Puromycin-sensitive aminopeptidase: An antiviral prodrug activating enzyme

Article abstract of DOI:10.1016/j.antiviral.2009.12.003

Cidofovir (HPMPC) is a broad-spectrum antiviral agent, currently used to treat AIDS-related human cytomegalovirus retinitis. Cidofovir has recognized therapeutic potential for orthopox virus infections, although its use is hampered by its inherent low oral bioavailability. Val-Ser-cyclic HPMPC (Val-Ser-cHPMPC) is a promising peptide prodrug which has previously been shown by us to improve the permeability and bioavailability of the parent compound in rodent models (Eriksson et al., 2008. Molecular Pharmaceutics 5, 598-609). Puromycin-sensitive aminopeptidase was partially purified from Caco-2 cell homogenates and identified as a prodrug activating enzyme for Val-Ser-cHPMPC. The prodrug activation process initially involves an enzymatic step where the l-Valine residue is removed by puromycin-sensitive aminopeptidase, a step that is bestatin-sensitive. Subsequent chemical hydrolysis results in the generation of cHPMPC. A recombinant puromycin-sensitive aminopeptidase was generated and its substrate specificity investigated. The k_cat for Val-pNA was significantly lower than that for Ala-pNA, suggesting that some amino acids are preferred over others. Furthermore, the three-fold higher k_cat for Val-Ser-cHPMPC as compared to Val-pNA suggests that the leaving group may play an important role in determining hydrolytic activity. In addition to its ability to hydrolyze a variety of substrates, these observations strongly suggest that puromycin-sensitive aminopeptidase is an important enzyme for activating Val-Ser-cHPMPC in vivo. Taken together, our data suggest that puromycin-sensitive aminopeptidase makes an attractive target for future prodrug design.

Full text of DOI:10.1016/j.antiviral.2009.12.003

Antiviral Research 85 (2010) 482–489
Contents lists available at ScienceDirect
Antiviral Research
journal homepage: www.elsevier.com/locate/antiviral
Puromycin-sensitive aminopeptidase: An antiviral prodrug activating enzyme
Ulrika Tehlera, Cara H. Nelsona, Larryn W. Petersonc, Chester J. Provodaa,
John M. Hilfingerb, Kyung-Dall Leea,∗, Charles E. McKennac, Gordon L. Amidona
a Department of Pharmaceutical Sciences and Center for Molecular Drug Targeting, College of Pharmacy, The University of Michigan,
428 Church St., Ann Arbor, MI 48109-1065, United States
b TSRL, Inc., 540 Avis Drive, Suite A, Ann Arbor, MI 48108, United States
c Department of Chemistry, University of Southern California, United States
a r t i c l e i n f o
a b s t r a c t
Article history:
Cidofovir (HPMPC) is a broad-spectrum antiviral agent, currently used to treat AIDS-related human
cytomegalovirus retinitis. Cidofovir has recognized therapeutic potential for orthopox virus infections,
although its use is hampered by its inherent low oral bioavailability. Val-Ser-cyclic HPMPC (Val-Ser-
cHPMPC) is a promising peptide prodrug which has previously been shown by us to improve the
permeability and bioavailability of the parent compound in rodent models (Eriksson et al., 2008.
Molecular Pharmaceutics 5, 598–609). Puromycin-sensitive aminopeptidase was partially purified from
Caco-2 cell homogenates and identified as a prodrug activating enzyme for Val-Ser-cHPMPC. The pro-
drug activation process initially involves an enzymatic step where the l-Valine residue is removed by
puromycin-sensitive aminopeptidase, a step that is bestatin-sensitive. Subsequent chemical hydrolysis
results in the generation of cHPMPC. A recombinant puromycin-sensitive aminopeptidase was gener-
ated and its substrate specificity investigated. The kcat for Val-pNA was significantly lower than that
for Ala-pNA, suggesting that some amino acids are preferred over others. Furthermore, the three-fold
higher kcat for Val-Ser-cHPMPC as compared to Val-pNA suggests that the leaving group may play an
important role in determining hydrolytic activity. In addition to its ability to hydrolyze a variety of sub-
strates, these observations strongly suggest that puromycin-sensitive aminopeptidase is an important
enzyme for activating Val-Ser-cHPMPC in vivo. Taken together, our data suggest that puromycin-sensitive
aminopeptidase makes an attractive target for future prodrug design.
Received 3 September 2009
Received in revised form
25 November 2009
Accepted 1 December 2009
Keywords:
Prodrug
Cidofovir
Puromycin-sensitive
Aminopeptidase
Bioavailability
Antiviral
© 2009 Elsevier B.V. All rights reserved.
1. Introduction
original pharmaceutical, biopharmaceutical, and pharmacokinetic
properties. Reliable and predictable in vivo activation is a criti-
Prodrugs of therapeutically active agents have rightfully been
receiving increased attention. The term prodrug describes chem-
icals with little or no pharmacological activity that undergo
biotransformation to yield a therapeutically active metabolite
(Albert, 1958). The chemical changes involved in creating pro-
drugs are usually designed to improve one or more physio-chemical
properties that are lacking in the parent drug. Thus, numerous pro-
drugs of therapeutic agents have been developed to improve their
cal aspect of the prodrug strategy; therefore, identification of the
mechanisms of their in vivo activation is important from a prodrug
design perspective. Recently our laboratory identified human vala-
cyclovirase as one of the enzymes responsible for activation of the
valyl ester prodrug forms of acyclovir (valacyclovir) and ganciclovir
(valganciclovir) (Kim et al., 2003). The valyl ester prodrugs have
previously been shown to significantly increase their parent drugs’
oral absorption (Curran and Noble, 2001; Perry and Faulds, 1996;
Pescovitz et al., 2000; Smiley et al., 1996). In the case of valacyclovir,
for example, the oral bioavailability of acyclovir was increased 3-
to 5-fold (Weller et al., 1993). It has been shown that the observed
increase in bioavailability of acyclovir when administered as its
prodrug valacyclovir is due to carrier-mediated intestinal absorp-
tion of valacyclovir via the human peptide transporter 1 (hPepT1)
(Balimane et al., 1998; Ganapathy et al., 1998; Han et al., 1998).
Analogous to the valacyclovir case mentioned above, we have
now identified a puromycin-sensitive aminopeptidase as one of
the activating enzymes of an amino acid-containing prodrug of
cidofovir.
Abbreviations: APP-S, puromycin-sensitive aminopeptidase; HPMPC, cidofovir,
(methyl
(S)-2-((S)-2-amino-3-methyl-butyrylamino)-3-[(S)-5-(4-amino-2-oxo-
2H-pyrimidin-1-ylmethyl)-2-oxido-1,4,2-dioxaphosphinan-2-yloxy]propanoate);
cHPMPC, cyclic cidofovir; Val-Ser-cHPMPC, l-valine-l-serine cyclic cidofovir;
Val-pNA, l-valine para-nitroanilide; Ala-pNA, l-alanine para-nitroanilide; EDC,
N-(3-dimethylaminopropyl)-Nꢀ-ethylcarbodiimide; TFA, trifluoroacetic acid;
IPTG, isopropyl ␤-d-thiogalactopyranoside; DMSO, dimethyl sulfoxide; PVDF,
polyvinylidene difluoride; PBS, phosphate-buffered saline.
∗
Corresponding author. Tel.: +1 734 647 4941; fax: +1 734 764 6282.
E-mail address: kdl.cmdt@umich.edu (K.-D. Lee).
0166-3542/$ – see front matter © 2009 Elsevier B.V. All rights reserved.
doi:10.1016/j.antiviral.2009.12.003

U. Tehler et al. / Antiviral Research 85 (2010) 482–489
483
nucleotide prodrug hydrolysis have been determined. In addition,
the APP-S prodrug activating pathway has been verified and will
be discussed. Given the importance of antiviral agents in pharma-
cotherapy, the identification of enzymes responsible for activation
of phosphonate-containing prodrugs can provide important new
targets for the design of more effective therapeutic agents.
2. Experimental procedures
2.1. Chemicals and reagents
Fig. 1. Chemical structures of cidofovir (1) and Val-Ser-cHPMPC (2).
Val-Ser-cHPMPC was synthesized as previously described
(Eriksson et al., 2008). Cell culture reagents, Benchmark Pro-
tein Ladder, 4–12% Bis-Tris polyacrylamide gels and SYPRO Red
protein gel stain were obtained from Invitrogen/Gibco. EAH
Sepharose 4B, Superdex-200, PD-10 and MonoQ columns were
purchased from GE Healthcare. Bestatin, trifluoroacetic acid (TFA)
and N-(3-dimethylaminopropyl)-Nꢀ-ethylcarbodiimide (EDC) were
purchased from Sigma–Aldrich. Other chemicals were either ACS
reagent grade, analytical or HPLC grade and purchased from
Thermo Fisher Scientific, Inc. unless otherwise noted.
Cidofovir (Vistide®, HPMPC, 1, Fig. 1) is an antiviral agent that
is clinically used for treatment of the AIDS-related herpes virus
infection, cytomegalovirus retinitis. It is a broad-spectrum antiviral
agent with therapeutic potential in the treatment of other her-
pes and DNA viruses, including polyoma-, papilloma-, adeno-, and
poxvirus infections (De Clercq, 1997; De Clercq and Holy, 2005).
Cidofovir is of particular interest due to its potential use as therapy
in the event of an outbreak of smallpox (De Clercq, 2002). Cur-
rently, the drawback of using cidofovir in a large-scale emergency
situation is its need for intravenous administration. The phospho-
nic acid group of cidofovir is ionized under physiological conditions
and contributes substantially to its observed low oral bioavailabil-
ity (<5%) (Cundy et al., 1996a, b; Wachsman et al., 1996). In the
event of a smallpox outbreak, it would be essential to be able to
conveniently administer effective dosages via the preferable oral
route. Our research has therefore been focused on synthesizing
orally available prodrugs of cidofovir that are efficiently activated
in vivo.
We have recently reported several examples of novel cyclic cid-
ofovir (cHPMPC) prodrugs incorporating dipeptides and ethylene
glycol-linked amino acids onto the cidofovir scaffold (Eriksson et
al., 2006, 2007, 2008; McKenna et al., 2005, 2006). One of our lead
prodrugs, Val-Ser-cHPMPC (2, Fig. 1), shows significantly enhanced
intestinal uptake (18.1% vs 2.2% for cHPMPC) in an in situ rat
perfusion model (Eriksson et al., 2008; McKenna et al., 2005). Inter-
estingly, the majority (≥90%) of Val-Ser-cHPMPC was found to
be converted to cHPMPC during in situ rat perfusion experiments
(Eriksson et al., 2008). In cell culture-based assays of antiviral activ-
ity, Val-Ser-cHPMPC (IC50 = 0.3 ␮M) performed nearly as well as
HPMPC (IC50 = 0.26 ␮M), worse than cHPMPC (IC50 < 0.1 ␮M) and
significantly better than ganciclovir (IC50 = 3 ␮M) against human
cytomegalovirus (Eriksson et al., 2008). Val-Ser-cHPMPC did not
perform nearly as well as either of the parent compounds against
vaccinia and cowpox viruses, although this observation may be
at least partly attributable to the lack of activating proteases in
vitro, as well as the significantly shorter (3 vs 10 days) incuba-
tion times required for assaying the poxviruses (Eriksson et al.,
2008). Since Val-Ser-cHPMPC undergoes substantial in vivo activa-
tion in the rat intestine, we are interested in investigating plausible
human activation pathway(s) for Val-Ser-cHPMPC. Herein we are
reporting the purification and identification of a prodrug activating
enzyme, puromycin-sensitive aminopeptidase (GenBank acces-
sion no. CAA68964), and show that it is responsible for the in
vitro activation of Val-Ser-cHPMPC in Caco-2 cell homogenates.
A recombinant puromycin-sensitive aminopeptidase (hereafter
abbreviated APP-S,1 for aminopeptidase, puromycin-sensitive) has
been produced and the kinetic constants (Km and kcat) of the
2.2. Cell culture
The human colon carcinoma cell line, Caco-2, was obtained
from the American Type Culture Collection (ATCC HTB37, pas-
sage numbers 33–59). The cells were routinely maintained in
Dulbecco’s Modified Eagle Medium (DMEM) containing 10% fetal
bovine serum, 25 mM d-glucose, 4 mM l-glutamine and 1 mM
sodium pyruvate. The cells were grown in 150 mm tissue culture
dishes, split every 6 days and seeded at 1.7 × 104 cells/ml. Cells to
be used in experiments were harvested 21 days after reaching con-
fluence. All cells were maintained in an atmosphere of 5% CO2 and
90% relative humidity at 37 ◦C.
2.3. Prodrug hydrolysis assay
The cell homogenates or subcellular fractions were preincu-
bated with 10 mM sodium phosphate buffer (pH 7.4) for 3 min at
37 ◦C. Prodrug was added at a final concentration of 250 ␮M to
initiate the enzymatic reaction. Samples were removed at prede-
termined time points (3–30 min) and quenched by the addition
of 1–1.5 volumes of 10% ice-cold trifluoroacetic acid (TFA). The
quenched samples were spun through 96-well 0.45 ␮m polyvinyli-
dene difluoride (PVDF) membranes (Unifilter, Whatman GF/B) at
1800 × g in a Jouan MR 22i tabletop centrifuge to remove the pre-
cipitates before HPLC analysis. The HPLC system (Waters) consisted
of a reverse-phase column (XTerra RP18, 5 ␮m, 4.6 mm × 250 mm),
a 515 pump, a 996 Photodiode Array UV detector and a 717
Plus Autosampler. The remaining prodrug and the production of
parent drug as well as intermediate activation products were ana-
lyzed using a mobile phase consisting of 17 mM phosphate buffer
containing 0.5 mM ion-pairing agent (tert-butyl ammonium dihy-
drogen phosphate) and 5% acetonitrile at the pH 7.3, with a flow
rate of 1 ml/min and detection by absorbance at 274 nm. The spe-
cific activity of the homogenate was expressed as nmol min−1 mg−1
of protein based on the disappearance of the prodrug.
2.4. Purification of APP-S from Caco-2 cells
2.4.1. Caco-2 cell lysis and differential centrifugation
All centrifugations and column protein purifications were per-
formed at 4 ◦C unless otherwise noted. Caco-2 cells from 20
confluent 150 mm-diameter dishes were washed once with
phosphate-buffered saline (PBS) and harvested on ice by scraping
1
While PSA is the most commonly used abbreviation for puromycin-sensitive
aminopeptidase, the authors chose instead to use APP-S in order to avoid confusion
with the most prevalent definition of PSA in the scientific literature, prostate-specific
antigen.

484
U. Tehler et al. / Antiviral Research 85 (2010) 482–489
with a rubber policeman using ice-cold PBS. The cells were pel-
leted 5 min at 1000 × g and the supernatant discarded. Caco-2 cell
homogenates were prepared in 10 mM phosphate buffer containing
0.25 M sucrose (pH 7.4) using a Dounce homogenizer. After achiev-
ing >90% cell lysis, as determined by visual inspection of trypan
blue-stained cells using a Zeiss Axiovert 25 inverted microscope,
differential centrifugation was used to achieve various cell frac-
tions. Briefly, the cell lysates were spun in an Allegra centrifuge
(Beckman Coulter) for 10 min at 3000 × g to give a pellet (P1) and a
supernatant (S1). S1 was then centrifuged for 20 min at 10,000 × g
→ (P2) and (S2); S2 was centrifuged 30 min at 25,000 × g → (P3) and
(S3); S3 was spun in a Beckman Coulter Optimax ultracentrifuge for
60 min at 100,000 × g → P4 and the final cytosolic fraction of the
cell homogenate, which was stored at −80 ◦C until use. No pro-
tease or esterase inhibitors were added to the buffers throughout
the purification scheme to avoid potential inhibition of the enzyme
of interest.
the Lowry method (Lowry et al., 1951) using a Bio-Rad protein assay
kit with bovine serum albumin as a standard.
2.5. Identification of the prodrug activating enzyme
The partially purified proteins from the active fractions of
the Superdex-200 size exclusion chromatography were further
resolved by SDS-PAGE and visualized by Coomassie blue or
SYPRO Red staining. Images of gels stained with SYPRO Red
were acquired using a Molecular Dynamics Typhoon 9200 imager
(GE Healthcare). The protein band having an apparent molec-
ular mass of ∼100 kDa was excised, digested with trypsin and
analyzed in an ABI 4800 Proteome Analyzer (TOF/TOF) mass
spectrometer (Applied Biosystems) at the Michigan Proteome Con-
sortium (www.proteomeconsortium.org), University of Michigan.
The obtained amino acid sequence was used as a query for searching
the non-redundant (nr) protein data base (NCBInr), and the protein
of interest was identified with high confidence as NCBI accession
number CAA68964 (Ion Score C.I. > 99%).
2.4.2. Ammonium sulfate precipitation
Further concentration of the protein in the cytosolic frac-
tion (final centrifugation supernatant; 174 mg total protein) was
achieved with an ammonium sulfate precipitation at 4 ◦C (pH 7.4).
The pellets containing the fractions responsible for the activation of
Val-Ser-cHPMPC were found in the concentration range between
45% and 70% of (NH4)2SO4. The positive fractions were pooled and
desalted using a PD-10 desalting column equilibrated with 50 mM
Tris, pH 8.5 (buffer A) (55 mg total protein).
2.6. Generation of recombinant APP-S
2.6.1. Subcloning of APP-S cDNA
Human APP-S cDNA (IMAGE clone ID 6059589) in the mam-
malian expression vector pCMV-SPORT6 (Open Biosystems) was
subcloned into the pET-28a vector (Novagen) for expression of
the N-terminally His-tagged construct in Escherichia coli. Briefly,
the APP-S cDNA was excised from pCMV-SPORT6 and ligated
into pET28a after digesting both with the restriction enzymes
EcoRI and XhoI (New England Biolabs) and purifying by elec-
trophoresis in a 1% agarose gel. The ∼2.8 kb band corresponding
to APP-S and the ∼5.4 kb band corresponding to pET-28a were
purified from the agarose using a QIAEX II gel extraction kit
(QIAGEN) and ligated using T4 DNA ligase (New England Bio-
labs). To shift the inserted cDNA to the correct reading frame,
one amino acid was inserted upstream of the APP-S cDNA
using the primers 5ꢀ-GGCCTCGCCGCGAATGCCGGAG AAGAGG 3ꢀ
and 5ꢀ-CTCTTCTCCGGCATTCG CGGCGAGGCC-3ꢀ (Integrated DNA
Technologies) and QuikChange Site Directed Mutagenesis kit
(Stratagene). The His-APP-S/pET-28a construct was then trans-
formed into E. coli strain BL21-RIPL (Stratagene) followed by
dideoxy sequencing (University of Michigan DNA Sequencing Core)
to confirm the nucleotide sequence of the recombinant His-APP-S.
2.4.3. Bestatin affinity chromatography
The bestatin EAH Sepharose affinity column was synthesized
according to the Acosta procedure (Acosta et al., 1998). Briefly, 6 ml
of the EAH Sepharose 4B matrix were washed sequentially with a
10-fold volume excess each of water, 0.5 M NaCl, then water again
(adjusted to pH 4.5 with HCl) and finally resuspended in 12 ml
H2O (pH 4.5), to which was added an ∼45-fold molar excess of
EDC (500 mg in 5 ml H2O). Bestatin (25 mg) was dissolved in 5 ml
H2O and added dropwise over 3 h to the reaction suspension, and
then incubated an additional 16 h at 25 ◦C with gentle agitation.
The bestatin-modified Sepharose 4B beads were washed in the fol-
lowing order with 100 ml of 0.1 M sodium acetate containing 0.5 M
NaCl (pH 4.0), 100 ml of 0.1 M Tris containing 0.5 M NaCl (pH 8.0),
and 100 ml distilled H2O, before packing of the column. The Caco-
2 cytosolic fraction (55 mg total protein) was run in the bestatin
column equilibrated with 50 mM Tris, pH 8.5 and eluted with a
stepwise gradient of 1 M NaCl (total 60 ml) in the same buffer at
1 ml/min. The active fractions, which eluted at a concentration of
0.25 M NaCl, were pooled and diluted with an equal volume of
20 mM Tris, pH 7 (buffer MQ) (5.1 mg total protein) and run in a
MonoQ anion exchange column equilibrated in the same buffer.
The MonoQ samples were eluted with 25 ml of MQ buffer followed
by a gradient from 0 to 1 M NaCl in MQ (total 35 ml) at 0.5 ml/min.
The active fractions were pooled together to generate 0.8 mg of
protein.
2.6.2. Recombinant APP-S expression and purification
His-APP-S protein expression was induced according to the
method of Sengupta et al. (2006) with modifications. Briefly,
BL21-RIPL cells containing His-APP-S/pET-28a were grown to sta-
tionary phase in LB broth at 37 ◦C, and then expanded until
cultures reached an optical density of 0.8–1.0 at 570 nm, at which
point His-APP-S expression was induced with 1 mM isopropyl
␤-d-thiogalactopyranoside (IPTG) at 18 ◦C for 18–20 h. Following
centrifugation at 6000 × g for 10 min at 4 ◦C, the cell pellet was
resuspended in 50 mM sodium phosphate, 300 mM sodium chlo-
ride, and 20 mM imidazole, pH 8 (wash buffer) containing 1 mg/ml
lysozyme (Sigma-Aldrich), followed by three cycles of freeze-thaw.
After a 10 min incubation at 37 ◦C, the homogenate was pulsed for
30 s with a probe sonicator (Model KT40, Kontes) followed by cen-
trifugation at 20,000 × g for 40 min. His-APP-S was purified from
the supernatant using Ni-NTA agarose (QIAGEN), followed by wash-
ing with ∼100 bed volumes (250 ml) wash buffer. The recombinant
APP-S was eluted with 100 U (∼15 ␮g) thrombin (GE Healthcare)
in 1 ml PBS pH 7.4 for 18 h at 25 ◦C with gentle agitation, which
also served to remove the His-tag. The thrombin and APP-S were
separated from the Ni-NTA agarose by spinning at 1500 × g for
5 min and the supernatant transferred to a clean tube, after which
Additional purifications from Caco-2 homogenates were con-
ducted to achieve a total of 1.5 mg of protein. The combined protein
samples were concentrated using a Centricon YM-3 centrifugal fil-
ter device (Amicon) and resolved in a Superdex-200 size exclusion
column with 50 mM sodium phosphate containing 0.15 M NaCl, pH
6.8 at 0.4 ml/min.
2.4.4. SDS-PAGE analysis of purification
All purification steps were examined by 10% SDS-PAGE with
Bio-Rad SDS-PAGE high molecular weight standards for the esti-
mation of the molecular weight as described by Laemmli (1970).
The hydrolytic activity of each fraction was measured as described
above (Section 2.3), and the protein concentrations were based on

U. Tehler et al. / Antiviral Research 85 (2010) 482–489
485
Table 1
Partial purification scheme of APP-S from Caco-2 cells. Caco-2 cell homogenate was sequentially purified by ammonium sulfate precipitation, bestatin affinity chromatography,
and MonoQ anion exchange chromatography. Following each purification step, the active fraction was incubated with Val-Ser-cHPMPC at 37 ◦C (10 mM sodium phosphate
buffer, pH 7.4) with aliquots taken at predetermined time points (3–30 min). Disappearance of the prodrug peak was monitored by HPLC (detection at 274 nm) and used to
calculate specific activity (nmol min−1 mg−1 protein) and half-life (min).
Purification step
Total protein (mg)
Specific activity (nmol min−1 mg−1 protein)
Half-life of Val-Ser-cHPMPC (min)
Purification (fold)
Cytosolic fraction
(NH4)2SO4 precipitation
Bestatin column
174
55
5.1
0.8
6.8
13.6
7.2
ND
4.6
1.0
1.7
ND
22
11.2
NDa
142.5
MonoQ column
a
ND, not determined.
the Ni-NTA agarose was washed 3 × 1 ml with PBS, pH 7.4 and all
four supernatants were combined. Thrombin was removed by incu-
bating with 400 ␮l p-aminobenzamidine agarose (Sigma-Aldrich;
binding capacity 4–8 mg thrombin) for 2 h at 25 ◦C, followed by
pelleting of the p-aminobenzamidine agarose at 1500 × g. Pro-
tein concentration was determined using the Pierce BCA assay
(Thermo Fisher Scientific), and relative purity determined by SDS-
PAGE followed by staining with Krypton Protein Stain (Thermo
Fisher Scientific). Gel images were acquired and quantified using
a Molecular Dynamics Typhoon 9200 imager and ImageQuant soft-
ware (GE Healthcare). Purified APP-S was aliquoted and stored at
−80 ◦C.
p-nitroaniline was determined using the Beer-Lambert equation
(ε405 = 9500 L mol−1 cm−1). Km and Vmax were calculated for each
substrate using GraphPad Prism 4.
3. Results
3.1. Purification of the prodrug activating enzyme
Since our initial perfusion studies showed that the activation
step most likely occurred in or in the vicinity of the epithelial cells
in the gastrointestinal tract of the rat (Eriksson et al., 2008), we
decided to use Caco-2 cells as our protein source for identification
of the human activating enzyme for Val-Ser-cHPMPC. The chem-
ical as well as the enzymatic hydrolysis of Val-Ser-cHPMPC was
investigated. The disappearance of Val-Ser-cHPMPC was observed
2.7. Prodrug hydrolysis by recombinant APP-S
2.7.1. HPLC assay of metabolites
Recombinant APP-S (30 ␮g/ml) was preincubated with and
without the inhibitor bestatin (Fluka, 20 ␮g/ml) in 10 mM HEPES,
100 mM NaCl (pH 7.4) for 5 min at 37 ◦C. Prodrug was added at
final concentrations ranging from 0.125 to 1 mM to initiate the
enzymatic reaction. Aliquots of 40 ␮l were removed at predeter-
mined time points (0–15 min) and quenched by the addition of
80 ␮l of 10% ice-cold TFA. The samples were prepared and analyzed
by HPLC essentially as described above for the prodrug hydrolysis
assays using Caco-2 samples, except that the HPLC conditions con-
sisted of an acetonitrile gradient (2–52%) mobile phase with a flow
rate of 1 ml/min and detection at 274 nm. The specific activity was
expressed as nmol min−1 mg−1 of protein based on the disappear-
ance of the prodrug.
and the half-life (t ) of the prodrug in the two systems (buffer and
½
homogenate) were calculated from assuming first-order kinetics
and the rate constant was derived from linear regression (r2 = 0.95)
of first-order plots of prodrug concentration vs time. The half-
life of Val-Ser-cHPMPC in the buffer system (10 mM PBS, pH 7.4)
was determined to be 108.2 min ( 13.3 min) compared to 13.6 min
(
3.6 min) for the cytosolic fraction of the Caco-2 cell homogenate
(2000 ␮g/ml protein concentration).
Interestingly, a new prodrug activation peak appeared in the
Caco-2 cell homogenate HPLC-chromatogram that was solely
observed in the Caco-2 cell system. The activation peak was ana-
lyzed by mass spectrometry and determined to correspond to
Ser-cHPMPC (data not shown), and later confirmed with purified
recombinant APP-S. The formation of this activation peak can be
inhibited by the addition of bestatin, an aminopeptidase inhibitor,
whereupon the half-life of the prodrug in the presence of cell
homogenate increased to 100.1 min ( 1.6 min). Preliminary data
showed that other protease inhibitors did not affect the half-life
of Val-Ser-cHPMPC in the presence of Caco-2 cell homogenates
2.7.2. LC–MS identification of the metabolites
Further identification of the APP-S hydrolysis products was
achieved on a Finnigan LCQ Deca XP Max mass spectrometer in
positive mode with a Finnigan Surveyor PDA Plus detector and
MS Pump Plus, all controlled using Xcalibur software. The samples
(20 ␮l injection volume) were resolved in a Varian Microsorb-MV C-
18 column (100-5, 250 mm × 4.6 mm) with UV detection at 274 nm.
The eluent was diverted immediately before the mass spectrome-
ter, such that only half of the flow was injected into the MS. The
mobile phases consisted of 0.1N ammonium acetate buffer, pH 5.5
containing either 0% acetonitrile (A) or 17.5% acetonitrile (B), run
at 1 ml/min. Mobile phase gradients consisted of 100% A/0% B for
5 min, 50% A/50% B at 6 min (or 25% A/75% B at 7 min for prodrug
alone samples), 20% A/80% B from 15 to 20 min.
to nearly the same extent as bestatin: 1 mM AEBSF (t = 56 min),
½
10 ␮g/ml aprotinin (t = 14 min), 10 ␮g/ml leupeptin (t = 25 min),
½
½
10 ␮g/ml pepstatin A (t = 18 min) or 10 ␮g/ml E-64 (t = 25 min),
½
½
consistent with previously published data (Sengupta et al., 2006).
We utilized these results and the knowledge that bestatin is a
reversible inhibitor to make our purification procedure from the
cytosolic fraction of the Caco-2 cell homogenate more efficient. By
incorporating an FPLC step containing a bestatin-modified column
(an in-house synthesized affinity column) in between the ammo-
nium precipitation and the MonoQ anion exchange column, the
partial purification procedure for APP-S was greatly enhanced.
The specific activity (SA) from the active fractions
obtained following the MonoQ column was determined to
2.7.3. APP-S hydrolysis of model substrates
The ability of APP-S to hydrolyze various amino acids was
tested using the chromogenic substrates l-valine p-nitroanilide and
l-alanine p-nitroanilide (Bachem). Recombinant APP-S was prein-
cubated in PBS, pH 7.4 with and without 20 ␮g/ml bestatin in a final
volume of 300 ␮l. The reaction was started by adding 0.05–1.6 mM
substrate dissolved in dimethyl sulfoxide (DMSO) and carried out at
37 ◦C. The production of p-nitroaniline was measured spectropho-
tometrically at 405 nm every 30 s for 15 min. The concentration of
be 142 nmol min−1 mg−1 protein, corresponding to
a 22-fold
purification from the initial cytosolic cell homogenate (Table 1).
Interestingly, when Val-Ser-cHPMPC was incubated for a prolonged
time period (2 h) at 37 ◦C in the presence of the active fractions
from the MonoQ column, the activation peak corresponding to
Ser-cHPMPC disappeared.

486
U. Tehler et al. / Antiviral Research 85 (2010) 482–489
Table 2
Kinetics of Val-Ser-cHPMPC hydrolysis by recombinant APP-S. Recombinant APP-S was incubated with Ala-pNA, Val-pNA or Val-Ser-cHPMPC at 37 ◦C in the presence or
absence of the inhibitor bestatin. Hydrolysis of the p-nitroanilide compounds was monitored spectrophotometrically at 405 nm every 30 s for 15 min. Hydrolysis of Val-Ser-
cHPMPC was determined by monitoring the disappearance of the prodrug peak by HPLC (detection at 274 nm) at t = 0, 1, 2, 3, 5, 10 and 15 min. Km and Vmax were calculated
using GraphPad Prism 4.0.
a
Substrate
Km (mM)
Vmax (nmol min−1 mg−1 protein)
kcat (min−1
)
kcat/Km (min−1 M−1
)
Ala-pNA
Val-pNA
Val-Ser-cHPMPC
0.51 0.14
0.28 0.19
0.85 0.33
5365 610
289 85
1873 400
1071
58
187
2.1 × 106
0.21 × 106
0.22 × 106
a
kcat values are calculated from Vmax values with the assumption that all enzyme molecules are catalytically active.
3.2. Identification of the prodrug activating enzyme
subsequently identified at amino acid 15 in APP-S’s N-terminus,
consistent with the ∼1.5 kDa MW difference observed in these
gels. The activity of purified recombinant APP-S was first analyzed
using the model substrates l-alanine p-nitroanilide (Ala-pNA) and
l-valine p-nitroanilide (Val-pNA) in the presence and absence of the
inhibitor bestatin. Val-pNA was shown to have a lower Km than Ala-
pNA (0.28 0.19 mM vs 0.51 0.14 mM, respectively) and a greater
than 18-fold lower Vmax (289 85.1 nmol min−1 mg−1 protein vs
5365 610.0 nmol min−1 mg−1 protein, respectively) as shown in
Table 2. APP-S hydrolysis of the model substrates was completely
inhibited by the addition of bestatin.
The active fractions following MonoQ and size exclusion chro-
matography (lanes 4–6, Fig. 2) mentioned above were collected
and analyzed by 10% SDS-PAGE and Coomassie staining. By
visually comparing the SDS-PAGE gels through the purification
process, deducting bands present in non-active fractions from
bands present in the active fractions, the band at ∼100 kDa was
identified exclusively to be present in the active fractions. An SDS-
PAGE gel was submitted to the Michigan Proteome Consortium
for MS/MS analysis, and the identity of the band was deter-
mined with high confidence to be APP-S (NCBI accession number
CAA68964).
3.3. Recombinant APP-S hydrolysis investigations
To confirm that APP-S is an activating enzyme of Val-Ser-
cHPMPC, the APP-S cDNA (Open Biosystems) was subcloned into
the bacterial expression vector pET-28a (Novagen). Recombinant
His-tagged APP-S was purified to >98% purity (Fig. 3) using Ni-NTA
agarose (QIAGEN) and thrombin cleavage to remove the His-tag.
The purified recombinant APP-S migrated as two bands in SDS-
PAGE when eluted from Ni-NTA by thrombin digestion (Fig. 3), but
as a single band when eluted with imidazole (data not shown).
Both thrombin-eluted bands were identified as APP-S by peptide
mass analysis at the University of Michigan Proteome Core, and
the presence of a potential thrombin-cleavable sequence, in addi-
tion to the expected pET-28a vector thrombin cleavage site, was
Fig. 2. Superdex-200 purification of puromycin-sensitive aminopeptidase from
Caco-2 cell homogenates. Active and non-active fractions from the Superdex-200
purification were analyzed by 10% SDS-PAGE, here stained with SYPRO Red. Lanes
4–6 contain active fractions hydrolyzing Val-Ser-cHPMPC, while lanes 2, 3, 7 and 9
are non-active fractions. Lane 8 corresponds to the pooled active MonoQ fractions
that were initially applied to the Superdex-200 column. Lanes 1 and 10 are size mark-
ers with molecular mass expressed in kDa. The band visible at ∼100 kDa in lanes 4–6
and 8 was exclusively present in the active fractions and its identity was determined
by tandem mass spectrometry and database searches to be puromycin-sensitive
aminopeptidase (APP-S).
Fig. 3. Purified recombinant APP-S. Recombinant human APP-S was expressed in
BL21-RIPL cells and purified using a Ni-NTA affinity column. APP-S was eluted via
cleavage by thrombin between the His-tag and the recombinant APP-S. Following
quantification of total protein using the BCA assay, 80 ng of APP-S was run in a 4–12%
Bis-Tris gel and stained with Krypton Protein Stain. Lane 1 contains BenchMark Pro-
tein Ladder and lane 2 contains the purified recombinant APP-S, which was purified
to >97% purity as determined by ImageQuant software analysis.

U. Tehler et al. / Antiviral Research 85 (2010) 482–489
487
similar to what was observed with Caco-2 homogenates (data not
shown).
The kcat for Val-pNA was approximately 18-fold lower (p < 0.06)
than the kcat for Ala-pNA and approximately three-fold lower
than the kcat for Val-Ser-cHPMPC. The kcat/Km values for Val-Ser-
cHPMPC cleavage by APP-S (0.22 × 106 M−1 min−1) are comparable
to those obtained for the para-nitroanilide derivatives of l-alanine
and l-valine as well as by others (Sengupta et al., 2006).
3.4. Activation pathways for Val-Ser-cHPMPC and its metabolites
Using LC–MS analysis it was observed that in the presence of
the recombinant APP-S, the peptide bond in Val-Ser-cHPMPC was
cleaved to remove the N-terminal amino acid (l-valine, Fig. 5)
to generate the intermediate, Ser-cHPMPC (3). When Val-Ser-
cHPMPC was incubated for a prolonged time period (2 h at 37 ◦C
in the presence of the MonoQ active fractions), the activation peak
corresponding to 3 disappeared while the peak corresponding to 1
steadily increased. When Val-Ser-cHPMPC was incubated in buffer
alone or with APP-S in the presence of the inhibitor bestatin the
peak corresponding to compound 3 was not present. Besides the
above-mentioned activating products, a minor species (≤15%) with
a mass of 480 (positive ion mode) was also observed (4). This mass
corresponds to the intact dipeptide conjugate attached to the par-
ent compound and not the cyclic version of cidofovir.
Fig. 4. Michaelis–Menten plot of Val-Ser-cHPMPC hydrolysis by APP-S. Recombi-
nant APP-S (30 ␮g/ml) was incubated in 10 mM HEPES, 100 mM NaCl, pH 7.4 at
37 ◦C with 0.125–1 mM Val-Ser-cHPMPC. Aliquots of 40 ␮l were removed at pre-
determined time points (0–15 min) and quenched by the addition of 80 ␮l of 10%
ice-cold TFA. The samples were analyzed by HPLC with a 2–52% acetonitrile gradient
mobile phase at 1 ml/min and detection at 274 nm. The concentration detected at
0 min was used as the initial concentration in the plot above to control for variabil-
ity between sample sets. The V0 was calculated and the plot was generated using
GraphPad Prism 4.0.
To determine the Km and Vmax of APP-S for Val-Ser-cHPMPC,
APP-S was incubated with a range of substrate concentrations
with aliquots withdrawn at predetermined time points. Using
HPLC to determine the concentration of Val-Ser-cHPMPC at var-
ious time points, V0 was calculated using the disappearance of
prodrug. The data from four independent experiments were plot-
ted (Fig. 4) and analyzed by non-linear regression (GraphPad
4. Discussion
By stepwise purification from the human intestinal cell line
Caco-2 and MS/MS analysis we have demonstrated that APP-S is
involved in activation of the antiviral prodrug Val-Ser-cHPMPC.
While our data do not exclude the possibility that other proteases
may be involved, the observation that bestatin, a known inhibitor
of APP-S (Constam et al., 1995; Sengupta et al., 2006), is able
to inhibit enzymatic hydrolysis of Val-Ser-cHPMPC in Caco-2 cell
homogenates further suggests that APP-S is important in the in vivo
Prism 4, GraphPad Software, Inc) to determine Vmax and Km
.
APP-S was shown to hydrolyze Val-Ser-cHPMPC with a Vmax of
1873 400 nmol min−1 mg−1 protein and a Km of 0.85 0.33 mM.
As expected, this hydrolysis was almost completely inhibited by
addition of bestatin. APP-S was not able to appreciably hydrolyze d-
Val-d-Ser-cHPMPC beyond that which was detected in buffer alone;
Fig. 5. Chemical structures of the observed activating metabolites obtained during the hydrolysis of Val-Ser-cHPMPC by recombinant APP-S. Samples from APP-S hydrolysis
of Val-Ser-cHPMPC were analyzed by LC/MS. It was found that the peak corresponding to 3 was present in the samples hydrolyzed by APP-S, but not present in the negative
control samples (prodrug in buffer alone or with APP-S and bestatin). Enzymatic (APP-S), as well as chemical hydrolysis, is involved in the overall prodrug activation process.

488
U. Tehler et al. / Antiviral Research 85 (2010) 482–489
activation of the prodrug. Furthermore, the relative inability of the
inhibitors aprotinin, leupeptin, and pepstatin A to reduce Val-Ser-
cHPMPC hydrolysis in Caco-2 cell homogenate (<10% inhibition
of specific activity, data not shown) is consistent with previous
reports that these are not potent inhibitors of APP-S (Constam et
al., 1995; Sengupta et al., 2006). Moreover, it has been reported
that APP-S prefers basic and hydrophobic amino acids and has
relatively low affinity for acidic residues (Constam et al., 1995;
Wagner et al., 1981), consistent with our observation that the valine
residue is efficiently cleaved from Val-Ser-cHPMPC in Caco-2 cell
homogenates. Finally, APP-S has been shown to hydrolyze a variety
of amino acid substrates, with the exception of: (i) those that have
acidic side chains, (ii) d-amino acid isomers, or (iii) N-terminally-
blocked amino acids (26). Consistent with these observations,
the recombinant APP-S was unable to hydrolyze the d-amino
acid version of 2, d-Val-d-Ser-cHMPC (data not shown). Further-
more, there was no significant hydrolysis of d-Val-d-Ser-cHPMPC
in Caco-2 homogenate as compared to buffer alone (data not
shown).
Puromycin-sensitive aminopeptidase has been extensively
studied and has been implicated in a number of physiological pro-
cesses, including normal cellular turnover (Bhutani et al., 2007;
Botbol and Scornik, 1983; Goldberg and Rock, 1992), cell cycle reg-
ulation (Constam et al., 1995), processing of antigenic peptides for
display on MHC class I (Stoltze et al., 2000; Towne et al., 2008),
and degradation of neuropeptides and brain function (Osada et al.,
1999; Schulz et al., 2001). However, to our knowledge, this is the
first reported finding that APP-S is able to hydrolyze an antiviral
prodrug. The broad tissue distribution of APP-S and other neu-
tral aminopeptidases, as well as their homology and expression in
a variety of species (Constam et al., 1995; McLellan et al., 1988;
Schulz et al., 2001; Tobler et al., 1997) can be advantageous to
ensure complete and rapid prodrug activation, as was previously
noted for Val-Ser-cHPMPC in situ (Eriksson et al., 2008). Addition-
ally, APP-S has been shown to have a broad substrate specificity
with preference for hydrophobic and basic amino acids (Hui et al.,
1983; Johnson and Hersh, 1990; Schnebli et al., 1979; Sengupta et
al., 2006; Wagner et al., 1981) making it an attractive target for
future prodrug design.
The recombinant APP-S was able to efficiently hydrolyze Val-
Ser-cHPMPC in vitro, reinforcing our hypothesis concerning APP-S’s
role in Val-Ser-cHPMPC activation in vivo. Similar to that observed
in Caco-2 cell homogenates, bestatin was able to inhibit APP-S
hydrolysis of Val-Ser-cHPMPC. The approximately 18-fold lower
kcat for Val-pNA as compared to Ala-pNA further confirmed that
some amino acids are preferred over others. Furthermore, the
three-fold higher kcat for Val-Ser-cHPMPC as compared to Val-pNA
suggests that the leaving group may play an important role in
determining hydrolytic activity. Interestingly, the Km value for the
hydrolysis of Val-Ser-cHPMPC by APP-S is in the sub-millimolar
range, suggesting that under in vivo conditions the conversion of
Val-Ser-cHPMPC is likely to occur well below saturating substrate
concentrations. The kcat/Km values for Val-Ser-cHPMPC cleavage by
APP-S (0.22 × 106 min−1 M−1) are comparable to those obtained for
the para-nitroanilide derivatives of l-alanine and l-valine as well
as by others (Sengupta et al., 2006), suggesting that the prodrug
hydrolysis is likely to occur with a reasonable efficiency even in
the presence of other substrates.
Acknowledgements
This research was supported by grant number GM007767
from NIGMS (the contents of this paper are solely the respon-
sibility of the authors and do not necessarily represent the
official views of NIGMS) and NIH grant U01 AI061457. Pro-
teomics data were provided by the Michigan Proteome Consortium
(www.proteomeconsortium.org) which is supported in part by
funds from The Michigan Life Sciences Corridor.
References
Acosta, D., Goni, F., Carmona, C., 1998. Characterization and partial purification of a
leucine aminopeptidase from Fasciola hepatica. Journal of Parasitology 84 (1),
1–7.
Albert, A., 1958. Chemical aspects of selective toxicity. Nature 182 (4633), 421–422.
Balimane, P.V., Tamai, I., Guo, A., Nakanishi, T., Kitada, H., Leibach, F.H., Tsuji, A., Sinko,
P.J., 1998. Direct evidence for peptide transporter (PepT1)-mediated uptake of a
nonpeptide prodrug, valacyclovir. Biochemical and Biophysical Research Com-
munications 250 (2), 246–251.
LC–MS analysis revealed that the peptide bond in Val-Ser-
cHPMPC was cleaved to remove the N-terminal valine to generate
the Ser-cHPMPC intermediate (3). This intermediate disappeared
with prolonged incubation (2 h at 37 ◦C) in the presence of the
MonoQ column active fractions while the peak associated with
cHPMPC (cyclic cidofovir, 1) correspondingly increased. One likely
mechanism for this relative instability of Ser-cHPMPC is nucle-
ophilic attack on the phosphodiester linkage by the primary amine
of serine that is produced after the removal of valine by APP-S.
Indeed, Lazarus et al. have previously proposed and demonstrated
such a mechanism to explain the intramolecular hydrolysis of
amine-containing phosphoryl esters (Lazarus and Benkovic, 1979;
Lazarus et al., 1980). Based on these findings, and the fact that the
major species found in rat plasma after a modified in situ single
pass perfusion is indeed cHPMPC (Eriksson et al., 2008), we sus-
pected that the activation of Val-Ser-cHPMPC occurs through both
enzymatic and chemical pathways. It is worth noting that cyclic
cidofovir itself undergoes a biotransformation to generate cidofovir
when exposed to endogenous cyclic cytidine 3ꢀ,5ꢀ-monophosphate
(cCMP) phosphodiesterase (Bischofberger et al., 1994; Mendel et
al., 1997), and can therefore be regarded as a prodrug of cido-
fovir. In addition, cyclic cidofovir has been reported to be less
nephrotoxic than cidofovir, while also exhibiting potent antiviral
activity (Bischofberger et al., 1994). Nevertheless, cyclic cidofovir
itself shows low oral bioavailability (Cundy et al., 1996a, b), indi-
cating the need to mask the residual phosphonate negative charge
present at physiological pH, which the reported Val-Ser-cHPMPC
prodrug has been designed to do.
Bhutani, N., Venkatraman, P., Goldberg, A.L., 2007. Puromycin-sensitive aminopep-
tidase is the major peptidase responsible for digesting polyglutamine sequences
released by proteasomes during protein degradation. EMBO Journal 26 (5),
1385–1396.
Bischofberger, N., Hitchcock, M.J.M., Chen, M.S., Barkhimer, D.B., Cundy, K.C., Kent,
K.M., Lacy, S.A., Lee, W.A., Li, Z.-H., et al., 1994. 1-[((S)-2-Hydroxy-2-oxo-
1,4,2-dioxaphosphorinan-5-yl)methyl]cytosine, an intracellular prodrug for
(S)-1-(3-hydroxy-2-phosphonylmethoxypropyl)cytosine with improved ther-
apeutic index in vivo. Antimicrobial Agents and Chemotherapy 38 (10),
2387–2391.
Botbol, V., Scornik, O.A., 1983. Peptide intermediates in the degradation of cellular
proteins. Bestatin permits their accumulation in mouse liver in vivo. The Journal
of Biological Chemistry 258 (3), 1942–1949.
Constam, D.B., Tobler, A.R., Rensing-Ehl, A., Kemler, I., Hersh, L.B., Fontana, A.,
1995. Puromycin-sensitive aminopeptidase, sequence analysis, expression, and
functional characterization. The Journal of Biological Chemistry 270 (45),
26931–26939.
Cundy, K.C., Bidgood, A.M., Lynch, G., Shaw, J.P., Griffin, L., Lee, W.A., 1996a. Phar-
macokinetics, bioavailability, metabolism, and tissue distribution of cidofovir
(HPMPC) and cyclic HPMPC in rats. Drug Metabolism and Disposition: The Bio-
logical Fate of Chemicals 24 (7), 745–752.
Cundy, K.C., Li, Z.H., Hitchcock, M.J., Lee, W.A., 1996b. Pharmacokinetics of cidofovir
in monkeys. Evidence for a prolonged elimination phase representing phospho-
rylated drug. Drug Metabolism and Disposition: The Biological Fate of Chemicals
24 (7), 738–744.
Curran, M., Noble, S., 2001. Valganciclovir. Drugs 61 (8), 1145–1150, discussion
1151–1152.
De Clercq, E., 1997. In search of a selective antiviral chemotherapy. Clinical Micro-
biology Reviews 10 (4), 674–693.
De Clercq, E., 2002. Cidofovir in the treatment of poxvirus infections. Antiviral
Research 55 (1), 1–13.
De Clercq, E., Holy, A., 2005. Case history: acyclic nucleoside phosphonates: a key
class of antiviral drugs. Nature Reviews. Drug Discovery 4 (11), 928–940.
Eriksson, U., Peterson, L.W., Kim, J.-s., Mitchell, S., Kijek, P., Hilfinger, J.M., Drach, J.C.,
Kashemirov, B.A., McKenna, C.E., 2006. Ehylene glycol-linked amino acid con-

U. Tehler et al. / Antiviral Research 85 (2010) 482–489
489
juagted of cyclic cidofovir: synthesis and biological activity. Antiviral Research
70, A58.
Eriksson, U., Hilfinger, J.M., Kim, J.S., Mitchell, S., Kijek, P., Borysko, K.Z., Breitenbach,
J.M., Drach, J.C., Kashemirov, B.A., McKenna, C.E., 2007. Synthesis and biological
activation of an ethylene glycol-linked amino acid conjugate of cyclic cidofovir.
Bioorganic & Medicinal Chemistry Letters 17 (3), 583–586.
Eriksson, U., Peterson, L.W., Kashemirov, B.A., Hilfinger, J.M., Drach, J.C., Borysko,
K.Z., Breitenbach, J.M., Kim, J.S., Mitchell, S., Kijek, P., McKenna, C.E., 2008. Ser-
ine peptide phosphoester prodrugs of cyclic cidofovir: synthesis, transport, and
antiviral activity. Molecular Pharmaceutics 5 (4), 598–609.
Ganapathy, M.E., Huang, W., Wang, H., Ganapathy, V., Leibach, F.H., 1998. Vala-
cyclovir: a substrate for the intestinal and renal peptide transporters PEPT1
and PEPT2. Biochemical and Biophysical Research Communications 246 (2),
470–475.
Goldberg, A.L., Rock, K.L., 1992. Proteolysis, proteasomes and antigen presentation.
Nature 357 (6377), 375–379.
Han, H., de Vrueh, R.L., Rhie, J.K., Covitz, K.M., Smith, P.L., Lee, C.P., Oh, D.M., Sadee, W.,
Amidon, G.L., 1998. 5’-Amino acid esters of antiviral nucleosides, acyclovir, and
AZT are absorbed by the intestinal PEPT1 peptide transporter. Pharmaceutical
Research 15 (8), 1154–1159.
Hui, K.S., Wang, Y.J., Lajtha, A., 1983. Purification and characterization of an
enkephalin aminopeptidase from rat brain membranes. Biochemistry 22 (5),
1062–1067.
Johnson, G.D., Hersh, L.B., 1990. Studies on the subsite specificity of the rat brain
puromycin-sensitive aminopeptidase. Archives of Biochemistry and Biophysics
276 (2), 305–309.
Mendel, D.B., Cihlar, T., Moon, K., Chen, M.S., 1997. Conversion of 1-[((S)-2-
hydroxy-2-oxo-1,4,2-dioxaphosphorinan-5-yl)methyl]cytosine to cidofovir by
an intracellular cyclic CMP phosphodiesterase. Antimicrobial Agents and
Chemotherapy 41 (3), 641–646.
Osada, T., Ikegami, S., Takiguchi-Hayashi, K., Yamazaki, Y., Katoh-Fukui, Y.,
Higashinakagawa, T., Sakaki, Y., Takeuchi, T., 1999. Increased anxiety and
impaired pain response in puromycin-sensitive aminopeptidase gene-deficient
mice obtained by a mouse gene-trap method. Journal of Neuroscience 19 (14),
6068–6078.
Perry, C.M., Faulds, D., 1996. Valaciclovir. A review of its antiviral activity, pharma-
cokinetic properties and therapeutic efficacy in herpesvirus infections. Drugs 52
(5), 754–772.
Pescovitz, M.D., Rabkin, J., Merion, R.M., Paya, C.V., Pirsch, J., Freeman, R.B., O’Grady,
J., Robinson, C., To, Z., Wren, K., Banken, L., Buhles, W., Brown, F., 2000. Val-
ganciclovir results in improved oral absorption of ganciclovir in liver transplant
recipients. Antimicrobial Agents and Chemotherapy 44 (10), 2811–2815.
Schnebli, H.P., Phillipps, M.A., Barclay, R.K., 1979. Isolation and characterization of an
enkephalin-degrading aminopeptidase from rat brain. Biochimica et Biophysica
Acta 569 (1), 89–98.
Schulz, C., Perezgasga, L., Fuller, M.T., 2001. Genetic analysis of dPsa, the Drosophila
orthologue of puromycin-sensitive aminopeptidase, suggests redundancy of
aminopeptidases. Development Genes and Evolution 211 (12), 581–588.
Sengupta, S., Horowitz, P.M., Karsten, S.L., Jackson, G.R., Geschwind, D.H., Fu, Y.,
Berry, R.W., Binder, L.I., 2006. Degradation of tau protein by puromycin-sensitive
aminopeptidase in vitro. Biochemistry 45 (50), 15111–15119.
Smiley, M.L., Murray, A., de Miranda, P., 1996. Valacyclovir HCl (Valtrex): an acyclovir
prodrug with improved pharmacokinetics and better efficacy for treatment of
zoster. Advances in Experimental Medicine and Biology 394, 33–39.
Stoltze, L., Schirle, M., Schwarz, G., Schroter, C., Thompson, M.W., Hersh, L.B.,
Kalbacher, H., Stevanovic, S., Rammensee, H.G., Schild, H., 2000. Two new
proteases in the MHC class I processing pathway. Nature Immunology 1 (5),
413–418.
Kim, I., Chu, X.Y., Kim, S., Provoda, C.J., Lee, K.D., Amidon, G.L., 2003.
Identification of
a human valacyclovirase: biphenyl hydrolase-like protein
as valacyclovir hydrolase. The Journal of Biological Chemistry 278 (28),
25348–25356.
Laemmli, U.K., 1970. Cleavage of structural proteins during the assembly of the head
of bacteriophage T4. Nature 227 (5259), 680–685.
Lazarus, R.A., Benkovic, S.J., 1979. Mechanism of hydrolysis of phospho-
rylethanolamine triesters. Multiple catalytic effects of an intramolecular
amino group. Journal of the American Chemical Society 101 (15), 4300–
4312.
Lazarus, R.A., Benkovic, P.A., Benkovic, S.J., 1980. Mechanism of hydrolysis of phos-
phorylethanolamine diesters. Intramolecular nucleophilic amine participation.
Journal of the Chemical Society 2, 373–379.
Lowry, O.H., Rosebrough, N.J., Farr, A.L., Randall, R.J., 1951. Protein measurement
with the Folin phenol reagent. The Journal of Biological Chemistry 193 (1),
265–275.
McKenna, C.E., Kashemirov, B.A., Eriksson, U., Amidon, G.L., Kish, P.E., Mitchell, S.,
Kim, J.S., Hilfinger, J.M., 2005. Cidofovir peptide conjugates as prodrugs. Journal
of Organometallic Chemistry 690 (10), 2673–2678.
McKenna, C.E., Kashemirov, B.A., Peterson, L.W., Eriksson, U., Saejung, K., Kim, J.-S.,
Mitchell, S., Kijek, P., Hilfinger, J.M., Drach, J.C., 2006. Cidofovir and Foscarnet
peptide prodrugs. Antiviral Research 70, A37.
McLellan, S., Dyer, S.H., Rodriguez, G., Hersh, L.B., 1988. Studies on the tissue dis-
tribution of the puromycin-sensitive enkephalin-degrading aminopeptidases.
Journal of Neurochemistry 51 (5), 1552–1559.
Tobler, A.R., Constam, D.B., Schmitt-Graff, A., Malipiero, U., Schlapbach, R., Fontana,
A., 1997. Cloning of the human puromycin-sensitive aminopeptidase and evi-
dence for expression in neurons. Journal of Neurochemistry 68 (3), 889–897.
Towne, C.F., York, I.A., Neijssen, J., Karow, M.L., Murphy, A.J., Valenzuela, D.M.,
Yancopoulos, G.D., Neefjes, J.J., Rock, K.L., 2008. Puromycin-sensitive aminopep-
tidase limits MHC class I presentation in dendritic cells but does not affect
CD8 T cell responses during viral infections. Journal of Immunology 180 (3),
1704–1712.
Wachsman, M., Petty, B.G., Cundy, K.C., Jaffe, H.S., Fisher, P.E., Pastelak, A., Lietman,
P.S., 1996. Pharmacokinetics, safety and bioavailability of HPMPC (cidofovir) in
human immunodeficiency virus-infected subjects. Antiviral Research 29 (2,3),
153–161.
Wagner, G.W., Tavianini, M.A., Herrmann, K.M., Dixon, J.E., 1981. Purification and
characterization of an enkephalin aminopeptidase from rat brain. Biochemistry
20 (13), 3884–3890.
Weller, S., Blum, M.R., Doucette, M., Burnette, T., Cederberg, D.M., de Miranda, P.,
Smiley, M.L., 1993. Pharmacokinetics of the acyclovir pro-drug valaciclovir after
escalating single- and multiple-dose administration to normal volunteers. Clin-
ical Pharmacology and Therapeutics 54 (6), 595–605.