Expression of Human Glandular Kallikrein, Hk2, in Mammalian Cells

[CANCERRESEARCH56.5397-5402.December1,1996] Expression of Human Glandular Kallikrein, hK2, in Mammalian Cells

Abhay Kumar, Amita S. Goel, Timothy M. Hill, Stephen D. Mikolajczyk, Lisa S. Millar, Kristine Kuus-Reichel, and MohammadS. Saedi' Hybritech Incorporated, San Diego. California 92196-9006

ABSTRACT regulated by androgens (12â€”14),and share 78% an sequence homol ogy (6, 15). Because of similarities between hK2 and PSA molecules, The human kallikrein family consists of three members, hKl, hK2, and it is speculated that measuring serum concentrations of hK2 may hK3 [prostate-specific antigen (PSA)J. PSA is a widely accepted marker enhance or complement the sensitivity of the PSA test for screening, for prostate cancer. The mRNAs for both hK2 and PSA are localized staging, and monitoring of pCa and may help distinguish between pCa predominantly to prostate epithelium and are regulated by androgens. In addition, hK2 has 78% amIno acid homology to PSA. Although similari and BPH. PSA is a chymotrypsin-like protease (16â€”19), whereas ties to PSA make hK2 a potential prostate cancer marker, they also protein modeling studies suggest that hK2 is a trypsin-like protease impede biochemical characterization of hK2 in those human tissues and (20, 2 1). This implies that the physiological role of hK2 in prostate is body fluids where PSA is abundant. To study the expression, biosynthesis, different from PSA. The lack of knowledge regarding the biosynthesis and processing ofhK2, recombinant hK2 was expressed in the adenovirus and protein processing of hK2 has impeded the understanding of its Induced Syrian hamster tumor cell line AV12-664 (AV12-hK2). Expres physiological role in the normal and cancerous prostate gland. sion of hK2 was analyzed by Western blots and ELISA using monoclonal Significant progress has been made in understanding the biology antibodies HK1G 464.3 [specific for prohK2 (phK2)] and HK1D 106.4 and biochemistry of PSA. PSA is postulated to liquefy semenogelin [specific for phK2 and mature hK2 (hK2)J. Western blot and ELISA and enhance male fertility (16, 17, 22). In contrast, very little is known analyses showed that phK2 was secreted into the media by AV12-hK2 cells on day 1 and was gradually converted to the mature form of hK2 by day about hK2 biology. Recently, hK2 was identified in the prostate 7. N-terminal amino acid sequencingverifiedthe Westernblot and ELISA cytosol and seminal plasma where it was uncornplexed or complexed data, This demonstrates for the first time that hK2 is secreted as phK2 and with PCI (14, 23); however, biological and biochemical properties converted to hK2 extracellularly. In addition, hK2 detected in day 4â€”7 have not been reported. AV12-hK2-spent media was enzymatlcally active. Recombinant hK2 was To study the biochemical and biological properties of hK2, the also expressed in human prostate carcinoma cell lines, PC3 (PC3-hK2) protein needs to be overexpressed in a heterologous system. This and DU14S (DU145-hK2), that do not express endogenous hK2 or PSA. approach provides a reliable source for hK2 that is devoid of PSA. We Shnilar to AV12-hK2 cells, both cell lines secreted phK2 that was con have previously reported expression of recombinant pphK2 in Esch verted to hK2 extracellularly. phK2 was the major form detected in the erichia coli that proved to be an invaluable immunogen (24). How spent media of PC3-hK2 cells, even after 7 days, indicating a slow con ever, to study biosynthesis, protein processing, secretion, and activity, version of phK2 to hK2. hK2 was the predominant form detected in the it is imperative to express hK2 in mammalian cells. Lovgren et al. spent media of DU145-hK2 starting on day 1, indicating the rapid con version of phK2 to hK2. In this study, we demonstrate that hK2 exists in (25) reported transient, low-level expression of hK2 in BHK-2 1 cells different forms and is secreted as phK2. phK2 is then converted to using the Semiliki Forest virus system to study cross-reactivity of enzymaticaily active hK2 extracellularly. antibodies used in immunoassays for free PSA and PSA. However, this system did not provide a stable source of hK2 which is necessary for studying the biosynthetic processing of the protein. In studies INTRODUCTION presented here, hK2 was cloned and stably expressed in Syrian ham pC a2 is the most commonly diagnosed malignancy in American ster tumor cells, AV 12, and human prostate carcinoma cell lines PC3 males, with an estimated 244,000 new cases and 40,400 deaths ex and DU14S. We demonstrate for the first time that hK2 is secreted pected in 1995, accounting for 36% and 14% of total new cases and into the spent media by mammalian cells as phK2 and converted to deaths of all cancer types, respectively (1). PSA has been widely enzymatically active hK2 extracellularly. These observations establish recognized as a reliable prognostic marker for this disease (2â€”4). the biosynthetic processing of hK2 and suggest that phK2 may exist Current PSA tests, however, lack the specificity to distinguish be in biological fluids and could be a useful marker for prostate cancer tween pCa and BPH, especially when serum PSA value is only and/or BPH. slighfly elevated (5). hK2 and PSA belong to the family of human glandular kallikreins. The kallikreins are a multigene family of serine MATERIALS AND METHODS proteases. Three kallikreins have been identified in humans, desig nated hKl, hK2, and hK3 (PSA; Refs. 5â€”8).Their genes have been Expression Vectors. A 0.8-kbDNA fragmentcodingforentirepphK2was mapped to chromosome 19 (9). hKl pancreatic/renal glandular kal PCR amplified using primers A (5'-ATATGGATCCATATGTCAGCATGT GGGACCTGGUCTCTCCA), primer B (5'-ATATGGATCCTCAGGGGTF likrein is produced primarily in the kidney, pancreas, and submandib GGCTGCGATGGT), and plasmid pVL1393 containing pphK2 cDNA (gift ular salivary gland (10). hK2 and PSA share many characteristics. from Dr. Charles Young, Mayo Clinic, Rochester, MN) as the template. The Both are primarily produced by the prostate epithelium (7, 11), are PCR products were cloned into the TA cloning vector (Invitrogen Corp., San Diego, CA) to obtain TA-hK2 and sequenced to ensure the fidelity of PCR. Received 5/13/96; accepted 10/16/96. Fig. I shows the hK2 expression vectors used in the study. pGTD-hK2 was Thecostsof publicationofthisarticleweredefrayedinpartby thepaymentofpage constructed by cloning the 0.8-kb hK2 insert from the TA-hK2 clone into the charges. This article must therefore be hereby marked advertisement in accordance with Bc!! site of pGT-d (gift from Dr. Brian Grinnell, Eli Lilly and Co., Imdianap 18 U.S.C. Section 1734 solely to indicate this fact. I To whom requests for reprints should be addressed, at Hybritech Incorporated. Box ohs, IN) under the control of the GBMT promoter (26). Other hK2 expression 269006, San Diego, CA 92196-9006. Phone: (619) 535-8706; Fax: (619) 457-5308; vectors, pLNS-hK2 and pLNC-hK2, were obtained by cloning the 0.8-kb insert E-mail: [email protected]. from the TA-hK2 clone into the HindIII-ClaI sites of pLNSX under the control 2 â€˜@â€˜@abbreviations used are: pCa, prostate cancer; BPH, benign prostatic hyperplasia; of SV4O promoter and HindIH-HpaI sites of pLNCX under the control of pphK2, preprohK2; phK2, prohK2; hK2, mature hK2; PSA, prostate-specific antigen; mAb, monoclonal antibody; aa, amino acid; pNA, p-nitroanilide; ACT, al-antichymo cytomegalovirus promoter, respectively. pLNSX and pLNCX were received trypsin; PCI, protein C inhibitor; IGF, insulin-like growth factor. from Dr. A. Dusty Miller (Fred Hutchinson Cancer Center, Seattle, WA; Ref. 5397

pGTD-hK2 bodies (1 : 1000 dilution of ascites) and secondary antibodies (goat anti-mouse horseradish peroxidase, 1:500; Jackson Immumosearch Laboratories, Inc., West Grove, PA) were used to probe the blots. The immunoreactive signals were detected using the enhanced chemilumimescence(Amersham, Bucking Bcll Bcll hamshire, England) system according to the manufacturer's instructions. To measure hK2 by ELISA, microtiter plates (Becton Dickinson Labware, Lincoln Park, NJ) were coated with 50 pJ of spent media overnight at 37Â°C. pLNS-hK2 The wells were washed with PBS + 0.1% Tweem20, blocked with 2% BSA in PBS, and incubated for 1 h at 37Â°C with 50 pJ of HK1D 106.4 or HK1G @ L@â€”M@NEO I SV4d4@ 4@LTR1 464.3. The wells were washed again with PBS + 0.1% Tween 20 and incubated for 1 h at 37Â°C with 50 pA of goat anti-mouse IgG Fcy antibodies Hind III Cla I coupled to horseradish peroxidase (1:500; Jackson Immunosearch Laborato ries). The wells were washed with PBS + 0.1% Tween 20 again, incubated pLNC-hK2 with o-phemylenediamine dihydrochloride (Sigma) for 15 mm, and quenched with 4 N sulfuric acid. The colonmetric reaction was measured at A490 with an ELISA reader (model EL31O;Biotek Instruments, Inc., Burlington, VT). All samples were assayed in triplicate. Spent media from AVI2 cells transfected with vector alone were used as negative control. Purified phK2'Â°'217â€•was used Hind III Hpa I to construct the standard curve. Fig. 1. Schematic representation of hK2 expression vectors. Arrows, transcription start Assay for the Measurement of hK2 Activity. Spent media from AV12- sites. hK2 cells was incubated with 1 mM pNA-derivatized peptide substrates (H- D-Pro-Phe-Arg-pNA, 5-2302; Pharmacia Hepar, Inc., Franklin, OH) in 200 mM Tris/5 mM EDTA (pH 8.0) at 37Â°C. The enzymatic activity of hK2 was 27). In all of the mammalian expression vectors, the orientation of the insert determined by hydrolysis of the chromogenic substrate, leading to an increase was confirmed by DNA sequencing. in absorbance at 405 mm. Cell Lines and Transfection. AV12-664 (ATCCCRL9595) and DU145 cells were cultured in DMEM (high glucose) containing 10% fetal clone RESULTS (Hyclone, Logan, UT). PC3 cells were cultured in minimal Eagle's medium containing 10% fetal clone. AVI2-664, DUI45, and PC3 cell lines were The cDNA for hK2 was cloned into pGT-d vector under the GBMT transfected, respectively, with pGTD-hK2, pLNS-hK2, and pLNC-hK2 cx promoter. This promoter is activated by the Ela tumor antigen. pression vectors using LipofectAMlNE (Life Technologies, Inc., Gaithersburg, AV12-664 cells are derived from an adenovirus type 12-induced MD) according to the manufacturer's instructions. Transfected AV12-664cells (AV12-hK2) were selected in 200 nst methotrexate (Sigma Chemical Corn tumor in Syrian hamster and constitutively express Ela tumor antigen. pany, St. Louis, MO). Transfected DUI45 (DU145-hK2) and PC3 (PC3-hK2) Thus, when pGT-d vector is transfected into the AV12 cells, the cells were selected in 400 p@g/ml G418 (Life Technologies, Inc.). Single-cell GBMT promoter is activated by the Ela antigen, leading to the clones were isolated for the study. Viability of cells was assessed by trypan overexpression of the protein cloned under this promoter. This host blue dye exclusion. vector system has been used for expression of other serine proteases Antibodies. Murine mAbs HK1D 106.4 (against a peptide corresponding such as human protein C and tissue plasminogen activator (26). To to 17â€”71 aa of hK2), HK1G 464.3 and HK1G 586.1 (against phK2'@'217â€•, study the expression of hK2 in mammalian cells, AV12 cells were prohK2 with an Ala to Val mutation at aa 2l7)@were obtained using standard transfected with pGTD-hK2 expression vector. Cells were selected in protocols of immunization and hybridoma production (28). ELISA and West 200 nrvimethotrexate for 2 to 3 weeks, and single-cell clones were ens blot analyses showed that HK1D 106.4and HK1G 586.1 mAbs recognized analyzed for hK2 expression using an ELISA. Spent media from five both phK2 and hK2, whereas HKIG 464.3 mAb recognized only phK2 and not the mature form of hK2 (see Fig. 2). randomly selected hK2 positive clones were analyzed with Western Purification of hK2. mAb HK1G 586.1 was coupled with GammaBind blot. A hK2-immunoreactive band migrating at â€”34kDa was detected Plus Sepharose (Pharmacia Biotech Inc., Piscataway, NJ) according to the in all of the clones (data not shown). AV12-hK2#27 was selected for published procedure (29). Tissue culture Petri dishes containing various clones further characterization and purification. grown to approximately 60-70% cell confluency were washed with HBSS To determine the identity of the protein that was secreted at day 1, (Life Technologies, Inc.) and serum-free HH4 medium (JRH Biosciences, spent media from AV12-hK2#27 was collected, concentrated, and Lenexa, KS) was added. Spent media was concentrated (YM- 10 membrane; affinity purified using HK1G 586.1 resin. N-terminal sequencing Amicon Inc., Beverly, MA) and incubated with the above resin at 4Â°Cfor 48 h analysis of the purified protein revealed a sequence: VPLIQSRIVG with gentle agitation. The resin was spun down and washed three times with GWEXEK. No competing sequence was evident from the profile of aa PBS and bound proteins were eluted with 1% SDS. The eluent was subjected released sequentially by the Edman degradation procedure. This se to SDS-PAGE on a 4â€”20% polyacrylamide gel and electroblotted to a poly quence corresponds with the sequence predicted for the NH2 terminus vinylidene difluoride membrane. The polyvinylidene difluoride membrane was stained with Coomassie blue, and the band corresponding to the appropriate of phK2. A similar purification protocol was used to purify the protein molecular weight was excised and subjected to protein sequencing using an from day 7 spent media of AV12-hK2#27 cells. N-terminal sequenc Applied Biosystems model 477A protein sequencer. phK2A2hl@'(ph}(@with ing analysis of the purified protein revealed a sequence, IVGG Ala to Val mutation at an 217) was purified using column chromatography and WEXEK, that corresponds with the sequence predicted for the NH2 converted to hK2@A@2hlvbymild trypsin treatment.3 terminus of hK2. Western Blot and ELISA Analyses. Spent media (prepared as described To study the biosynthesis of hK2 in mammalian cells, spent media in â€œPurification of hK2â€•) from resistant clones were collected at specified days, from AV12-hK2#27 was collected for 8 consecutive days and eval concentrated 10-fold using Centricon 10 (Amicon, Inc.), and subjected to uated for the presence of hK2. The samples were concentrated and SDS-PAGE on a 4â€”20%gel(Bio-Rad, Inc., Melville, NY). After electrophore subjected to SDS-PAGE. The proteins were transferred to nitrocellu sis, proteins were electroblotted onto nitrocellulose membranes. Primary anti lose membrane and probed with either HK1D 106.4 or HK1G 464.3 mAbs (Fig. 2). As also shown in Fig. 2, HK1D 106.4 recognizes both 3 S. D. Mikolajczyk, L. S. Millar, A. Kumar, and M. S. Saedi. Human glandular kallikrein (hK2) shows arginine-restricted specificity and forms complexes with plasma phK2 and hK2, whereas HK1G 464.3 recognizes only phK2 since its protease inhibitors, submitted for publication. epitope lies in â€”7to +7 region of hK2 protein (data not shown). 5398

@ <,b'\ ,@, â€˜@, b'\ @ Ã§@Â°\ Ã§@@\ a I 11111111 bii ii@iiii

@ â€” â€” â€” 98kD â€” 64kD â€” 5OkD @ 0â€¢ â€”.@ 36kD 3OkD I6kD

Primary antibody Primary antibody HKID 106.4 HKIG 464.3 (phK2 + hK2) (phK2) Fig. 2. Western blot analysis of AV I2-hK2#27 spent media. At â€”60-70% confluency, AV 12-hK2#27 cells were washed with HBSS, and serum-free HH4 media was added. Up to 8 days, spent media were withdrawn each day, concentrated, and subjected to SDS-PAGE on a 12% gel. Proteins were electroblotted and probed with HK1D 106.4 (a) and HK1G 464.3 (b) mAbs as described in â€œMaterialsandMethods.â€•phK2A2I7V(f,()ng) and hK2@@2I7v(f@()ng)were used as controls. Arrow, position of hK2. Right, molecular weight markers. kD, kilodaltons.

Expression of hK2 (â€”34kDa) peaked by day 3 and plateaued there To demonstrate that hK2 detected in the spent media of AV12- after as detected by HK1D 106.4 mAb (Fig. 2). Two other immuno hK2#27 is enzymatically active, a spectrophotometric assay was de reactive bands migrating at â€”70kDa and â€”90kDa were also detected veloped using the commercially available chromogenic substrate, from day 4 onward. On the other hand, when the same samples were H-D-Pro-Phe-Arg-pNA (5-2302). Data shown in Fig. 3 indicate no blotted and probed with HK1G 464.3, a gradual reduction in the level hK2 activity in the spent media of AV12-hK2#27 cells at day 1 or day of phK2 expression was detected starting at day 4 (Fig. 2). By day 8, 3. However, hK2 activity was detected in the spent media at day 4 and very low levels of phK2 were found in the spent media. This result increased until day 7. Since phK2 is enzymatically inactive, the above shows that the pro form of hK2 is being secreted into the media by results support our previous observations and again indicate that phK2 AVI2-hK2#27 and is gradually converted to mature form of hK2 is secreted by the AV12-hK2#27 cells and is gradually converted to extracellularly. We did not observe â€”70-and â€”90-kDa bands with HK1G 464.3 mAb, indicating that these bands are either homooli gomers of hK2 or hK2 covalently complexed with yet unknown protein(s). Although the identity of these bands is not known at this time, they serve as markers for the presence of hK2 in the spent media. Purified phK2'@217â€•(phK2with Ala to Val mutation at aa 217) was used as the positive control in this experiment. This protein is stable to purification and can be converted from phK2â€•@'217â€•to hK2A2hl@'by trypsin digestion.3 To study the relationship between extracellular conversion of phK2 to hK2 and viability of AV12-hK2#27 cells in culture, cell viability was determined by trypan blue exclusion and correlated with hK2 expression. Expression of phK2 and hK2 in the spent media was measured with an ELISA using both HK1D 106.4 and HK1G 464.3 mAbs. As shown in Fig. 3 (inset), the number of viable cells peaked at 3.8 X iO@in culture by day 3 and gradually decreased thereafter. By ,< day 8, the number of viable cells was reduced to <1.0 X l0@. The LI) b expression of both phK2 (measured by HK1G 464.3) and hK2 (meas ured by HK1D 106.4) also peaked by day 3. However, the expression .@ of phK2 declined thereafter, whereas the expression of hK2 plateaued ..fl U after day 4. This result indicates that phK2 is secreted by AV1 2- hK2#27 and a fraction of it is converted extracellularly to hK2 by day 4. Moreover, it shows that conversion of phK2 to hK2 correlates with a decrease in cell viability, suggesting that intracellular proteases released by the dying cells are responsible for this conversion. Ex pression of phK2 was highest when cells were most viable (day 3). A decrease in phK2 paralleled a decline in cell viability, suggesting that w phK2 is secreted by these cells, as opposed to being released follow Days in culture ing cell death and lysis. In addition, the rise in hK2 corresponded to the drop in phK2, indicating that the pro form of hK2 was being Fig. 3. Appearance of enzymatically active hK2 correlates with decreasing viability of AV 12 cells. At â€”60-70% confluency, AV 12-hK2#27 cells were washed with HBSS, and converted to the mature form of hK2 over time. The reason for â€”50% serum-free HH4 media were added. Up to 8 days. spent media were withdrawn each day decline in total hK2 concentration from day 3 to day 4 is not under and the hK2 concentration was measured by an ELISA using HK1D 106.4 (0) and HK1G 464.3 (0) mAbs. Enzymatic activity (A) was measured on chromogenic substrates as stood at present. It is possible that the sudden release of unstable absorbance units/mm. Inset, viable cells () were counted each day using the trypan blue proteases from dying cells degrade hK2 in the culture media. dye exclusion procedure. 5399

123456 analyzed for hK2 expression using Western blotting. Spent media of PC3-hK2 were collected at day 7, concentrated, subjected to SDS PAGE, and electroblotted onto nitrocellulose membranes. Blots were j I I I I I probed with HK1D 106.4 and HK1G 464.3 mAbs (Fig. 5). Both HK1D 106.4 and HK1G 464.3 mAbs detected a major band of â€”34 kDa in the spent media of PC3-hK2 cells, indicating that phK2 is 98kD present in the spent media of this cell line even after 7 days. To 64kD confirm these observations, the â€”34-kDa immunoreactive band was \ 5OkD partially purified from day 7 spent media of PC3-hK2 cells by affinity 36kD 3OkD purification. N-terminal analysis revealed a sequence of VPLIQS I6kD RIVGGWEXEK, verifying it to be phK2. Spent media of DU14S-hK2 was collected at day 1 and analyzed as described above (Fig. 5). HK1D 106.4 detected a stronger band of â€”34kDa compared to Primary antibody HK1G 464.3. This result indicated that predominantly hK2 is present HKID 106.4 in DU145-hK2 spent media even at day 1. Western blot analysis of (phK2 + hK2) PC3-hK2 and DU14S-hK2 cell lysates revealed negligible levels of

Fig. 4. Coincubation of serum and cell-free AV12-hK2#27 spent medium with those of either phK2 or hK2 (data not shown). A time course experiment AV 12-pGTD and AV 12 cells. Approximately 1 ml of spent media (day 1) of AV 12- similar to the one described for AV12-hK2#27 (Fig. 2) was also hK2#27 was coincubated with an equal quantity of either spent media (day 7) of AV12 or conducted for PC3-hK2 and DU14S-hK2 cells. The Western blot spent media (day 5) of AV l2-pGTD cells at 37Â°Cfor48 h. The media were concentrated and subjected to SDS-PAGE on a 12% gel. Proteins were electrobloued and probed with patterns did not change significantly from day 1 to day 7, indicating HK1D 106.4.Lanes I and 2, day 7 and day 1 spentmediaof AV12-hK2#27cells, that predominantly phK2 was present in PC3-hK2 spent media even respectively; J..ine 3, spent media (day 7) of AV12 cells; Lane 4, spent media (day 5) of after 7 days, whereas predominantly hK2 was present in DU145-hK2 AV I2-pGTD; Lane 5, spent media (day I) of AV I2-hK2#27 cells coincubated with spent media (day 7) of AVI2 cells; Lane 6, spent media (day I) of AVI2-hK2#27 cells spent media starting from day 1 (data not shown). coincubated with spent media (day 5) of AVI2-pGTD cells. Right, molecular weight markers. kD, kilodaltons. DISCUSSION

hK2 starting from day 3. Day 7 spent media of AV12 or AV12-pGTD Serine proteases are produced as prepro proteins inside the cell. cells showed no detectable activity in this assay, indicating that After passage through the secretory pathway, the signal peptide (pre) hK2-like proteases are not secreted by AV12 cells (data not shown). is cleaved, yielding the pro form of the protein. Although pro peptides Moreover, in the process of hK2 purification (see above), removal of are generally cleaved inside the cell (e.g. , tissue plasminogen activa hK2 from day 7 spent media of AV I2-hK2#27 by passage through the tor, protein C, and tumor necrosis factor), there are exceptions (e.g., HK1G 586.1 immunoaffimity column also removed â€”90% of the renin, trypsin, and chymotrypsin) which are secreted as pro proteins enzymatic activity detected in the spent media. These results indicate and cleaved extracellularly (reviewed in Refs. 30â€”32).The following that the enzymatic activity in the spent media of AV12-hK2#27 as four lines of evidence presented in this article demonstrate that hK2 is detected in our assay was specific to hK2 and not due to other expressed as prohK2 in mammalian cells and is converted to the proteases released from dying cells. mature form extracellularly: (a) presence of phK2 at day 1 and hK2 at To ascertain whether phK2 is converted to hK2 extracellularly, an day 7 as determined by Western blot and enzymatic analyses of spent equal volume of day 1 spent media of AV12-hK2#27 cells was media of AVI2-hK2#27 cells; (b) purification and sequencing of incubated with that of either AV12 (day 7 spent media) or AV12- pGTD (day 7 spent media) for 48 h. The mixtures were then concen trated, electrophoresed, and electroblotted as described previously. @ The blot was probed with HK1D 106.4 mAb (Fig. 4). As is clearly S/@,p visible in Lanes 5 and 6, the â€”70-kDahK2 immunoreactive band was ,@y, present in AV12-hK2#27 supernatants upon incubation with either spent media of AV12 or AV12-pGTD cells. When day 1 spent media allil bull of AVI2-hK2#27 was incubated alone or with plain media, the â€”70-kDa band was not present. Appearance of an â€”70-kDa immu â€” 64kD noreactive band indicates that a fraction of phK2 was converted to â€” 5OkD @ hK2. This suggests that intracellular proteases released into the media ______â€” 36kD @ by the lysed AV12 cells may be responsible for the initial conversion â€” J@ 0' â€¢ __ 3OkD of phK2 to hK2. â€” I6kD Three other independently isolated AV12-hK2 clones were exam med with Western blot (using HK1D 106.4 and HK1G 464.3 mAb) â€” 4kD for expression of phK2 and hK2. Results showed that the patterns of phK2 and hK2 expression were the same in all clones, indicating that Primary antibody Primary antibody the above results were not limited to a single AV12-hK2 clone (data HKID 106.4 HKIG 464.3 not shown). (phK2 + hK2) (phK2) To study the expression of hK2 in human prostate carcinoma cell lines, pLNC-hK2 (hK2 under the control of the cytomegalovirus Fig. 5. Expression of hK2 in PC3-hK2 and DU145-hK2 cells. At 60â€”70%confluency, cells were washed once with PBS and complete media were replaced with serum-free promoter) was transfected into PC3 cells, and pLNS-hK2 (hK2 under HH4 media. Serum-free media were collected after 7 days for PC3-hK2 and after 1 day the control of the SV4O promoter) was transfected into DU145 cells for DU145-hK2 cells, concentrated, and subjected to SDS-PAGE on a 12% gel. Proteins were electroblotted and probed with HK1D 106.4 (a) and HK1G 464.3 (b) mAbs. (Fig. 1). PC3-hK2 and DU145-hK2 cells were selected in G418 (400 PhK2A2t7v(@ mg)and @.v@A217V(@Jmg)were used as controls. Arrow, position of hK2. @g/ml)-containing media for 2 to 3 weeks, and resistant clones were Right, molecular weight markers. kD, kilodaltoms. 5400

Downloaded from cancerres.aacrjournals.org on September 24, 2021. © 1996 American Association for Cancer Research. MAMMALIAN EXPRESSION OF hK2 phK2 from day 1 spent media and hK2 from day 7 spent media of other proteins. hK2 has been shown to complex with ACT3 and PCI AV12-hK2#27 cells; (c) detection of gradually increasing hK2 activ (23). In an attempt to identify â€”70-and 90-kDa bands, samples of the ity in spent media of AV12-hK2#27 cells from day 4 onward; and (d) spent media were probed with anti-ACT or anti-PCI antibodies on similar expression patterns of recombinant hK2 in Syrian hamster Western blots. The results indicated that â€”70-and 90-kDa bands were tumor cells (AV12) and human prostate carcinoma cell lines, PC3 and not complexes of hK2 with these ACT or PCI (data not shown). DU14S. The evidence that phK2 is secreted in the spent media of two Although these bands are represented as a minor fraction in the spent human prostate carcinoma cell lines (DU145 and PC3) indicates that media of AV12-hK2#27 cells, they nonetheless remain a useful phK2 may be present in biological fluids and thus may be a relevant marker for the presence of hK2. marker for screening, staging, and monitoring progression of the A considerable amount of work has been done to elucidate the prostatediseases. physiological role of PSA. PSA is a chymotrypsin-like enzyme and phK2 is converted to hK2 by protease cleavage and release of a the native substrates for PSA are semenogelins. One function for PSA short propeptide (7 an) ending in arginine. The presence of phK2 in in seminal fluid is to dissolve the seminal clot, thereby increasing the the day 1 spent media of AV12-hK2#27 cells at first indicated that sperm motility (16). Enzymatic activity of PSA may also be related to AV12 cells may not have the enzymatic machinery to convert phK2 proliferation, migration, and metastasis of prostatic cancer cells. For to hK2. However, when the cells were left in culture for longer example, PSA has been shown to hydrolyze interleukin 2, fibrinogen, periods of time, hK2 was detected. The appearance of hK2 correlated gelatin, insulins A and B, myoglobin, and ovalbumin in vitro (34). with a decrease in viability of the AV12-hK2#27 cells. This indicated PSA can proteolytically modulate activities of transforming growth that although proteases responsible for conversion of phK2 to hK2 are factor f3 and cell surface receptors. Recently, PSA has been reported present in the cells, they may be compartmentalized away from phK2. to cleave the IGF-binding protein 3, resulting in increased IGF activ Upon cell lysis, such proteases become accessible and may be respon ity, a function which may be important for the regulation of epithelial sible for the initial conversion of phK2 to hK2. This observation was tissue growth in the prostate (35). There is no report in the literature further supported by the conversion of phK2 in day 1 spent media of indicating the possible physiological role of hK2. As predicted by AV12-hK2#27 cells to hK2 upon coincubation with day 7 spent media molecular modeling studies, we have shown hK2 to have trypsin-like ofeither AV12 or AV12-pGTD cells (Fig. 4). Based on reactivity with activity, suggesting hK2 has a different physiological role than PSA. the HK1G 464.3 mAb, it seems that a major fraction of phK2 was still The reagents developed in this study will be useful in deciphering the left in these samples (data not shown), suggesting that proteases role of hK2 in the normal prostate and in prostate cancer. responsible for initial conversion of phK2 to hK2 are not stable in the The above results collectively suggest that hK2 is expressed as the extracellular environment under the experimental conditions used pro form in mammalian cells and is converted to the enzymatically here. It is interesting to note that day 7 spent media of AV12 or active mature form extracellularly by as of yet an unknown mecha AV12-pGTD cells had no detectable activity on chromogenic sub nism. These results also suggest that prohK2 may be present in the strate (data not shown), indicating that AV12 cells lack any hK2-like biological fluids and therefore could be a useful prognostic marker for proteases. It is speculated that the process of phK2 to hK2 conversion the prostate diseases. The cell lines described in this report constitute is initiated by as yet unknown non-hK2 like proteases that appear to a valuable source of phK2 and enzymatically active hK2 to study their be either unstable or present at low levels or both. The process of biological roles in greater detail. conversion from phK2 to hK2 is an important step in understanding the regulation of hK2 and its role in prostate development. Thus, more ACKNOWLEDGMENTS research is necessary to better characterize this process and to identify the protease(s) responsible for hK2 activation. We thank Craig Fenstermaker and Michelle Cass for excellent technical The hK2 detected in the spent media of AV12-hK2#27 from day 4 support; Cindy Evans, Jan Payne, Saul Levine, and Steve Burch for generating onward was enzymatically active on a chromogenic substrate (Fig. 3). mAbs; and Dr. Judith Finlay for providing HK1G 586.1 affinity column and The increased activity correlated with a decrease in phK2 concentra protocols. tion and increase in hK2 concentration in the spent media. Day 1 and day 3 spent media that contained mainly phK2 had no detectable REFERENCES activity. Based on molecular modeling studies, hK2 is predicted to 1. Wingo, P. A., Tong, T. and Bolden, S. Cancer statistics. CA CancerJ. Clin., 45: 8â€”30, have a trypsin-like specificity (20, 21). The enzymatic activity of hK2 1995. 2. Catalona, W. J., Smith, D. S., Ratliff, T. L., Dodds, K. M., Coplen, D. E., Yuan, J. J., is specific for selected arginines and is different from the substrate Petros, J. A., and Andriole, G. L. Measurement of prostate specific antigen in serum specificity of trypsin.3 Thus, it may be possible to generate hK2- as a screening test for prostate cancer. N. Engl. J. Med., 324: 1156â€”1161, 1991. specific substrates that could be used in activity-based diagnostic 3. Osterling, J. E. Prostate specific antigen: a critical assessment of the most useful tumor marker for adenocarcinoma of the prostate. J. Urol., 145: 907â€”923,1991. assays to detect hK2 in bodily fluids. 4. Labrie, F., Dupont, A., Suburu, R., Cusan, L., Tremblay, M., Gomez, J-L., and hK2 is also secreted as phK2 from PC3-hK2 and DU145-hK2 cells; Emond, J. Serum prostate specific antigen as pre-screening test for prostate cancer. however, the process of extracellular conversion from phK2 to hK2 J. Urol., 147: 846â€”852, 1992. 5. McConnack, R. T., Rittenhouse, H. G., Finlay, J. A., Sokoloff, R. L., Wang, T. J., appears to be more efficient for DU145 cells than for PC3 cells (Fig. Wolfert, R. L., Lilja, H., and Oesterling, J. E. Molecular forms ofpmstate-specific antigen 5). The reason for this observation is not known at present. It may be and the human kallikrein gene family: a new era. Urology, 45: 729â€”744,1995. 6. Schedlich, L. J., Bennetts, B., and Morris, B. J. Primary structure of a human because DU145 cells secrete more varieties or quantities of proteases glandular kallikrein gene. DNA, 6: 429â€”437, 1987. in the media compared to PC3 cells. Differences in expression of 7. Moms, B. J. hGK-1. A kalhikrein gene expressed in human prostate. Clin. Exp. urokinase and plasminogen activator inhibitor 1 have been observed Pharmacol. Physiol., 16: 345â€”351,1989. 8. Young, C. Y. F., Andrews, P. E., and Tindall, D. J. Expression and androgenic between PC3 and DU145 cells (33). Furthermore, we have observed regulation of human prostate-specific kallikreins. J. Androl., 16: 97â€”99,1995. that PC3 cells survive for more than 2 weeks in serum-free media, 9. Riegman, P. H. J., Vlietstra, R. J., Suurmeijer, L., Cleutjems, C. B. J. M., and whereas DU145 cells lose viability rapidly without serum, thereby Trapman, J. Characterization of the human kallikrein locus. Genomics, 14: 6â€”11, 1992. liberating proteases more rapidly into the media. 10. Fukushima, D., Kitamura, N., and Nakanishi, S. Nucleotide sequence of cloned We detected two hK2 immunoreactive bands (â€”70and 90 kDa) in cDNA for human pancreatic kallikrein. Biochemistry, 24: 8037â€”8043, 1985. 11. Chapdelaine, P., Paradis, G., Tremblay, R., and Dube, J. High level of expression in the spent media of AV12-hK2#27 cells. These proteins could be the prostate of a human glandular kallikrein mRNA related to prostate-specific homooligomers of mature hK2 or complexes of mature hK2 with antigen. FEBS Leu., 236: 205â€”208,1988. 5401

12. Clements, J. The glandular kallikrein family of enzymes: tissue specific expression human prostate-specific glandular kallikreim, hK2, in E. coli and generation of and hormonal regulation. Endocr. Rev., 10: 393â€”419, 1989. antibodies. Mol. Cell. Endocrinol., 109: 237â€”241,1995. 13. Young, C. Y. F., Andrews, P. E., Montgomery, B. T., and Tindall, D. J. Tissue 25. Lovgren, J., Piironen, T., Overmo, C., Dowell, B., Karp, M., Petterson, K., Lilja, H., specific and hormonal regulation of human prostate-specific glandular kallikrein. and Lundwall, A. Production of recombinant PSA and hK2 and analysis of their Biochemistry,31:818â€”824,1992. immunologic cross-reactivity. Biochem. Biophys. Res. Commun., 231: 888â€”895, 14. Grauer, L., Charlesworth, M. C., Saedi, M. S., Finlay, J. A., Liu, R-S., Kuus-Reichel, 1995. K.,andTindall,D.J. IdentificationofhumanglandularkallikreinhK2fromLNCaP 26. Berg, D. T., McClure, D. B., and Grinnell, B. W. Ela-responsive mammalian cells. J. Androl., 17: 353â€”359,1996. hotlvector system for the stable high-level expression of secreted proteins. Nucleic 15. Lilja, H. Structure, function and regulation ofthe enzyme activity of prostate-specific Acids Res., 20: 5485â€”5486, 1992. antigen.WorldJ.Urol.,11:188â€”191,1993. 27. Miller, A. D., and Rosman, G. J. Improved retroviral vectors for gene transfer and 16. Lilja, H. A kallikrein-Iike serine protease in prostatic fluid cleaves the predominant expression. Biotechniques, 7: 980â€”990, 1989. seminal vesicle protein. J. Clin. Invest., 76: 1899â€”1903, 1985. 28. Yokoyama, W. M. Production of monoclonal antibodies. In: J. E. Coligan, A. M. 17. Lilja, H., Oldbring, J., Rannevik, G., and Laurell, C-B. Seminal vesicle-secreted Kruisbeek, D. H. Margulies, E. M. Sevach, and W. Strober (eds.), Current Protocols proteins and their reactions during gelation and liquefaction of human semen. J. Clin. in Immunology, pp. 2.5.1â€”2.5.17,New York: John Wiley and Sons, 1991. Invest.,80:281â€”285,1987. 29. Schneider, C, Newman, R. A., Sutherland, D. R., Aster, U., and Greaves, M. F. A 18. Lilja, H., Abrahamsson, P-A., and Lundwall, A. Semenogelin. the predominant one-step purification of membrane proteins using a high efficiency immunomatrix. protein in human semen. J. Biol. Chem., 264: 1894â€”1900, 1989. J.Biol.Chem.,257: 10766â€”10769,1982. 19. Christensson, A., Laurell, C-B., and Lilja. H. Enzymatic activity of prostate-specific antigen and its reactions with extracellular serine proteinase inhibitors. Eur. J. Bio 30. Fisher, J. M., and Scheller, R. H. Prohormone processing and the secretory pathways. chem.,194:755â€”763,1990. J.Biol.Chem.,263: 16515â€”16518,1988. 20. Vihinen, M. Modeling of prostate-specific antigen and human glandular kallikreinâ€¢ 31. Gierasch, L. M. Signal sequences. Biochemistry, 28: 923â€”930,1989. structures. Biochem. Biophys. Res. Commun., 204: 1251â€”1256,1994. 32. . Harris, R. Processing of pro-hormone precursor proteins. Arch. Biochem. Biophys., 21. Bridon, D. P., and Dowell, B. L. Structural comparison of prostate-specific antigen 275:315â€”333,1989. and human glandular kallikrein using molecular modeling. Urology, 45: 801â€”806, 33. Van Veldhuizen, P. J., Sadasivan, R., Wyatt, A., and Herdler, J. The plasminogen 1995. activator system in human prostatic cancer cell lines. Proc. Am. Assoc. Cancer Ret., 22. Lilja, H., and Lundwall, A. Molecular cloning of epididymal and seminal vesicular 35: 608, 1994. transcripts encoding a semenogelin related protein. Proc. Nat]. Acad. Sci. USA, 89: 34. Watt, K. W. K., Lee, P-J., M'TimKulu, T., Chan, W-P., and Loor, R. Human 4559â€”4563,1992. prostate-specific antigen: structural and functional similarity with serine proteases. 23. Deperthes, D., Chapdelaine, P., Tremblay, R. R., Brunet, C., Berton, J., Hebert, J., Proc. NatI. Acad. Sci. USA, 83: 3166â€”3170, 1986. Lazure, C., and Dube, J. Y. Isolation of prostatic kallikrein hK2, also known as 35. Kanety, H., Madjar, Y., Dagan, Y., Levi, J., Papa, M. J., Pariente, C., Goldwasser, B., hGK-l, in human seminal plasma. Biochim. Biophys. Acta, 1245: 31 1â€”316,1995. and Karasik, A. Serum insulin-like growth factor-binding protein-2 (IGFBP-2) is 24. Saedi, M. S., Cass, M. M. J., God, A. S., Grauer, L. S., Hogen, K. L., Okaneya, T., increased and IGFBP-3 is decreased in patients with prostate cancer: correlation with Griffin, B. Y., Klee, G. G., Young, C. Y. F., and Tindall, D. J. Overexpression of a serum prostate-specific antigen. J. Clin. Endocrinol. Metab., 77: 229â€”233, 1992.

5402

Abhay Kumar, Amita S. Goel, Timothy M. Hill, et al.

Cancer Res 1996;56:5397-5402.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/56/23/5397

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/56/23/5397. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.