Ž .Immunopharmacology 36 1997 221–227

Expression of human tissue kallikrein in rat salivary glands andits secretion into circulation following adenovirus-mediated gene

transfer

Cindy Wang, Caroline Chao, Lee Chao, Julie Chao )

Department of Biochemistry and Molecular Biology, Medical UniÕersity of South Carolina, Charleston, SC 29425-2211, USA

Abstract

Replication-deficient adenovirus Ad.CMV-cHK, expressing human tissue kallikrein under the control of the cy-tomegalovirus enhancerrpromoter, was introduced into rat salivary glands via a direct intracapsular injection. A singleinjection of Ad.CMV-cHK at a dose of 4=109 pfu resulted in a sustained expression of human tissue kallikrein in ratsalivary glands. The level of immunoreactive human tissue kallikrein in rat sera was the highest at 1 day post gene deliverywhen both salivary glands were injected and decreased in a time-dependent manner after gene delivery. Human tissuekallikrein levels in sera increased concomitantly with the amount of adenovirus used in direct salivary injection. Thedetection of human tissue kallikrein in sera after gene delivery into salivary glands provided direct evidence indicating thatrat salivary glands secrete locally synthesized human tissue kallikrein to the systemic circulation. The direct injection ofsalivary glands with replication-deficient adenovirus could provide a systemic route for gene delivery for studying salivarygland function and development. Targeted gene delivery to the salivary gland may provide the means to express therapeuticproteins in saliva and the systemic circulation.

Keywords: Tissue kallikrein; Adenovirus; Gene delivery; Salivary gland; ELISA; Rat

1. Introduction

Ž .Tissue kallikreins E.C. 3.4.21.35 are a subgroupof serine proteinases that release vasoactive kininpeptides from kininogens. Kinins have been shownto mediate vasodilation, vascular permeability,smooth muscle relaxation and contraction and glu-

Ž .cose transport Bhoola et al., 1992; Margolius, 1995 .In addition to processing vasoactive kinins, tissuekallikrein has been implicated in the processing of

) Corresponding author. Tel.: q1-803-7924321; fax: q1-803-7924322.

growth factors and peptide hormones. Tissuekallikrein is widely distributed in various tissues andin a variety of body fluids, such as serum, saliva,urine, bile and sweat. However, little is known aboutthe source of circulating tissue kallikrein. Studieshave indicated that several tissues may contribute tocirculating tissue kallikrein by secreting locally syn-thesized tissue kallikrein into the systemic circula-tion. In the salivary gland the release of tissuekallikrein into the circulation was greatly enhanced

Ž .by alpha-adrenergic stimulation Berg et al., 1985 .The kidney was demonstrated as another importantprobable source of circulating tissue kallikrein, be-

0162-3109r97r$17.00 Copyright q 1997 Elsevier Science B.V. All rights reserved.Ž .PII S0162-3109 97 00025-8

( )C. Wang et al.r Immunopharmacology 36 1997 221–227222

cause the venous effluent of the isolated rat kidneyperfused with an artificial medium contains tissue

Ž .kallikrein Roblero et al., 1976; Vio et al., 1983 .Additional sources of circulating tissue kallikreinhave been identified as vascular kallikrein. An en-zyme with kininogenase activity has been isolated

Žfrom rat arteries Nolly and Lama, 1982; Nolly et al.,.1985 . Circulating human neutrophils at the end of

their life cycle could release tissue kallikrein as theyŽ .disintegrate Figueroa et al., 1989 . These studies

suggest that several tissues may be capable to secret-ing tissue kallikrein into the circulation. However,there is no direct evidence supporting the hypothesisthat tissue kallikrein is synthesized locally and thensecreted continuously into the systemic circulation invivo.

In vivo gene transfer is a new technology with thepotential for treating a broad range of acquired and

Ž .inherited diseases Crystal, 1995 . Replication-defi-cient adenoviruses represent an efficient and safemethod of in vivo gene transfer. Gene transfer tovarious tissues has been accomplished with an aden-

Žovirus-mediated gene transfer approach Berns and.Giraud, 1995 . Adenovirus-mediated gene transfer

provides a promising tool for the study of salivarygland function and gene regulation. Recently, genetransfer to the salivary gland has been achieved bothin vitro and in vivo. Retrograde ductal injection ofthe replication-deficient adenoviral vector into ratsalivary glands in vivo resulted in the expression ofan exogenous gene in acinar and ductal cellsŽ .Mastrangeli et al., 1994 . It has been demonstratedthat the uptake and internalization of naked plasmidDNA in the salivary gland can be facilitated by

Ž .adenovirus infection O’Connell et al., 1995 .In the studies described in this report, we demon-

strate that a single intracapsular injection of a repli-cation-deficient adenovirus encoding human tissuekallikrein produces human tissue kallikrein in ratsalivary gland and serum in a dose-dependent man-ner. The results suggest that intracapsular injectionof replication-deficient adenoviruses could be a use-ful tool in studying the physiological function andpotential therapeutic effect of recombinant proteinsin oral cavity. It could also facilitate the developmentof an alternative in vivo gene transfer approach todeliver recombinant proteins to the systemic circula-tion.

2. Methods

2.1. Preparation of replication-deficient adenoÕiralÕector Ad.CMV-cHK

Plasmid CMV-cHK was constructed as previouslyŽ .described Chao et al., 1996 , in which the expres-

sion of human tissue kallikrein cDNA was under thecontrol of the cytomegalovirus enhancerrpromoterand was followed by the bovine growth hormonegene polyadenylation signal sequence. The transcrip-tion unit of CMV-cHK-polyA, including the cy-tomegalovirus enhancerrpromoter, human tissuekallikrein cDNA and the bovine growth hormonegene polyadenylation signal sequence, was releasedfrom the CMV-cHK plasmid with NaeIrNruI diges-tions. Plasmid pAd.CMV-cHK was constructed byinserting the released fragment into the adenoviralshuttle vector pAdLink.1 at an EcoRV site. ThepAd.CMV-cHK plasmid DNA was purified using a

Ž .Qiagen plasmid DNA kit Qiagen, Chatsworth, CAand the purified DNA was sent to the Institute forHuman Gene Therapy, Wistar Institute, Philadelphiafor generating adenovirus Ad.CMV-cHK harboringthe CMV-cHK-polyA transcription unit. Adenovirusharboring the LacZ gene under the control of the

Žcytomegalovirus enhancerrpromoter Ad.CMV-.lacZ was obtained from the Institute for Human

Gene Therapy, Wistar Institute, Philadelphia, PA.

2.2. Direct saliÕary deliÕery of adenoÕiral ÕectorsAd.CMV-cHK and Ac.CMV-lacZ

Ž .Male Sprague-Dawley rats 9 weeks, 280 g wereŽpurchased from Sprague-Dawley Harlan Indianapo-

.lis . Rats were anesthetized by intraperitoneal injec-Ž .tion of pentobarbital 50 mgrkg body weight . The

salivary glands were exposed by a central, longitudi-nal cut at the neck. Adenoviral vector Ad.CMV-cHKŽ 94=10 pfu diluted in 200 ml of 10 mM Tris–HCl,

.pH 7.4, 0.1 mM MgCl , 10% glycerol was directly2

injected into each salivary gland using a 30-gaugeneedle. Control rats received the same amount ofadenoviral vector Ad.CMV-lacZ. At 1 to 5 days postgene delivery, salivary glands and blood were col-lected from tail vein to examine the expression ofhuman tissue kallikrein.

( )C. Wang et al.r Immunopharmacology 36 1997 221–227 223

2.3. Preparation of tissue extracts

ŽRats were anesthetized with pentobarbital 50.mgrkg body weight and blood was collected via

cardiac puncture. Rats were subsequently perfusedwith normal saline until tissues appeared to beblood-free. The salivary glands were removed,minced and homogenized in 10 mM sodium phos-phate, pH 7.4, containing 0.14 M sodium chlorideŽ . Ž .PBS Chao et al., 1990 . The salivary extract wascentrifuged at 10,000=g for 30 min and the super-natant was collected for immunoassays. Total proteinwas determined by Lowry protein assay using bovine

Ž .serum albumin as the standard Lowry et al., 1951 .

( )2.4. Enzyme-linked immunosorbent assay ELISA

For detection of immunoreactive recombinant hu-man tissue kallikrein, an ELISA was performed as

Ž .previously described Wang et al., 1995 . Briefly,human tissue kallikrein standard and samples wereadded into each well, in duplicate, of a 96-wellmicrotiter plate which was coated with non-labeledhuman tissue kallikrein IgG. After incubation at378C for 1.5 h followed by washing, biotin-labeledhuman tissue kallikrein IgG antibody was added andthe plate was incubated at 378C for 1 h.Peroxidase–avidin was added and incubated at 378C

for 1 h before substrate addition. After a 30 mincolor reaction, the plate was read using an ELISAreader.

3. Results

3.1. Preparation of adenoÕirus Ad.CMV-cHK ex-pressing human tissue kallikrein

Fig. 1 shows the schematic drawing of CMV-cHKand pAd.CMV-cHK plasmid constructs. PlasmidDNA construct CMV-cHK contains the full-length

Ž .human tissue kallikrein cDNA coding region cHKplaced in the eukaryotic expression vector pcDNA3.The expression of human tissue kallikrein is underthe control of the cytomegalovirus enhancerrpromo-ter. pAd.CMV-cHK was constructed by inserting thehuman tissue kallikrein transcription unit into theadenovirus shuttle vector pAdLink.1 at the EcoRVsite. Ad 0–1 m.u. and Ad 9–16 m.u. representadenoviral sequences required for homologous re-combination with the viral genome and pAT153represents the vector sequence. Human embryonickidney 293 cells transduced with Ad.CMV-cHK wereprovided by the Institute of Human Gene Therapy,Wistar Institute, Philadelphia. The clone expressinghuman tissue kallikrein was identified by an ELISA.

Ž .Fig. 1. Human tissue kallikrein plasmid constructs. CMV-cHK is a eukaryotic expression vector pcDNA3 containing a full-length humanŽ . Ž .tissue kallikrein cDNA coding region cHK . The cHK is under the control of a cytomegalovirus enhancerrpromoter pCMV and followed

Ž .by the bovine growth hormone gene polyadenylation signal sequence BGH pA . pAd.CMV-cHK is an adenoviral shuttle plasmidŽ . Ž .pAdLink.1 containing the human tissue kallikrein transcription unit including the cytomegalovirus enhancerrpromoter pCMV , a

Ž . Ž .full-length human tissue kallikrein cDNA coding region cHK and polyadenylation signal sequence BGH pA .

( )C. Wang et al.r Immunopharmacology 36 1997 221–227224

Fig. 2. Enzyme-linked immunosorbent assay of recombinant hu-man tissue kallikrein in rat salivary extract and sera at 1 day postgene delivery. The standard curve of human tissue kallikreinranges from 0.4 to 25.0 ngrml. Plots of serial dilutions of ratsalivary extracts and sera are parallel to the human tissue kallikreinstandard curve indicating immunological identity.

The clone which expresses a high level of humantissue kallikrein was chosen for preparing a largequantity of adenovirus.

3.2. Expression of human tissue kallikrein in ratsaliÕary glands after direct injection of Ad.CMV-cHK

Immunoreactive human tissue kallikrein levels insalivary gland extract of rats receiving human tissuekallikrein gene delivery were measured with anELISA specific for human tissue kallikrein. Im-munoreactive human tissue kallikrein detected insalivary gland extracts displayed parallel curves tothe human tissue kallikrein standard, indicating their

Ž .immunological identity Fig. 2 . No immunoreactivehuman tissue kallikrein was detected in salivary glandextracts of control rats receiving Ad.CMV-lacZ,

Ždemonstrating the specificity of the assay data not.shown . In rat salivary extracts, immunoreactive hu-

man tissue kallikrein levels reached a peak at 1 dayŽ .post gene delivery 4.47"1.07 ngrmg protein with

lower levels at 2 days, 3 days and 5 days post geneŽdelivery 1.93"0.39, 1.22"0.45 and 0.49"0.12

. Ž .ngrmg protein, respectively Fig. 3A . These re-sults suggest that direct salivary injection of theadenoviral vector carrying the human tissue kallikreinexpression unit can result in a local expression ofhuman tissue kallikrein in rat salivary glands.

3.3. Secretion of human tissue kallikrein into thesystemic circulation after intracapsular injection ofAd.CMV-cHK into saliÕary glands

Immunoreactive human tissue kallikrein levels inserum of rats receiving human tissue kallikrein genedelivery were measured with an ELISA specific forhuman tissue kallikrein. Immunoreactive human tis-sue kallikrein detected in serum displayed parallelcurves to the human tissue kallikrein standard, indi-

Ž .cating their immunological identity Fig. 2 . No im-munoreactive human tissue kallikrein was detectedin serum of control rats receiving Ad.CMV-lacZ

Fig. 3. Time-dependent expression of human tissue kallikrein inŽ . Ž .rat salivary gland A and sera B . The immunoreactive levels of

recombinant human tissue kallikrein in rats at different times afterreceiving Ad.CMV-cHK adenovirus injection were measured withan ELISA specific for human tissue kallikrein. Immunoreactivehuman tissue kallikrein was not detected in salivary extracts andsera of control rats receiving Ad.CMV-lacZ adenovirus. Data areexpressed as mean"SEM, ns3 or 4.

( )C. Wang et al.r Immunopharmacology 36 1997 221–227 225

Fig. 4. Dose-dependent expression of human tissue kallikrein insera of rats receiving Ad.CMV-cHK adenovirus injection in bothsalivary glands and a single salivary gland. The immunoreactivelevels of recombinant human tissue kallikrein in rats at thedifferent times after receiving Ad.CMV-cHK adenovirus injec-tions were measured with an ELISA specific for human tissuekallikrein. Data are expressed as mean"SEM, ns3 or 4.

Ž .data not shown . In rat serum, immunoreactive hu-man tissue kallikrein levels reached a peak at 1 day

Ž .post gene delivery 5.39"0.13 ngrml with lowerlevels at 3 days and 5 days post gene deliveryŽ .3.51"0.57 and 1.37"0.39 ngrml, respectivelyŽ .Fig. 3B . The pattern of immunoreactive humantissue kallikrein levels in rat serum was similar tothat observed in rat salivary gland extracts afterintracapsular injection of adenovirus Ad.CMV-cHK.

To demonstrate the dose-dependent expression ofhuman tissue kallikrein, immunoreactive humankallikrein levels in serum were evaluated betweenrats receiving intracapsular injections of adenovirusAd.CMV-cHK in both salivary glands and rats re-ceiving an intracapsular injection of adenovirusAd.CMV-cHK in a single salivary gland. At one daypost gene delivery, immunoreactive human kallikreinlevels in serum of rats receiving intracapsular injec-tions of adenovirus Ad.CMV-cHK in both salivary

Ž .glands 5.39"0.13 ngrml were about two-foldhigher than in rats receiving an intracapsular injec-tion of adenovirus Ad.CMV-cHK in a single salivary

Ž .gland 2.23"0.17 ngrml . Similarly, at 3 days postgene delivery, immunoreactive human tissuekallikrein levels in serum of rats receiving intracap-sular injections of adenovirus Ad.CMV-cHK in both

Ž .salivary glands 3.51"0.57 ngrml were abouttwo-fold higher than in rats receiving an intracapsu-

lar injection of adenovirus Ad.CMV-cHK in a singleŽ . Ž .salivary gland 1.64"0.31 ngrml Fig. 4 .

4. Discussion

Adenovirus has been used as an efficient vectorfor sustained in vivo expression of various genes indifferent target tissues e.g. liver, skeletal muscle and

Ž .lung Berns and Giraud, 1995 . In this study, weassessed the potential application of an adenoviralvector in salivary-targeted gene delivery. Our studydemonstrates that the adenovirus vector Ad.CMV-cHK is an efficient means to produce immunoreac-tive human tissue kallikrein in normal adult rats invivo. The immunoreactive human tissue kallikrein isa direct consequence of human tissue kallikreincDNA expression. Following a single intracapsularinjection of replication-deficient adenovirus, humantissue kallikrein levels increase dramatically with apeak at 1 day post injection and gradually decreasefrom 2 to 5 days post gene transfer in the rat salivarygland and the systemic circulation. This study pro-vides direct evidence supporting the hypothesis thatcirculating tissue kallikrein, in part, is synthesizedlocally and then secreted continuously into the sys-temic circulation in vivo.

A recent study suggests that a small amount oftissue kallikrein is translocated continuously fromsubmandibular parenchyma to the blood and suchmovement is enhanced to similar extents duringcomparable parasympathetic or sympathetic stimula-

Ž .tion Garrett et al., 1995 . The notion that salivarytissue kallikrein is continuously released into thecirculation is further supported by the finding thatsubmandibular sialoadenectomy reduces serum tissue

Ž .kallikrein levels Rabito et al., 1982 . In addition, anongoing movement of proteins including tissuekallikrein into the intraglandular lumina was ob-

Žserved in the absence of external stimulation Garrett.et al., 1996 . Collectively, it appears that there is a

constant movement of salivary tissue kallikrein intothe systemic circulation and the movement is regu-lated by parasympathetic or sympathetic impulses.Consistent with these findings, this study demon-strates that the salivary gland-targeted gene deliveryof tissue kallikrein results in the presence of humantissue kallikrein in the systemic circulation. The

( )C. Wang et al.r Immunopharmacology 36 1997 221–227226

molecular mechanism of tissue kallikrein secretionawaits further investigations.

Exposure of salivary glands in vivo to adenoviruscarrying an alpha 1-antitrypsin expression cassetteresulted in expression of a1-antitrypsin in the sali-

Žvary gland and secretion in the saliva Mastrangeli et.al., 1994 . Likewise, immunoreactive human tissue

kallikrein was detected in the saliva after intracapsu-Ž .lar injection of Ad.CMV-cHK data not shown .

These results indicate that recombinant gene con-structs can be effectively delivered into the salivarygland and the recombinant protein is being synthe-sized and secreted through both endocrine and ex-ocrine pathways.

Although direct salivary injection of adenovirusprovides a simple and effective approach for salivarygland-targeted gene delivery, this expression of aforeign gene is short-lasting. The immunoreactivehuman tissue kallikrein levels decreased with timebecause this vector is not integrated in the genomeŽ .Stratford-Perricaudet et al., 1990 . However, pro-longed expression is possible under certain condi-tions. A recent study showed that a single injectionof adenoviral vector to the left ventricular my-ocardium of athymic nude rats resulted in a sustainedgene expression for at least 120 days without signifi-

Ž .cant inflammatory reactions Quinones et al., 1996 .It is concluded that direct injection of replication-

deficient adenovirus could reliably deliver genes intothe salivary gland and circulation. The movement ofhuman tissue kallikrein synthesized in the salivarygland to the systemic circulation provides directevidence that tissue kallikrein in the circulation issupplied, at least in part, by the salivary gland.Whether the amount of human tissue kallikrein in thecirculation is sufficient to have any systemic effect isnot clear. Previous studies have demonstrated that exvivo and in vivo gene transfer approaches result in adelivery of secreted proteins to the systemic circula-

Ž .tion Barr and Leiden, 1991; Tripathy et al., 1994 .In this study, the use of adenoviral vectors has beenshown to be a powerful new approach to transfergenes into the salivary gland to study salivary glandfunction and development. The salivary gland-targeted gene delivery technology should have amajor impact on the study of salivary gland geneexpression, as well as provide a means to expresstherapeutic proteins in the salivary gland.

Acknowledgements

This work was supported by National Institute ofHealth Grants HL 29397 and HL 56686.

References

Barr, E., Leiden, J.M., 1991. Systemic delivery of recombinantproteins by genetically modified myoblasts. Science 254,1507–1509.

Berns, K.I., Giraud, C., 1995. Adenovirus and adeno-associatedvirus as vectors for gene therapy. Ann. N.Y. Acad. Sci. 772,95–104.

Berg, G., Johansen, L., Nustad, K., 1985. Enzymatic activity ofrat submandibular gland kallikrein released into blood. Am. J.Physiol. 249, H1134–H1142.

Bhoola, K.D., Figueroa, C.D., Worthy, K., 1992. Bioregulation ofkinins: Kallikreins, kininogens and kininases. Pharmacol. Rev.44, 1–80.

Chao, J., Chai, K.X., Chen, L.M., Xiong, W., Chao, S.,Woodley-Miller, C., Wang, L., Lu, H.S., Chao, L., 1990.Tissue kallikrein-binding protein is a serpin. I. Purification,characterization and distribution in normotensive and sponta-neously hypertensive rats. J. Biol. Chem. 265, 16394–16401.

Chao, J., Jin, L., Chen, L.M., Chen, V.C., Chao, L., 1996.Systemic and portal vein delivery of human kallikrein genereduced blood pressure in hypertensive rats. Hum. Gene Ther.7, 901–911.

Crystal, R.G., 1995. Transfer of genes to humans: Early lessonsand obstacles to success. Science 270, 404–410.

Figueroa, C.D., Maciver, A.G., Bhoola, K.D., 1989. Identificationof a tissue kallikrein in human polymorphonuclear leukocyte.Br. J. Haematol. 72, 321–328.

Garrett, J.R., Chao, J., Proctor, G.B., Wang, C., Zhang, X.S.,Chan, K.M., Shori, D.K., 1995. Influences of secretory activi-ties in rat submandibular glands on tissue kallikrein circulatingin the blood. Exp. Physiol. 80, 429–440.

Garrett, J.R., Zhang, X.S., Proctor, G.B., Anderson, L.C., Shori,D.K., 1996. Apical secretion of rat submandibular tissuekallikrein continues in the absence of external stimulation:Evidence for a constitutive secretory pathway. Acta Physiol.Scand. 157, 299–304.

Lowry, O.H., Rosebrough, N.J., Farr, A.L., Randall, R.J., 1951.Protein measurement with the folin phenol reagent. J. Biol.Chem. 193, 265–275.

Margolius, H.S., 1995. Kallikreins and kinins. Some unansweredquestions about system characteristics and roles in humandisease. Hypertension 26, 221–229.

Mastrangeli, A., O’Connell, B., Aladib, W., Fox, P.C., Baum,B.J., Crystal, R.G., 1994. Direct in vivo adenovirus-mediatedgene transfer to salivary glands. Am. J. Physiol. 266, G1146–G1155.

Nolly, H., Lama, M.C., 1982. Vascular kallikrein: A kallikrein-likeenzyme present in vascular tissue of rat. Clin. Sci. 63, 249–251.

Nolly, H., Scicli, A., Scicli, G., Carretero, O.A., 1985. Characteri-

( )C. Wang et al.r Immunopharmacology 36 1997 221–227 227

zation of a kininogenase from rat vascular tissue resemblingtissue kallikrein. Circ. Res. 56, 816–821.

O’Connell, B.C., Tenhagen, K.G., Lazowski, K.W., Tabak, L.A.,Baum, B.J., 1995. Facilitated DNA transfer to rat submandib-ular gland in vivo and GRP-CA gene regulation. Am. J.Physiol. 31, G1074–G1078.

Quinones, M.J., Leor, J., Kloner, R.A., Ito, M., Patterson, M.,Witke, W.F., Kedes, L., 1996. Avoidance of immune responseprolongs expression of genes delivered to the adult rat my-ocardium by replication-defective adenovirus. Circulation 94,1394–1401.

Rabito, S.F., Scicli, A.G., Carretero, O.A., 1982. Immunoreactiveglandular kallikrein in rat plasma: A radioimmunoassay for itsdetermination. Am. J. Physiol. 242, H602–H611.

Roblero, J., Croxatto, H.R., Garcia, R., Corthorn, J., De Vito, E.,1976. Kallikrein-like activity in perfusates and urine of iso-lated rat kidneys. Am. J. Physiol. 231, 1383–1389.

Stratford-Perricaudet, L.D., Leverro, M., Chasse, J., Perricaudet,M., Briand, P., 1990. Evaluation of the transfer and expressionin mice of an enzyme-encoding gene using an adenovirusvector. Hum. Gene Ther. 1, 241–256.

Tripathy, S.K., Goldwasser, E., Lu, M.M., Barr, E., Leiden, J.,1994. Stable delivery of physiologic levels of recombinanterythropoietin to the systemic ciuculation by intramuscularinjection of replication-defective adenovirus. Proc. Natl. Acad.Sci. USA 91, 11557–11561.

Vio, C.P., Churchill, I., Rabito, S.F., Terragno, A., Carretero,O.A., Terragno, N.A., 1983. Renal kallikrein in venous efflu-ent of filtering and non filtering isolated kidney. Adv. Exp.Med. Biol. 156B, 897–905.

Wang, C., Chao, L., Chao, J., 1995. Direct gene delivery ofhuman tissue kallikrein reduces blood pressure in sponta-neously hypertensive rats. J. Clin. Invest. 95, 1710–1716.