Download pdf - A cannabinoid analogue of Δ9-tetrahydrocannabinol disrupts neural development in chick

Original Article

A Cannabinoid Analogue of D9-TetrahydrocannabinolDisrupts Neural Development in Chick

Delphine Psychoyos,1� Basalingappa Hungund,2,3,4 Thomas Cooper,2,3,4 and Richard H. Finnell1

1Center for Environmental and Genetic Medicine, Institute of Biosciences and Technology, Texas A&M Health Science Center,Houston, Texas

2Department of Psychiatry, College of Physicians & Surgeons, Columbia University, New York, New York3New York State Psychiatric Institute, New York, New York

4Nathan S. Kline Institute for Psychiatric Research, Orangeburg, New York

Marijuana is the most commonly abused illicit drug by pregnant women. Its major psychoactive constituent, D9-THC(D9-tetrahydrocannabinol), crosses the placenta and accumulates in the f!tus, potentially harming its development. Inhumans, marijuana use in early pregnancy is associated with miscarriage, a fetal alcohol-like syndrome, as well aslearning disabilities, memory impairment, and ADHD in the offspring. Classical studies in the 1970 s have reacheddisparate conclusions as to the teratogenic effects of cannabinoids in animal models. Further, there is very little knownabout the immediate effects of D9-THC on early embryogenesis. We have used the chick embryo as a model in order tocharacterize the effects of a water-soluble D9-THC analogue, O-2545, on early development. Embryos were exposed tothe drug (0.035 to 0.35 mg/ml) at gastrulation and assessed for morphological defects at stages equivalent to 9–14somites. We report that O-2545 impairs the formation of brain, heart, somite, and spinal cord primordia. Shorterincubation times following exposure to the drug show that O-2545 interferes with the initial steps of head process andneural plate formation. Our results indicate that the administration of the cannabinoid O-2545 during earlyembryogenesis results in embryotoxic effects and serves to illuminate the risks of marijuana exposure during thesecond week of pregnancy, a time point at which most women are unaware of their pregnancies. Birth Defects Res (Part B)83:477–488, 2008. r 2008 Wiley-Liss, Inc.

Key words: cannabinoids; anencephaly; brain malformation; CNSdevelopment; neural plate; neural folds; neural tube defects; chick embryo;animal model

INTRODUCTION

Marijuana is the most commonly abused illicit drug bypregnant women (NIDA, 2001; WHO, 1997). Its majorpsychoactive constituent, D9-THC (Gaoni and Mechou-lam, 1971), crosses the placenta and accumulates in thef!tus (Blackard and Tennes, 1984), potentially harmingits development. Marijuana use in early pregnancy isassociated with miscarriage (Day and Richardson, 1991),an increased prevalence of congenital malformations(Hecht et al., 1968; Carakushansky et al., 1969; Hingsonet al., 1982), as well as learning disability, memoryimpairment, and ADHD in the exposed offspring(Goldschmidt et al., 2004; Fried et al., 2005; Nolandet al., 2005). However, very little is known aboutmechanisms underlying the adverse effects of D9-THCin early pregnancy. Miscarriage itself could result fromseveral causes, including failure of blastocyst implanta-tion due to a nonreceptive endometrium, as demon-strated in rodent models (Paria et al., 1995), or secondaryto teratogenic effects during early organogenesis (weeks2–3 of human gestation). At such early stages, any severemalformations resulting from the marijuana exposure

would be lethal and result in spontaneous abortion, andthus the embryo would not be detectable for clinicalstudies. The adverse effects of marijuana use duringpregnancy are aggravated by the fact that the potency ofmarijuana preparations, in terms of contents of itspsychoactive constituent D9-THC, has increased nearly25-fold since 1970, when the content of D9-THC inmarijuana was 1.25%; it now reaches 15%–30% in somepreparations (NIDA, 2001). Thus, the modern cannabissmoker may be exposed to doses of D9-THC many timesgreater than his or her counterpart in the 1960 s and 1970 s(WHO, 1997). This fact is important, because the effects ofD9-THC are dose related, and most of the research on

Published online in Wiley InterScience (www.interscience.wiley.com)DOI: 10.1002/bdrb.20166

Additional supporting information may be found in the online version ofthis article.

*Correspondence to: Delphine Psychoyos, Center for Environmentaland Genetic Medicine, Institute of Biosciences and Technology, TexasA&M Health Science Center, Houston, Texas 77030.E-mail: [email protected] 7 July 2008; Accepted 15 August 2008

Birth Defects Research (Part B) 83:477–488 (2008)& 2008 Wiley-Liss, Inc.

cannabis was carried out in the 1970 s using doses thatreflected cannabis intake at the time (WHO, 1997).

Classical studies in rodents show that the develop-mental stage at which D9-THC is administered is acritical factor in determining the degree of D9-THCembryotoxicity. The period of greatest susceptibility tothe embryotoxic effects of D9-THC occurs during earlyorganogenesis (E6.0–8.0 in mouse). During this period,D9-THC administration results in a high incidence ofembryonic death and congenital malformations. Theincidence of fetal loss varies between 40% and 100%,depending on the route of administration, the species,

and the dosage used (Harbison and Mantilla-Plata, 1972;Mantilla-Plata et al., 1973; Uyeno, 1973; Fleischman et al.,1975; Wright et al., 1976). Surviving embryos presentwith malformations in the enteric and nervous systems(holoprosencephaly, anencephaly, exencephaly, cleft pa-late, degenerating spinal cord, and spina bifida), defec-tive skeletogenesis (absent or reduced sternebrae, extraor fused ribs, abnormal vertebral ossification) (Geber andSchramm, 1969; Mantilla-Plata et al., 1975; Jonega, 1976,1977; Rosenkrantz, 1978), and cognitive deficiencies inthe adult (Gianutsos and Abbatiello, 1972; Antonelli,2005). By contrast, exposure to D9-THC prior to

Fig. 1. Dose related effects of cannabinoid O-2545. (A) Structure of O-2545: The terminal carbon atom of the side chain of D9-THC issubstituted by an imidazole group in O-2545. This substitution results in a CB1 agonist which produces the same spectrum ofpharmacological effects in the mouse model as D9-THC. (B–G) Representative embryos treated with vehicle alone (B–D; PBS), or O-2545(+CB) at 0.035 mg/ml (E), 0.07 mg/ml (F), and 0.35 mg/ml (G). Embryos were treated at stages (HH) 4+ (B,E), 4 (C,F) and 4� (D,G).Control embryos reached the equivalent of stages 11 (B), 8+ (C), and 10� (D). Whole-mount in situ hybridization was performed withKrox20 (B,E), or Otx2 and Delta-1 (C,D,F,G). Red arrow in (E): failure of the neural tube to close in treated embryo (see text). All embryosare shown in dorsal view. Abbreviations: aip, anterior intestinal portal; fb, primordium for the forebrain; fg, margin of the foregut; hn,Hensen’s node; mb, primordium for the midbrain; hb, primordium for the hindbrain; nt, neural tube; psm, presomitic mesoderm; s,somites. Scale bar 500 mm in C,D,F,G; 400 mm in B,E.

478 PSYCHOYOS ET AL.

Birth Defects Research (Part B) 83:477–488, 2008

organogenesis (peri-implantation stages), results in 100%embryo death before the embryo can reach organogen-esis (Paria et al., 1995; Persaud and Ellington, 1967,1968a,b). Administration of D9-THC after organogenesis(E10–14) no longer results in a high incidence of death orcongenital malformations (Mantilla-Plata et al., 1975;Banerjee et al., 1975; Haley et al., 1975; Harbison et al.,1977; Fleischman et al., 1980).

There is clearly a developmental period of suscept-ibility to the embryotoxic effects of D9-THC. Thiswindow of susceptibility coincides with the period ofearly organogenesis. In the present studies, we havetargeted this developmental window using the chickembryo in culture model system (stages 31 to 12). Thissystem enables the continuous monitoring of drug-induced teratogenic effects over time. Thus, any mal-formations occurring during the initial stages of embryo-nic development can be easily detected without resultingin embryo death (Gebhardt, 1972; Kotwani, 1998).Furthermore, the separation of the embryo from maternalinfluences permits the evaluation of a response attributedsolely to the effect of the drug applied to the embryo(Wilson, 1978; Jelinek, 1982). Finally, the chick embryorepresents a long-standing model of embryonic verte-brate development (Romanoff, 1960; Balinsky, 1975;Stern, 2002), with close similarities to the human embryoduring early organogenesis (Romanoff, 1960; Balinsky,1975; Rahilly and Muller, 1987; Sulik, 2008).D9-THC is highly lipophilic and thus requires solubi-

lisation with either DMSO or surfactant agents (Tween80t, Emulphort), all of which are potentially embry-otoxic. To overcome this problem, we have used a water-soluble analogue of D9-THC, termed O-2545 (Martinet al., 2006). In this cannabinoid, the terminal carbon ofthe side chain of D9-THC is substituted by an imidazolegroup (Fig. 1A). This substitution results in a cannabi-noid receptor 1 (CB1) agonist that produces the samespectrum of pharmacological effects in the mouse modelas D9-THC (Martin et al., 2006). Yet, O-2545 is soluble inPBS and thus readily applicable to the chick embryo inculture.

METHODS

Chick Embryo Culture and Drug Administration

Chick embryos (Charles River, CT, USA) were ex-planted in culture (New, 1955; Psychoyos and Stern,1996a,b), and grouped according to developmental stage(Hamburger and Hamilton, 1951): group 1 (stage 31 to4�; 39 controls, 63 treated), group 2 (stage 4 to 41; 39controls, 36 treated), group 3 (stage 5 to 6; 38 controls, 42treated) and group 4 (stage 7 to 8–; 8 controls, 13 treated).These stages correspond to days 13–19 (weeks 2 to 3) ofhuman pregnancy (Nishimura et al., 1974). Embryoswere treated with a single exposure of O-2545 in PBS(10 ml) at a given dose: Initially, O-2545 was solubilized inPBS at 0.035 mg/ml and kept in aliquots at �201C. Forsubsequent experiments, O-2545 was diluted with thesame vehicle to obtain concentrations of 0.070 mg/mland 0.35 mg/ml and kept in aliquots at �201C. O-2545was thawed immediately before use for all experiments.The drug or vehicle alone for synchronous controls wasgently placed just above the surface of the embryos(ventral side up), using a 10ml Lambda tip (Corning LifeSciences, Corning, NY, USA). Embryos were cultured for

18–22 hr at 381C, or until controls (vehicle alone) hadreached stages 9–11. Embryos were fixed overnight in 4%PFA in depc-PBS, examined for morphological defects aspreviously described (Psychoyos and Stern, 1996b), andprocessed for immunohistochemistry or whole-mount insitu hybridization.

Endpoints, Morphometric Analysis, andMorphological Score

Each embryo was given a morphological score basedon the following criteria: (1) abnormal brain primordia;(2) abnormal foregut and/or heart primordia, as visua-lized by abnormal anterior intestinal portal, abnormaldorsal and ventral margins of the foregut, abnormalprecardiac mesoderm, and abnormal heart, wheneverpresent; (3) abnormal neural tube posterior to the levelequivalent to the hindbrain/first somite; (4) fragmentedor absent notochord; and (5) abnormal somitogenesis.For each embryo, 1 point was given for each abnormality,and the points were cumulative. Normal embryos scored0; abnormal embryos scored 1 or above. Embryos withthe highest number of malformations scored 5.

Short-Term Incubation Following Exposure toCannabinoid

Embryos in culture were exposed to 0.070 mg/ml O-2545, as described above (47 embryos in total; n 5 31 withdrug; n 5 16 with vehicle alone). Embryos were incu-bated for a shorter period of time (4–5 hr or 6–8 hr insteadof 18–22 hr), after which they were fixed, grosslyanalyzed, and processed as above. Morphometric criteriaincluded presence of a thickened primitive streak andHensen’s node, as well as changes in the morphologyof the nascent neural plate and head fold (whenidentifiable).

Double Whole-Mount in situ Hybridization

In situ hybridization was performed as previouslydescribed (Wilkinson, 1992; Henrique et al., 1995), withminor modifications: Following rehydration from metha-nol, embryos were incubated in 50 mg/ml Proteinase Kfor a shorter period (1 min for stages up to 8, 2 min forstage 9, and 3 min for stages 10–12). For the analysis oftwo genes, embryos were hybridized simultaneouslywith digoxigenin (DIG)- and fluorescein (fluorescein-12-UTP, FITC)-conjugated probes (Roche Diagnostics). Afterhybridization at 661C, embryos were washed in hybridi-zation buffer, followed by washes in 100mM maleatebuffer containing 0.1% Tween 20 (MABT). Embryos werethen blocked in MABT containing 2% Blocking Reagent(Roche Diagnostics) and 20% sheep serum (JacksonImmunoResearch), and incubated overnight with anti-DIG antibody coupled with alkaline phosphatase (1:2000;Roche Diagnostics) at 41C. After several washes inMABT, anti-DIG antibody-coupled alkaline phosphataseactivity was revealed with BCIP (5-bromo-4-chloro-3-indolyl phosphate)/NBT (nitroblue tetrazolium chloride)substrate (Roche Diagnostics) in staining solution (0.1 MTris-HCl pH9.5; 0.1 M NaCl). In order to inactivatealkaline phosphatase activity, embryos were pinned ona sylgard-coated dish containing PBS, fixed overnight in4% PFA, placed in a scintillation vial with 100 mM EDTAin MABT buffer, and incubated at 651C for 30 min. The

479CANNABINOID ANALOGUE


embryos were subsequently washed in MABT, andblocked in the same blocking buffer as above. Embryoswere incubated in alkaline phosphatase-coupled anti-FITC-antibody (1:500) overnight (41C), and washed inMABT for up to 3 hr. Embryos were washed in Trisbuffer (100 mM Tris-HCl pH 8.45 with 0.1% Tween-20),and the color reaction was performed with Vector RedSubstrate Kit (Vector Research) in Tris buffer. Embryoswere transferred to PBS, fixed in 4% paraformaldehyde,and further examined morphologically prior to beingprocessed for photography. The following probes wereused at 200ng/ml: Delta-1 (Henrique et al., 1995); Krox20(Irving et al., 1996); Lef-1 (Schmidt et al., 2004); Noggin(Connolly et al., 1997); Otx2 (Bally-Cuif et al., 1995); Pax6(Goulding et al., 1993); Shh (Levin et al., 1995); Sox2 (Rexet al., 1997); to determine alterations in expressionpatterns secondary to the drug treatment.

Photomicroscopy and Histology

Embryos were photographed as whole mount pre-parations with a stereo microscope (model MZ9.5; LeicaMicrosystems Ltd, Switzerland), equipped with a 5.0megapixel CDD 35 mm digital camera (model DFC480;Leica Microsystems Ltd, Switzerland). Image acquisitionwas controlled via the MetaView software. Embryoswere processed for sectioning as described previously(Psychoyos and Stern, 1996b). To ensure signal preserva-tion, embryos were dehydrated in absolute methanol, inpropan-2-ol and cleared in 1,2,3,4-tetrahydronaphtalene(Sigma Aldrich) prior to xylene and wax treatment.10 mm Sections were photographed on an uprightmicroscope (model Axioskop 40 A Pol; Carl Zeiss MicroImaging, Inc., Thornwood, NY, USA) equipped witha 1.4 megapixel CDD digital camera (model DigitalEclipse DXM 1200; Nikon Inc., Melville, NY, USA).Image acquisition was controlled via ACT-1 software(Nikon Inc.).

RESULTS

Embryotoxic Effects of Cannabinoid O-2545

Stage 31 to 8� embryos were exposed either to vehicleor O-2545 at concentrations ranging from 0.035 mg/ml to0.35 mg/ml, and cultured for 18–22 hrs, or until thecontrol embryos had reached stages 9–12. Results aresummarized in Table 1: At the lowest dose of O-2545used (0.035 mg/ml), O-2545–induced malformations in

77.7% of the embryos (42/54; all stages combined;p 5 0.0001; Table 1). A typical example is shown in Fig.1E, in which the brain is subdivided into identifiableprimordia; yet, the forebrain lacks subdivision intoprosencephalon and separate optic vesicles. There is noclear boundary between midbrain and hindbrain pri-mordia. The anterior walls of the neural tube at the levelof the midbrain and hindbrain fail to oppose each other(red arrow in Fig. 1E). At a slightly higher dosage(0.07 mg/ml), O-2545–induced malformations in 82.8%embryos (58/70; all stages combined; p 5 0.0001; Table 1).At this dose, the embryos still appear viable. A typicalexample is shown in Fig. 1F: In this embryo, neurulationis severely disrupted, resulting in a poorly formed brainwith no clear forebrain, midbrain, and/or hindbrainprimordia. There are no visible heart and foregutprimordia. The neural tube fails to close along theanteroposterior axis of the embryo. There are no somitesvisible despite Delta-1 expression in presomitic meso-derm. At slightly higher dosage (0.35 mg/ml), O-2545severely impedes embryonic development (30/30; allstages combined; p 5 0.0001; Table 1). A representativeexample is depicted in Fig. 1G, in which the embryo lacksany morphological structures (such as brain, neural tube,somites). It is not clear whether at this higher dose,treated embryos would have survived past the culturewindow (18–22 hrs).

Stage-Specific Embryotoxicity to CannabinoidO-2545

Morphological examination of the embryos followingtreatment with O-2545 shows that at all stages at whichO-2545 is administered, they are responsive to itsteratogenic effects (Table 1). At a cannabinoid dose of0.035 mg/ml, stage 5 to 6 embryos are highly sensitive toits teratogenic effects (18/19 cases; p 5 0.0001; Table 1).At slightly higher cannabinoid dosage (0.07 mg/ml),stages 31 to 41 embryos are most responsive to theeffects of O-2545 (42/45 cases; stages 31 to 4� and 4 to 41

combined). Finally, at a 0.35 mg/ml dose, all stages areequally responsive to the effects of cannabinoid O-2545.It is not clear why older embryos (stage 5 to 6) are moreresponsive than younger embryos at the lowest doseused in these experiments. Analysis of the morphologicalscore (Supplementary Fig. 1) shows that at 0.07 mg/ml,there is a greater number of embryos displaying themaximal score (5) following O-2545 treatment at stages

Table 1Teratogenic Effect of O-2545 Following Initial Exposure at Stages 31 to 8� and Culture for Up to 18–22 hr

Dose (mg/ml) 0 0.035 0.07 0.35Doses 0.035 mg/ml and0.07 mg/ml combined

Stageb na na na na na

31 to 4- 5/39 (12.8%) 11/18 (61.1%)� 25/27 (92.7%)� 18/18 (100.0%)� 26/45 (57.7%)�

4 to 41 5/39 (12.8%) 7/9 (77.7%)� 17/18 (94.4%)� 9/9 (100.0%)� 24/27 (88.8%)�

5 to 6 4/38 (10.55) 18/19 (94.7%)� 15/20 (75.0%)� 3/3 (100.0%)�� 33/39 (84.5%)�

7 to 8 0/8 (0.0%) 6/8 (75.0%)�� 1/5 (20.0%) No exps 7/13 (53.8%)��

combined 14/124 (11.3%) 42/54 (77.7%)� 58/70 (82.8%)� 30/30 (100.0%)�

aNo. abnormal/total no. embryos, also given as a % of all treated embryos within each category.bStage at which treatment with O-2545 or vehicle began (in HH).p-Values are based on Chi-squared tests with Yates correction (�P, 0.0001, ��P, 0.001, ��P, 0.015; ��P, 0.0389).



31 to 41 (31/44 embryos) than following treatment atstages 5 to 6 (2/20 embryos). This suggests that at earlystages (31 to 41), more primordia are affected by O-2545than at later stages (5 to 6), implying that youngerembryos are more vulnerable to the adverse effect ofO-2545 than older ones. We subsequently assessed therange of malformations resulting from cannabinoidO-2545 exposure during development. We found thatO-2545 is capable of inducing a range of embryonicmalformations (Table 2):

Anterior Neural Pattern Defects

We assessed brain malformations following cannabi-noid treatment using morphological observation, aswell as the expression pattern of the anterior neuralmarkers Krox20, Otx2, Pax6, and Sox2. We find that thedegree of brain malformations observed in treatedembryos increases with dosage: At lower O-2545concentration (0.035 mg/ml), 46.3% treated embryoshave neural fold malformations (p 5 0.0001), comparedto 72.5% following treatment with a slightly higher doseof O-2545 (0.07 mg/ml). At a higher dose of O-2545(0.35 mg/ml), 96.6% embryos present neural foldmalformations.

In less severe cases of brain anomalies followingO-2545 treatment, the neural folds fail to elevate andfuse, comparable to craniorachischisis in mammaliansystems (red arrow in Fig. 1E). In most of the cases ofembryos with abnormal brains, the brain is present, butpoorly segmented into forebrain, midbrain, and hind-brain primordia (Fig. 2B,C). In other embryos presentingabnormal brains, the different primordia of the brainpresent severe hypoplasia; neurulation is severely

disrupted, resulting in a poorly formed brain with noclearly discernible brain primordia (Fig. 2D,F); In thoseembryos, there is no clear organization along thedorsoventral axis of the embryo, and no separation ofthe eye field into two separate domains (Fig. 3E,K).

Analysis of gene expression indicates that the spatio-temporal expression of the anterior neural markersKrox20 and Otx2 is maintained even within the abnormalbrain of treated embryos (Fig. 1E,F; Fig. 2). Thespatiotemporal expression of the anterior neural markerPax6 is also conserved in most cases of treatedembryos, despite severely abnormal brain morphology(Fig. 3F,G). There are some exceptions where Pax6expression in treated embryos is not down-regulatedalong the ventral midline of the anterior neural tube (Fig.3E). This suggests that in some cases, cannabinoidO-2545 impedes the separation of anterior neural tissueinto two independent eye fields. Finally, analysis of theresults show that O-2545 exerts most of its teratogeniceffects on brain development when administered atstages 31 to 41, stages at which anterior neuronalprecursors migrate anteriorly to form the anterior neuralplate (Table 2).

Defects in the Neural Tube Posterior toHindbrain/1st Somite

The majority of embryos treated with O-2545 exhibitneural tube defects (Table 2). These are classified intothree groups: In group A, the treated embryos present aneural tube which is atrophied, either unilaterally (Fig.2C) or bilaterally (Fig. 2D) with respect to the neuraltube. In Fig. 2C, the left side of the neural tube appearscontorted, whereas the right side appears atrophied at

Table 2Types of Morphological Defects Observed Following Treatment with O-2545

Stagea Dose (mg/ml) Total Neural foldsb Neural tubec Somitesd

31 to 4� 0 39 4 (10.3%) 1 (2.6%) 0 (0.0%)0.035 18 7 (38.8%)� 6 (33.3%)�� 7 (38.8%)�

0.07 27 22 (81.5%)�� 21 (77.7%)� 22 (81.5%)�

0.35 18 18 (100.0%)� 18 (100.0%)� 18 (100.0%)�

4 to 41 0 39 2 (5.2%) 0 (0%) 1 (2.6%)0.035 18 4 (44.4%) 3 (33.3%)�� 6 (66.6%)��

0.07 17 16 (94.1%)� 13 (76.5%)�� 16 (94.2%)�

0.35 9 9 (100.0%)� 9 (100.0%)� 9 (100.0%)�

5 to 6 0 38 4 (10.5%) 0 (0%) 0 (0%)0.035 19 11 (57.9%)�� 7 (36.8%)�� 10 (52.6%)�

0.07 20 11 (55.5%)�� 4 (20.0%)�� 10 (50.0%)�

0.35 3 2 (66.6%) 2 (66.6%)�� 3 (100.0%)�

7 to 8 0 8 0 (0%) 0 (0%) 0 (0%)0.035 8 3 (37.5%) 1 (12.5%) 6 (75.0%)1.9 5 1 (20.0%) 0 (0%) 0 (0%)0.35 No exps – – –

All stages 0 124 10 (8.1%) 1 (0.8%) 1 (0.8%)0.035 54 25 (46.3%)� 17 (31.5%)� 29 (53.7%)�

0.07 69 50 (72.5%)� 38 (55.1%)� 48 (69.6%)�

0.35 30 29 (96.6%)� 29 (96.6%)� 30 (100.0%)�

aStage at which treatment with O-2545 or vehicle began (in HH). The results are presented as the number of embryos presentingabnormalities in the anterior nervous system (b), the neural tube posterior to the level of the first somite (c), and the somites (d). Eachresult is also given as a % of all treated embryos within each category. p-Values are based on Chi-squared tests with Yates correction(�P, 0.0001, ��P, 0.0004, ��P, 0.0008, ��P, 0.0043, ��P, 0.0208, ��P, 0.0328).



the level posterior to Otx2 expression. In Fig. 2D, theneural tube appears contorted and atrophied on bothsides. In group B, the neural tube of treated embryos isentirely missing posterior to the level of the presumptivehindbrain (Fig. 2F,H). In those embryos, there is noneural tube visible posterior to the level of the pre-sumptive hindbrain. Finally, in group C, the treatedembryos present with a neural tube in which the neuralfolds fail to elevate along the dorsoventral axis (Fig.3E,G). In those embryos, the atrophied neural tube hasthe morphology of a neural plate, and no neural folddevelopment has yet occurred.

We subsequently analysed the expression pattern ofPax6 in these embryos: We find that following treatmentat stage 31 to 4�, Pax6 expression was severely down-regulated along most of the length of the neuraltube, except for a hemilateral Pax6 positive segment(Fig. 3E–G); By contrast, when the embryos were treatedwith O-2545 at slightly later stage (stage 41), Pax6expression was not downregulated along the neuraltube (Fig. 3H). It is not known from these experimentswhether the lack of Pax6 expression observed inembryos depicted in Fig. 3E–G is attributable to thelack of somites and/or notochord in these embryos.Finally, we also noted that most neural tube defectsobserved herein occur when embryos are treated atstages 31 to 41 (Table 2), indicating that gastrulation

stages are the most sensitive to O-2545–inducedembryotoxicity.

Reduced Paraxial Mesoderm and AbnormalSomite Patterning

O-2545 exposure induced somite defects in a majorityof treated embryos (Table 2). In a majority of those cases,there are no somites visible on either side of the neuraltube (60/123; Fig. 4C; see also Fig. 1F). In other cases oftreated embryos with abnormal somites, somitogenesisoccurs partially (16/123): In those embryos, one side ofthe neural tube has no visible somites (with theoccasional exception of 1–2 anterior somites); the otherside of the neural tube has somites, some of which maybe atrophied (Fig. 4F, I and L). In those embryos, Lef-1expression is either absent on the side of the embryodevoid of somites (blue arrow; Fig. 4F) or very weak anddiffuse (blue arrow; Fig. 4I). It is not clear from theseexperiments whether the lack of Lef-1 expression on oneside of the neural tube denotes (1) the presence ofunsegmented mesoderm which has failed to develop intosomites and which lacks Lef-1 expression, or (2) the lackof unsegmented mesoderm, and the concomitant pre-sence of lateral plate mesoderm. Transverse sectionstaken through similarly treated embryos (Fig. 4M)suggest that at least in some cases, the ‘unsegmented’

Fig. 2. O-2545 inhibits the formation of the nervous system. Representative embryos treated with vehicle alone (A,E,G), 0.035 mg/mlO-2545 (B–D), or 0.07 mg/ml O-2545 (F,H). Embryos were treated at stages 4� (G,H), 4+ (A,B,E,F) and 5 (C,D). Control embryo reachedthe equivalent of stage 9+ (A,E) and 11�(G). Whole-mount in situ hybridization was performed with Delta-1(E-H), Otx2 (A–D,E,F), andSox2. (G,H); Arrow in (C) indicates posterior extent of Otx2 expression. All embryos are shown in dorsal view, except (D), which isshown in ventral view. Abbreviations as in Fig. 1. Scale bar, 650mm (E), 500mm (G-H), 300mm (A–D).



tissue adjacent to the neural tube is, in fact, lateral platemesoderm, and not unsegmented somite mesoderm. Thiswould indicate that O-2545 interferes not only with thecondensation of presomitic mesoderm into segmentalunits (somites), but also with the initial allocation ofmesoderm into presomitic mesoderm. Experimentsdescribed below further investigate this possibility.

Early Effects of Cannabinoid O-2545 on NeuralPlate and Presomitic Mesoderm Formation

We sought to determine whether the embryonicdefects described above can be at least partly explained

by an initial impairment of neural plate, head process,and presomitic mesoderm morphogenesis as a result ofthe cannabinoid treatment. To investigate this possibility,stage 31 to 7 embryos were treated with 0.07 mg/mlO-2545, cultured for up to 6–8 hr (by which stage controlembryos had developed to between the head process to3rd somite stages), and analyzed for gene expression ofNoggin and Shh (head process markers; Levin et al., 1995;Connolly et al., 1997), Sox2 (neural plate marker; Rexet al., 1997), Pax6 (nascent neural fold marker; Gouldinget al., 1993), and Delta-1 (presomitic marker; Henriqueet al., 1995). We found that O-2545 produced a majorinsult to the neural plate of treated embryos. In contrast

Fig. 3. Effect of O-2545 on neural tube formation. Representative embryos treated with vehicle alone (A–D), or 0.035 mg/ml O-2545(E–H). Embryos were treated at stages 3+/4� (A–C,E–G) and 4+ (D,H). Control embryos reached the equivalent of stages 9 (A), 10 to 10+

(B,C) and 11 (D). Whole-mount in situ hybridization was performed with Pax6 and Shh. (A–H) shown in dorsal view. Abbreviations asin Fig. 1. Scale bar 600 mm (A–E), 500 mm (F,G), 400 mm (H), 200 mm (I), 125 mm (J–N).



to control embryos in which Sox2 correlates with thepresumptive neural plate domain just anterior to theexpression domain of Delta-1, the Sox2 expressiondomain appeared shrunken in treated embryos, andoverlapped with the domain of Delta-1 expression (Fig.5G). The neural plate rudiment appeared disorganizedand formed many blebs (Fig. 5G–I). Thus, the brainabnormalities observed in drug-treated embryos follow-ing 16–22 hr of initial treatment can be at least partlyexplained by an immediate effect of this cannabinoid onneural plate formation. O-2545 is also shown to affecthead process formation, as evidenced by the absence ofShh expression (Fig. 5J), as well as the formation of neuralfolds, based on the weak pattern of expression of Pax6(Fig. 5J).

We subsequently explored whether O-2545 impairspresomitic mesoderm formation. Gene expression datarevealed that in control embryos, expression of Delta-1 isdispersed beyond the primitive streak proper, expandinginto the regions immediately adjacent to it (Fig. 5A). Bycontrast, in O-2545 treated embryos, expression of Delta-1is strictly confined to the boundaries of the primitivestreak (Fig. 5F). Furthermore, the intensity of Delta-1signal in treated embryos appears much stronger than inpaired control embryos (compare Fig. 5F [treated] to 5A[control]). It is not known whether this intensity is due toa failure of cells to emigrate from the primitive streak. Weobserved that in treated embryos, the cell layers withinthe primitive streak (ectoderm, mesoderm, and endo-derm) are not linear, as is the case in the controls (thelatter shown in Fig. 5M), but appear contorted, as if themigrating cells are not able to exit the primitive streakand are forced to remain there (Fig. 5P). The contortions,which we observe within the primitive ridges, appear toform as a result of the accumulation of cells within theprimitive ridges (Fig. 5P). Perhaps this accumulation inthe primitive streak can partially explain the phenotypeobserved at the later developmental stages, namely thefailure to form somitic mesoderm. Experiments areunderway to determine whether this is the case, usingcarbocyanine dyes to assess morphogenetic movementsas previously reported (Psychoyos and Stern, 1996a) inthe drug-treated embryos. These results suggest thatinhibition of the morphogenetic movements duringgastrulation and neurulation may in part explain theteratogenic effects of O-2545 observed at later stages(9–11). To summarize, observation of embryos in thehours after O-2545 administration indicate that the cellmovements and proliferation characteristic of gastrula-tion may be impeded by cannabinoid exposure,

Fig. 4. Abnormal somitogenesis in cannabinoid treated em-bryos. Representative embryos treated with vehicle alone(A,D,G,J), or O-2545 at 0.035 mg/ml (B,H,K) or 0.07 mg/ml(E). Embryos were treated at stages 3+/4� (A,B,G,H), 5 (D,E)and 7 (J,K). Control embryos reached the equivalent of stages 9-11. Whole-mount in situ hybridization was performed with Otx2(A,B,J,K), or Lef-1 (D-H). (C,F ,I and L) Higher magnification ofthe boxed area at level of presumptive somite mesoderm asindicated in (B, E, H and K respectively). (M) transverse sectionat level indicated in (L); Blue arrow in (F,I,L,M) denotes absenceof somite on hemilateral side of the anteroposterior axis. Allembryos are shown in dorsal view. Abbreviations as in Fig. 1;lpm, lateral plate mesoderm (splanchnic and somatic mesodermprimordium). Scale bar, 700 mm (G), 650 mm (D,E), 500 mm(A,B,H,J), 400 mm (K), 150 mm (C,F,I,L), 70 mm (M).

3



subsequently leading to neural and somite deficienciesseen in these embryos at stages 9–11. O-2545 appears toinhibit the migration of mesodermal cells, and to causethe formation of a small neural plate, subsequentlyresulting in anterior CNS and somite malformations.

DISCUSSION

This study shows that administration of synthetic D9-THC analogue, cannabinoid receptor CB1 specific ago-nist O-2545, to a chick embryo prior to neurogenesis andsomitogenesis, results in abnormalities in the developingbrain and somites, indicating that cannabimimetic drugsare indeed capable of interfering with the initial stages ofCNS and musculoskeletal development. The presentexperiments illustrate for the first time the effects of acommonly abused recreational drug on early embryomorphology and endogenous gene expression, both inthe short as well as the long term. In particular, we show

that cannabinoid O-2545 causes down-regulation of Pax6expression within the nascent neural tube. Finally, wedemonstrate that the short-term developmental effects ofcannabinoid O-2545 include failure of the presomiticmesoderm to emigrate from the primitive streak andabnormal neural plate formation.

The concentrations of cannabinoid O-2545 used for thepresent study were selected because at the volume used(10ml), all three doses (equivalent to 30–292 mg/kg for a12 mg embryo) correspond to levels approximately 60- to116-fold higher than levels of D9-THC detected inhumans following inhalation of a single marijuanacigarette with 3.55%–20% D9-THC content (McGilveray,2005), and thus reflects the concentrations of cannabinoidfound in chronic cannabis users. Because the embryos arecultured, it is possible to determine the exact cannabi-noid dosage to which the embryo per se is exposed, incontrast with rodent studies, which require administra-tion of the drug through the mother, thereby limiting the

Fig. 5. Short term effects of O-2545. Representative embryos treated with vehicle alone (A–E) or O-2545 at 0.07 mg/ml (F–J). Embryoswere treated at stages 3+ to 4 (A–D, F–I) or 5+ (E,J) and cultured for up to 6–8 hr. Control embryos reached the equivalent of stages 5 to8� (3 somites). Whole-mount in situ hybridization was performed with Noggin (A,F), Delta-1 (A,B,F,G), Sox2 (B-D,G-I), Pax6 and Shh(E,J) probes. (K-P) Transverse sections through (B,G) at levels indicated in (B,G). (A–J) shown in dorsal view. Abbreviations: anp,anterior neural plate; fg, margin of the foregut; hf, head fold (subcephalic pocket); hm, head mesenchyme; n, notochord; nf, anteriorneural folds; Hensen’s node (hn); pg, primitive groove of the primitive streak; pr, primitive ridges of the primitive streak; ps, primitivestreak; psm, prospective mesoderm; vm, ventral midline of the neural tube. Scale bar, 800 mm (F), 600mm (B,G), 350 mm (E,J), 200 mm(A,D,H) 170 mm (C,I,N,O), 100mm (P), 70 mm (L,M,K).



ability to know the actual amount to which thedeveloping embryo is exposed. The findings presentedherewith corroborate classical studies, which havedescribed a high percentage of malformations, particu-larly holoprosencephaly and musculoskeletal defects, inthe offspring of D9-THC-treated pregnant rodents (Geberand Schramm, 1969; Mantilla-Plata et al., 1975; Jonega,1976, 1977; Rosenkrantz, 1978). The defects observed inthese classical studies can now partly be explained by thedirect interference of cannabinoids with the formation ofsomite and brain primordia. Classical studies on canna-binoid teratogenicity have also noted a high level ofembryonic death in rodent embryos harvested on E14–E17 following administration of cannabinoids at E6–E8(reviewed in Fleischman et al., 1975). In the presentstudy, chick embryos in culture are collected 18–22 hrpost–drug administration. This short period of develop-ment is equivalent to mouse embryos developedto E8.5. This approach therefore allows us to visualizeembryonic malformations prior to an embryonic stage atwhich resorption would have occurred in mouse.Thus, the present study extends these earlier classicalfindings by providing a possible mechanism forthe observed embryotoxicity and teratogenicity of can-nabinoids. Our results show that cannabimimetic O-2545impairs the formation of the neural plate, whichappears disorganized, with formation of multiple em-bryonic blebs. Similar bleb formation has been observedon the neural plate of ethanol-treated mouse embryos(Nakatsuji and Johnson, 1984), suggesting thatperhaps alcohol and cannabinoids interfere with earlydevelopment via some shared mechanisms. Thispossibility is supported by the finding that somechildren who have been exposed to cannabinoidsthrough maternal use of cannabis during pregnancypresent fetal alcohol syndrome–like defects (Hingson,1982).

The neural and somite developmental defects inducedby cannabimimetic O-2545 correlate with the expressionpattern of cannabinoid CB1 receptor in early embryos,which includes the neural tube of E11 rat, and thepresomitic mesoderm and hindbrain of stage 10 chickembryos (Buckley et al., 1998; Christiansen et al., 2001;Begbie et al., 2004). The present experiments do notdetermine whether exogenous cannabinoids, such asO-2545 and D9-THC, mediate their embryotoxic effectsthrough disruption of an endogenous cannabinoid CB1receptor system (Matsuda et al., 1990). Yet, in support ofthis hypothesis, exogenous cannabinoids have beenshown to bind to cannabinoid CB1 receptor in theembryonic mouse brain and during implantation (Anto-nelli et al., 2005; Paria and Dey, 2000). Furthermore,recent studies have determined a role for endogenouscannabinoids in later stages of neurogenesis (retina andhippocampal neural progenitor proliferation) (Straikeret al., 1999; Aguado et al., 2007), suggesting thatcannabinoid CB1 receptor signaling might have a rolein earlier stages of nervous system formation. Thus, ourfuture studies will focus on determining whethercannabinoids interfere with early chick developmentvia interference with endogenous cannabinoid CB1receptor system signaling. Our preliminary studies usingRT-PCR already indicate CB1R expression in chickembryos at stage HH4 (gastrulation), 8� (3 somite) and 111 (12 somite).

The direct implication of our findings in the chickembryo to the human condition must be interpreted withextreme caution. First, there have been no conclusivestudies so far as to the teratogenicity of D9-THC (6–8) inhuman. Instead, recent reports show that the adverseeffects of D9-THC might be more subtle in human than inanimal models: Those recent studies report the presenceof ‘‘subtle’’ neurobehavioral effects as opposed to frankteratogenic effects (Goldschmidt et al., 2004; Fried et al.,2005; Noland et al., 2005). Second, the cannabinoidconcentrations required to induce malformations in thedeveloping embryo are rather high, even though we donot know the corresponding effective intraembryonicconcentration in humans. We believe, however, that thisanalysis complements experiments investigating thebasis of cannabinoid-induced embryotoxicity in mam-malian models.

Finally, the results presented herein indicate thatstage 31 to 6 embryos are particularly responsive tothe embryotoxic effect of D9-THC analogue O-2545.Exposure to this cannabimimetic drug during thisshort developmental window results in malformationswhich are so severe by stage 8–12, that embryoswould most likely not have survived past the cultureperiod, a phenotype similar to resorptions observed inmouse following cannabimimetic administration atE6.0–E8.0 (Persaud and Ellington, 1967; Fleischmanet al., 1975; Rosenkrantz, 1978; Fleischman et al., 1980;Rosenkrantz, 1983; Abel, 1984). This developmentaltime window corresponds to week 2.5 of humangestation (Nishimura et al., 1974; Butler and Juurlink,1987; Harkness and Baird, 1997). Most women areunaware they are pregnant at this point of gestationand thus of the risks of concomittant use of marijuana.Yet, we know that there is a high incidence of earlymiscarriage observed in marijuana smokers (Piomelli,2004). Thus, our studies could provide a possibleexplanation for this adverse and embryotoxic effect ofmarijuana.

Because the basic molecular and cellular mechanismsthat control early organogenesis are evolutionary con-served amongst species, the knowledge gained byanalyzing the mechanism of cannabinoid-mediatedembryotoxicity during early chick embryogenesis canbe directly applicable to our understanding of human D9-THC–induced teratogenesis. Our approach using thechick embryo as a model, provides valuable informationfor better understanding of risks of marijuana exposureduring the very early stages of human pregnancy. This isparticularly relevant to current considerations of mar-ijuana legalization and therapeutic applications. Finally,the research presented herein has implications wellbeyond the field of marijuana toxicity; it could serve asa basis for teratogenic studies involving other drugs ofabuse, such as cocaine, methamphetamines, and opiates,or combinations thereof.

Conflict of Interest Statement: The authors declarethere are no conflicts of interest.

ACKNOWLEDGMENTS

This paper is dedicated to the memory of Bill Martin. Weare grateful to Bill Martin (Virginia CommonwealthUniversity) for O-2545, and to Vid Persaud (Universityof Manitoba) and Raphael Mechoulam (Hebrew



University Jerusalem) for comments. We thank LaureBally-Cuif, James Briscoe, Martyn Goulding, DomingosHenrique, Andrew Lassar, Robin Lovell-Badge, AndreaMunsterberg, Cliff Tabin and David Wilkinson. Thiswork was supported by the Margaret M. Alkek Founda-tion to RHF. DP is recipient of Award 1F32DA021977-01A1 from NIDA.

References

Abel EL. 1983. Marihuana. Tobacco. Alcohol and Reproduction. BocaRaton: CRC Press.

Aguado T, Romero E, Monory K, Palazuelos J, Sendtner M, Marsicano G,Lutz B, Guzman M, Galve-Roperh I. 2007. The CB1 cannabinoidreceptor mediates excitotoxicity-induced neural progenitor prolifera-tion and neurogenesis. J Biol Chem 282:23892–23898.

Antonelli T, Tomasini MC, Tattoli M, Ferraro L. 2005. Prenatal exposure tothe CB1 receptor agonist WIN 55,212-2 causes learning disruptionassociated with impaired cortical NMDA receptor function andemotional reactivity changes in rat offspring. Cereb Cortex 15:2013–2020.

Balinsky BI. 1975. An Introduction to Embryology. New York: WBSaunders.

Bally-Cuif L, Gulisano M, Broccoli V, Boncinelli E. 1995. c-Otx2 isexpressed in two different phases of gastrulation and is sensitive toretinoic acid treatment in chick embryo. Mech Dev 49:49–63.

Banerjee BN, Galbreath C, Sofia RD. 1975. Teratologic evaluation ofsynthetic delta-9-tetrahydrocannabinol in rats. Teratology 11:99–101.

Begbie J, Doherty P, Graham A. 2004. Cannabinoid receptor, CB1,expression follows neuronal differentiation in the early chickembryo. J Anat 205:213–218.

Blackard C, Tennes K. 1984. Human placental transfer of cannabinoids. NEngl J Med 311:797–799.

Buckley NE, Hansson S, Harta G, Mezey E. 1998. Expression of the CB1and CB2 receptor messenger RNAs during embryonic developmentin the rat. Neuroscience 82:1131–1149.

Butler H, Juurlink BHJ. 1987. An atlas for staging mammalian and chickembryos. Florida: CRC Press.

Carakushansky RL, Neu RL, Gardner LI. 1969. Lysergide and cannabis aspossible teratogens in man. Lancet 1:150–151.

Christiansen J, Coles E, Robinson V, Pasini A, Wilkinson, DG. 2001.Screening from a subtracted embryonic chick hindbrain cDNAlibrary: identification of genes expressed during hindbrain, midbrainand cranial neural crest development. Mech Dev 102:119–133.

Connolly DJ, Patel K, Cooke J. 1997. Chick Noggin is expressed in theorganizer and neural plate during axial development, but offers noevidence of involvement in primary axis formation. Int J Dev Biol41:389–396.

Day NL, Richardson G. 1991. Prenatal marijuana use: epidemiology,methodological issues and infant outcome. In: Chasnoff IJ, editor.Clinics in Perinatology. Philadelphia: WB Saunders. p 77–92.

Fleischman RW, Hayden DW, Rosenkrantz H, Braude MC. 1975.Teratologic evaluation of delta-9-tetrahydrocannabinol in mice,including a review of the literature. Teratology 12:47–50.

Fleischman RW, Naqvi RH, Rosenkrantz H, Hayden DW. 1980. Theembryotoxic effects of cannabinoids in rats and mice. J EnvironPathol Toxicol 4:471–482.

Fried PA, Watkinson B, Gray R. 2005. Neurocognitive consequences ofmarihuana—a comparison with pre-drug performance. NeurotoxicolTeratol 27:231–239.

Gaoni Y, Mechoulam R. 1971. The isolation and structure of delta-1-tetrahydrocannabinol and other neutral cannabinoids from hashish.J Am Chem Soc 93:217–224.

Geber WF, Schramm LC. 1969. Effect of marihuana extract on fetalhamsters and rabbits. Toxicol Appl Pharmacol 14:276–282.

Gebhardt DE. 1972. The use of chick embryo in applied teratology. In:Woolam DHM, editor. Advances in Teratology. London: AcademicPress. p 97–111.

Gianutsos G, Abbatiello ER. 1972. The effect of prenatal cannabis sativaon maze learning ability in the rat. Psychopharmacology 27:117–122.

Goldschmidt L, Richardson GA, Cornelius MD, Day NL. 2004. Prenatalmarijuana and alcohol exposure and academic achievement at age10. Neurotoxicol Teratol 26:521–532.

Goulding MD, Lumsden A, Gruss P. 1993. Signals from the notochordand floor plate regulate the region-specific expression of two Paxgenes in the developing spinal cord. Development 117:1001–1016.

Haley SL, Wright PL, Plank JB, Keplinger ML, Braude MC, Calandra JC.1975. The effect of natural and synthetic delta-9-tetrahydrocannabi-nol on fetal development. Toxicol Applied Pharmacol 25:450–457.

Hamburger V, Hamilton HL. 1951. A series of normal stages in thedevelopment of the chick. J Morphol 88:49–92.

Harbison RD, Mantilla-Plata B. 1972. Prenatal toxicity, maternal distribu-tion and placental transfer of tetrahydrocannabinol. J Pharmacol ExpTher 180:1446–1453.

Harbison RD, Mantilla-Plata B, Lubin DJ. 1977. Alteration of delta-9-tetrahydrocannabinol-induced teratogenicity by stimulation andinhibition of its metabolism. J Pharmacol Exp Ther 202:455–465.

Harkness LM, Baird DT. 1997. Morphological and molecular character-istics of living human fetuses between Carnegie stages 7 and 23:developmental stages in the post-implantation embryo. Hum ReprodUpdate 3:3–23.

Hecht F, Beals RK, Lees MH, Jolly H, Roberts P. 1968. Lysergic-acid-diethylamide and cannabis as possible teratogens in man. Lancet2:1087.

Henrique D, Adam J, Myat A, Chitnis A, Lewis J, Ish-Horowicz D. 1995.Expression of a Delta homologue in prospective neurons in the chick.Nature 375:787–790.

Hingson R, Alpert JJ, Day N, Dooling E, Kayne H, Morelock S,Oppenheimer E, Zuckerman B. 1982. Effects of maternal drinkingand marijuana use on fetal growth and development. Pediatrics70:539–546.

Irving C, Nieto MA, DasGupta R, Charnay P, Wilkinson DG. 1996.Progressive spatial restriction of Sek-1 and Krox-20 gene expressionduring hindbrain segmentation. Dev Biol 173:26–38.

Jelinek R. 1982. Use of chick embryo in screening for embryotoxicity.Teratog Carcinog Mutagen 2:255–561.

Jonega MG. 1976. A study of teratological effects of intravenous,subcutaneous, and intragastric administration of delta 9-tetra-hydrocannabinol in mice. Teratology 36:151–162.

Jonega MG. 1977. Effects of delta 9-tetrahydrocannabinol on hamsterfetuses. J Toxicol Environ Health 2:1031–1040.

Kotwani A. 1998. Use of chick embryo in screening for teratogenicity.Indian J Physiol Pharmacol 42:189–204.

Levin M, Johnson RL, Stern CD, Kuehn M, Tabin C. 1995. A molecularpathway determining left-right asymmetry in chick embryogenesis.Cell 82:803–814.

Mantilla-Plata B, Clewe CL, Harbison RD. 1973. Teratogenic andmutagenic studies of -9-tetrahydrocannabinol in mice. Fed Proc32:746–756.

Mantilla-Plata B, Clewe CL, Harbison RD. 1975. 9-tetrahydrocannabinol-induced changes in prenatal growth and development of animals.Toxicol Appl Pharmacol 33:333–340.

Martin BR, Wiley JL, Beletskaya I, Sim-Selley, LJ, Smith FL, Dewey WL,Cottney J, Adams J, Baker J, Hill D., et al. 2006. Pharmacologicalcharacterization of novel water-soluble cannabinoids. PharmacolExp Ther 318:1230–1239.

Matsuda LA, Lolait SJ, Brownstein MJ, Young AC, Bonner TI. 1990.Structure of a cannabinoid receptor and functional expression of thecloned cDNA. Nature 346:561–564.

McGilveray IJ. 2005. Pharmacokinetics of cannabinoids. Pain Res Manag10:15A–22A.

Nakatsuji N, Johnson KE. 1984. Effects of ethanol on the primitive streakstage mouse embryo. Teratology 29:369–375.

New DAT. 1955. A new technique for the cultivation of the chick embryoin vitro. J Embryol Exp Morph 3:326–331.

NIDA, National Institute on Drug Abuse. 2001. National HouseholdSurvey on Drug Abuse Volume I: Summary of National Findings—Prevalence & Correlates of Alcohol, Tobacco, and Illegal Drug Use.Bethesda: NIDA. http://www.oas.samhsa.gov/nhsda/2k1nhsda/vol1/toc.htm.

Nishimura H, Tanimura T, Semba R, Uwabe C. 1974. Normal develop-ment of early human embryos: observation of 90 specimens atCarnegie stages 7 to 13. Teratology 10:1–5.

Noland JS, Singer LT, Short EJ, Minnes S, Arendt RE, Kirchner, HL, BearerC. 2005. Prenatal drug exposure and selective attention in pre-schoolers. Neurotoxicol Teratol 27:429–438.

O’Rahilly R, Muller F. 1987. Developmental Stages in Human Embryos,Carnegie Institution of Washington, Washington Publication 637.

Paria BC, Das SK, Dey SK. 1995. The pre-implantation mouse embryo is atarget for cannabinoid ligand-receptor signaling. Proc Natl Acad SciUSA 92:9460–9464.

Paria BC, Dey SK. 2000. Ligand-receptor signaling with endocannabi-noids in pre-implantation embryo development and implantation.Chem Phys Lipids 108:211–220.

Persaud TV, Ellington AC. 1967. Cannabis in early pregnancy. Lancet2:1306.

Persaud TV, Ellington AC. 1968a. Teratogenic activity of cannabis resin.Lancet 2:406–407.

Persaud TV, and Ellington AC. 1968b. The effects of cannabis sativa L.(Ganja) on developing rat embryos—preliminary observations. WestIndian Med J 4:232–234.



Piomelli D. 2004. THC: moderation during implantation. Nature Med10:19–20.

Psychoyos D, Stern CD. 1996a. Fates and migratory routes of primitivestreak cells in the chick embryo. Development 122:1523–1534.

Psychoyos D, Stern CD. 1996b. Restoration of the organizer after radicalablation of Hensen’s node and the anterior primitive streak in thechick embryo. Development 122:3263–3273.

Rex M, Orme A, Uwanogho D, Tointon K, Wigmore PM, Sharpe PT,Scotting PJ. 1997. Dynamic expression of chicken Sox2 and Sox3genes in ectoderm induced to form neural tissue. Dev Dyn 209:323–332.

Romanoff AL. 1960. The Avian Embryo: Structural and FunctionalDevelopment. New York: MacMillan.

Rosenkrantz H. 1978. Effects of cannabis on fetal development of rodents.Adv Biosc 22:479–499.

Rosenkrantz H. 1983. Cannabis, marijuana and cannabinoid toxicologicalmanifestations in man and animals. In Cannabis and Health Hazards(Fehr KO and Kalant H editors). Toronto: Addiction ResearchFoundation. p 91–175.

Schmidt M, Patterson M, Farrell E, Munsterberg A. 2004. Dynamicexpression of Lef/Tcf family members and beta-catenin duringchick gastrulation, neurulation, and early limb development. DevDyn 229:703–707.

Stern CD. 2002. Induction and initial patterning of the nervoussystem—the chick embryo enters the scene. Curr Opin Genet Dev12:447–451.

Straiker A, Stella N, Piomelli D, Mackie K, Karten HJ, Maguire G. 1999.Cannabinoid CB1 receptors and ligands in vertebrate retina:localization and function of an endogenous signaling system. ProcNatl Acad Sci USA 96:14565–14570.

Sulik K. 2008. Carnegie Stages – Scanning Electron Micrography, TheUniversity of New South Wales, Sydney, Australia. http://embryology.med.unsw.edu.au/wwwhuman/Stages/Stagesem.htm.

Uyeno ET. 1973. delta-9-tetrahydrocannabinol administered duringpregnancy of the rat. Proc. West Pharmacol Soc 16:614–667.

WHO, World Health Organization. 1997. Epidemiology of Cannabis Use.In: Programme on Substance Abuse. Cannabis: A Health Perspectiveand Research Agenda. Geneva: WHO. http://whqlibdoc.who.int/hq/1997/WHO_MSA_PSA_97.4.pdf.

Wilkinson D. 1992. In situ Hybridization: A Practical Approach. Oxford:IRL Press. p 75–83.

Wilson JG. 1978. Survey of in vitro systems: Their potential uses interatogenicity screening. In: Wilson JG, Fraser FC, editors. Handbookof Teratology. New York: Plenum Press. Vol 4, p 135–158.

Wright PL, Smith SH, Keplinger ML, Calandra LC, Braude MC. 1976.Reproductive and teratologic studies with D-9-tetrahydrocanna-binol and crude marihuana extract. Toxicol Appl Pharmacol 38:223–235.

