed in the cytoplasm of neutrophils that had migrated into the epidermis and had accumulated in the psoriatic scales (I'iji. S and dl. In the epidermis the neutrophils orientated either individually or in clusters (Fig. 6).
239
of the enzyme and substrate into the acinar lumen probably occurs by exocytosis. The structural organisa-
EPIDERMtS
Kininogeii. A distrihution similar to that described for the normal skin was observed in the hiopsies from psoriatic patients.
Discussion Following the description of a bradykinin-forming enzyme in human sweat by Fox and Hilton.' several investigators proposed a role for kinins in sweat gland function. However, the precise segment of the gland and the cells that contain the enzyme(s) and the substrate {kininogen) necessary for the formation of kinins were not clearly delineated. Using immunocytochemical techniques we have now localized tissue kallikrein and its substrate, kininogen, in the dark cells of the secretory fundus of sweat glands. Furthermore, using doubleimmunostaining procedures it was found that tissue kallikrein and kininogen were located in distinct regions of the fundus. In the junctional zone the enzyme and substrate were each seen in separate dark cells, which were intermingled (Fig. 7). A similar anatomical relationship has been found in the human kidney.'' (lur observation that the secretory granules containing tissue kallikrein or kininogen are orientated to the apical end of the cell suggests that the primary secretion
KININOGEN KALLIKREIN
SECRETORY FUNDUS OR ACINUS
[•'igurc 7. i)i;igrummulic rcprcscntiilion of the striicturiil and (.clluliir iippciiriincf i)f Ihc luiniiin swi'iit yhuid. BM. biisiil TncTuhnine; M. myoepitholial cell: CC. dear cell: DC, dark cell; T), tight junction; CA, cannalicutus.
Figure 1. Tissuf kallikrein, normal skin. Intense itnmunostaining in the dark cells (DC) of the secretory fundus. Clear cells {CO and myoepithelial cells (M) remain unstained. KISS method x 2 JOl). Figure 2. Tissue kallikrein, normal skin. Arrows indicate immunoreactivity visualised along the luminal microvilli of the duct cells. PAP method X 2000. Figure J. Kininogen. normal skin. Immunoreactive kininogen is contained in dark celts (DC) of the secretory fundus. Arrows point to unstained DC which showed immunoreactivity for tissue kallikrein in the next section. Immunoreactive kininogen is also observed in the lumen of a capillary (C). CC. clear cells: M, myoepithelia! cell. PAl' method x 1000, Figure 4. Tissue kallikrein-kininogen double immunostaining. normal skin, [unctional region of the secretory fundus in which the dark cells containing tissue katlikrein (TK. brown) und kininogen (KNC), purple) intermingle. PAP method x 12()fl. Figure S. Tissue kallikrein, psoriatic skin. Dense neutrophil infiltrates (arrows) invade the epidermis (R) and psoriatic scales (S). Most of the neulrophils contain immunoreactive tissue kallikrein. PAP method x 500. I'igure fi. Tissue kallikrein. psoriatic skin. Immunoreactive tissue kallikrein is localized in the cytoplasmic granules of neutrophils (arrows). PAP method X25OO.
240
M.T.POBLETEff al.
tion of the dark cells and intercellular tight junctions"' is such that during active sweat formation hackward transport of macromolecules'" into the interstitial tissue space is unlikely"'. In our study we found no immunoreactive tissue kallikrein or kininogen beyond the outer surface of the basal lamina of the secretory epithelium, During prolonged stimulation dark cells that cannot sustain secretion are discarded from the secretory epithelium.'" Disintegration iff dark cells causes large intercellular gaps to appear, thereby exposing the myoepithelial cells of the secretory coil to the contractile action of kinins. in addition to the neuronaliy released acetylcholine. In the recovery phase ibilowing intense sweating, the dead cells are replaced by direct division of remaining epithelial cells. The known mitogenic property of kinins''' -' may induce mitotic activity in the secretory epithelium after recovery from sweating. The iontophoretic application of hradykinin produces an abrupt fall of sweat osmolalityr- Moreover. kalHdin (lys-bradykinin) increases sodium absorption across cultured sweat epithelium in a manner that seems to be independent of the generation of prostaglandins.-' However, these effects were only observed when the peptide was added to the basolateral and not to the apical face of the monoiayers.-' During thermal stimulation of sweat glands, small clear vesicles appear on the luminal surface that constitute a resorptive mechanism for the movement t)f molecules to theserosai side.-"* In addition, the intercellular spaces between the luminal cells of the duct widen during sweating. If. therefore, kinins are formed in the duct lumen, permeation to the serosal side of the iuminal cells could modulate the excretion of sodium and water by the gland, assuming they survive the action of kininases. A kininase with properties similar to those of angiotensln I converting enzyme (kininase} has been recently partially purified from human sweat by Hibino ct ai' Sideness-prelerence (hasoiateral versus apical) for hradykinin and kallidln in ion transport has also been described using isolated cortical collecting duct.s from rabbit kidneys. At this level kinins inhibit sodium absorption and arginine vasopressin-mediated water resfirption.^"'-''' Furthermore, the formed kinins could enhance the contractile movements of the microvilli in order to increase propulsion of sweat to the epidermis. One of the characteristic histopathological features of psoriasis is the presence of a dense Infiltrate of polymorphonuclear leucocytes, that may result in microabscess formation within the stratum corneum.-' I'soriatic scales contain leucotactic substances that prohahly correspond to complement-associated peptides generated hy the action of proteolytic enzymes.•'"-'* We have shown that circulating neutrophils which infiltrate the
epidermis of these patients contain immunoreactive tissue kallikrein.'" In addition, neutrophils possess receptors for H-kininogen and contain immunoreactive I.- and H-kininogens." '- Kininogen is increased in several inflammatory diseases including rheumatoid arthritis'' and psoriasis. '^ As kininogens are synthesized hy hepatocytes in the liver, cytokines or similar mediators produced locally in inflamed tissue may be responsible for switching on the hepatic synthesis and secretion of the kininogen into the circulation. Kinins are believed to act as mitogens. stimulating DNA synthesis and cell division.'''"'' Intradermal occlusions of the sweat gland duct unit'^ "' together with the migration of neutrophils into pustules and microabscesses may result in an increaf^ed formation of the peptide. The higher kinin levels could in part account for the increased mitotic activity seen in the psoriatic epidermis. It is therefore possible that kinins contribute to the pathogenesis of psoriasis, and locally applied kinin antagonists may prove of therapeutic value.
A c k n o wied gm en ts We thank the Psoriasis Association for flnancial support. The study was approved by the Bristol Royal Infirmary Ethical Ct)mmittee,
References 1 Fox RH, Hilton SM. Bradykinin formation in human skin as a factor in heat vasodilatation. / Phiisiol 195K: 142: 219-J2. 2 Fraki JK, lansen Cl\ HopsLi-Havu VK, Human sweat kallikrein, Acta Demi Vviu-reol {Storkh) 1970: 50: ?21-f). i Schachter M. Kullikreins (kiiijnoti;cnases)—a group of scrine proteases with bioregulalory actions, Phiirmucol Rev 1980: 11: 1-17. 4 Prcwiii ilB, McConnell D|, Downey | A. Is a kininogenase necessary for human sweating? iMicel ] 9 7 i : ii: 744, 5 Toki N. Yamura T, Kinin-forming enzyme in human skin: The purification and characlerization of a kinin furming ennyme I tnvfsl I\'rmatol 197'*: 7J: 297-J(12, b Mann K, Lipp B. tJrunst | el al. Determination of kallikrein by radioimmunoasskiy in human body lluids, A(ii-nts Actions 19S(); 10: J29-i4, 7 Hibino T, Takemura T, Sato K, Demonstration of glandular kallikrein and angiotensin converting ennyme ikininase-II) in human eccrine sweat. / invest Derinato} 1988; 90: 569, 8 Maylield RK. Sens VA. |affa A. Margolius S, Studies of sweat kallikrein in normal human subjects. Adv Exp Med Hiol iKiiiim V> 9 Hibino T, Isaki S, Kimura H el
