Cellular Localization Domains of a Rabbit and a Human Carboxylesterase: Influence on Irinotecan (CPT-11) Metabolism by the Rabbit Enzyme

Philip M. Potter, Judith S. Wolverton, Christopher L. Morton, et al. Cancer Res 1998;58:3627-3632.

Updated Version

Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/58/16/3627

Citing Articles

This article has been cited by 15 HighWire-hosted articles. Access the articles at: http://cancerres.aacrjournals.org/content/58/16/3627#related-urls

E-mail alerts Reprints and Subscriptions Permissions

Sign up to receive free email-alerts related to this article or journal. To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at pubs@aacr.org. To request permission to re-use all or part of this article, contact the AACR Publications Department at permissions@aacr.org.

Downloaded from cancerres.aacrjournals.org on November 8, 2011 Copyright 1998 American Association for Cancer Research

[CANCER RESEARCH 58. 3627-3632,

August 15, IW8]

Cellular Localization Domains of a Rabbit and a Human Carboxylesterase: Influence on Irinotecan (CPT-11) Metabolism by the Rabbit Enzyme1
Philip M. Potter, Judith S. Wolverton, Christopher L. Morton, Monika Wierdl, and Mary K. Danks2
Department of Molecular Pharmacology, St. Jude Children's Research Hospital. Memphis, Tennessee 38105

ABSTRACT
Enzyme activation of prodrugs to improve the therapeutic index of specific anticancer agents is an attractive alternative to current chemo therapy regimens. This study addresses the potential for activating irinotecan (CPT-11) with recombinant carboxylesterases (CEs). CEs are a ubiquitous class of enzymes thought to be involved in the detoxification of xenobiotics. Their primary amino acid sequence indicates that these pro teins should be localized to the endoplasmic reticulum. By PCR-mediated mutagenesis of a rabbit liver and a human alveolar macrophage CE cDNA, expression in Cos? cells, and subsequent immunohistochemical localization, we have determined that an 18-amino acid N11,-term inul hydrophobic signal peptide is responsible for the localization of these proteins to the endoplasmic reticulum. By similar approaches, we have demonstrated that the COOH-terminal amino acids HIL prevent secre tion of the proteins from the cell. Enzymatic activity was lost by removing the NH2-terminal domain; however, active enzyme could be detected in the culture media of cells expressing the COOH-terminally truncated proteins. Secretion of CEs lacking the six COOH-terminal amino acids could be prevented with brefeldin A, confirming that these truncated enzymes were processed and released from cells by endoplasmic reticulum-mediated exocytosis. Double-truncation mutant enzymes lacking both NH2- and COOH-terminal sequences demonstrated immunostaining pat terns similar to those of the NH2-terminally truncated proteins and also lacked CE activity. In all cases, metabolism of the classic esterase sub strate o-nitrophenyl acetate predicted the sensitivity of cells expressing the rabbit CE to the anticancer agent CPT-11. In addition, the secreted enzyme sensitized Cos? cells to this drug, indicating that protein associ ation with a lipid bilayer is not required for substrate metabolism.

INTRODUCTION CEs3 are a family of ubiquitous enzymes that are thought to be involved in the detoxification of xenobiotics ( 1). They are present in virtually all organisms ranging from bacteria to mammals (Refs. 2-4 and references therein). Because many of these enzymes are abun dantly expressed, numerous proteins have been purified to homoge neity, and the cDNAs encoding these CEs have been isolated (5, 6). Characteristic motifs that have been identified in the primary amino acid sequence include the B-l and B-2 domains, the former of which contains a serine that is labeled by the reaction of these proteins with radioactive inhibitor molecules, suggesting that it is part of the cata lytic active site (7). Additionally, an 18-residue highly hydrophobic signal sequence is present at the NH2 termini of many of these CEs. It has been presumed, by homology to other signal peptides, that this sequence is responsible for the localization of these enzymes within the ER of cells (8).
Received 4/10/98; accepted 6/17/98. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. ' Supported by NIH Grants CA-66124. CA-63512, CA-23099, and CA-76202, the Cancer Center Core Grant P30-CA-21765. and by the American Lebanese Syrian Asso ciated Charities. 2 To whom requests for reprints should be addressed, at Department of Molecular Pharmacology, St. Jude Children's Research Hospital, 332 North Lauderdale, Memphis, TN 38105. Phone: (901)495-3440; Fax: (901) 521-1668: E-mail: mary.danks@stjude.org. 3 The abbreviations used are: CE. Carboxylesterase: CPT-11. irinotecan. 7-ethyl-10[4-(l-piperidino)-l-piperidino]carbonyloxycamptothecin: ER, endoplasmic reticulum: HA, influenza hemaglutinin: HRP, horseradish peroxidase; o-NPA, w-nitrophenyl acetate: SN-38, 7-ethyl-IO-hydroxycamptothecin; RIPA, radioimmunoprecipitation assay.

The generation of reagents for the identification of specific CEs in mammalian cells has been unsuccessful, largely due to the high degree of homology among the family members (8, 9). Consequently, the in situ analysis of particular CE molecules has been complicated. We have recently isolated a rabbit liver CE that can activate the prodrug CPT-11 to generate the potent topoisomerase I inhibitor SN-38 (10). CPT-11 has demonstrated significant antitumor activity in human tumor xenografts grown in immune-deprived mice (11), and this agent is currently under evaluation for the treatment of human tumors (12, 13). To facilitate analysis of the CE encoded by this rabbit liver cDNA as well as a human homologue of this gene, we previously epitope-tagged and expressed these proteins in mammalian cells. We localized the human and rabbit CEs predominantly to the ER by both immunofluorescence image cytometry and gold-enhanced electron microscopy (14). Previously published results have demonstrated that enzyme/prodrug combinations, such as Herpes simplex virus thymidine kinase/ ganciclovir and Escherichia coli cytosine deaminase/5-fluorocytosine, suggest that the generation of a bystander effect may be therapeutically beneficial (15, 16). We reasoned, therefore, that a CE that is secreted from the cell might result in localized extracellular activation of CPT-11 and possibly in enhanced tumor cytotoxicity. However, because information concerning the cellular processing of CEs is limited, we undertook a series of studies to determine whether the coding sequences of the enzymes could be modified to alter the subcellular localization of the proteins. We therefore performed PCRmediated mutagenesis of potential enzyme localization domains of the cDNAs. This study identifies the motifs responsible for subcellular localization of CEs and their retention within the ER and monitors the effect of these modifications on the metabolism of both a simple substrate. o-NPA. and anticancer prodrug CPT-11.

MATERIALS

AND METHODS

Cell Lines and Plasmids. Cos7 monkey kidney cells were obtained from American Type Culture Collection, and Rh30 human rhabdomyosarcoma cells were derived from a bone marrow aspirate of a patient at St. Jude Children's Research Hospital (17). Both cell lines were grown in 90% DMEM containing 10% PCS and 2 mM glutamine under an atmosphere of 90% air: 10% CO, at 37C. The mammalian expression vectors pCIneo and pIRESneo were obtained from Promega (Madison, WI) and Clontech (Palo Alto. CA), respectively. The plasmid constructs used in this study and the CEs encoded by these cDNAs are described in Table 1 and Fig. 1. Epitope-tagged cDNAs encoding a human alveolar macro phage CE and a rabbit liver CE have been described previously (14). Immunohistochemical Reagents. Anti-influenza HA and anti-HA anti body conjugated to HRP were purchased from Santa Cruz Biotechnology (Santa Cruz. CA) and Boehringer Mannheim (Indianapolis, IN), respectively. The anti--tubulin antibody was obtained from ICN (Costa Mesa, CA). Texas Red and FITC-conjugated rat immunoglobulins were purchased from Jackson ImmunoResearch Laboratories (West Grove, PA), and the DNA stain Hoechst 33342 was obtained from Sigma (St. Louis, MO). Construction of Deleted cDNAs. All deletions at the 5' and 3' ends of the cDNAs were performed by PCR using Pfu polymerase (Stratagene, La Jolla, CA). This enzyme was chosen because it has a much lower error rate than Taq polymerase (18). Briefly, oligonucleotides 24-30 residues in length were designed that created artificial methionine initiation or termination codons. PCR was performed using the tagged full-length cDNA as a template under the following conditions: denaturation at 94Cfor 45 s; annealing at the appro-

3627

Downloaded from cancerres.aacrjournals.org on November 8, 2011 Copyright 1998 American Association for Cancer Research

METABOLISM

OF CPT-ll

BY A RABBIT CARBOXYLESTERASE

piiate temperature for 1 min; and extension at 72Cfor 1 min and 30 s. The initial 4 cycles of annealing were performed at 50C,and the subsequent 19 cycles were performed at 56C.The reduced annealing temperatures in the early amplification cycles allowed the formation of the mutations at the ends of the template. All oligonucleotides created EroRI sites within the amplified sequence to allow subsequent manipulation of the DNA fragment. Fragments were ligated into pUC9, and the termini were sequenced to confirm the generation of the mutations. Assay for CE Activity. Whole-cell extracts were prepared by sonication of the cells in minimal volumes of 50 mM HEPES (pH 7.4) on ice, and CE activity was determined by the conversion of o-NPA to o-nitrophenol as described previously (10, 19). A minimum of four assays were performed for each sample, and protein concentrations were determined with the Biorad protein assay reagent (Biorad, Hercules, CA) using BSA as a standard. Enzyme activities in medium were calculated per milliliter of sample. SDS-PAGE Analysis and Western Analysis. Cell sonicates were sepa rated in 8% denaturing acrylamide gels as described by Laemmli (20). Gels were either stained in Coomassie Blue R-250, or proteins were transferred by electroblotting to Immobilon-P (Millipore, Bedford, MA) for Western analysis (21). To facilitate the detection of epitope-tagged proteins present on Immo bilon-P membranes, the filters were incubated with the appropriate antibody, and specific immunoreactivity was detected by chemiluminescence (enhanced chemiluminescence: Amersham. Arlington Heights, IL). To remove antibodies from filters to allow subsequent reprobing, membranes were stripped in 62.5 mM Tris (pH 6.7), 2% SDS, and 100 mM 0-mercaptoethanol at 55Cfor 45 min. To confirm approximate equal loading of gels, membranes containing cell extracts were incubated with an anti-/3-tubulin antibody. Immunoprecipitations of Proteins from Cell Culture Medium. To de tect the presence of tagged secreted proteins, culture medium was removed from actively growing cells and centrifuged at 1,000 x g for 5 min. The supernatant was transferred to a new tube, and the procedure was repeated twice. After centrifugation at 14,000 X g for 5 min, 900 /il of medium were diluted to 1 ml with 100 d IOx RIPA buffer [IX RIPA buffer = 50 mM of Tris (pH 7.4), 150 mM NaCl, 1 mM EGTA, 1% NP40, and 0.25% sodium deoxycholate], and 25 /I f protein A/G-agarose beads (Santa Cruz Biotech o nology) were added. The samples were rocked for 4 h at 4C,and the beads were removed by centrifugation at 14,000 x g for 5 min. Two tg anti-HA of monoclonal antibody were added to the cleared supernatant, and the samples were inverted repeatedly at 4Covernight. Protein A/G-agarose beads (25 /xl) were added to the samples and rocked at 4Cfor 6 h. After centrifugation at 14,000 X g for 10 s, the medium was removed, and the pellet was washed with 1 ml of 1X RIPA buffer. This procedure was repeated six times, the beads were resuspended in 40 p.] of SDS-PAGE loading buffer, and the samples were heated at 95C for 3 min. Proteins were applied to 8% SDS-PAGE and transferred to Immobilon-P membranes by electroblotting. HA-tagged proteins were detected with an anti-HA-HRP-conjugated antibody and visualized by enhanced chemiluminescence. Control samples consisted of medium spiked

N pCIRABFL, pCIHUMCAR pIRESRABFL

B-2

B-1

TEHIEL

iosa

pCIRABHA, pCIHUMHA pIRESRABHA

HA

pCIRABHAN,

pCIHUMHAN

pCIRABHAC,

pCIHUMHAC

pClRABHANC,

pCIHUMHANC

KM

Fig. 1. Schematic representation of the modified CEs used in this study. B-1 and B-2. CE motifs; H. //indiII; HA. influenza HA epitope tag; N. 18-residue NH,-lentiinal hydrophobic signal peptide; TEHIEL. six COOH-terininal amino acids.

with cell extracts expressing the HA-tagged proteins and medium from cells transfected with plasmids encoding untagged proteins. Transfection of Mammalian Cells. Cos7 cells (IO7) were electroporated with 20 ig plasmid DNA in a volume of 200 /I f PBS using a Biorad of o electroporator and a capacitance extender (Biorad). Optimized conditions for electroporation were achieved using 260 V and 960 /xF. After transfection, cells were plated into 75-cm2 flasks in fresh media and harvested by trypsinization after 48 h. For stable transfection of Rh30, cells were electroporated under similar conditions, except that the voltage was reduced to 180 V. After 48 h, transfectants were selected in media containing 500 /ig/ml G418 for 10 days before analysis. Growth inhibition assays were performed as described previ ously (10, 22, 23). Fluorescence Immunohistochemical Localization of Proteins. Cells were plated on glass slides and allowed to attach overnight. After the removal of media, cells were fixed in 2% paraformaldehyde. and the nuclear membrane was permeabilized by incubation in 0.25% Triton X-100. After extensive washing in PBS, antibody diluted to 100 ng/ml in PBS containing 1% BSA was applied to the cells for 1 h. Antibody localization was determined by incuba tion with a Texas Red or a FITC-conjugated immunoglobulin. Cell DNA was stained with Hoechst 33342 to define the nuclear area for each cell. Cells were analyzed on a Zeiss Axiovert 135TV microscope connected to a Silicon Graphics Crimson workstation as described previously (24).

RESULTS
Table 1 Plasmida ilescribetl in this study

Name
pCInco pIRESneo pCIHDMCAR pC'IHUMIIA pC'IHUMHAN pCIHUMHAC pCIHUMHANC pCIRABFl. pCIRABHA pCIRABHAN pCIRABHAC pClRABHANC pIRESRABFL pIRESRABHA

Description Mammalian expression vector Mammalian expression vector pCInco containing human alveolar macrophage CE cDNA pCInco containing tagged human alveolar macrophage CE cDNA (Ref. 14) pClneo containing NHylcrminally truncated lagged human

alveolar macrophage CH cDNA pCInco containing COOH-lcrminally truncated lagged human alveolar macrophage CE cDNA pCInco containing both NHr and COOH-terminally truncated tagged human alveolar macrophage CE cDNA pCInco containing rahhil liver CE cDNA pClneo containing tagged rabbit liver CE cDNA (Ref. 14) pCInco containing NH,-lcrminally truncated tagged rabbit liver CE cDNA pC'Ineo containing C(X)H-terminally truncated tagged rabbit liver CE cDNA pClneo containing both NH,- and COOH-lerminally lagged rabbit liver CE cDNA pIRESnco containing rabbit liver CE cDNA pIRESnco containing tagged rabbit liver CE cDNA truncated

Deletion of the NI I, iminai Hydrophobic Signal Peptide of i the Rabbit and Human CE. To determine whether the residues at the NH2 terminus of the CEs affected subcellulur localization, we deleted nucleotides encoding the first 18 amino acids and created an artificial methionine initiation codon in each cDNA by PCR. After DNA sequencing to confirm the integrity of the constructs, plasmids pCIHUMHAN and pCIRABHAN (see Table 1 and Fig. 1) were transfected into Cos? cells. Cell sonicates were prepared for both Western analysis and CE activity assays. The Western analysis shown in Fig. 2A indicates that the proteins can be expressed in cells and are stable. However, very little CE activity was observed in cell sonicates, indicating that this NH2-terminal domain is crucial for enzyme func tion (Table 2). Immunohistochemical staining of these cells with the anti-HA antibody demonstrated diffuse cytoplasmic staining consist ent with a failure to localize to a particular cellular organelle. This pattern is very different from the typical ER staining observed in cells transfected with the tagged full-length CE cDNAs in pCIHUMHA and pCIRABHA (Figs. 3 and 4: Ref. 14). No CE protein was detected 3628

Downloaded from cancerres.aacrjournals.org on November 8, 2011 Copyright 1998 American Association for Cancer Research

METABOLISM

OF CRT-11 BY A RABBIT CARBOXYLESTERASE

98 HA Ab 64 60 36 98 64

present in the culture medium (Table 2), and truncated HA-tagged protein was detected in aliquots of the same medium by immunoprecipitation (Fig. 2B). Immunohistochemistry of the COOH-terminally truncated proteins confirmed their localization to the ER (Figs. 3 and 4), demonstrating a similar fine network-like staining pattern to the full-length tagged CEs. To confirm that the protein was being actively secreted from transfected Cos? cells, we removed the medium and replaced it with serum-free culture medium. After 4 h, aliquots of this medium were removed for CE activity assays. Approximately 6- and 9-fold more CE activity was detected in serum-free medium harvested from Cos? cells expressing either pCIHUMHAC or pCIRABHAC, respectively (Table 2), than in medium harvested from cells trans fected with cDNAs encoding either full-length or NH-,-terminally truncated proteins. Additionally, both the rabbit and human HAtagged COOH-terminally truncated proteins could be detected by immunoprecipitation from the culture medium (Fig. 2B). Brefeldin Treatment of Cos? Cells Expressing COOH-termi nally Truncated CEs. To confirm that extracellular secretion of the COOH-terminally truncated rabbit and human CEs was occurring via the ER, we exposed transiently transfected Cos? cells to 10 ju.g/ml brefeldin in serum-free medium for 4 h. Brefeldin is a fungal-derived antibiotic that causes reversible loss of the ER in mammalian cells (25, 26). After brefeldin treatment, cell sonicates contained significantly greater CE activity (Table 2) than did untreated Cos? pCIRABHAC cells, concom itant with reduced levels of CE activity in the culture media. Immunofluorescence staining of brefeldin-treated cells demonstrated a punctate staining pattern consistent with a disruption of ER integrity (Fig. 5/4), and immunoprecipitation and Western analysis of culture media confirmed the lower levels of secreted CE (Fig. 5fi). Deletion of Both the NH2-terminal and COOH-terminal Do mains. cDNAs encoding CEs lacking both the 18-residue NH2-terminal signal peptide and the 6 COOH-terminal amino acids were constructed by PCR and ligated into pCIneo to create pCIRABHANC and pCIHUMHANC (see Table 1 and Fig. 1). After transfection into Cos? cells, cell sonicates were assayed for CE activity, and tagged proteins were ana lyzed by Western analysis. In addition, the presence of CE enzyme in cell culture medium was determined by both activity assays and immunopre cipitation. Fig. 2A demonstrates the expression of proteins of the expected size in Cos? cell extracts, indicating that the proteins were stable. How ever, no CE enzyme activity could be detected in identical extracts or in the cell culture medium (Table 2). Immunohistochemistry of cells trans fected with pCIRABHANC or pCIHUMHANC demonstrated a diffuse staining pattern similar to that of cells transfected with pCIRABHAN and pCIHUMHAN (Figs. 3 and 4).
Table 2 CE activity in ceil extracts and culture media of transiently trtin.sfecled

Tubulin

___

__.^ _

SO 36

B
98 64 20 secs 50 36 98 64 5 min
50

36
Fig. 2. A. Western analysis of truncated CEs expressed in Cos7 cells. Whole-cell extracts derived from Cos7 cells after transient transfection with the indicated plasmids were subjected to immunoblotting using an anti-HA antibody (HA) or anti--tubulin antibody. B, Western analysis of immunoprecipitations of cell culture media from Cos7 cells after transient transfection with the same plasmids. The membrane was probed with an anti-HA-HRP-conjugated antibody and exposed to film for 20 s or 5 min.

in the culture medium as determined by Western and immunoprecipi tation analyses (Fig. 2B) or enzyme activity assays (Table 2). Deletion of the Six COOH-terminal Amino Acids of the Rabbit and Human CE. To determine whether the COOH-terminal amino acids were responsible for sequestering the CEs within the ER and preventing secretion from the cell, we removed the nucleotides en coding these residues and added an artificial termination codon to the tagged CE cDNAs by PCR. This procedure truncated both proteins by six amino acids, deleting the TEHIEL residues. After confirmation of construct integrity by DNA sequencing, the cDNAs were ligated into pCIneo to create pCIHUMHAC and pCIRABHAC (see Table 1 and Fig. 1). Each plasmid was transfected into Cos?, and after 48 h, cells were subjected to Western analysis and immunohistochemistry, and CE activities were determined. The cell culture medium was also assayed for enzyme activity and analyzed by Western analysis. Table 2 indicates the CE activity of cell extracts and cell culture medium, and Fig. 2A demonstrates the presence of the tagged COOH-termi nally truncated proteins in cell sonicates. Enzyme activity was also

Cos? cells
CE activity was determined spectrophotomelrically -NPAto nitrophenol at 420 nm. activitySampleCell by monitoring the conversion of

CE enzyme + serum (4 h) extract serum (48 h) (fimol/min/ml) (/mol/min/mg)Media (/xmol/min/ml)Media

+Cos? Cos? +Cos? +Cos? +Cos? +Cos? +Cos? +Cos? +Cos? +Cos? +Cos?

1.736.2 1.933.7 2.8109.837.5N37.432.7188.13.6"1.3}1.81.14.5"37.9

BrefpClRABHApCIRABHANpCIRABHACpCIRABHANCpCIRABHAC +

1.5NDND"3.94.425.24.613.04.14.839.34.80.20.11 + Bref6.8347.36.5453.55.7825.1785.37.7918.36.80.420.20.318.80.468.415.80.226.80.418 +pCIneopCIHUMHApCIHUMHANpCIHUMHACpCIHUMHANCpCIHUMHAC 1.0" 44.638.3 " ND. not determined. bP < 0.001 using ANOVA. ' Bref, brefeldin treatment, 10 fig/ml for 4 h.

3629

Downloaded from cancerres.aacrjournals.org on November 8, 2011 Copyright 1998 American Association for Cancer Research

METABOLISM

OF CPT-II

BY A RABBIT CARBOXYLESTERASE

Phase

DMA

Anti - HA

Negative Control

HA

HAN

HAC

HANC

Fig. 3. Fluorescence detection of truncated epilope-tagged rabbit CEs expressed in Cos7 cells. Phase-contrast photomicrographs and fluorescence images of the same cell stained with the anti-HA antibody (HA) to detect CEs and Hoechst 33342 (DNA) are shown. Negative controls are similar cells stained with an isotype-matched primary antibody. 3630

Downloaded from cancerres.aacrjournals.org on November 8, 2011 Copyright 1998 American Association for Cancer Research

METABOLISM

OF CPT-ll

BY A RABBIT CARBOXYLESTERASE

HAN

MAC

HANC

Fig. 4. Immunofluorescence photomicrograph of Cos7 cells after transfection with the indicated plasmids encoding epitope-tagged human CEs.

- brefeldin

brefeldin

Because the efficacy of transfection and hence protein expression can vary in transient assays with Cos? cells, we ligated the tagged rabbit cDNA into pIRESneo to generate pIRESRABHA and trans fected Rh30 cells. As expected, cells expressing either the full-length (pIRESRABFL) or the tagged (pIRESRABHA) rabbit CE were sen sitized to CPT-ll (Table 4). These data confirm that the tagged enzyme metabolizes CPT-11 to SN-38 in situ. DISCUSSION This study presents the novel observations that the NH2-terminal

98 64 60 36
Fig. 5. A, immunochemical localization of the COOH-terminally truncated human CEs in Cos? cells after treatment with 10 fig/ml brefeldin for 4 h. B, immunoprecipitation of cell culture media harvested from Cos? cells expressing the COOH-terminally truncated rabbit and human CEs (pClRABHAC and pCIHUMHAC) after treatment with brefeldin (+B).

Effect of Protein Truncation on CTP-11-mediated Cytotoxicity. Because the human alveolar macrophage CE does not metabolize CPT-ll,4 the effect of the NH2- and COOH-terminal deletions on CPT-11-mediated cytotoxicity was determined by transfecting Cos? cells with the various rabbit mutant cDNAs. After transfection with plasmids pCIRABHA, pCIRABHAN, pCIRABHAC, or pCIRABHANC, HA-tagged proteins were detectable in cell extracts by Western analysis (Fig. 2A); additionally, CE activity was present in pCIRABHA and pCIRABHAC cells (Table 2). The IC50 values of Cos7 cells transfected with the various plasmids to CPT-ll are indicated in Table 3. As can be seen, the IC5() values for Cos? cells transfected with pCIRABFL (the full-length untagged rabbit liver CE) or pCIRABHA or pCIRABHAC are approximately 3-20-fold lower than those of cells electroporated with pCIneo, pCIRABHAN, or pCIRABHANC. These data corroborate the pattern of CE enzyme activities in these cell extracts as determined by the metabolism of o-NPA (Table 2). No sensitization was seen in cells expressing the human alveolar macrophage CE (pCIHUMCAR).
4 M. K. Danks and P. M. Potter. Comparison of the efficiency of CPT-11 activation by a rabbit and a human carboxylesterase preparation. for use in enzyme/prodrug therapy, manuscript in

amino acids target a rabbit liver CE and a human alveolar macrophage CE to the ER, and that the COOH-terminal residues prevent secretion from the cell. Removal of the 18 hydrophobic NH2-terminal residues also resulted in loss of enzyme activity as monitored by the metabo lism of o-NPA and CPT-11-mediated cytotoxicity, even though the protein was detected by Western analysis of cell extracts. Because, no CE activity was detected in cells expressing the NH2-terminally truncated proteins, we presume that enzyme processing within the ER is essential for catalytic activity. Additionally, because the COOHterminally truncated secreted proteins were active, association with a lipid bilayer is not a requirement for a functional enzyme, but possibly posttranslation modification, such as the addition of oligosaccharides or the formation of disulfide bonds by isomerases, is required. Whereas the exact purpose of modification is unclear, it is probably involved in protein folding. Because inhibitors of glycosylation such as tunicamycin do not interfere with transport to the Golgi apparatus or ER, it is unlikely that the addition of these carbohydrate moieties is required for protein transport through these organd-es (27). Addi tionally, proteins can still be actively secreted from cells by exocytosis in the presence of these inhibitors (28). Our data suggest that modi fication by the ER is required to generate active CEs, regardless of whether the enzyme is retained in the cell or secreted. Removal of the six COOH-terminal amino acids, TEHL, resulted in secretion of the protein from cells (Fig. 2B and Table 2). Because FCS contains detectable amounts of CE activity, as indicated by the assays of serum-containing medium (Table 2), we replaced the medium with a minimal volume of serum-free DMEM. After 4 h, CE activity assays confirmed the presence of active enzyme in the media, approximately 6-9-fold greater than that of the samples analyzed from cells transfected
Table 3 /Cso values of transientlv transfecied Cos7 cells after treatment with CPT-11 Cells were exposed to CPT-11 for 2 h, 48 h after transfection. After 3 days, a time equivalent to three cell doublings, the cell number was determined. IC5() values were calculated from sigmoidal dose response curves using the Graphpad Prism program. PlasmidpCIneo pCIRABFL pCIRABHA pCIRABHAN pCIRABHAC pCIRABHANC pCIHUMCARIC5() (MM)20 0.9 0.95 6.0 28 2.4 21 60

3631

Downloaded from cancerres.aacrjournals.org on November 8, 2011 Copyright 1998 American Association for Cancer Research

METABOLISM

OF CPT-ll

BY A RABBIT CARBOXYLESTERASE

Table 4 ICW values of Iransfecleil Rh30 cells after treatment with CPT-1 Cells were exposed to CRT-11 for 2 h, and the cell number was determined 3 days later. The ICSO values were calculated from sigmoidal dose response curves using the Graphpad Prism program. Plasmid pIRESneo pIRESRABFL plRESRABHA 1C, >IM) ( 41.7 7.1 1.19 0.55 1.05 0.22

with the full-length cDNAs. This suggests that these proteins conform to the empirical rules suggested by Munro and Pelham (29), who hypoth esized that the electrostatic charge and/or structure of the COOH-terminal amino acid residues caused proteins to be retained in the ER. Although the exact mechanism of this retention is unclear, it is apparent that these amino acids prevent secretion. Brefeldin treatment of cells transiently transfected with pCIRABHAC or pCIHUMHAC resulted in the loss of integrity of the ER and the accumulation of the tagged enzymes in discrete vesicles (Fig. 5). Interestingly, levels of CE activity in the extracts of cells treated with brefeldin increased, presumably due to the prevention of secretion of the enzyme. Simultaneously, levels of CE in the cell culture media were reduced. The prevention of secretion of the CEs by brefeldin treatment did not result in the inactivation of the protein in the cells, suggesting that the glycosylation and/or disulfide bond formation (and hence intracellular activation of the CE) can occur in the Golgi apparatus. Previous reports have indicated that the activity of CEs may be dependent on the glycosylation of the protein (30). Our data are in agreement with these studies, indicating that if the rabbit liver and human alveolar macrophage CEs are not targeted to the ER by the NH2-terminal signal peptide, then the proteins are inactive. Because the addition of carbohydrate residues occurs within the Golgi and ER, we presume that glycosylation and/or disulfide bond formation is essential for the correct processing and folding of the mature protein. In contrast, removal of the COOH-terminal residues does not affect enzyme activity but allows secretion of the proteins from the cell. In summary, removal of the NH,-terminal signal peptide resulted in the formation of catalytically inactive proteins, whereas removal of the six COOH-terminal residues resulted in the secretion of active rabbit and human CEs. Because the rabbit CE can convert CPT-11 to SN-38 (10), the potential application of the secreted enzyme for an enhanced by stander effect with CPT-11 is under investigation. For cells expressing an intracellular CE, CPT-11 activation is dependent on diffusion of the drug across the cell membrane for conversion to SN-38. However, with a secreted CE, CPT-11 activation should occur in the extracellular fluid, and the SN-38 generated should readily diffuse into the surrounding cells, resulting in cytotoxicity even in cells that cannot activate CPT-11. We are currently assessing the efficacy of both the ER-localized and secreted rabbit liver CE to produce a bystander effect in combination with CPT-11 in both cells in culture and xenografts. Additionally, we are generating baculovirus expressing COOH-terminally truncated CEs to allow rapid isolation and purification of large amounts of these enzymes for structural studies.

ACKNOWLEDGMENTS
We thank Dr. Pat McGovern (Pharmacia Upjohn, Kalamazoo, MI) for the gifts of CPT-11 and SN-38 and David Whipple for excellent technical assist ance. We also thank Clayton Naeve and the Center for Biotechnology sequencing and oligonucleotide synthesis. for DNA

REFERENCES
I. Cashman, J. R.. Perroti. B. Y. T., Berkman. C. E.. and Lin. J. Pharmacokinetics molecular detoxification. Environ. Health Perspect., 104: 23-40. 1996. and 3632

McGhee. J. D.. Birchall. J. C.. Chung. M. A.. Cottrell. D. A., Edgar. L. G.. Svendsen, P. C.. and Ferrari. D. C. Production of null mutants in the major intestinal esterase gene (ges-1) of the nematode Caemirhabdilis e/egiins. Genetics. 125: 505-514, 1990. Hwang. C-S.. and Kolaltukudy, P. E. Molecular cloning and sequencing of thioesterase B cDNA and stimulation of expression of the thioesterase B gene associated with hormonal induction of peroxisome proliferation. J. Biol. Chem., 268: 14278-14284. 1993. Miller. S. B.. Main. A. R.. and Rush, R. S. Purification and physical properties of oligomeric and monomeric carboxylesterases from rabbit liver. J. Biol. Chem., 255: 7161-7167, 1980. Alexson, S. E. H., Finlay, T. H.. Hellman. U., Svensson. L. T., Diczfalusy, U., and Eggertsen, G. Molecular cloning and identification of a rat serum carboxylesterase expressed in the liver. J. Biol. Chem.. 269: 17118-17124. 1994. Munger. J. S.. Shi. G-P.. Mark. E. A.. Chin, D. T.. Gerard, C., and Chapman, H. A serine esterase released by human alveolar macrophages is closely related to liver microsomal carboxylesterases. J. Biol. Chem., 266: 18832-18838, 1991. DiPersio. L. P.. Fontaine. R. N.. and Hui. D. Y. Identification of the active site serine in pancreatic cholesterol esterase by chemical modification and site-specific mutagenesis. J. Biol. Chem., 265: 16801-16806. 1990. Satoh, T.. and Hosokawa. M. Molecular aspects of carboxyleslerase isoforms in comparison with other esterases. Toxicol. Lett.. 82/83: 439-445, 1995. Shibala. F.. Takagi. Y.. Kitajima. M.. Kuroda, T.. and Omura, T. Molecular cloning and characterization of a human carboxylesterase gene. Genomics, 17: 76-82. 1993. Potter. P. M.. Pawlik. C. A.. Morton. C. L.. and Danks, M. K. Isolation and partial characterization of a cDNA encoding a rabbit liver carboxyleslerase that activates the prodrug CPT-ll. Cancer Res. 58: 2646-2651, 1998. Thompson, J.. Zamhoni. W. C.. Cheshire. P. J.. Richmond, L.. Luo, X.. Houghton, J. A.. Stewart. C. F., and Houghton. P. J. Efficacy of oral irinotecan against neuroblastoma xenografts. Anticancer Drugs, 8: 313-322, 1997. 12. Siichenmyer. W. J.. Rowinsky, E. K.. Donehower. R. C.. and Kaufmann. S. H. The current status of camptothecin analogues as antilumor agents. J. Nati. Cancer Inst., 85: 271-291, 1993. Rowinsky. E. K.. Grochow, L. B.. Ettinger. D. S., Sartorius, S. E.. Lubejko, B. G., Chen. T-L.. Rock. M. K.. and Donehower. R. C. Phase 1 and pharmacological study of the novel topoisomerase I inhibitor 7-ethyl-10-[4-(l-piperidino)-l-piperidino]carbonyloxycamptothecin (CPT-ll) administered as a 90-minute infusion every 3 weeks. Cancer Res., 54: 427-436. 1994. Potter, P. M.. Wolverton, J. S., Morton, C. L.. Whipple. D. O., and Danks, M. K. In .v//ubcellular localization of epitope-tagged human and rabbit carboxylesterases. s Cytometry. 32: 223-232, 1998. Namba. H.. Tagawa, M.. Iwadate, Y.. Kimura. M.. Sueyoshi. K.. and Sakiyama, S. Bystander effect-mediated therapy of experimental brain tumor by genetically engi neered tumor cells. Hum. Gene Then, 9: 5-11, 1998. 16. Trinh. Q. T., Austin. E. A., Murray. D. M.. Knick, V. C., and Huber, B. E. Enzyme/prodrug gene therapy: comparison of cytosine deaminase/5-fluorocytosine versus thymidine kinase/ganciclovir enzyme/prodrug systems in a human colorectal carcinoma cell line. Cancer Res., 55: 4808-4812. 1995. 17. Douglass. E. C.. Valentine, M.. Etcubanas, E., Parham. D. M.. Webber. B. L.. Houghton. P. J.. Houghton, J. A., and Green. A. A. A specific chromosomal abnor mality in rhabdomyosarcoma (RMS). Cytogenet. Cell Genet., 45: 148-155, 1987. Lundberg, K. S., Shoemaker, D. D.. Adams, M. W., Short, J. M., Sorge, J. A., and Mathur. E. J. High-fidelity amplification using a thermostable DNA polymerase isolated from Pyrococcus furiosas. Gene (Amst.), 108: 1-6. 1991. 19. Beaufay, H., Amar-Costesec. A.. Feytmans. E.. Thines-Sempoux, D.. Wibo, M., Robbi. M.. and Berthet. J. Analytical study of microsomes and isolated subcellular membranes from rat liver. I. Biochemical methods. J. Cell Biol.. 61: 188-200. 1974. 20. Laernmli, U. K. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature (Lond.), 227: 680-685, 1970. expression of the 21. Morton, C. L.. and Potter, P. M. Rhabdomyosarcoma-specific Herpes simplex virus thymidine kinase gene confers sensitivity to ganciclovir. J. Pharmacol. Exp. Then, in press. 1998. 22. Conter. V., and Beck, W. T. Acquisition of multiple drug resistance by CCRF-CEM cells selected for different degrees of resistance to vincristine. Cancer Treat. Rep., 68: 831-839, 1984. 23. Danks. M. K.. Yalowich. J. C., and Beck, W. T. Atypical multiple drug resistance in a human leukemic cell line selected for resistance to teniposide (VM-26). Cancer Res., 47: 1297-1301. 1987. 24. Danks. M. K., Garren. K. E., Marion. R. C., and Whipple, D. O. Subcellular redistribution of DNA (opoisomerase I in anaplaslic astrocytoma cells treated with topotccan. Cancer Res.. 56: 1664-1673. 1996. 25. Misumi, Y.. Misumi, Y., Miki, K.. Takatsuki, A., Tamura, G., and Ikehara, Y. Novel blockade by brefeldin A of intracellular transport of secretory proteins in cultured rat hepatocytes. J. Biol. Chem., 26/: 11398-11403, 1986. 26. Miller. S. G.. Carnell. L.. and Moore. H. H. Post-Golgi membrane traffic: brefeldin A inhibits export from distal Golgi compartments to the cell surface but not recycling. J. Cell. Biol.. 118: 267-283. 1992. 27. Hogstorp, H.. and Carlin. G. Studies on biosynthesis of a 2-antiplasmin in rat liver cells. Thromb. Res.. 49: 307-318. 1988. 28. Damsky. C. H.. Levy-Benshimol. A.. Buck. C. A., and Warren. L. Effect of tunicamycin on the synthesis, intracellular transport and shedding of membrane glycoproteins in BHK cells. Exp. Cell Res., 119: 1-13, 1979. 29. Munro. S., and Pelham. H. R. B. A C-terminal signal prevents secretion of luminal ER proteins. Cell, 48: 899-907, 1987. 30. Chris. D. L.. McBride. O. W., and Gonzalez. F. J. Glycosylation-dependent activity of baculovirus-expressed human liver carboxylesterases: cDNA cloning and characterization of two highly similar enzyme forms. Biochemistry. 32: 11606-11617. 1993.

Downloaded from cancerres.aacrjournals.org on November 8, 2011 Copyright 1998 American Association for Cancer Research