Académique Documents
Professionnel Documents
Culture Documents
321–333, 2006
Advance Access publication April 5, 2006 doi:10.1093/molehr/gal036
Cannabis use in pregnancy is associated with a range of obstetrical conditions. The molecular mechanisms underlying these
effects have not been elucidated but are attributed to the actions of delta-9-tetrahydrocannabinol (D9-THC). In this study, concen-
trations of D9-THC equivalent to those found in the serum of cannabis users, i.e. ~20 mM, inhibited proliferation and activated a
restricted tight transcriptional programme in the BeWo trophoblast cell line. Employing genome-wide expression profiling meth-
ods, we found that the pattern of gene expression differs from that described in the placenta of patients with fetal growth restric-
tion (FGR), associated with either hypoxia or discordant dichorionic twins, or of patients with pre-eclampsia. It was also dissimilar
to the patterns obtained from the transcriptome of other tissues, such as the mouse brain, treated with D9-THC. The expression of
transcription factors, such as thyroid hormone receptor-b1 (TRb1), and transcriptional co-repressors, such as histone deactylase
3 (HDAC3), was affected by D9-THC in a dose-dependent manner, whereby 15 mM D9-THC caused a 2.8-fold inhibition of TRb1
expression, but a 3.5-fold increase in HDAC3 expression. These data were confirmed by end-point RT–PCR analyses and under-
pin the observed D9-THC-induced inhibition of BeWo cell proliferation. Genes encoding for growth, apoptosis, cell morphology
and ion exchange pathways were modulated by 15 mM D9-THC. This study may provide insight into the mechanisms underlying
the effects of D9-THC and cannabis use upon placental development during pregnancy.
© The Author 2006. Published by Oxford University Press on behalf of the European Society of Human Reproduction and Embryology. All rights reserved. For
Permissions, please email: journals.permissions@oxfordjournals.org 321
M.Khare et al.
Cohran et al., 1996; Kudo et al., 2004) and in toxicology studies with ethanol applied to the RNeasy mini columns, subsequently purified and
(Burres and Cass, 1987; Thibault et al., 2000; Nomura et al., 2004). concentrated through ethanol precipitation. RNA integrity was assessed using
Additionally, these cells have been used in genome-wide analyses agarose gel electrophoresis and Agilent 2100 Bioanalyser (Agilent Technolo-
(Aronow et al., 2001; Saito et al., 2001; Endo et al., 2002; Kudo et al., gies, UK, South Queensferry, UK). RNA from two separate experiments per-
formed in triplicate was pooled and samples submitted to the MRC
2004) that have identified putative novel targets for the mechanism of
Geneservice (Babraham, Cambridge, UK) for further purification and produc-
FGR that were subsequently confirmed in animal studies to be of sig- tion of biotinylated cRNA. Biotinylated cRNA was prepared from 10 μg of
nificant experimental and clinical value (Loiselle et al., 2004; Nomura total RNA according to Affymetrix protocols. The integrity of the labelled
et al., 2004; Ohara et al., 2004). The BeWo cell line thus provides an cRNA and fragmentation products were assessed on the Agilent 2100 bioana-
appropriate model in which to study some aspects of human trophob- lyser. Next, 15 μg of biotinylated cRNA fragments were hybridized to human
last physiology and pathophysiology without the aspect of inter- HU133_plus 2 microarray chips overnight and then stringently washed, stained
patient variability and other confounding variables. In this study, we and scanned using a GeneArray scanner (Agilent Technologies, Palo Alto, CA,
have therefore employed this cell line to determine whether Δ9-THC USA), according to Affymetrix protocols.
would specifically affect cytotrophoblast cell survival and gene tran-
scription and whether extrapolation of such effects may be used to Gene expression analysis
explain the reported in-vivo effects of Δ9-THC upon clinical condi- Fluorescence data were corrected for background fluorescence, reduced in
tions affecting feto-placental development. intensity values before normalization against internal standards. All arrays
were scaled to the same target intensity and were normalized to housekeeping
genes on the U133_plus 2 arrays and analysed using dChip Analyzer software
Materials and methods version 1.4 (Harvard School of Public Health and Dana-Farber Cancer Research
Institute, Boston, MA, USA, 2004; http://biosun1.harvard.edu/complab/dchip/)
Materials
using a 2-fold change in expression with an a of 0.90 and a b of 0.05 being the
The human choriocarcinoma cell line BeWo (ECACC 86082803, European statistical cut-off points for real expression changes from duplicate chips.
Collection of Cell Cultures, Salisbury, Wiltshire, UK) was maintained in Hierarchical clustering was used to obtain gene expression patterns and ontology
Ham’s F12 medium (Invitrogen, Paisley, UK) supplemented with 10% fetal information.
calf serum (FCS) (Invitrogen) and cultured at 37°C in humidified atmosphere
of 5% CO2 in air. Δ9-THC (Sigma-Aldrich, Poole, Dorset, UK) was diluted in
End-point RT–PCR
ethanol and stored under nitrogen at –20°C. Oligonucleotide primers were syn-
thesized (Sigma-Genosys Ltd., Haverhill, Suffolk, UK) and desalted before To confirm the data from the microarray experiment, we performed end-point
reconstitution in sterile dH2O at 200 pmol/μl and stored at –20°C. RT–PCR upon the samples used in these experiments. One microgram of total
RNA was reversed transcribed with avian myeloblastosis virus-reverse tran-
scriptase (AMV-RT; Promega, Southampton, UK) at 42°C for 1 h in the pres-
Cell culture ence of 5 units of RNase inhibitor (RNasin; Promega). A minus RT reaction
BeWo cells were plated into Nunc 6-well multi-well plates in triplicate for was obtained by omitting the AMV-RT enzyme. At the end of the reaction, the
each data point (Fisher Scientific, Loughborough, Leicestershire, UK) at a den- enzymes were denatured by heating at 95°C for 5 min and the cDNA was stored
sity of 8 × 105 cells per well (Nomura et al., 2004). After 48 h culture, the at –20°C. One microlitre of cDNA was subject to PCR using 10 pmol/μl of spe-
medium was exchanged with one that contained 5% FCS and 15 μM Δ9-THC cific primers for histone deactylase 3 (HDAC3), thyroid hormone receptor-β1
[final concentration of 0.1%(v/v) ethanol]. The control consisted of culture (TRβ1) and glyceraldehyde-3-phosphatedehydrogenase (GAPDH) using the
medium containing 0.1%(v/v) ethanol. Cells were cultured with media plus annealing temperatures given in Table I. The RNA from the dose-ranging study
additives for 1 h to allow non-specific binding of Δ9-THC to plastic to reach was treated with RNAse-free DNAse 1 (Promega) before reverse transcription
equilibrium. The medium was then replaced and culture continued for 48 h, and PCR amplification. PCR products were resolved on 3% agarose gels and
with fresh medium replaced at 24 h. Photomicrographs were obtained using a stained with ethidium bromide (2 μg/ml; ICN Biomedicals, Basingstoke, UK)
Nikon Eclipse TE2000-U inverted microscope equipped with a DN-100 digital for 15 min, before being destained with dH2O for 30 min. Gel images were cap-
camera image capture system, and percentage confluency was measured by tured using a Syngene GeneGenius system (Syngene, Cambridge, UK)
image analysis of pixels representing cells and substratum. equipped with GeneSnap version 6 gel documentation software. Product densi-
To examine the effect of Δ9-THC on cell survival and cell proliferation indi- ties were assessed using the Scion Image beta version 4.0.2 software (Scion
ces, we plated BeWo cells to Nunc 96-well plates (Fisher Scientific) at either Corporation, Frederick, MD, USA, 1999; http://www.scioncorp.com).
4 × 104 or 1 × 104 cells per well, respectively, in 200 μl of normal growth medium
and allowed them to proliferate for 48 h until the higher density cultures reached
Data analysis
confluency. The medium was then changed to one that contained 0–30 μM
Δ9-THC for 1 h to allow for non-specific binding. The medium was then Subtraction of the densitometry values obtained from the –RT lanes from those
replaced and culture continued for 48 h, with fresh medium replaced at 24 h. from the +RT lanes, and then division by the relative GAPDH transcript levels,
After 48 h, cell numbers were assessed using the Cell Proliferation and Apop- was used to correct the PCR data. These data were then normalized to the
tosis Kit II (Roche Diagnostics Ltd., Lewes, East Sussex, UK) as per the man- untreated control and expressed relative to the untreated controls. All data were
ufacturer’s instructions with measurements taken on a Multiskan Ascent then analysed for differences using one-way ANOVA with Tukey’s honest sig-
ELISA plate reader (Labsystems Oy, Helsinki, Finland) with the detection fil- nificant difference (HSD) test with the InStat version 3.0 software package
ter set at 420 nm and the reference set at 620 nm. Cell numbers were obtained (GraphPad Software, San Diego, CA, USA, 1998; http://www.graphpad.com).
by calibration against a standard curve of untreated BeWo cell numbers grown Statistical significance was accepted when P < 0.05.
in parallel (Taylor et al., 2002). To make direct comparisons between cultures,
we then converted the cell numbers to a percentage of the untreated control. Results
The effect of Δ9-THC dose on gene expression was obtained by culturing
BeWo cells with doses of Δ9-THC ranging from 0.3 to 30 μM in 0.1% ethanol
Effect of D9-THC on gross BeWo cell growth, survival
for 48 h, as described above. and morphology
BeWo cells were plated such that cultures achieved approximately
Total RNA extraction, cRNA synthesis and microarray 70–80% confluency after 48 h, the point at which Δ9-THC treatments
hybridization were initiated. Confluency was achieved within 24 h of the subse-
Total RNA was extracted using a combination of TRIZOL reagent (Invitro- quent 48 h culture period (data not shown). Cultures treated at 48 h
gen), RNeasy mini kit (Qiagen Ltd., Crawley, UK) and ethanol precipitation. with a range of Δ9-THC demonstrated an inhibitory effect on the
Briefly, aqueous total RNA from the TRIZOL reagent procedure was mixed BeWo cell cultures (Figure 1A and B), but only at concentrations in
322
Effects of D9-THC upon BeWo cells
mRNA species Oligonucleotide sequence Product size (bp) Annealing Accession number
temperature (°C)
GAPDH
671
5′ AGAACATCATCCCTGCCTC689 347 60 X01677*
1017
3′ GCCAAATTCGTTGTCATACC998
HDAC3
1576
5′ GGGACATTATTGGCAGTG1593 233 55 U75697
1808
3′ GGATTCAGGTGTTAGGGAG1790
TRβ1
59
5′ GCGATTTCCTTCTGGTTG76 314 and 168 55 NM-000461.2
372
3′ AGTGCTTCGGTTTGTCCC355
Figure 1. The dose-dependent effect of Δ9-tetrahydrocannabinol (Δ9-THC) upon BeWo cell growth, morphology and survival. (A) Photomicrographs of BeWo
cells treated with 0.1% ethanol (control) or the indicated concentrations of Δ9-THC for 48 h. Areas not covered with cells are arrowed. Data are representative of
five experiments performed in triplicate. Scale bar = 120 μm (B) Graph showing the dose-dependent effect of Δ9-THC on BeWo cell culture confluency. Data are
mean ± SEM of four experiments performed in triplicate; *P < 0.05 one-way ANOVA with Tukey’s honest significant difference (HSD) test. (C) Graph showing
the dose-dependent effect of Δ9-THC on inhibiting BeWo cell numbers when the cells were grown to 25% confluency (sub-confluent) before treatment compared
with a lack of effect on cell numbers grown to confluency. Data are the mean ± SEM of two experiments performed in quadruplicate; **P < 0.001 one-way
ANOVA with Tukey’s HSD test.
excess of 3 μM Δ9-THC, where confluency was significantly reduced remained stable (1.32 ± 0.065 × 105 compared with 1.22 ± 0.063 × 105
from ∼95 to ∼70% (*P < 0.05; one-way ANOVA with Tukey’s HSD cells per well; P = 0.37, n = 8 ANOVA with Tukey’s HSD test) and
test; n = 4). Cultures treated with 15 and 30 μM Δ9-THC did not reach confirmed the observation that the effect of Δ9-THC on BeWo cul-
confluence (Figure 1A and B). There did not appear to be any tures was not due to increased cell death. Sub-confluent cultures
increased cell death or failure to attach to the substratum as evidenced treated with Δ9-THC demonstrated a significant dose-dependent inhi-
by the lack of increase of shedding of cells in the spent medium. Anal- bition of cell numbers at concentrations above 3 μM Δ9-THC after 48 h
ysis of the effect of Δ9-THC on BeWo cell survival and proliferation (Figure 1C). These cultures failed to exceed the 30% confluency level
revealed that cultures that reached confluency before treatment (the untreated cultures reached a cell density of 3.7 ± 0.31 × 104 cells
323
M.Khare et al.
per well from an initial density of 1 × 104 cells per well). However, the very high degree of reproducibility across replicate GeneChip
15 μM Δ9-THC inhibited cell proliferation so that the final cell density experiments (Irizarry et al., 2003; Cope et al., 2004). In this study, to
was 2.25 ± 0.19 × 104 cells per well after Δ9-THC treatment (P < ensure consistent data, we adopted both biological (experiments per-
0.001; n = 8; ANOVA with Tukey’s HSD test). Further morphological formed four times in triplicate) and technical replicates (hybridizations
examination of cell cultures treated with 15 μM Δ9-THC indicated performed twice) for these studies, whereas others adopt either bio-
that the inhibitory effect resulted from a decrease in growth rather than logical (Hoffman et al., 2004) or technical replicates (Kudo et al.,
involution of the culture after achieving confluency (Figure 2). 2004), but seldom both (Kothapalli et al., 2002). Therefore, total RNA
from two sample pools of the control and 15 μM Δ9-THC-treated cul-
Microarray analysis: identification of D9-THC-regulated tures (n = 12) were used to generate biotinylated target cRNA and
transcripts hybridized to HU133A_Plus 2 arrays, which represented approximately
56 000 characterized transcripts and expressed sequence tags (ESTs).
The use of microarray analysis facilitates the global assessment of
Scaling factors, noise and percentage of present and absent calls showed
transcriptional profiles of cells. A number of studies have established
only minor variations between arrays, an essential requirement when
comparing multiple arrays. These replicate gene chip experiments
revealed an overall correlation (r2 = 0.89) with an average of only
0.04% of probe sets showing more than a two-fold change between the
replicate experiments. The number of genes induced and repressed on
microarray 1 at the two-fold level were 304 and 105, respectively,
whereas the number induced and repressed on microarray 2 were 220
and 50, respectively, representing a mean error measurement rate of
only 0.5%. When the cut-off was reduced to 1.6-fold, the number of
genes induced and repressed on microarray 1 were 1504 and 1127,
respectively, whereas the number induced and repressed on microarray
2 were 2291 and 5670, respectively, representing a mean error measure-
ment rate of 26.3%. Under these conditions, it was considered that
26.3% error lacked precision for detailed transcriptome analysis and the
conventional two-fold cut-off point was used (Cope et al., 2004).
Focusing on genes showing an increased level of expression on both
arrays and using a two-fold change cut-off, we identified 134 genes of
which the top 20 genes are mentioned in Table II. Focusing on genes
showing a decreased level of expression on both arrays with a two-fold
(50%) change cut-off, we identified only 18 genes (Table II).
Table II. Putative target genes of Δ9-tetrahydrocannabinol (Δ9-THC) identified by microarray analysis
Probe set Gene listing Fold 1 Fold 2 Median fold % Change Accession
change number
Up-regulated genes
227062_at Trophoblast-derived non-coding RNA/plectin 1 6.81 2.07 5.23 423 AU155361
229713_at cDNA FLJ13267 fis, clone OVARC1000964 7.79 3.00 3.69 269 AW665227
224726_at DAPK-interacting protein 1 5.63 2.57 3.58 258 W80418
224568_x_at Metastasis associated in lung adenocarcinoma transcript 1/ 6.51 2.45 3.47 247 AW005982
MALAT-1/HDAC3
244804_at Sequestosome 1 4.74 2.63 3.44 244 AW293441
208610_s_at Serine/arginine repetitive matrix 2 4.93 2.13 3.42 242 AI655799
225640_at Hypothetical gene supported by AK091718 (LOC401504), mRNA 4.94 2.56 3.38 238 AA875998
239451_at Tumour rejection antigen (gp96) 1 4.49 2.33 3.38 238 AI684643
234605_at Homo sapiens cDNA: FLJ21233 fis, clone COL00789 3.51 3.15 3.37 237 AK024886
239274_at Transcribed sequence with weak similarity to protein 4.99 2.18 3.29 229 AV729557
NP_062553.1 hypothetical protein FLJ11267
1554178_a_at Hypothetical protein MGC39518 5.11 2.69 3.26 226 BC039295
1558080_s_at Hypothetical protein LOC144871 3.88 2.17 3.21 221 BG913589
213593_s_at Transformer-2α 3.50 2.15 3.20 220 AW978896
239742_at Tubby like protein 4 4.14 2.59 3.19 219 H15278
1561097_at Clone IMAGE:5270855, mRNA 2.48 3.84 3.14 214 BC040602
237469_at Unknown protein; UG = Hs.242998; EST 3.23 2.52 3.07 207 T96523
233819_s_at Zinc finger protein 294 4.79 3.27 3.04 204 AK023499
236524_at cDNA FLJ37319 fis, clone BRAMY2018027 3.20 2.16 3.03 203 AA737437
242528_at cDNA FLJ31668 fis, clone NT2RI2004916 2.08 5.98 3.03 203 AI473887
212468_at Sperm-associated antigen 9 4.28 2.63 2.99 199 AK023512
223940_x_at Metastasis associated in lung adenocarcinoma transcript 1/ 5.93 2.21 2.96 196 AF132202
MALAT-1/HDAC3
227740_at Kinase interacting with leukemia-associated gene (stathmin) 4.91 3.32 2.95 195 AW173222
236327_at Transcribed sequences 3.56 2.75 2.94 194 AA767373
201730_s_at Translocated promoter region (to activated MET oncogene) 4.31 3.42 2.93 193 BF110993
201996_s_at SMART/HDAC1-associated repressor protein 4.88 2.09 2.87 187 AL524033
222413_s_at Myeloid/lymphoid or mixed-lineage leukemia 3 4.46 3.09 2.87 187 AW137099
222576_s_at Eukaryotic translation initiation factor 2C, 1 4.70 4.08 2.86 186 AW071829
213067_at Myosin, heavy polypeptide 10, non-muscle 5.76 2.41 2.85 185 AI382123
211944_at HbxAg transactivated protein 2 3.65 2.41 2.80 180 BE729523
216550_x_at Ankyrin repeat domain 12 3.61 2.87 2.78 178 X80821
Down-regulated genes
211106_at Suppressor of Ty 3 homolog (S. cerevisiae) −3.49 −218.65 −89.12 98.8 AF064804
204133_at RNA, U3 small nucleolar interacting protein 2 −2.49 −152.67 −69.80 98.6 NM_004704
214729_at TWIST neighbour −2.06 −62.76 −22.89 95.6 AA400421
216588_at Ribosomal protein L7 −2.62 −4.87 −4.07 75.4 AL031577
212444_at Retinoic acid induced 3 −3.20 −1.92 −3.80 73.7 AA156240
212838_at Dynamin-binding protein −2.25 −4.83 −3.15 68.3 AB023227
209907_s_at Intersectin 2 −3.06 −4.41 −2.80 64.3 AF182198
229657_at Thyroid hormone receptor, b [erythroblastic leukemia viral –5.28 –2.98 –2.79 64.2 BF431989
(v-erb-a) oncogene homolog 2, avian]
242229_at N-acyl-phosphatidylethanolamine-hydrolyzing phospholipase D .2.06 −5.39 −2.66 63.4 W01715
241617_x_at EST moderately similar to 810024C cytochrome oxidase I −3.86 −4.27 −2.64 62.1 BE961949
1555433_at Solute carrier family 39 (zinc transporter), member 14 −2.90 −3.07 −2.64 62.1 BC015770
234179_at Homo sapiens cDNA: FLJ23200 fis, clone KAIA38871. .2.67 −7.00 −2.57 61.1 AK026853
238493_at Zinc finger protein 506 −2.05 −4.71 −2.53 60.5 AI559570
244752_at Hypothetical protein LOC220929 −2.09 −4.63 −2.49 59.8 AI563915
238283_at Hypothetical protein LOC151658 −2.07 −3.84 −2.42 58.7 AI685344
222341_x_at Transcribed sequences −3.50 −2.71 −2.10 52.4 AW973235
224321_at Transmembrane protein with EGF-like and two follistatin-like −2.11 −2.00 −2.07 51.7 AB004064
domains 2
207350_s_at Vesicle-associated membrane protein 4 −2.00 −2.04 −2.02 50.5 NM_003762
LOC, LocusLink; UG, Unigene; EST, expressed sequence tag; fold 1, mean fold change array 1; fold 2, mean fold change array 2; median fold change, median fold
change for both arrays; % change, % movement from 100%. The top 20 most up-regulated and the 18 down-regulated probe sets are shown. Genes studied to
confirm the microarray data are shown in bold.
shown to be decreased by 4.61-fold (78.3%) in experiment 1 and by Effect of D9-THC dose on HDAC3 and TRb1 mRNA levels
3.36-fold (70.2%) in experiment 2 (Figure 4C); the mean was 3.90- Treatment of BeWo cells with Δ9-THC did not show a dose-dependent
fold (74.3%), which compared favourably with the 5.28-fold (81.1%) increase on HDAC3 mRNA expression, but a threshold effect at 3 μM
and 2.98-fold (66.5%) reductions in the two microarrays, respectively. (Figure 5). At this dose, HDAC3 mRNA levels increased by 1.55-fold
The median change in TRβ1 levels was 2.79-fold (64.1%; Table II). and reached a maximum of 1.70-fold change at 15 μM (Figure 5). By
Variant G (Frankton et al., 2004) was not observed in any BeWo extract. contrast, there was a dose-like effect on the repression of TRβ1 that
325
M.Khare et al.
Figure 4. Comparison of the transcript levels in the microarray and the end-point RT–PCR of the pooled RNA samples used in the microarray experiments.
HDAC3 mRNA levels determined by RT–PCR (A) from control Array 1 and Array 2 (+RT; lanes 1 and 5), control Array 1 and Array 2 (–RT; lanes 2 and 6), 15 μM
Δ9-tetrahydrocannabinol (Δ9-THC) Array 1 and Array 2 (+RT; lanes 3 and 7), 15 μM Δ9-THC Array 1 and Array 2 (–RT; lanes 4 and 8). (B) Graphical representa-
tion of HDAC3 transcript levels in the microarray (fluorescence) compared with those of the end-point RT–PCR (corrected for GAPDH levels). TRβ1 mRNA levels
determined by RT–PCR (C) from control Array 1 and Array 2 (+RT; lanes 3 and 5), control Array 1 and Array 2 (–RT; lanes 4 and 6), 15 μM Δ9-THC Array 1 and
Array 2 (+RT; lanes 1 and 7), 15 μM Δ9-THC Array 1 and Array 2 (–RT; lanes 2 and 8). (D) Graphical representation of TRβ1 transcript levels in the microarray
(fluorescence) compared with those of the end-point RT–PCR (combined long and short-form transcripts, corrected for GAPDH levels). Data are mean ± SD from
15 fluorescence points in duplicate (microarray) and two data points for the end-point RT–PCR (C).
The microarray analysis also revealed the regulation of several tran- that several transcripts that might be considered relevant could have
scripts for which the predicted protein has no known function (Table III). been omitted from these analyses in an attempt to obtain precision and
identify the most likely mediators involved in Δ9-THC-induced BeWo
cell-growth inhibition. The tight regulation of the transcriptome is not
Discussion limited to this study, as similar tight transcriptional control mecha-
Employing concentrations of Δ9-THC that are detected in patients nisms have been demonstrated in neurons undergoing apoptosis that
using cannabis for recreational use (Cherlet and Scott, 2002), we have do not modulate ‘classical’ apoptosis proteins (Desagher et al., 2005).
demonstrated that 3–30 μM Δ9-THC inhibited BeWo cell proliferation The effects of Δ9-THC upon BeWo cells also appeared to be cell
in a dose-dependent manner. Conversely, the same doses of Δ9-THC specific in that although expression of CPNE6 (copine VI), the lipid
showed no significant effect on confluent cell cultures indicating no metabolizing, neuronal form of the copine vesicle transport family of
toxic or pro-apoptopic effect of Δ9-THC upon BeWo cells. Through calcium-sensing proteins, and LlGI 1, a cytoskeletal protein that
microarray analysis, we have also demonstrated distinct changes in associates with non-muscle myosin II heavy chain to control asym-
the pattern of transcription by the BeWo cells in response to 15 μM metrical cellular polarization in neuroblasts (Klezovitch et al., 2004),
Δ9-THC. However, only a relatively low number of transcripts were was repressed in both BeWo cells and the brain of Δ9-THC-treated
consistently modulated that suggested a tight transcriptional control mice (Parmentier-Batteur et al., 2002), no overlap in the expression
be manifested in the Δ9-THC-treated BeWo cultures. The reason for pattern of other genes was noted. The altered transcripts belonged to
this is not readily apparent but may be part-way due to the conven- seven functional groups with the largest number being transcriptional
tional two-fold cut-off point used in microarray analyses. However, regulators, such as TRβ1, their co-regulators, such as HDAC3 (Table III),
although ‘raising the bar’ to 1.6-fold increased the number of tran- suggesting that one of the major effects of Δ9-THC on the BeWo cell
scripts that were regulated by Δ9-THC in the BeWo cell, an unaccept- is the regulation of gene transcription and RNA processing.
able significant increase in the amount of experimental variability As cannabis use during pregnancy is associated with FGR, the present
from 0.5 to 26.3% was also introduced which would lead to false- microarray data were compared with such data derived from placental
positive identification of non-regulated transcripts. These data mean tissue obtained from conditions associated with FGR. Interestingly,
327
M.Khare et al.
328
Effects of D9-THC upon BeWo cells
Table III. Classification of differentially expressed genes and their putative functions
329
M.Khare et al.
330
Effects of D9-THC upon BeWo cells
LOC, LocusLink; UG, Unigene; EST, expressed sequence tag; fold 1, mean fold change array 1; median fold change, median fold change for both arrays;
% change, % movement from 100%.
Of note was the down-regulation in expression of N-acyl- similar effects are observed in vivo, these data suggest that Δ9-THC
phosphatidylethanolamine-hydrolyzing Phospholipase D, the enzyme use during human pregnancy may similarly inhibit trophoblast prolif-
responsible for the production of anandamide, the principal endog- eration and that placentae during early development when the cytotro-
enous cannabinoid with major effects on human reproductive function phoblastic populations predominate may be more sensitive to the
(Wenger et al., 1999; Habayeb et al., 2002; Maccarrone et al., 2002). adverse effects of marijuana use and lead to a failure to achieve full
As anandamide and the exocannabinoid Δ9-THC exhibit different placental development, and hence fetal growth. The lack of cell prolif-
affinities for the cannabinoid and vanilloid receptors, in vivo exposure eration and migration associated with early placental development
to Δ9-THC may further exacerbate its effects by switching from anan- might therefore go some way to explain some of the clinical observa-
damide to Δ9-THC action. Also of note was the suppression of expression tions of miscarriage and placental abruption associated with marijuana
of an uncharacterized lipoprotein lipase (Garnica and Chan, 1996), use in early pregnancy and also explain why marijuana use in late
because recently it has been suggested that lipoprotein lipase defi- pregnancy is not anti-gestational (Conner, 1984; Hatch and Bracken,
ciency can lead to FGR (Magnusson et al., 2004) or fetal death (Tsai 1986; Zuckerman et al., 1989). Although the regulatory network that
et al., 2004). underlies the temporal control of transcript expression and gene
We have identified a restricted alteration in the expression of genes function remains to be determined, the identification of Δ9-THC-
in BeWo cells exposed to Δ9-THC. The alterations in the expression modulated trophoblast transcripts is likely to shed light on the mechanisms
of a number of characterized genes are consistent with its effects upon underlying trophoblast response to marijuana use in pregnancy.
the morphology of cells in culture, whereby the cultures were pre-
vented from achieving confluency not through increased syncytial Acknowledgements
formation or culture involution, but through inhibition of cell prolifer- The authors thank H. Longland and E. Beighton of the RNA Laboratory, MRC
ation. Because BeWo cells are considered an appropriate model for Geneservice, Babraham Bioincubator, Babraham, Cambridge, for the produc-
human in vivo trophoblast action and function (Sullivan, 2004) and tion of cRNA, hybridization of probes to Affymetrix genechips and the
331
M.Khare et al.
production of scanned images and Dr E. Halligan, Genomic Instability Habayeb OMH, Taylor AH, Evans MD, Cooke MS, Taylor DJ, Bell SC and
Research Group, Department of Cancer Studies and Molecular Medicine, for Konje JC (2004) Plasma levels of the endocannabinoid anandamide in
assistance with the dChip software. women – a potential role in pregnancy maintenance and labor? J Clin
Endocrinol Metab 89,5482–5487.
Hall LL, Bicknell GR, Primrose L, Pringle JH, Shaw JA and Furness PN
References (1998) Reproducibility in the quantification of mRNA levels by RT-PCR-
ELISA and RT competitive-PCR-ELISA. Biotechniques 24,652–658.
Aronow BJ, Richardson BD and Handwerger S (2001) Microarray analysis of Harclerode J (1980) The effect of marijuana on reproduction and development.
trophoblast differentiation: gene expression reprogramming in key gene NIDA Res Monogr 31,137–166.
function categories. Physiol Genomics 6,105–116. Hatch EE and Bracken MB (1986) Effect of marijuana use in pregnancy on
Bailey JR, Cunny HC, Paule MG and Slikker W (1987) Fetal disposition of fetal growth. Am J Epidemiol 124,986–993.
delta 9-tetrahydrocannabinol (THC) during late pregnancy in the rhesus Helliwell RJ, Chamley LW, Blake-Palmer K, Mitchell MD, Wu J, Kearn CS
monkey. Toxicol Appl Pharmacol 90,315–321. and Glass M (2004) Characterization of the endocannabinoid system in
Barber KJ, Franklyn JA, McCabe CJ, Khanim FL, Bulmer JN, Whitley GS and early human pregnancy. J Clin Endocrinol Metab 89,5168–5174.
Kilby MD (2005) The in vitro effects of tri-iodothyronine on epidermal Hoffman K, Dixon DN, Greene WK, Ford J, Taplin R and Kees UR (2004) A
growth factor-induced trophoblast function. J Clin Endocrinol Metab microarray model system identifies potential new target genes of the proto-
90,1655–1661. oncogene HOX11. Genes Chromosomes Cancer 41,309–320.
Brauner-Osborne H, Jensen AA, Sheppard PO, Brodin B, Krogsgaard-Larsen P Hohn HP, Parker CR Jr, Boots LR, Denker HW and Hook M (1992) Modula-
and O’Hara P (2001) Cloning and characterization of a human orphan family C tion of differentiation markers in human choriocarcinoma cells by extracel-
G-protein coupled receptor GPRC5D. Biochim Biophys Acta 1518,237–248. lular matrix: on the role of a three-dimensional matrix structure.
Burres NS and Cass CE (1986) Density-dependent inhibition of expression of Differentiation 51,61–70.
syncytiotrophoblastic markers by cultured human choriocarcinoma BeWo Hu D, Mayeda A, Trembley JH, Lahti JM and Kidd VJ (2003) CDK11
cells. J Cell Physiol 128,375–382. complexes promote pre-mRNA splicing. J Biol Chem 278,8623–8629.
Burres NS and Cass CE (1987) Inhibition of methotrexate-induced differentia- Hutchings DE and Dow-Edwards D (1991) Animal models of opiate, cocaine,
tion of cultured human choriocarcinoma BeWo cells by thymidine. Cancer and cannabis use. Clin Perinatol 18,1–21.
Res 47,5059–5064. Irizarry RA, Bolstad BM, Collin F, Cope LM, Hobbs B and Speed TP (2003)
Calabrese EJ (2005) Challenging dose–response dogma. The Scientist Summaries of Affymetrix GeneChip probe level data. Nucleic Acids Res
19,22–23. 31,e15.
Chan S, Murray PG, Franklyn JA, McCAbe CJ and Kilby MD (2004) The use Ishii M, Merrill AE, Chan Y-S, Gitelman I, Rice DPC, Sucov HM and Maxson
of laser capture microdissection (LCM) and quantitative polymerase chain REJ (2003) Msx2 and Twist cooperatively control the development of the
reaction to define thyroid hormone receptor expression in human ‘term’ neural crest-derived skeletogenic mesenchyme of the murine skull vault.
placenta. Placenta 25,758–762. Development 130,6131–6142.
Cherlet T and Scott JE (2002) Tetrahydrocannabinol (THC) alters synthesis Jho SH, Vouthounis C, Lee B, Stojadinovic O, Im MJ, Brem H, Merchant A,
and release of surfactant-related material in isolated fetal rabbit type II cells. Chau K and Tomic-Canic M (2005) The Book of Opposites: the role of the
Drug Chem Toxicol 25,171–190. nuclear receptor co-regulators in the suppression of epidermal genes by retin-
Cohran V, Fang J, Milio L, Smith CH and Fant M (1996) Type I insulin-like oic acid and thyroid hormone receptors. J Invest Dermatol 124,1034–1043.
growth factor receptors in the BeWo choriocarcinoma cell during cell differ- Jin Y, Blue EK, Dixon S, Shao Z and Gallagher PJ (2002) A death-associated
entiation. Placenta 17,313–320. protein kinase (DAPK)-interacting protein, DIP-1, is an E3 ubiquitin ligase
Conner CS (1984) Marijuana and alcohol use in pregnancy. Drug Intell Clin that promotes tumor necrosis factor-induced apoptosis and regulates the
Pharm 18,233–234. cellular levels of DAPK. J Biol Chem 277,46980–46986.
Cope LM, Irizarry RA, Jaffee HA, Wu Z and Speed TP (2004) A benchmark Jones R (1980) Human effects: an overview. NIDA Res Monogr 31,54–80.
for Affymetrix GeneChip expression measures. Bioinformatics 20,323–331. Kenney SP, Kekuda R, Prasad PD, Leibach FH, Devoe LD and Ganapathy V
Cunliffe T (2004) Histone deacetylase 1 is required to repress Notch target (1999) Cannabinoid receptors and their role in the regulation of the serot-
gene expression during zebrafish neurogenesis and to maintain the produc- onin transporter in human placenta. Am J Obstet Gynecol 181,491–497.
tion of motoneurones in response to hedgehog signalling. Development Kilby MD, Verhaeg J, Gittoes N, Somerset DA, Clark PMS and Franklyn JA
131,2983–2995. (1998) Circulating thyroid hormone concentrations and placental thyroid
Dennedy MC, Friel AM, Houlihan DD, Broderick VM, Smith T and Morrison hormone receptor expression in normal pregnancy and pregnancy complicated
JJ (2004) Cannabinoids and the human uterus during pregnancy. Am J by intrauterine growth restriction. J Clin Endocrinol Metab 83,2964–2971.
Obstet Gynecol 190,2–9. King JC (1997) Substance abuse in pregnancy. A bigger problem than you
Desagher S, Severac D, Lipkin A, Bernis C, Ritchie W, Le Digarcher A and think. Postgrad Med 102,135–150.
Journot L (2005) Genes regulated in neurons undergoing transcription- Klezovitch O, Fernandez TE, Tapscott SJ and Vasioukhin V (2004) Loss of
dependent apoptosis belong to signaling pathways rather than the apoptotic cell polarity causes severe brain dysplasia in Lgl1 knockout mice. Genes
machinery. J Biol Chem 280,5693–5702. Dev 18,559–571.
Endo H, Okamoto A, Ozawa M, Shinya M, Akiyama Y and Tanaka T (2002) Knight EM, James H, Edwards CH, Spurlock BG, Oyemade UJ, Johnson AA,
Expression profile of intrauterine growth restriction (IUGR). Placenta West WL, Cole OJ, Westney LS and Westney OE (1994) Relationships of
23,A1–A53,P97. serum illicit drug concentrations during pregnancy to maternal nutritional
Felder CC and Glass M (1998) Cannabinoid receptors and their endogenous status. J Nutr 124,973S–980S.
agonists. Annu Rev Pharmacol Toxicol 38,179–200. Kothapalli R, Yoder SJ, Mane S and Loughran TP Jr (2002) Microarray
Fisher SE, Atkinson M and Chang B (1987) Effect of delat-9-tetrahydrocan- results: How accurate are they? BMC Bioinformatics 3,22.
nibinol on the in vitro uptake of alpha-amio isobutyric acid by term placen- Kudo Y, Boyd CAR, Sargent IL, Redman CWG, Lee JM and Freeman TC (2004)
tal slices. Pediatr Res 21,104–107. An analysis using DNA microarray of the course of gene expression during
Frankton S, Harvey CB, Gleason LM, Fadel A and Williams GR (2004) Multi- syncytialization of a human placental cell line (BeWo). Placenta 25,479–488.
ple messenger ribonucleic acid variants regulate cell-specific expression of Lazar MA (1993) Thyroid hormone receptors: multiple forms, multiple possi-
human thyroid hormone receptor β1. Mol Endocrinol 18,1631–1642. bilities. Endocr Rev 14,184–193.
Garnica AD and Chan WY (1996) The role of the placenta in fetal nutrition Leuschner JT, Harvey DJ, Bullingham RE and Paton WD (1986) Pharmacoki-
and growth. J Am Coll Nutr 15,206–222. netics of delta-9-tetrahydrocannabinol in rabbits following single or multi-
Girnita L, Girnita A and Larsson O (2003) Mdm2-dependent ubiquitination ple intravenous doses. Drug Metab Dispos 14,230–238.
and degradation of the insulin-like growth factor 1 receptor. Proc Natl Acad Li H, Wood CL, Getchell TV, Getchell ML and Stromberg AJ (2004) Analysis
Sci USA 100,8247–8252. of oligonucleotide array experiments with repeated measures using mixed
Guenther MG and Lazar MA (2003) Biochemical isolation and analysis of a models. BMC Bioinformatics 5,209.
nuclear receptor corepressor complex. Methods Enzymol 364,246–257. Loiselle FB, Morgan PE, Alvarez BV and Casey JR (2004) Regulation of the
Habayeb OMH, Bell SC and Konje JC (2002) Endogenous cannabinoids: human NBC3 Na+/HCO3- cotransporter by carbonic anhydrase II and PKA.
Metabolism and their role in reproduction. Life Sci 70,1963–1977. Am J Physiol Cell Physiol 286,C1423–C1433.
332
Effects of D9-THC upon BeWo cells
Maccarrone M, Bisogno T, Valensise H, Lazzarin N, Fezza F, Manna C, Robbins MJ, Michalovich D, Hill J, Calver AR, Medhurst AD, Gloger I,
Di Marzo V and Finazzi-Agro A (2002) Low fatty acid amide hydrolase and Sims M, Middlemiss DN and Pangalos MN (2000) Molecular cloning and
high anandamide levels are associated with failure to achieve an ongoing characterization of two novel retinoic acid-inducible orphan G-protein-cou-
pregnancy after IVF and embryo transfer. Mol Hum Reprod 8,188–195. pled receptors (GPRC5B and GPRC5C). Genomics 67,8–18.
Magnusson AL, Waterman IJ, Wennergren M, Jansson T and Powell TL Roh C-R, Budhraja V, Kim H-S, Nelson DM and Sadovsky Y (2005) Microarray-
(2004) Triglyceride hydrolase activities and expression of fatty acid binding based identification of differentially expressed genes in hypoxic term human
proteins in the human placenta in pregnancies complicated by intrauterine trophoblasts and in placental villi of pregnancies with growth restricted
growth restriction and diabetes. J Clin Endocrinol Metab 89,4607–4614. fetuses. Placenta 26,319–328.
Mannavola D, Moeller LC, Beck-Peccoz PB, Persani L, Weiss RE and Refetoff Saito S, Nomura K, Tanuma A and Abiko Y (2001) Identification of altered
S (2004) A novel splice variant involving the 5′ untranslated region of thyroid gene expression in the trophoblast model BeWo cells by caffeine using
hormone receptor β1 (TRβ1). J Endocrinol Invest 27,318–322. cDNA microarray technology. Med Biol 143,171–173.
Maruo T, Matsuo H and Mochizuki M (1991) Thyroid hormone as a biologi- Stephanou A and Handwerger S (1995) Retinoic acid and thyroid hormone
cal amplifier of differentiated trophoblast function in early pregnancy. regulate placental lactogen expression in human trophoblast cells. Endo-
Acta Endocrinol 125,58–66. crinology 136,933–938.
McGavin MK, Badour K, Hardy LA, Kubiseski TJ, Zhang J and Siminovitch Storey KB and Storey JM (2004) Metabolic rate depression in animals:
KA (2001) The intersectin 2 adaptor links Wiskott Aldrich Syndrome pro- Transcriptional and translational controls. Biol Rev Camb Philos Soc
tein (WASp)-mediated actin polymerization to T cell antigen receptor 79,207–233.
endocytosis. J Exp Med 194,1777–1787. Sullivan MH (2004) Endocrine cell lines from the placenta. Mol Cell Endocri-
Nickel BE and Cattini PA (1991) Tissue-specific expression and thyroid- nol 228,103–119.
hormone regulation of the endogenous placental growth hormone variant Taylor AH, Al-Azzawi F, Pringle JH and Bell SC (2002) Inhibition of
and chorionic sommatomammotropin genes in a human choriocarcinoma endometrial carcinoma cell growth using antisense estrogen receptor oligo-
cell line. Endocrinology 128,2353–2359. deoxyribonucleotides. Anticancer Res 22,3993–4003.
Nomura K, Saito S, Ide I, Kamino Y, Sasahara H, Nakamoto T and Abiko Y Thibault C, Lai C, Wilke N, Duong B, Olive MF, Rahman S, Dong H, Hodge
(2004) Caffeine suppresses the expression of the Bcl-2 mRNA in BeWo cell CW, Lockhart DJ and Miles MF (2000) Expression profiling of neural cells
culture and rat placenta. J Nutr Biochem 15,342–349. reveals specific patterns of ethanol-responsive gene expression. Mol
Ohara N, Tsujino T and Maruo T (2004) The role of thyroid hormone in tro- Pharmacol 52,1593–1600.
phoblast function, early pregnancy maintenance, and fetal neurodevelop- Tsai EC, Brown JA, Veldee MY, Anderson GJ, Chait A and Brunzell JD
ment. J Obstet Gynaecol Can 26,982–990. (2004) Potential of essential fatty acid deficiency with extremely low fat diet
Paria BC, Song H, Wang X, Schmid PC, Krebsbach RJ, Schmid HHO, in lipoprotein lipase deficiency during pregnancy: a case report. BMC
Bonner TI, Zimmer A and Dey SK (2001) Dysregulated cannabinoid sig- Pregnancy Childbirth 4,27.
naling disrupts uterine receptivity for embryo implantation. J Biol Chem Uitto J and Pulkkinen L (1996) Molecular complexity of the cutaneous
276,20523–20528. basement membrane zone. Mol Biol Rep 23,35–46.
Park B, Gibbons HM, Mitchell MD and Glass M (2003) Identification of the Wallace DF, Pedersen P, Dixon JL, Stephenson P, Searle JW, Powell LW and
CB1 cannabinoid receptor and fatty acid amide hydrolase (FAAH) in the Subramaniam VN (2002) Novel mutation in ferroportin1 is associated with
human placenta. Placenta 24,990–995. autosomal dominant hemochromatosis. Blood 100,692–694.
Parliamentary Office of Science and Technology, U.K. (1996) Common illegal Wenger T, Toth BE, Juaneda C, Leonardelli J and Tramu G (1999) The
drugs and their effects: Cannabis, Ecstacy, Amphetamines and LSD. House effects of cannabinoids on the regulation of reproduction. Life Sci
of Commons, London. 65,695–701.
Parmentier-Batteur S, Jin K, Xie L, Mao XO and Greenberg DA (2002) DNA Wolf G (2002) The regulation of the thyroid-stimulating hormone of the
microarray analysis of cannabinoid signaling in mouse brain in vivo. Mol anterior pituitary gland by thyroid hormone and by 9-cis-retinoic acid. Nutr
Pharmacol 62,828–835. Rev 60,374–377.
Porcella A, Gessa GL and Pani L (1998) Delta9-tetrahydrocannabinol Xu H, Firulli AB, Zhang X and Howard MJ (2003) HAND2 synergistically
increases sequence-specific AP-1 DNA binding activity and fos-related anti- enhances transcription of dopamine-beta-hydroxylase in the presence of
gens in the rat brain. Eur J Neurosci 10,1743–1751. Phox2a. Dev Biol 262,183–193.
Reimer T, Koczan D, Gerber B, Richter D, Thiesen HJ and Friese K (2002) Yen PM (2001) Physiological and molecular basis of thyroid hormone action.
Microarray analysis of differentially expressed genes in placental tissue of pre- Physiol Rev 81,1097–1142.
eclampsia: up-regulation of obesity-related genes. Mol Hum Reprod 7,674–680. Zuckerman B, Frank DA, Hingson R, Amaro H, Levenson SM, Kayne H,
Rishi AK, Zhang L, Boyanapalli M, Wali A, Mohammad RM, Yu Y, Fontana Parker S, Vinci R, Aboagye K, Fried LE et al. (1989) Effects of maternal
JA, Hatfield JS, Dawson MI, Majumdar AP and Reichert U (2003) Identifi- marijuana and cocaine use on fetal growth. New Engl J Med 320,762–768.
cation and characterization of a cell cycle and apoptosis regulatory protein-1
as a novel mediator of apoptosis signaling by retinoid CD437. J. Biol. Submitted on May 25, 2005; resubmitted on December 15, 2005; accepted on
Chem. 278,33422–33435. March 1, 2006
333
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.