AM580, a stable benzoic derivative of retinoic acid, has powerful... leukemia selective cyto-differentiating effects on acute promyelocytic

From www.bloodjournal.org by guest on October 21, 2014. For personal use only.
1996 87: 1520-1531
AM580, a stable benzoic derivative of retinoic acid, has powerful and
selective cyto-differentiating effects on acute promyelocytic leukemia
cells
M Gianni, M Li Calzi, M Terao, G Guiso, S Caccia, T Barbui, A Rambaldi and E Garattini
Updated information and services can be found at:
http://www.bloodjournal.org/content/87/4/1520.full.html
Articles on similar topics can be found in the following Blood collections
Information about reproducing this article in parts or in its entirety may be found online at:
http://www.bloodjournal.org/site/misc/rights.xhtml#repub_requests
Information about ordering reprints may be found online at:
http://www.bloodjournal.org/site/misc/rights.xhtml#reprints
Information about subscriptions and ASH membership may be found online at:
http://www.bloodjournal.org/site/subscriptions/index.xhtml
Blood (print ISSN 0006-4971, online ISSN 1528-0020), is published weekly by the American
Society of Hematology, 2021 L St, NW, Suite 900, Washington DC 20036.
Copyright 2011 by The American Society of Hematology; all rights reserved.
From www.bloodjournal.org by guest on October 21, 2014. For personal use only.
AM580, a Stable Benzoic Derivative of Retinoic Acid, Has Powerful and
Selective Cyto-Differentiating Effects on Acute Promyelocytic
Leukemia Cells
By Maurizio Gianni, Marco Li Calzi, Mineko Terao, Giovanna Guiso, Silvio Caccia, Tiziano Barbui,
Alessandro Rambaldi. and Enrico Garattini
All-trans retinoic acid (ATRA) is successfully used in the cytodifferentiating treatment of acute promyelocytic leukemia
(APL). Paradoxically, APL cells express PML-RAR, an aberrant formof the retinoicacid receptortype a (RARa) derived
from theleukemia-specific t(15;17) chromosomal translocation. We show here that AM580, a stable retinobenzoic derivative originally synthesizedas a RARa agonist, is a powerful
inducer of granulocytic maturation in NB4, an APL-derived
cell line, and in freshly isolated APL blasts. After treatment
of APL cells with AM580 either alone or in combination with
granulocyte colony-stimulating factor (G-CSF), thecompound induces granulocytic maturation, as assessed by determination of the levels of leukocyte alkaline phosphatase,
CD11b. CD33, and G-CSF receptor mRNA, at concentrations
that are 10- t o 100-fold lower than thoseof ATRA necessary
t o produce similareffects. By contrast, AM580 is notso effective as ATRA in modulating theexpression of these differentiation markers in the HL-60 cell line and in freshly isolated
granulocytes obtained from theperipheral blood of chronic
myelogenous leukemia patients during thestable phase of
the disease. In NB4 cells, t w o other synthetic nonselective
RAR ligands are capable of inducingLAP asmuch as AM580,
whereas RARP- or RARy-specific ligands are totally ineffective. These results show thatAM580 is more powerful than
ATRA in modulating the expression of differentiation antigens only in cells in which PML-RAR is present. Binding experiments, using COS-7 cells transiently transfected with
PML-RAR and the normal RARa, show that AM580 has a
lower affinity than
ATRA for bothreceptors. However, in the
presence of PML-RAR, the synthetic retinoid isa much better
transactivator
of
retinoic acid-responsive element-containing promoters than the natural
retinoid, whereas, in the
presence of RARa, AM580 and ATRA have similar activity.
This may explain the strong cyto-differentiating potential of
AM580 in PML-RAR-containing leukemic cells.
0 1996 b y The American Society of Hematology.
T
HECYTO-DIFFERENTIATINGagentall-rmns
retifreshlyisolated APL blastsas well as on cultures of the
noic acid (ATRA)’ is
successfully used in the treatment APL-derived NB4” and HL-60” cell lines.
In basal conditions, APL cells express PML-RAR, RARa,
of acute promyelocytic leukemia (APL; M3 in the Frenchandthe retinoic acid related receptors known asRXRs,’?
American-British [FAB] classification).’.’ In this type of diswhereas expression of RARP and RARy is not o b ~ e r v e d . ’ ~
ease, ATRAcircumventsthe
differentiation block of the
leukemic blasts, causing them to mature from the promyelo- ATRA-induced granulocytic differentiation leads to the appearance of RARP mRNAs.14 Because ATRA doesnot show
cyte stage along the granulocytic pathway! The exquisite
any selectivity for the RAR subtypes””’ and it can interact
sensitivity of APL blasts to the cyto-differentiating action
with RXRs viaisomerization to 9-cis retinoic acid ( 9 4 s
of ATRA is paradoxical. In fact, as a consequence of the
RA), it is entirely possible thatactivation of any oneof these
typical t( 15;17) chromosomal translocation, APL cells exreceptors may underlie the differentiation process.
ATRApress PML-RAR, a functionally altered form of the nuclear
retinoic acid receptor known as RARcY.~”” ATRA-dependent dependent maturation of APL cells is incomplete and can
beenhanced by the addition of other cyto-differentiating
granulocyticdifferentiationisreproducible
on cultures of
agents, suchas granulocyte colony-stimulating factor (GCSF)14 and cell-permeable CAMP analogs.18-2’ With respect
From the Molecular Biology Unit, Centro Catullo e Danielu Borto this, we recently showed that leukocyte alkaline phosphagomainerio and the Laboratory of Drug Metabolism, Istituto di Ritase (LAP), a specific and restrictive marker for the postmicerche Farmacologiche “MarioNegri”, Milano; the Division of
totic
is not expressed
at
significant levels
Hematology, Ospedali Riuniti di Bergamo, Largo Barozzi, Bergamo;
in APL blasts on treatmentwith ATRAalone.However,
andthe Istituto di Ricerche Farmacologiche “Mario Negri” sede
incubation of the leukemic cells with combinations between
di Bergamo, Bergamo, Italy.
the retinoid andthetwoother
cyto-differentiating agents
Submitted June 14, 1995; accepted September 21, 1995.
leads
to
a
dramatic
induction
of
the
enzyme.l4.’’
Supported in part by Grants from the Consiglio Nazionale delle
In an attemptto definewhetherstimulation
of specific
Ricerche (CNR), Progetto Finalizzato ‘Ingegneria Genetica”, P m RAR isoformsresultsingranulocyticmaturation
of APL
getto Finalizzato “Biotecnologie e Biostrumentazione’ ’ and from the
cells, we evaluated the effectsof a series of compounds with
Associazione per la Ricerca contro il Cuncro (AIRC). The precious
different specificities for RARs on the expression of LAP
financial contribution ofthe Associazione Paolo Belli, Lotta allu
Leucemia to A.R. is also acknowledged. M.L.C. is a recipient of a
andother myeloidmaturationmarkers,
in the absence or
fellowship from “La via di Nutale”.
presence of G-CSF. In this report, we show that AM580, a
Address reprint requests to Enrico Garattini, MD, Molecular Biolretinobenzoicderivativeoriginally
developed as a RARaogy Unit, Centro Catullo e Daniela Borgomainerio, Istituto di Ricerspecific agonist, has a selective cyto-differentiating activity
che Farmacologiche “Mario Negri”, via Eritrea, 62. 20157 Milano,
on PML-RAR-containing APL blasts. In this leukemic cell
Italy.
type, thecompoundis much more potentthan ATRA in
The publication costs of this article were defrayed in part by page
inducing granulocyticdifferentiation. This selectiveaction
charge payment. This article must therefore be hereby marked
is correlated with the ability of AM580 to transactivate reti“advertisement” in accordance with 18 U.S.C.section 1734 solely to
noic-acid-responsive-element (RARE)-containing genes in
indicate this fact.
the presence of PML-RAR much better than in the presence
0 1996 by The American Society of Hematology.
of RARa.
0006-4971/96/8704-0028$3.00/0
1520
Blood, Vol 87, No 4 (February 15). 1996: pp 1520-1531
From www.bloodjournal.org by guest on October 21, 2014. For personal use only.
CYTO-DIFFERENTIATINGEFFECTSOF
1521
AM580 ON APL
Northern blotanalysis.
Total RNA was prepared from NB4,
HL-60, and freshly isolated APL cells according to a modification
Cell culture conditions and reagents. The NB4 APL cell line’
of the guanidium isothiocyanate/CsCl method and used for Northern
was a kind gift of Dr Michel Lanotte (Unit6 INSERM 301, “Geblot analysis.” The probes used for Northern blot analysis were a
netique cellulaire et moleculaire des leucemies”). HL-60 leukemic
full-length human liverhonekidney-type (LIBIK-type) alkaline
cells were obtained from the American Type Culture Collection
phosphatase cDNA (ATCC),’& the cDNA coding for the G-CSF
(ATCC; Rockville, MD). Cells were routinely seeded at 4 X lo5/
receptor,” and the cDNA coding for the human glucose-6-phosphate
mL in RPMI 1640 containing 10% fetal calf serum (FCS; GIBCOdehydrogenase (G6PDH):’ The various probes were labeled to a
BRL, Gaithersburg, MD). Cultures were free from mycoplasma as
specific radioactivity of 1 to 2 X 10’ cpndpg by using hexanucleotide
assessed using the Hoechst 33258 fluorescent dye system (Farbwerke
primers and [3zP]dCTP.3’
Hoechst AG, Frankfurt, Germany). ATRA and 8-bromo-CAMP(8Transacrivarion experiments. Simian COS-7 fibroblasts were
Br-CAMP) were purchased from Sigma (St Louis, MO). The
obtained from ATCC and routinely passaged in Dulbecco’s modified
compounds 4-(5,6,7,8-tetrahydro-5,5,8,8-tetramethyl-2-naphtamido) Eagle’s medium (DMEM) containing 10% FCS. Transient transfecbenzoic acid (AM580) and 4-(5,6,7,8-tetrahydro-5,5,8,8-tetra- tion experiments were performed according to a standard calciummethyl-2-anthraceny1)benzoic acid (CD367) were obtained from
phosphate coprecipitation procedure3’ using the following plasmids:
CIRD-Galderma (Valbonne, France). Ethyl p-[(E)-2-(5,6,7,8-tetpSG5-RARa9 (from Dr P.G. Pelicci, Perugia, Italy) and pSG5-PMU
rahydro-5,5,8,8-tetramethyl-2-naphtyl)-l-propenyll-benzoic acid
RARa9 containing the human RARa and PML-RAR cDNAs, respec(TTNPB) and 9 4 s RA) were obtained from Hoffmann-LaRoche
tively, under the control of the SV-40 early T-antigen promoter
(Nutley, NJ). Stock solutions of the various retinoids (lo-’ mom)
enhancer; TRE-TK-CAT, containing the chloramphenicol acetyl
were prepared in dimethylsulfoxide under dimmed light and stored
transferase gene under the control of a palindromic synthetic RARE
at -80°C and protected from light until use. Recombinant human
placed in front of the viral thymidine kinase gene’; pSVALBK (deG-CSF (specific activity, 10’ Ulmg protein) was from Amgen Inc
nominated LAP CAT in this report), containing approximately 4.5
(Thousand Oaks, CA).
kb of the 5‘ flanking region of the upstream exon of the LIBK-type
Preparation of freshly isolated APL and chronic myelogenous
alkaline phosphatase gene placed in front of the CAT reporterT3
leukemia (CML)cells. Peripheral blood leukocytes from 5 CML
(from Dr T. Kadesch, Philadelphia, PA); and pnlsLACZ containing
patients during the stable phase of the disease were purified to 90%
the bacterial @-galactosidasegene under the control of the early Thomogeneity from buffy coats by Ficoll-Hypaque gradient centrifuantigen of the SV40 enhan~er-promoter~~
(from Dr A. Weisz, Napoli,
gation and sedimentation on dextran, as described e 1 s e ~ h e r e . l ~ ~Italy).
~ ~ Briefly, 50 ng each of RARa or PML-RAR and RXRa were
Almost pure preparations of APL cells were obtained by dextran
cotransfected with 1 pg of the reporter gene (TRE-TK CAT or LAP
sedimentation of bone marrow aspirates from 4 patients with a classiCAT), 0.5 pg of pnlsLACZ, and pBluscript to I O pg. After leaving
cal form of APL (M3 according to the FAB classification) presenting
the DNA coprecipitate in contact with cells for 16 hours, fresh
the typical t( 15- 17) chromosomal translocation. These cell preparamedium (10% charcoal stripped fetal calf serum to eliminate endogetions consisted of more than 90% APL blasts as assessed by morpholnous retinoids) alone or fresh medium containing ATRA or AM580
ogy. Leukemic cells were resuspended in RPMI 1640 containing
was added and cells were further incubated for 36 hours. At the end
10% FCS and cultured in this medium in the various experimental
of the treatment, cells were harvested and processed for determinaconditions described in the report.
tion of CAT and P-galactosidase activity. CAT and P-galactosidase
Analysis of cell surface markers. NB4, HL-60, or APL cells
activities were measured on cell extracts according to standard procewere seeded at a concentration of 4 X ldlmL in RPMI 1640 condures” and as described,3brespectively. The results are expressed
taining 10%FCS and incubated for 4 days in the presence of medium
as relative CAT activity, which is the ratio of CAT activity produced
alone or medium containing retinoids and/or G-CSF. The number
by the reporter in cpm of acetylated chloramphenicol divided by
of CDllb+ and CD33+ cells and the mean associated fluorescence
the P-galactosidase activity produced by pnlsLACZ expressed in
were quantitated using a FACScan analyzer (Becton Dickinson,
absorbance units at 420 nm.
Mountain View, CA). Determination of surface markers was perPreparation of nuclear extracts and binding experiments. Nuformed by a direct immunofluorescence assay using the following
clear extracts from RARa, PML-RAR, and mock transfected COSfluorescence conjugated monoclonal antibodies purchased from Bec7 cells were prepared as described by Nervi et
Briefly,the cell
ton Dickinson: phycoerythrin (PE)-conjugated Leu15 (IgG2a; antipellet was resuspended in 5 mL PTG (5 mmol/L sodium phosphate,
CD1 Ib) and Leu M9 (IgG1; anti-CD33). Simultest Control (PE conpH 7.4, 10 mmol/L thioglycerol, 10%glycerol, aprotinin [0.15pmoU
jugates IgGl and IgG2a; Becton Dickinson) was used as a negative
L], leupeptin [2 mmol/L], and phenylmethylsulfonylfluoride [O. 1
control.14
mmol/L]; Boheringer) and homogenized with a Dounce homogeMeasurement of LAP activity. Approximately 1 X lo6NB4, HLnizer. The pellet, containing the cell nuclei, was washed twice with
60, and APL fresh leukemic cells or CML neutrophils were har2 mL of PTG buffer. The nuclear pellet was extracted in 10 mL
vested, pelletted by centrifugation at 400g for 10 minutes, washed
of a buffer containing 10 mmoUL Tris-HC1,pH 8.5, 2 mmol/L
once with 0.9% NaCl, and centrifuged again. The washed cell pellet
dithiothreitol, 10% glycerol, 0.8 m o m KCl, and the same protease
was resuspended in homogenization buffer (1 m m o E MgC12, 1
inhibitors as above; incubated for 1 hour in ice; and then centrifuged
mmol/L CaCI’, 20 mmol/L ZnCl,, 0.1 m o m NaCl, 0.1% [voUvol]
at 130,OOOg for 30 minutes. The resulting supernatant is referred to
Triton X-100, 50 mmol/L Tris/HCl, pH 7.4) and disrupted by vigoras the nuclear extract. Nuclear extracts were used either immediately
ous pipetting. The homogenate was used for the LAP assay, which
or after storage at -80°C.
was performed with p-nitrophenol phosphate (Sigma) as substrate
In routine binding assays, nuclear extracts (200 to 400 pL) were
according to the instructions of the manufacturer. LAP activity was
incubated with 1 nmol/L [3H]CD367(CIRD-Galderma; specific acnormalized for the content of protein in the sample. Proteins were
tivity, 52.8 Cilmmol) in the presence or absence of varying concenmeasured according to the Bradford methodz6 using BSA fraction
trations of CD367, AM580, or ATRA as cold competitors. The
V (Sigma) as a standard. One unit of LAP activity is defined as the
preparations were incubated for 18 hours at 4°C and directly loaded
amount of enzyme capable of transforming I nmol of substrate in
on a Superose 6HR 10/30 (Pharmacia, Uppsala, Sweden). Chroma1 minute at 25°C. Enzyme assays were performed in conditions of
tography was performed in isocratic conditions using PTG buffer
linearity relative to the substrate and tothe concentration of proteins.
containing 0.4 mol/L KC1 at room temperature at a flow rate of 0.4
MATERIALS AND METHODS
’
From www.bloodjournal.org by guest on October 21, 2014. For personal use only.
1522
9-CISRA
GlANNl ET AL
TTNPB
ATRA
CD367
c0437
/
H
9
on
mL/min. Fractions (0.4 mL each) were collected and the radioactivity
in each fraction was determined by liquid scintillation counting after
the addition of 5 mL Filtercount (Packard, Meriden, CT). Although
most of the saturation or competition binding assays were performed
with this methodology, we also performed experiments with a faster
technique that gives similar results and allows the simultaneous
handling of a large number of sample^.^' Briefly, nuclear extracts
were loaded over PDlO desalting columns (Pharmacia) previously
washed consecutively with 15 mL phosphate-buffered saline and 10
mL PTG buffer containing 0.4 m o m KC1 and eluted with the same
buffer atroom temperature. The first 2 mLof the eluted buffer
were discarded, and the following 2.5 mLwas collected andthe
radioactivity measured by liquid scintillation counting as above. Linear least square analysis of the scatchard plot was performed with
the aid of the computer program BDATA-EMF (Vanderbilt Medical
Center, Nashville,
The retinoid concentrations that inhibit
50% of the total specific CD367 binding were calculated using the
nonlinear least squares regression analysis program ALLFIT (Vanderbilt Medical Center).”
Determination of intracellular levelsof AM580 and ATRA in NB4
and HL-60 cells. ATRA and AM580 were extracted from NB4
and HL-60 cells with acetonitrile, after adding the internal standard
and quantified by high-performance liquid chromatography with UV
detection as d e s ~ r i b e d . ~ Drug-free
.~’
cells and cells containing
known amounts of ATRA or AM580 were analyzed concurrently
with each set of samples. Standard calibration curves were constructed by linear least squares regression analysis of the plot of the
peak-height ratios between the compounds and the internal standard
versus their concentrations in biologic samples. The limit of detection, precision, and reproducibility were as previously described for
plasma analysis?’ Area under the curve (AUC) values were calculated by computer-assisted integration of the intracellular levels of
AM580 or ATRA measured at various time after the addition of the
two compounds to the culture medium.
RESULTS
Cyto-differentiating effects of AM580, ATRA, and other
retinoids in NB4 and HL-60 cells. We determined the effects of AM580, CD2019, and CD437, three specific ligands
for RARa, RARP, and RARy,” respectively, on the granulocytic maturation of NB4 cells. In addition, wetested
CD367” and TTNPB;’ two RAR-specific ligands with no
selectivity for the various isoforms of this type of receptor,
and 9-cis RA, an isomer of ATRA that interacts with both
RAR and RXR i s o f o r m ~ .The
~ ~ chemical structure of the
various retinoids is shown in Fig 1 . Granulocytic maturation
of NB4 and HL-60 cells was initially studied by determining
the levels of LAP enzymatic activity after treatment with the
various retinoids in the presence of optimal concentrations
Fig 1. Chemical structure of ATRA and the other
retinoids used in this study. The chemical structure
of ATRA, 9-cis RA, AM580, lTNPB, CD367, CD2019,
and CD437 is indicated.
of G-CSF.I4 In our experimental conditions, when cells are
treated with retinoids alone (at any given concentration) or
G-CSF alone (10 ng/mL), no increase in LAP expression is
observed. As shown in Fig 2A, in the presence of G-CSF,
treatment of NB4 cells with ATRA, at concentrations between
and
mom, leads to a dose-dependent and
robust increase in LAP enzymatic activity. Maximal induction of the enzyme is observed after incubation with 10”
m o m ATRA. A similar dose-response curve is evident when
ATRA is substituted by 9 4 s RA. Interestingly, TTNPB,
CD367, and AM580 have very similar dose-response curves
and cause maximal induction of LAP at concentrations that
are two logarithms lower than those of ATRA and 9-cis
RA necessary to produce the same effect. Furthermore, the
maximal level of induction is approximately threefold higher
after exposure to the three synthetic retinoids than after exposure to ATRA or 9-cis RA.At
mol& ATRAandall
the other synthetic retinoids increase LAP enzymatic activity
submaximally, which is the consequence of a mild cytotoxic
effect. The compound CD2019 induces significant LAP enzymatic activity only at the highest doses testedand
lo-’ mom), and the level of induction is several fold lower
than that caused by ATRA or 9 4 s RA. CD437 is totally
inactive in this assay (data not shown). As shown in Fig 2B,
in the presence of G-CSF (10 ng/mL), treatment of HL-60
cells with ATRA at concentrations between 10“” and 1 0 h~
mom causes an almost linear and dose-dependent increase
in the levels of LAP activity. However, at optimal concentrations of ATRA, the absolute amounts of LAP induced in
HL-60 are threefold lower than those observed in NB4 cells.
Up to
mom, a similar LAP induction profileis observed when HL-60 cells are grown in the presence of GCSF and 9 4 7 RA, TTNPB, or AM.580. CD367 is slightly
more effective than ATRA or the other three retinoids at
10”’ and lo-’ m o m , although this effect is lost at higher
concentrations. At
m o m , ATRA is
more active than
all the other retinoids and this higher potency is maintained
at lo-’ mol&, although, at this concentration, a submaximal
increase in LAP is observed. Up to lo-’ mom, CD2019
and CD437 (data not shown) are totally inactive in modulating the levels of LAP enzymatic activity. Taken together,
the data indicate that ATRA, in combination with G-CSF,
is the most effective retinoid in inducing LAP in HL-60.
In further experiments, we focused our attention on
AM580 and we characterized the activity of this retinoid
relative to that of ATRA. This choice was suggested by the
From www.bloodjournal.org by guest on October 21, 2014. For personal use only.
CYTO-DIFFERENTIATINGEFFECTS OF AM580 ON APL
1523
Fig 2. Induction of NB4 and HL-60 LAP enzymatic activity by various retinoids in combination with G-CSF. NB4 (A) or HL-60 (B) were seeded
at 4 x lo6 cells/mL and treated with G-CSF (10 ng/mL) or with G-CSF and the indicated concentrations of ATRA (A),9 4 s RA (W), CD2019 (A),
lTNPB (U), CD367 (0)
CD2019
.
( 01,or AM580 (0)for 4 days. Cells were collected and processed for the determination of LAP enzymatic
activity. Treatment of NB4 or HL-60 cells for 4 days with thevarious retinoids at theindicated concentrations inthe absence of G-CSF did not
result in theexpression ofdetectable levels ofLAP activity (c10 mU/mg protein). Each experimental value is the mean 2 SD of three separate
cultures.
fact that CD367 and TTNPB are very strong and toxic retinoids in vivo,' whereas AM580 has a better toxicologic profile (unpublished results).
In APL cells, we showed that LAP is induced not only
by the combination of ATRA and G-CSF but also by the
combination between the retinoid and CAMP analogs?'
Thus, expression of LAP was studied in NB4 and HL-60
cells, in the presence of various concentrations of AM580
or ATRA and an optimal concentration of 8Br-CAMP (1
mmoVL). Inthese experimental conditions, treatment of NB4
cells with AM580 results in the induction of LAP at concentrations 100-fold lower than those necessary to obtain the
same effect with ATRA. In HL-60 cells the same combinations are ineffective2' (data not reported). This shows that
induction of LAP enzymatic activity by AM580 is independent of the second cyto-differentiating stimulus used in combination with the retinoid.
Figure 3A shows that, in the presence of G-CSF (10 ng/
mL), induction of LAP enzymatic activity in NB4 cells by
AM580 and ATRA is the result of an increase in the steadystate levels of the corresponding transcript. Consistent with
the results obtained by measurement of LAP activity,
AM580 is substantially more active than ATRA in causing
accumulation of the corresponding mRNA. In the presence
of G-CSF, AM580 (at
m o m ) produces a remarkable
induction in LAP mRNA, whereas, in the same experimental
conditions and at the same concentration, ATRA is totally
ineffective. At a concentration of10"' m o m , AM580 and
ATRA, in combination with G-CSF, induce almost the same
level of LAP transcript. The Northern blot does not show
LAP mRNA upregulation on treatment of cells with lo-'
m o m CD2019. As shown by Fig 3B, in HL-60 cells, LAP
mRNA accumulation is more dramatic in the presence of
ATRA + G-CSF than in the presence of AM580 + G-CSF.
In fact, at
m o m and lo-' m o m , ATRA causes a 10fold and threefold higher upregulation of LAP transcript than
AM580. This finding is again in line with what observed at
the level of LAP enzymatic activity.
To evaluate the effects of AM580 and ATRA on other
granulocytic maturation-associated markers, we determined
the level of expression of CD1 lb, CD33, and G-CSF receptor mRNA in NB4 and HL-60 cells. As shown in Fig 4 (left
panels), CD1 l b is not expressed in NB4 in basal conditions,
whereas CD33 is present on the majority of blasts. After
treatment with ATRA for 4 days, the number of NB4 cells
expressing C D l l b increases, starting from
m o m , and
is maximal at
m o l L Relative to what is observed with
ATRA, the dose-response curve of AM580 for the surface
expression of CD1 l b is shifted towards the left by more than
2 logs. In fact, the number of CD1 lb+ cells is already high
at
mol/L and is maximal at
mol/L.
In the presence
of ATRA and AM580, G-CSF enhances the surface expression of CD1 l b relative to what is observed with the retinoids
alone. In the case of AM580, this effect is evident only at
10"' mol/L, whereas, in the case of ATRA, enhancement
is observed at levels greater than lo-' m o m . Notice that,
even in the presence of G-CSF, maximal upregulation of
CD 11b by ATRA is lower than that observed in the presence
of A M 5 8 0 alone. After 4 days in the presence of G-CSF,
AM580 is more potent than ATRA in decreasing the amounts
of CD33 present on the surface of NB4 cells, although the
difference between the effects caused by the two retinoids
is less remarkable than that observed on CD1 l b expression.
Data on CD33 levels after treatment of NB4 with G-CSF
(10 ng/mL), ATRA (10"' to
mom), or AM580 (10"'
tomol/L)
alone are not documented, because they are
not significantly different from what is observed in control
conditions. As shown in the right panels of Fig 4, in basal
conditions, HL-60 cells are CDllb- and CD33+. ATRA
causes maximal induction of CD1 1b at
m o m , whereas
AM580 effectively upregulates the surface marker only at
m o m . However, even at this concentration, the number
From www.bloodjournal.org by guest on October 21, 2014. For personal use only.
GlANNl ET AL
1524
.
LAP
. E
a
~
.
LAP
Fig 3. Effects of ATRA and other retinoids in combination with G-CSF on the levels of LAP mRNA in NB4 and HL-60 cells. NB4 (A) or HL60 (B) cells (4 x 105/mL) were treated for4 days with medium alone (medium), medium containing 10 nglmL G-CSF, and medium containing
the indicatedconcentrations of ATRA, CD2019, or AM580, each alone or in combination with G-CSF. Total RNA was extracted andloaded (10
p g for each lane) on a 1% formaldehydelagarose gel. RNA was transferred by capillarity onto a nylon membrane that was
subsequently
processed for Northern blot analysis. The same membrane was sequentially hybridized with LAP and GGPDH cDNAs. The molecular weight
of LAP mRNA is approximately2.5 kb and that ofG6PDH is approximately2.6 kb. Autoradiograms were quantitated by densitometry
and the
intensity of each LAP mRNA signal was normalized for the intensity of the relative G6PDH mRNA signal. Results are indicated above each
autoradiogram and are expressed as the percentage of the highest value (AM580 10.' mol/L + G-CSF in the case of NB4 and ATRA
moll
L + G-CSF in thecase of HL-60 cells) taken as 100.
of CD1 Ib' cells is significantly lower than that observed in
the presence of ATRA. In this cell line. surface expression
of CD33 is left unaltered after challenge with the combinations of G-CSF andATRA or AM580. In bothNB4and
HL-60 cells, the mean associated fluorescence of CD1 Ib'
and CD33 cells is very similar in all the experimental conditions tested (data not shown).
In NB4 cells, maximal expression of G-CSF receptor
mRNA is observed on treatment with ATRA a t high concentrations." The addition of G-CSF to the medium containing
ATRA does not further enhance the accumulation of the
transcript." Figure SA shows that 4 days of treatment of NB4
cells with AM580 (at IO-' mol/L) leads to an approximately
sixfold increase in the steady-state levels of the transcript
coding for the G-CSF receptor. whereas. at the same concentration, ATRA does not significantly upregulate thebasal
level of expression of this mRNA. At
mol/L, ATRA
and AM580 are both effective in upregulating the expression
of the G-CSF receptor transcript, whereas G-CSF (at 10 ng/
mL) and CD2019 (at
and 10" mol/L), which are used
as negative controls. are totally inactive in this respect. In
HL-60 cells (Fig SB), only ATRA at 1 0-5mol/L iscapable of
significantly augmenting (approximately S-fold) the relative
amounts of the cytokine receptor mRNA. G-CSF (at 10 ng/
mL) and AM580 (at IO-' and
mol/L) do not change
the basal level of expression of the mRNA.
Taken together, the data so far presentedshowthat
L
AM580 is a more powerful cyto-differentiating agent than
ATRA in NB4. whereas it islesspotentthanthenatural
retinoid in HL-60 cells. The selective cyto-differentiating
action of AM580 on NB4 is not accompanied by a similar
selectivity on cell growth. In fact. at the concentrations tested
(IO-' mol/L for NB4and 10"' moVL for HL-60), AM580
and ATRA show equivalent levels of growth inhibition in
bothcell lines. In addition. pharmacodynamic effects do
not explain the phenomenon. because the peak intracellular
levels, as well as the rate of accumulation and disappearance
of the two compounds, are similar in both cell lines (data not
shown). Indeed. over a 4-day period. the calculated AUCs of
AM580 and ATRA are 13.4and 10.5 ng/105 cells per day
for NB4 and 12.4 and 6.1
@IO5 cells per day for HL-60,
which does not correlate with the relative biologic activity
of AM580 and ATRA in the two cell types.
C?.to-differentintinR~ efects
ofAM580 and ATRA in,freshly
isolated A P L andCML cells. AsNB4 expresses PMLRAR. whereas HL-60 does not express it. the data obtained
in the two APL experimental models suggest thatAM580
cyto-differentiating activity may be higher in cells containing
the aberrant retinoic acid receptor. To support this hypothesis. we studied LAP enzymatic activity in freshly isolated
APL cells and CML granulocytes challenged in vitro with
AM580andATRA alone or in combination with G-CSF.
CML granulocytes represent a useful source of myeloid cells
lackingPML-RAR. because they express R A R a (unpub-
From www.bloodjournal.org by guest on October 21, 2014. For personal use only.
CYTO-DIFFERENTIATINGEFFECTSOF
-
1525
AM580 ONAPL
NB4
l00
HL60
100 l
v)
Q
50 -
n
0
8
a
00
-10
-8
-6
-4
3
0
-10
-8
-6
-4
100
50
0
Fig 4. Effect of AM580, ATRA, and the combination of the two compounds with G-CSF on the expression of the myeloid surface markers
CDllb and CD33. NB4 cells (left panels) and HL60 cells (right panels) were seeded at 4 x lo6 cellslmL and incubated for 4 days in medium
containing increasing concentrationsof ATRA (triangles) or AM580 (circles) either in the absence (open symbols) or in the presence of 10 ngl
mL G-CSF (solid symbols). The number of cellspositive for the expression of CD1l b (upper panels)and CD33 (lower panels) were quantitated
by flow cytometry. For each experimental point, appropriate controls consistingof cells decorated with an irrelevant monoclonal antibody of
the same isotype
~. were performed. In these controls, less than 2% of the cells show background fluorescence;thus the results are not presented
in the figure.
lished results) and respond to the challenge with combinations of ATRA and G-CSF with an induction of LAP.I4
As shown in Fig 6, in the presence of 10 ng/mL G-CSF,
AM580 maximally induces LAP enzymatic activity at concentrations that are one logarithm lower than those necessary
to obtain the same effect with ATRA. However, individual
variability in the shape of the dose-response curve and in
the concentration of the two retinoids causing maximal induction of LAP is evident. Indeed, APL cells from patients
no. 2 and 3, respectively, are the most and the least sensitive
to the effects of both AM580 and ATRA. After treatment
of CML granulocytes for 3 days with 10 ng/mL G-CSF
and increasing concentrations of ATRA or AM580, a dosedependent induction of LAP is evident. In all the cases analyzed, ATRA is equally effective (patients no. l and 4) or
more potent (patients no. 2, 3, and 5 ) than AM580 in inducing LAP enzymatic activity. In freshly isolated APL cells,
induction of LAP by the combination of G-CSF and ATRA
or AM580 is the consequence of an increased accumulation
of the corresponding transcript, as shown by Northern blot
experiments performed on RNA extracted from the blasts of
APL patient no. 4 (data not shown).
As shown in Fig 7, AM580 is more effective than ATRA
in inducing the surface expression of C D l l b in APL cells
derived from patient no.1.In fact, AM580 at
m o m is
sufficient to cause the appearance of CD1 l b on the majority
of APL cells and higher concentrations of the retinoid do
not give rise to a further recruitment of positive cells or an
increase in the mean associated fluorescence. By contrast,
for ATRA, a concentration of lo-* m o m is necessary to
induce maximal expression of C D l l b on the majority of
APL cells. In this particular APL case, CD33 is not modulated by either AM580 or ATRA, and CD1 l b is a more
sensitive differentiation marker than LAP. With respect to
the last point, C D l l b is upregulated at concentrations of
both AM580 and ATRA that are lower than the respective
concentrations of the two retinoids necessary to induce LAP
in combination with G-CSF (compare Fig 7 with Fig 6). A
similar analysis was conducted on a second APL case (patient no. 4), where maximal induction of CD1 l b was observed at
m o m , in the presence of AM580, and at
m o m in the presence of ATRA (data not shown).
Binding of AM580 and ATRA in the presence of RARa
and PML-RAR. The selective cyto-differentiating action of
AM580 on APL cells may be related to a peculiar capacity
of the retinoid to interact with PML-RAR. To test this hypothesis, we evaluated the ability of the retinobenzoic compound to displace the binding of CD367 on PML-RAR and
RARa, because direct binding experiments are prevented by
the lack of availability of radiolabeled AM580. CD367 was
selected as the ligand, because it binds to the various forms
of RAR with the same affinity as ATRA. In addition, this
synthetic compound is more stable and gives lower nonspecific binding than the natural retin~id.'~
Binding experiments
were performed on nuclear extracts of COS-7 cells
transfected with the cDNA encoding PML-RAR or RARa.
The free form of radiolabeled CD367 was separated from
that boundto PML-RAR or RARa on size exclusion chroma-
From www.bloodjournal.org by guest on October 21, 2014. For personal use only.
GlANNl ET AL
1526
50L
0
G-CSF-R
GGPDH
(JI)o.()[email protected]
GGPDH
5
cp
0
7-
Q
U
Fig 5. Effects of AM580, ATRA. CD2019, and G-CSF on the levels of G-CSF receptor mRNA in NB4 and HL-60 cells. NB4 (left panel) or HL60 (right panel) cells (4 x 105/mLI were treated for 4 days with medium alone (medium), medium containing 10 nglmL G-CSF, and medium
containing the indicated concentrations of AM580, ATRA, or CD2019. Total RNA was extracted and loaded ( l 0 p gfor each lane) on a 196
formaldehydelagarose gel. RNA was transferred by capillarity onto a nylon membrane that was subsequently processed for Northern blot
analysis. The same membrane was sequentially hybridized with the G-CSF receptor (G-CSF-R) and G6PDH cDNAs. The molecular weight of
the G-CSF receptor mRNA is approximately3.0 kb and that of
GGPDH is approximately2.6 kb. Autoradiograms were quantitated by
densitometry and the intensity of each G-CSF receptor mRNA signal was normalized for the intensityof the relative G6PDH mRNA signal. Results are
indicated above each autoradiogram and are expressed as the percentage of the highest value (AM580 10 mol/L in the case of NB4 and
ATRA
mollL in the case of HL-60 cells) taken as100.
tography. Typicalchromatogramsareshown
in Fig 8A.
RARa-bound CD367 elutes in a discrete peak at an apparcnt
molecular mass of around 50 to 60 kD (fractions No. 3136). as assessed by calibrationof the column with appropriate proteins of known molecular weight (data not shown).
This is the expectedmolecular weightof RARa.37..1.1Thc
elution profile of PML-RAR-bound CD367 is more complex and consists of two peaks. A major peak. representing
approximately 80% of the bound radiactivity. which elutes
with the void volume of the column (fractionsno. 2 I through
27), and a minor one that elutes at a molecular mass of
approximately 100 kD (fractions no. 30 through 34). Thc
first PML-RAR peak represents aggregates of thercceptor,37.4-1.45 and the
second
one is probably
the
monomeric
form of the p~otein.'~..'.'We observed that the proportion of
the two PML-RAR species is always the samc. as determined
in a large series of independent experiments. In addition. the
saturation curve of radiolabeled CD367 is the same for the
two peaks (see below). Finally. CD367-associated radioactwopeaks ofprotein by cold
tivity is displacedfromthe
CD367, ATRA. and AM580 with :I similar Ki. Thus. determination of PML-RAR binding constants of the vnrious rctinoids is thesame, regardlessof the factthat analysis is
conducted on peak 1. peak 11, or the combinations of the two
peaks. The nuclear extract obtained from untransfected COS7 cells bind less than I % of the CD367 bound to that derived
from cells trnnsfected with PML-RAR or RARa. Radiolabeled CD367 is completely displacedfrom both retinoid
receptors by a 200-fold excess of the cold ligand. With the
use of thisbindingassay.
we generatedsaturation curves
that showed that the Kd for the binding of tritiated CD367
to both PML-RAR and RARcu is the same and is approxiof radiolabeled
mately I nmol/L.Usingthisconcentration
CD367. the binding of this retinoid to PML-RAR and RARa
was displaced by increasing concentrations of ATRA or
AMS80. TypicalHill's plots forATRAandAM580
are
shown i n Fig 8R. The Ki (mean 2 SD of 3independent
experiments) of AM580 and ATRAfor the inhibition of
CD367 binding to PML-RAR and RARa. respectively. are
shown in Table 1 . ATRA has an affinity for PML-RAR that
isalmostfourfold
lowcr than that for RARcu. whereas no
statisticallysignificant difference between thebindingof
AM580 to the two receptors is observed. In both the cases
of PML-RAR and RARcu. the Ki of AM580 is higher than
that of ATRA. These data strongly suggest that ATRA is a
slightlybetter ligand to both PML-RARand RARa than
AM580 and that both compounds bind better to the aberrant
than the wild-type receptor.
From www.bloodjournal.org by guest on October 21, 2014. For personal use only.
1527
CYTO-DIFFERENTIATINGEFFECTSOFAM580ONAPL
Transactivation potentialof AM580 and ATRA in thepresence of RARa and PML-RAR. The transactivation potential
of AM580 and ATRA in the presence of both PML-RARand
RARa was tested on 3 different RARE-containing promoters
placed in front of the bacterial CAT gene used as a reporter.
The three promoters contain different types of RARE: 2
copies of a palyndromic artificial RARE in the case of TRETK CAT: an ill-defined but functional RARE in the case
of LAP CAT:'
and 2 copies of the RARP2 RARE in the
case of &RARE CAT.47Notice that the LAP promoter is
directly upregulated by ATRA, even though the LAP enzyme
and the respective transcript do not accumulate in NB4 cells
400] PT1
400
l
300 -
300
200.
200
-8
-6
PT2
PT 1
0
0
4
0 -10
-8
-6
600.
400.
200-10
-8
120] P T 3
-6
0-
0
-10
-8
-6
0
-10
-8
-6
I
80.
40,
0-
el
0
50
100
h
loo:
0
0
-10
0
75
25
CML
APL
100
-10
-8
-6
lWMl
7500
t;
PT 5
Fig 6. Effects of AM580 and ATRA alone or in combination with
G-CSF on LAP enzymatic activity in APL blasts andCML granulocytes.
APL blasts (left panels) from 4 patients or CML granulocytes (right
panels,)from 5 patients were seeded at 4 x lo5cells/mL and treated
concenfor 4 days with medium, G-CSF ( l 0 ng/mL), and the indicated
trations of AM580 (circles) or ATRA (triangles) either in the absence
(open symbols) or in the presence (solid symbols) of 10 ng/mL GCSF. Cells were collected and
processed for the determination LAP
of
enzymatic activity. Each experimental value is the mean 2 SD of
three separate cultures.
Fig 7. Effect of ATRA and AM580 on theexpression of the myeloid
surface markers C D l l b and CD33 in freshly isolated APL cells.Leukemic cells (4 x 106/mL) isolated from1 APL patient (patient no. 1 in
Fig 6 ) were incubated for 4 days in medium alone (upper panel) or
in medium containing the indicatedconcentrations of ATRA (triangles) or AM580 (circles). The number of cells positive for theexpression ofC D l l b (solid symbols) and CD33 (open symbols) were quantitated by flow cytometry. For each experimental point, appropriate
controls consisting ofcells decorated with an irrelevant monoclonal
antibody of the same isotype wereperformed. In these controls, less
than 2% of the cells show background fluor-nce;
thus the results
are not presented in the figure.
unless optimal concentrations of G-CSF or CAMP analogs
are simultaneously present in the growth medium (Fig 2 and
Sat0 et aIz4). To perform these experiments, COS-7 cells
were transiently transfected with PML-RAR or RARa in the
presence of the appropriate reporter and subsequently treated
with increasing concentrations of AM580 or ATRA. Typical
results of these experiments are shown in Fig 9. In the presence of PML-RAR (left panels), AM580 has a better transactivating potential than ATRA on the three reporter genes,
whereas the two retinoids have similar activity in the presence of RARa (right panels). A summary of the results
obtained in several experiments are shown in Table 1. If
results are expressed as EC5,,, ie, the concentration of the
retinoid giving half of the maximal induction of each promoter, it is evident that ATRA transactivates each promoter
in a similar fashion either in the presence of PML-RAR or
in the presence of RARcu. By contrast, PML-RAR mediates
the activity of AM580 better than RARa, regardless of the
promoter taken into consideration. For LAP CAT, TRE-TK
CAT and &RARE CAT, the transactivation potential of
AM580 is approximately 8-, 20-, and 3-fold higher, respectively, in the presence of PML-RAR than in the presence of
RARa.
AM580 selectivity for the transactivation of PML-RAR
From www.bloodjournal.org by guest on October 21, 2014. For personal use only.
GlANNl ET AL
1528
8ooo ]
RARa
6000 -
4000 h
2
ii
2000-
v
U
c
3
0
0-
a
0
10
20
30 40
50
60
(D
m
c3
0
-10
-9
-8
-7
log[ATRA] (M)
-6
-5
0
-10
-9
-8
-7
log[AM580] (M)
-6
-5
0
0
0
1
m
-
2000 -
1000 -
0
l
0
10
20 30 40
Fraction No.
50
60
Fig 8. Effects of AM580 and ATRA o n t h e b i n d i nof
g radiolabeled CD367 t o PML-RAR and RARa. (A) Size exclusionHPLC analysis of nuclear
extracts prepared from COS-7 cells transiently transfected with RARa or PMLRAR expression vectors. Transient transfection of COS-7 cells
with either n o DNA (A),pSG5/RARa ( W , or PSGS/PML-RAR ( W ) was performed with the calcium phosphate coprecipitation method. Nuclear
extracts (200 pL in t h e case of RARa and400 pL in t h e case of PML-RAR) were incubatedwith 1 nmolfL t3H1CD367 in the absence (W and 0)
or in the presence (A)of a 200-fold excess of unlabeled CD367 for 18 hours at 4°C. The samples were then fractionated over a superose
6HR
10/30 size exclusion column (Pharmacia) using
PTG buffer containing 0.4 moVL KC1 as eluent at a flow rate of0.4 mL/min. RARa-boundCD367
is collectedin fractions 31 through36, and PML-RAR-boundCD367 is collectedin t w o distinct peaks at fractions
21 through 27 and 30 through
34. Free CD367 is recoveredin fractions 43 through 47. (B) Competition binding curves of 13H1CD367 with ATRA and AM580. Nuclear extracts
of COS-7 cells transfected with PML-RAR (A)or RARa (0)
were incubated with 1 nmol/L radiolabeled CD367 in the presence of the indicated
concentrations of cold ATRA (upper panel) or AM580 (lower panel).
relative to RARa seems to be an intrinsic characteristic of
this compound, although a low EC,,, for the transactivation
of RARE-containing genes in the presence of the aberrant
retinoic acid receptor is shared by TTNPB and CD367. In
fact,inthepresence
of PML-RAR,TTNPBandCD367
transactivate TRE-TK CAT, withECSosof 0.4 and 0.5 X
lo-' mol/L, respectively, whereas, in the presenceof RARa,
the twocompoundstransactivatethesame
reporter gene,
withEC5(]s of 0.03and 0.3 X
mol/L, respectively(resultsarethe mean of twoexperiments, withECSovalues
varying less than 20% between each other). Thus, TTNPB
shows inverseselectivity relative to AM580, whereas CD367
is a nonselective transactivator of PML-RAR and RARa.
DISCUSSION
In this report, we show that AM.580, a synthetic benzoic
derivative of ATRA, originally developed as a RARa-speC;fic agonist,'5.1h.4X 1s considerably more active than the
natu'
ral retinoid in causing granulocyticmaturation of APL cells.
This is observed on a series of differentiation markers that
can be modulated by retinoids alone or by the combination
of these compounds and other differentiating agents. Cytodifferentiation is observed at concentrations of AM580 that
are atleast onelogarithmlower
than those necessary to
obtain similar results with ATRA. In addition, the level of
induction of a series of markers attained at low concentrations of the retinobenzoic derivative cannot beobtained even
with high concentrations of ATRA. AM580 shows cosiderable cell specificity, because it is more active than ATRA in
NB4 and freshly isolated APL promyelocytes and muchless
active than the naturalretinoid in HL-60cellsandCML
granulocytes. Notably, ATRA is more powerful than AM580
in two other nonhematopoietic cell lines. In F9 teratocarcinoma cells, ATRA is more effectivethan AM580 in inducing
tissue plasminogen activator and L/B/K-type alkaline phosphatase.'"'' Furthermore, in L929 fibroblasts, when associated to CAMP analogs, thenatural retinoid is more powerful
than the synthetic compound in inducing L/B/K-type alkaline
From www.bloodjournal.org by guest on October 21, 2014. For personal use only.
CYTO-DIFFERENTIATINGEFFECTSOF
1529
AM580 ON APL
phosphatase activity (unpublished results). Whereas freshly
isolated APL promyelocytes and NB4 cells express PMLRAR besides the normal RARa protein, CML granulocytes,
HL-60, F9, and L929 cells do not synthesize PML-RAR.
Taken together, all these results show that, whereas AM580
is an active retinoid in cells containing the normal RARa,
the potential of this compound is remarkably enhanced in
PML-RAR-expressing cells.
PML-RAR is believed to have an important role in causing
the granulocytic maturation of APL cells observed in the
presence of ATRA." Thus, it is conceivable that the aberrant
retinoic acid receptor may represent the primary target for
the activity of AM580 in APL cells. Our data show that both
ATRA and AM580 have better binding affinities for PMLRAR than for RARa. However, these differences in affinity
are small. In addition, the Ki of ATRA for CD367 binding
to PML-RAR is lower than that of AM580. This suggests
that differences inPML-RARaffinity
do not explain the
different ability of the two retinoids to induce granulocytic
maturation in APL cells.
We observe a strong correlation between the transactivation of PML-RAR by AM580 and the cyto-differentiating
effect of the retinoid on APL cells. In fact, whereas three
different reporter genes are similarly activated by ATRA
both in the presence of RARa and PML-RAR, AM580 activates the same reporters much better in the presence of the
aberrant receptor. Furthermore, AM580 transactivates PMLRAR more potently and at significantly lower concentrations
than ATRA, although the strength of the effect depends on
the reporter taken into consideration. The specific interaction
between AM580 and PML-RAR may, at least partially, explain why this compound is such an active cyto-differentiating agent in APL cells and why its activity is superior to
Table 1. Binding Constant and Transactivation Potential
of ATRA and AM580
(ECw)
Affinity
(Ki)
ATRA
PML-RAR
3.4
RARa
12.6
AM580
PML-RAR
15.0
RARa
32.5
2 1.4
2 3.1*
Lap CAT
1.30 2 0.60
3.80 ? 1.00
2 5.2
0.16
2 0.05
0.4
2 10.8
1.30
2 0.49t
TRE-TK CAT
(ECd
pI-RARE CAT
IECd
4.3 ? 1.3
5.5 2 1.4
7.0 2 3.5
9.3 2 0.9
2 0.1
2.8
7.8 2 3 . l t
2 1.9
9.0 2 1.0t
The affinity (Ki) of ATRA and AM580 for PML-RAR and RARa is
defined as the concentration (in nanomoles) of each compound producing 50% displacement in the [3H]CD367 binding to each retinoic
acid receptor, determined as detailed in the Materials and Methods
from curves similar tothose shown in Fig 8. The transactivation poten,,
ie, the concentration (in
tial of AM580 and ATRA is given in EC
nanomoles) of each compound necessary to produce 50% of themaximal activation of each CAT construct, determined from plots similar
to those shown inFig 9. Eachvalue is the mean 2 SD of three independent experiments.
* Significantly higher ( P< .05) relative to the respective ECsovalue
observed in the presence of PML-RAR, as assessed by the Tukey's
test after one-way analysis of variance.
t Significantly higher ( P< ,011 relative to the respective ECsovalue
observed in the presence of PML-RAR,asassessed by the Tukey's
test after one-way analysis of variance.
RARa
PML-RAR
50
40
LAP
LAP
40
30
30
20
10
0
h
c
.-
g
0
m
100
20
r
0
-11
TRE-TK
-9
-7
10
150
80
60
100
>
.-c
m
-
40
50
2
0
2
0
Q)
20
-11
9
800
600
600
400
400
200
200
-9
-7
-5
TRE-TK
0 0 c& -10
800
0
0
-8
-6
82 RARE
0 0 ;;L;.^
-10
-8
-6
Fig 9. Effects of AM580 and ATRA on the activity ofRARE-containing promoters in the presenceofPML-RAR and RARa. COS-7
cells were transiently cotransfected with LAP CAT (containing the
alkaline phosphatase gene upstream promoter), TRE-TKCAT(containing an artificial inverted repeat RARE placed in front of the viral
thymidine kinase promoter), &RARECAT (containing a natural direct
repeat RARE placed in front of the viral thymidine kinase promoter),
pnlsLACZ (containing the bacterial pgalactosidase gene under the
control of a constitutive promoter), and the indicated form of RAR.
Sixteen hours after transfection, medium waschanged and the incubation continued for a further 36 hours with theindicated concentrations of AM580 (0)or ATRA 10).At the end of each treatment, cells
were harvested and processed for the measurement of CAT and pgalactosidase activity. The results are the mean 2 SD of three replicate dishes and are expressed as relative CAT activity, which is the
ratio of CAT activity produced by the reporter in cpm of acetylated
chloramphenicol divided by the p-galactosidase activity expressed
in absorbance units at 420 nm.
that of ATRAin this particular leukemic cell type. Like
AM580, CD367 and TTNPB are much more powerful than
ATRA in inducing granulocytic maturation of NB4 cells.
Interestingly, in the presence of PML-RAR, CD367 and
TTNPB transactivate TRE-TK CAT at concentrations that
are almost identical to those necessary to obtain the same
effects withAM580. Thus, strong interaction with PMLRAR at low concentrations and potent cyto-differentiating
activity inNB4
cells are features common to AM580,
TTNPB, and CD367. Although interaction with PML-RAR
may be the basis for the pharmacologic activity of the three
retinobenzoic agents in APL cells, this does not rule out the
possibility that other molecular mechanisms have a role.
With respect to this, the three compounds are RAR-selective
ligands, unlike ATRA and 9-cis RA, which activate both
RARs and RXRs (the effect of ATRA onRXRs is the consequence of spontaneous or enzymatic isomerization to the 9-
From www.bloodjournal.org by guest on October 21, 2014. For personal use only.
GlANNl ET AL
1530
cis isomer). The lack of interaction with RXRs may be of
significance for the superiority of retinobenzoic derivatives
on ATRA and9-cis RA in APL cells, especially in consideration of the fact that RXR activation does not seem to be
required forthegranulocytic
differentiation of myeloid
cells.’”
In APL cells, three major breakpoint clusters on chromosome 15 (bcr-type I, 11, and 111), which results in the expression of three distinct types of PML-RAR fusion proteins,
have been recognized.” Although the number of cases analyzed is small, the selectiveaction of AM580 on PML-RAR
containing APL blast seems to be independent of the type
of aberrant retinoic acid receptor expressed. In fact, the series
of APL cases analyzed includes four bcr-type I (NB4 cells,
patients no. 2, 3, and 4) and one bcr-type 111 (patient no. 1)
t( 15; 17) chromosomal rearrangements. It would be important
toestablishwhether
the threetypes of PML-RARsare
equally effective in transactivating RARE-containing genes
in the presence of AM580. Similarly, studies arerequired to
investigate whether the
retinobenzoic analog is activein APL
relapse cases showing biologicresistance to ATRA. With
respect to this last point, it is worthwhile mentioning that
the compound is totally ineffective in inducing granulocytic
differentiation (unpublished observations) of arecently in
vitro developed ATRA-resistantNB4 clone.” However, this
is probably expected, as this cell line does not express the
PML-RAR fusion protein,” which seems to be thetarget for
the activity of AM580.
Regardless of the mechanism of action, AM580 is much
more interesting than CD367 and TTNPB from a perspective
therapeuticalpoint of view. In vivo, the retinoicacid-mimetic activity of the
two latter compounds is very strong;
however, bothCD367 and TTNPB show significant systemic
(unpublished observations), probably as a consequence of their promiscous ability to interact with RARa,
p, and y . The toxicologic profile of AM580 in animals and
humans is not yet completely known, although it looks favorable. This probably reflects the ability of the compound to
significantlyinteract only with RARa innormalanimals.
The results obtained with AM580 show that it is possible to
develop retinoic acid derivatives that preferentially interact
with PML-RAR and are more effective than ATRA in causinggranulocyticdifferentiation
of APL cells. Suchcompounds may show lower toxicity and a higher therapeutic
index than ATRA in the treatment of this type of leukemia.
ACKNOWLEDGMENT
We thank Dr M. Lanotte (Unit6 INSERM 301, “Genetique cellulaire et moleculaire des leucemies”, Centre G. Hayem, Hopital St
Louis, Paris, France) for supplying us with the NB4 acute promyelocytic cell line. We are grateful to Dr P.C. Pelicci (Policlinico Monteluce, Universith di Perugia, Perugia, Italy) for the kind gift of the
plasmids pSG5-RARa and pSG5-PML-RAR and the reporter construct TRE-TK-CAT. We thank Dr C. Carlberg (University of Geneva, Geneva, Switzerland) and Dr A. Weisz (University of Naples,
Naples, Italy) for providing us with the reporter construct DRS[wt]CAT and pnlsLACZ, respectively. Finally, we are also grateful to
Prof S. Garattini, Dr M. D’Incalci, DrA. Mantovani, and Dr M.
Salmona for the critical reading of the manuscript. This workis
dedicated to thememory of Prof Alfred0 Leonardi, the late Secretary
General of the Istituto di Ricerche Farmacologiche “Mario Negri”.
REFERENCES
1. Hong WK, Itri LM: Retinoids andhuman cancer, in Sporn
MB, Roberts AB, Goodman DS (eds): The Retinoids: Biology.
Chemistry and Medicine (ed 2). New York, NY, Raven. 1994, p
597
2. Huang ME, YeYC, Chen SR, ChaiJR. Lu JX, Zhoa L, Gu
LJ, Wang ZY: Use of all-trans retinoic acid in the treatment of acute
promyelocytic leukemia. Blood 72567, 1988
3. Castaigne S, Chomienne C, Daniel MT, Ballerini P, Berger R,
Fenaux P, Degos L: All-trans retinoic acid as a differentiation therapy for acute promyelocytic leukemia. I. Clinical results. Blood
76: 1704, 1990
4. Breitman TR, Collins SJ, Keene BR: Terminal differentiation
of human promyelocytic leukemic cells in primary culture in response to retinoic acid. Blood 57:1000, 1981
5. Larson RA, Kondo K, Vardiman JW, Butler AE, Golomh HM,
Rowley JD: Evidence for a 15;17 translocation in every patient with
acute promyelocytic leukemia. Am J Med 76:827, 1984
6. de ThC H, Chomienne C, Lanotte M, Degos L, Dejean A: The
t (15;17) translocation of acute promyelocytic leukemia fuses the
retinoic acid receptor-alpha gene to a novel transcribed locus. Nature
347558, 1990
7. Borrow .l, Goddard AD, Sheer D: Molecular analysis of acute
promyelocytic leukemia breakpoint cluster region on chromosome
17. Science 249:1577, 1990
8. Longo L, Pandolfi PP, Biondi A, Rambaldi A, Mencarelli A.
Lo Coco F, Diverio D, Pegoraro L, Avanzi G, Tabilio A, Zangrilli
D, Alcalay M,Donti E, Grignani F, PelicciPG: Rearrangements
and aberrant expression of the retinoic acid receptor (Y gene in acute
promyelocytic leukemia. J Exp Med 172: 1571. 1990
9. Pandolfi PP, Grignani F, Alcalay M, Mencarelli A, Biondi A,
Lo Coco F, Grignani F, Pelicci PG: Structure and origin of theacute
promyelocytic leukemia myVRARa cDNA and characterization of
its retinoid-binding and transactivation properties. Oncogene 6: 1285.
1991
IO. Grignani F,Fagioli M, Alcalay M, Longo L, Pandolfi PP,
Donti E, Biondi A, Lo Coco F, Grignani F, Pelicci PG: Acute promyelocytic leukemia: From genetics to treatment. Blood 83:10, 1994
I 1. Lanotte M, Martin-Thouvenin V, Najman S, Ballerini P, Valensi F, Berger R: NB4, a maturation inducible cell line with t( 15;17)
marker isolated from human acute promyelocytic leukemia (M3).
Blood 77:1080, 1991
12. Breitman TR, Selonick SE, Collins SJ: Induction of differentiationofthehuman
promyelocytic leukemia cell line (HL-60) by
retinoic acid. Proc Natl Acad Sci USA 77:2936, 1980
13. Weis K, Rambaud S, Lavau C, Jansen J, Carvalho T, CarmoFonseca M, Lamond A, Dejan A: Retinoic acid regulates aberrant
nuclear localization of PML-RARa in acute promyelocytic leukemia
cells. Cell 76545, 1994
14. Gianni M, Terao M, Zanotta S, Barhui T, Rambaldi A, Garattini E: Retinoic acid and granulocyte-colony stimulating factor synergistically induce leukocyte alkaline phosphatase in acute promyelocytic lekemia cells. Blood 83:1909, 1994
15. Martin B, Bernardon J-M, Cavey MT, Bernard B, Carlavan
l, Charpentier B, Pilgrim WR, Shroot B, Reichert U: Selective synthetic ligands for human nuclear retinoic acid receptors. Skin Pharmacol 557, 1992
16. Delescluse C, Cavey MT, Martin B, BernardBA, Reichert
U, Maignan J, Darmon M, Shroot B: Selective high affinity retinoic
acid receptor (Y or p - y ligands. Mol Pharmacol 40556, 1991
17. Graupner G, Malle G, Maignan J, Lang G, Prunieras M, Pfahl
M: 6”substituted naphtalene-2-carboxylic acid analogs, a new class
From www.bloodjournal.org by guest on October 21, 2014. For personal use only.
CYTO-DIFFERENTIATING EFFECTS OF AM580 ON APL
of retinoic acid receptors subtype-specific ligands. Biochem Biophys
Res Commun 179:1554, 1991
18. Olsson I, Breitman T, Gallo R: Priming of human myeloid
leukemic cell lines HL-60 and U-937 with retinoic acid for differentiation effects of cyclic adenosine 3’:5’ monophosphate-inducing
agents and a T-lymphocyte derived differentiation factor. Cancer
Res 42:3928, 1982
19. Ruchaud S, Duprez E, Gendron MC, Houge G, Genieser HG,
Jastorff B, Doskeland SO, Lanotte M: Two distinctly regulated
events, priming and triggering, during retinoid-induced maturation
and resistance of NB4 promyelocytic leukemia cell line. Proc Natl
Acad Sci USA 91:8428, 1994
20. Gianni M, Li Calzi M, Terao M, Rambaldi A, Garattini E:
Tyrosine kinases but not CAMP-dependent protein kinase mediate
the induction of leukocyte alkaline phosphatase by granulocyte-colony-stimulating factor and retinoic acid in acute promyelocytic leukemia cells. Biochem Biophys Res Commun 208:846, 1995
21. Gianni M, Terao M, Norio P, Barbui T, Rambaldi A, Garattini
E: ATRA and CAMP cooperate in the expression of leukocyte alkaline phosphatase in acute promyelocytic leukemia cells. Blood
85:3619, 1995
22. Rambaldi A, Terao M, Bettoni S, Tini ML, Bassan R, Barbui
T, Garattini E: Expression of leukocyte alkaline phosphatase gene
in normal and leukemic cells: Regulation of the transcript by granulocyte colony-stimulating factor. Blood 76:2565, 1990
23. Pedersen B: Functional and biochemical phenotype in relation
to cellular age of differentiated neutrophils in chronic myeloid leukemia. Br J Haematol 51:339, 1982
24. Sat0 N, Asano S, Urabe A, Ohsawa N, Takaku F Induction
of alkaline phosphatase in neutrophilic granulocytes, a marker of
cell maturity, from bone marrow of normal individuals by retinoic
acid. Biochem Biophys Res Commun 131:1181, 1985
25. Rambaldi A, Terao M, Bettoni S, Bassan R, Battista R, Barbui
T, Garattini E: Differences in the expression of alkaline phosphatase
mRNA in chronic myelogenous leukemia and paroxysmal nocturnal
hemoglobinuria polymorphonuclear leukocytes. Blood 73: 1113,
1989
26. Bradford M: A rapid and sensitive method for the quantitation
of microgram quantities of protein utilizing the principle of proteindye binding. Anal Biochem 72:248, 1976
27. Rambaldi A, Young DC, Griffin JD: Expression of the MCSF (CSF-I) gene by human monocytes. Blood 69:1409, 1987
28. Weiss MJ, Henthom PS, Lafferty MA, Slaughter C, Raducha
M, Harris H: Isolation and characterization of a cDNA encoding a
human liverhonekidney-type alkaline phosphatase. Proc Natl Acad
Sci USA 83:7182, 1986
29. Larsen A, Davis T, Curtis BM, Gimpel S, Sims J, Cosman
D, Park L, Sorensen E, March CJ, Smith C: Expression cloning of
a human granulocyte colony stimulating factor receptor: A structural
mosaic of hemopoietin receptor immunoglobulin and fibronectin domains. J Exp Med 172:1559, 1990
30. Persico MG, Viglietto G, Martini G, Toniolo D, Paonessa G,
Moscatelli C, Dono R, Vulliamy T, Luzzatto L, D’Urso M: Isolation
of human glucose-6-phosphate dehydrogenase (G6PD) cDNA
clones: Primary structure of the protein and unusual 5’ non-coding
region. Nucleic Acids Res 14:251 I, 1986
3 I . Feinberg A, Vogelstein B: A technique for radiolabeling DNA
restriction endonuclease fragments to high specific activity. Anal
Biochem 132:6, 1983
32. Graham FL, Van Der Eb A J : A new technique for the assay
of infectivity of human adenovirus 5 DNA. Virology .52:456, 1973
33. Kiledjian M, Kadesch T Analysis of the human liverhone/
kidney alkaline phosphatase promoter in vivo and in vitro. Nucleic
Acids Res 18:957, 1990
1531
34. Ambrosin0 C, Cicatiello L, Cobellis G, Addeo R, Sica V,
Bresciani F, Weisz A: Functional antagonism between the estrogen
receptor and Fos in the regulation of c-fos protooncogene transcription. Mol Endocrinol 7:1472, 1993
35. Seed B, Sheen J-Y: A simple phase-extraction assay for chloramphenicol acyltransferase activity. Gene 67:271, 1988
36. Cazzaniga G, Seldin MF, Terao M, Lo Schiavo P, Galbiati F,
Segalla F, Garattini E: Isolation, characterization and chromosomal
mapping of the mouse xanthine dehydrogenase gene. Genomics
23:390, 1994
37. Nervi C, Poindexter EC, Grignani F, Pandolfi, PP, LoCoco
F, Avvisati G, Pelicci PG, Jetten AM: Characterization of the PMLRARa chimeric product of the acute promyelocytic leukemia-specific t(15;17) translocation. Cancer Res 52:3687, 1992
38. Rodbard D: Mathematics of hormone-receptor interaction. I.
Basic principles, in O’Malley BW, Means AR (eds): Receptors for
Reproductive Hormones. New York, NY, Plenum, 1973, p 289
39. De Lean A, Munson PJ, Rodbard D: Simultaneous analysis
of families of sigmoidal dose curves: Application to bioassay, radioligand assay, and physiological dose-response curves. Am J Physio1 235:E97, 1978
40. Guiso G, Rambaldi A, Dimitrova B, Biondi A, Caccia S:
Determination of orally administered all-trans retinoic acid in human
plasma by high-performance liquid chromatography. J Chromatogr
B 656:239, 1994
41. Dimitrova B, Caccia S, Garattini E, Guiso G: Determination
of the retinobenzoic acid derivative Am580 in rat plasma by highperformance liquid chromatography. J Chromatogr BiomedAppl
667:301, 1995
42. Crettaz M, Baron A, Siegenthaler G, Hunziker W: Ligand
specificities of recombinant retinoic acid receptors RARa and
RARP. Biochem J 272:391, 1990
43. Levin AA, Sturzenbecker LJ, Kazmer S, Bosakowski T, Huselton C, Allenby G, Speck J, Kratzeisen C, Rosenberger M, Lovey
A, Grippo JF: 9-cis retinoic acid stereoisomer binds and activates
the nuclear receptor RXRa. Nature 355:359, 1992
44. Dermime S, Grignani F, Clerici M, Nervi C, Sozzi G, Talamo
GP, Marchesi E, Formelli F, Parmiani G, Pelicci PG, GambacortiPasserini C: Occurence of resistance to retinoic acid in the acute
promyelocytic leukemia cell line NB4is associated with altered
expression of the pml/RARa protein. Blood 82:1573, 1993
4.5. Perez A, Kastner P, Sethi S, Lutz Y, Reibel C, Chambon
P PMLRAR homodimers: Distinct DNA binding properties and
heterodimeric interaction with RXR. EMBO J 12:3171, 1993
46. Heath JK, Suva LJ, Yoon K, Kiledjian M, Martin TJ, Rodan
GA: Retinoic acid stimulates transcriptional activity from the alkaline phosphatase promoter in the immortalized rat calvarial cell line,
RCT-I. Mol Endocrinol 6:636, 1992
47. Schrader M, Wyss A, Sturzenbecker LJ, Grippo JF, LeMotte
P, Carlberg C: RXR-dependent and RXR-independent transactivation by retinoic acid receptors. Nucleic Acids Res 21:1231, 1993
48. Kagechika H, Kawachi E, Hashimoto Y, Shudo K: Retinobenzoic acids. 2. Structure-activity relationships of chalcone-4-carboxylic acids and flavone-4’-carboxylic acids. J MedChem 32:834,
1989
49. Gianni M, Zanotta S, Terao M, Garattini S, Garattini E: Effects of synthetic retinoids and retinoic acid isomers on the expression of alkaline phosphatase in F9 teratocarcinoma cells. Biochem
Biophys Res Commun 196:252, 1993
50. Nagy L, Thomazy VA, Shipley GL, Fesus L, Lamph W,
Heyman RA, Chandraratna RAS, Davies PJA: Activation of retinoid
X receptors induces apoptosis in HL-60 cell lines. Mol Cell Biol
1.5:3540, 1995