Article pubs.acs.org/crt
Transcriptomic Analysis of Thalidomide Challenged Chick Embryo Suggests Possible Link between Impaired Vasculogenesis and Defective Organogenesis Vimal Veeriah,†,¶ Pavitra Kumar,†,¶ Lakshmikirupa Sundaresan,†,‡ Zeenath Mafitha,‡ Ravi Gupta,§ Uttara Saran,† Jeganathan Manivannan,†,⊥ and Suvro Chatterjee*,†,‡ †
Vascular Biology Laboratory, AU-KBC Research Centre and ‡Department of Biotechnology, Anna University, Chennai, Tamil Nadu 600044, India § SciGenom Laboratories, Cochin, Kerala 682037, India S Supporting Information *
ABSTRACT: Since the conception of thalidomide as a teratogen, approximately 30 hypotheses have been put forward to explain the developmental toxicity of the molecule. However, no systems biology approach has been taken to understand the phenomena yet. The proposed work was aimed to explore the mechanism of thalidomide toxicity in developing chick embryo in the context of transcriptomics by using genome wide RNA sequencing data. In this study, we challenged the developing embryo at the stage of blood island formations (HH8), which is the most vulnerable stage for thalidomide-induced deformities. We observed that thalidomide affected the early vasculogenesis through interfering with the blood island formation extending the effect to organogenesis. The transcriptome analyses of the embryos collected on sixth day of incubation showed that liver, eye, and blood tissue associated genes were down regulated due to thalidomide treatment. The conserved gene coexpression module also indicated that the genes involved in lens development were heavily affected. Further, the Gene Ontology analysis explored that the pathways of eye development, retinol metabolism, and cartilage development were dampened, consistent with the observed deformities of various organs. The study concludes that thalidomide exerts its toxic teratogenic effects through interfering with early extra-embryonic vasculogenesis and ultimately gives an erroneous transcriptomic pattern to organogenesis. malformations have also been documented.7 The prevalence of ocular deficits in humans has been reported to be 25%, which majorly includes coloboma.8 The effects of thalidomide exposure on a developing embryo revolutionized drug toxicological testing.9 The anti-angiogenic property of thalidomide has been shown to be responsible for the adverse effects on development.10 Although few genome-based studies have demonstrated the effects of thalidomide on expression of genes involved in vascular development, limb and embryonic development,11−13 the exact mechanism of action of its teratogenicity at molecular level is not well understood. Only three protein targets of thalidomide namely soluble guanyl cyclase (sGC),14 cereblon,15 and tubulin16 have been reported so far. Its immunomodulatory properties are largely attributed to its modulation of TNF-α.2 Also, thalidomide has been demonstrated to perturb the BMP/Dkk1/Wnt signaling resulting in developmental defects including limb truncations and microphthalmia, which were counteracted upon Dkk1 or Gsk3 inhibition.17 Thalidomide has been shown to cause limb
1. INTRODUCTION Thalidomide was originally introduced into the market for the management of morning sickness in pregnant women. This resulted in the birth of children with various deformities ranging from amelia to microphthalmia leading to the withdrawal of the drug from the market.1 However, drug repositioning for the effective treatment of several pathological conditions resulted in significant attention and re-emergence of thalidomide.2,3 In 1998, FDA approved its therapeutic switching to the treatment of erythema nodosum leprosum4 followed by approval for multiple myeloma in 2006. In addition, thalidomide seems to be a promising candidate for the treatment of Crohn’s disease, bone marrow and blood cancers, inflammatory disorders affecting skin such as cutaneous lupus and scleroderma, HIV-related mouth, and throat ulcers.5,6 A comprehensive account on developmental toxicology of thalidomide based on the available literature has been published by Miller and Stromland that includes a wide range of deformities including limb, ocular anomalies and heart defects, kidney malformations, spine and chest structural abnormalities, central nervous system defects; genital and gastrointestinal © 2017 American Chemical Society
Received: July 19, 2017 Published: September 11, 2017 1883
DOI: 10.1021/acs.chemrestox.7b00199 Chem. Res. Toxicol. 2017, 30, 1883−1896
Article
Chemical Research in Toxicology
rate were observed at the dose of 60 μg, hence the optimum treatment dosage of thalidomide was selected as 60 μg for further experiments unless specified otherwise. Standardization of treatment time point was done by treating HH1−12-staged embryos with 60 μg of thalidomide, and it was found that HH8 as the most sensitive stage to thalidomide toxicity. Blood island parameters were studied between HH10 and HH15 since they differentiate in this time window. Organogenesis processes of eye, heart, and limb are complete on sixth day of incubation (HH29). Hence, work plan for these organs were made accordingly. Histology of liver was performed on 11th day of incubation since the size and shape of the hepatocytes stabilize during 11−12th day of development. 3.3. Morphological Analysis. Embryos treated with vehicle control or thalidomide at HH8 stage were dissected out on sixth day of incubation, and morphological deformities were tabulated. Images for all the deformities were taken using Olympus camera (Olympus, India) attached to a stereo microscope. The term retarded growth was used for the embryos having reduced body size and weight compared to embryos treated with vehicle control. 3.4. Alcian Blue Staining. Alcian blue was used to stain the cartilage in developing chick embryonic limbs to visualize the number of digits. Limbs were isolated from HH29 staged thalidomide or vehicle control treated chick embryos and stained with Alcian blue as described previously.23 In brief, the isolated limbs were fixed in Bouin’s solution for 2 h at room temperature and washed eight times within 24 h of incubation with 0.1% NH4OH in 70% ethanol. The samples were washed twice for 1 h each with 5% glacial acetic acid, then stained with staining solution (Alcian blue in 5% acetic acid) for 2 h. Limbs were washed again two times for 1 h first with 5% acetic acid and then 100% methanol. Samples were stored in benzoate/benzyl alcohol solution (2:1). All the images were taken with Olympus camera attached to stereo microscope. 3.5. Assessment of Blood Islands and Angiogenesis. Benzidine staining was used to visualize blood islands and primary plexus of blood vessels at HH10, HH12, and HH15 stages as described previously.26 Briefly, the embryonic membrane was removed from yolk sac and incubated in benzidine solution (3, 3′- dimethyl benzidine, 0.2% acetic acid, and 1% hydrogen peroxide) for 15 min in dark. After incubation, membranes were washed with 100% methanol, mounted, and the images were taken using Olympus camera attached to stereo microscope. Total numbers of blood island spots and primary plexus were counted manually. In another experiment, normal eggs (without prior treatment at HH8 stage) were opened at HH15+ stage and the yolk sac vasculature was treated with thalidomide or vehicle control using paper disc method and followed for angiogenesis up to 8 h by imaging at every 2 h under stereomicroscope (Figure 3F). Then blood vessel junctions were quantified using Angioquant software. 3.6. Hemoglobin Content at Blood Island Stage. Hemoglobin content at HH 10+ stage, in dissected extraembryonic membrane, was measured colorimetrically.27 Briefly, the yolk sac membrane was carefully dissected out and homogenized. Then the homogenized samples were centrifuged at 2500 rpm and 4 °C for 10 min. The supernatant was collected and measured at 540 nm in colorimeter. 3.7. Blood Vessel Density of Cephalic Region. Chick embryos were dissected at HH29 stage from eggs pre-treated with thalidomide or vehicle control. Then images of cephalic region of the embryos were taken using Olympus camera attached with a stereo microscope. Angiogenesis observed in the cephalic region was quantified using angioquant software. 3.8. Effect of Thalidomide on Angiogenesis - Aortic Ring Assay. Chick aorta was dissected from HH37 staged embryos, and arches were cut into length-wise similarly sized small rings. The aortic rings were washed carefully several times with 1X PBS. 12-well plate was coated with Matrigel, and each ring was placed in individual wells. Rings were treated with vehicle control or 40 μg/mL thalidomide and incubated at 37 °C for 36 h. Images were taken using an Olympus microscope at a total magnification of 200×. The number of tubes was counted manually.23
anomalies by modulating the PTEN/Akt pathway mediated apoptosis.18 In mammals, vasculogenesis occurs in parallel with hematopoiesis; it is the process of de novo formation of blood vessels by angioblasts and endothelial progenitors, which are morphologically characterized as blood islands of a developing embryo.19,20 Chicken is an established model for teretaogenic and angiogenic studies. A series of studies have documented thalidomide implications in embryo development, limb defects, ocular defects, and growth retardation similar to the humans by using chicken embryo development model.10 In chick embryo, vasculogenesis begins in the area vasculosa, which supplies the developing embryo with blood and nutrients. Several studies have demonstrated the significant role of blood islands in embryonic development where improper vasculogenesis leads to aberrant angiogenesis resulting in mid-gestation death or improper embryonic development.21,22 Our previous observations identified that blood islands fail to combine into a primary vascular plexus by thalidomide treatment, leading to subsequent effects.23 Recent findings also have shown that thalidomide exposure perturbs a general program of morpho-regulatory processes in the monkey embryo including down-regulation of pathways for vasculature after acute administration.10 We hypothesized that during embryonic development, thalidomide inhibited early vasculogenesis at blood island stage affecting the associated gene expression pattern ultimately resulting in embryopathy. Hence, we investigated the consequential effects of thalidomide treatment at the stage of blood island formation (HH8) on the transcriptome of embryos collected on the sixth day of incubation along with morphological observations.
2. MATERIALS Thalidomide was purchased from Sigma-Aldrich (India). S(-) confirmation of thalidomide attributes to teratogenic effects.24 However, thalidomide spontaneously shifts between the R and S conformations when it dissolves in biological fluid with pH 6. Therefore, racemic mixture of thalidomide was used in the present study. Dulbecco’s modified Eagle’s medium (DMEM) was purchased from PAN-Biotech. Fetal bovine serum and TriZol were from Invitrogen Life technologies. DMSO and 3,3′,5,5′-tetramethyl benzidinedihydrochloride (Benzidine) were procured from Amresco. Alcian blue 8GX was obtained from Himedia. All other chemicals were of reagent grade and obtained commercially.
3. METHODS 3.1. Chicken Embryos and Treatment Schedules. Brown leghorn (Gallus gallus) fertilized eggs were obtained from Poultry Research Station, Potheri, Chennai. They were incubated at 37 °C in a humidified incubator and staged according to Hamburger and Hamilton (HH) stages of chick development.25 All the experiments were performed on chick embryos between HH 1 and HH 38 stages. A small hole was created in the egg shell over the air sac using sterile needle, and all the treatments were administered in air sac as a single dose using a sterile tip. The eggs were then sealed with a sterile Mediplast tape and incubated further until the day of experiment. All experiments were performed with prior approval from the Institutional review committee. 3.2. Thalidomide Treatment and Optimum Concentration. Thalidomide (1 mg) was dissolved in 50 μL of DMSO, and then it was diluted to make the volume 1 mL with 1X PBS. Considering an average egg volume as 60 mL, we prepared thalidomide solution in such concentrations that upon injecting 20 μL of solution final exposing amounts of thalidomide were 10, 20, 30, 40, 50, 60, 70, 80, 90, and 100 μg. The 5% DMSO in 1X PBS was used as vehicle control. The maximum number of deformities and relatively higher survival 1884
DOI: 10.1021/acs.chemrestox.7b00199 Chem. Res. Toxicol. 2017, 30, 1883−1896
Article
Chemical Research in Toxicology
Figure 1. Early thalidomide treatment caused wide ranges of birth defects in chick embryos. (A) Schematic representation of drug administration and following experiments. (B) Chick embryos at HH stage 8 were treated with different concentrations of thalidomide and DMSO control. The percentage of abnormal embryos was calculated at HH 29 stage. n = 50; ∗p < 0.05 - 60 μg versus 70 μg and #p < 0.05 - 50 μg versus 60 μg of thalidomide embryos for each treatment. SE derived from three independent experiments. (C) Chick embryos at different HH stages were treated with 60 μg of thalidomide or DMSO and the percentage of abnormal embryos was calculated at HH29. n = 50; #p < 0.05 HH stage 7 versus 8 and ∗p < 0.05 HH stage 8 versus 9. 3.9. Heart Weight Measurement. Hearts from chick embryos of HH29 stage pre-treated with thalidomide or vehicle control were dissected and the wet weight was measured. 3.10. Histopathology of Eye and Liver. Fertilized eggs were treated with vehicle control or thalidomide as described previously. Eggs from each group were opened on sixth day and 11th day for isolating eye and liver, respectively. Eye and liver samples were fixed in 10% neutral buffered formalin for 24 h. Then 5 μm thick sections were cut and the formalin fixed tissue was embedded in paraffin after dehydration using different concentrations of ethanol and xylene. Then 5 μm thick sections were cut using microtome, and the samples were rehydrated after deparaffinization. Rehydrated samples containing slides were dipped into hematoxalin for 3−5 min, rinsed with distilled water, and kept under tap water for 5−7 min to develop the stain. Sections were destained by acid ethanol for 15 min and rinsed twice with tap water followed by deionized water. Samples were stained with eosin for 40 s and dehydrated using 3 × 5′ 95% ethanol, 3 × 5′ 100% ethanol, and 3 × 15′ xylene. Slides were mounted, dried overnight, and images were taken using microscope. 3.11. Eye Lens Transparency Assay. Fertilized eggs were treated with vehicle control or thalidomide at HH8 stage and incubated in a humidified incubator at 37 °C. On the 11th day of incubation, embryos were taken out from the eggshell and eyes were dissected out. Lenses were isolated from vitreous body, kept in PBS buffer, and washed three times. Then each lens was placed over a stripped transparent sheet and imaged using 4× objective using bright field microscope.28 The light intensity was calculated using ImageJ Software. 3.12. Transcriptome Sequencing and Analysis. Total RNA was isolated from chick embryo of stage HH29 (pre-treated with vehicle or thalidomide at 24 h) using TRIzol Reagent. Transcriptome sequencing was carried out using IlluminaHiSeq 2500 platform, and
the data were submitted to GEO (Accession No. GSE69159). TopHat (v2.0.8) with default parameters was used to align the sequence reads to the reference genome of chicken (Galgal4) downloaded from Ensembl Release 75 databases. Later, Cuffdiff (v2.2.0) program was used to identify the differentially expressed genes. In our previous study with this RNA sequencing data, we focused specifically on the effects of thalidomide on eye development and the role of nitric oxide in shielding the teratogenic effects of thalidomide.29 3.13. Validation of Sequencing Data by RT-PCR. To validate the RNA-sequencing data, six most modulated genes namely ASL1, KLF2, SOX14, GATA4, MPO, and SERPINA-1 were chosen, and RTPCR was performed to check their expression with β-actin as the internal control. The primer sequences of ASL1, KLF2, GATA4, and β-actin were obtained from published articles, and we designed the primers for SOX14, MPO, and SERPINA-1. The annealing temperature, primer sequences, and cycle number have been summarized in Suppl. Table 3. 3.14. Thalidomide Modulated Pathways and Functions. Huang et al. indicated that notable portions of up- or down-regulated genes are involved in certain interesting biological processes rather than being randomly spread throughout all possible biological processes.30 The differentially expressed genes of the current study served as input for Database for Annotation, Visualization and Integrated Discovery program (DAVID; http://david.abcc.ncifcrf.gov/ ). “Tissue expression” functions in DAVID program were used to predict the possible tissue/organ specific enriched expression of thalidomide targeted genes.30 The lower the P-value, the more significant is the correlation; we followed the recommended P-value cutoff of 0.05. Moreover, to elucidate the functional pathways of modulated genes obtained from the RNA sequencing data, we submitted the differentially expressed gene list to GeneCodis (genecodis.cnb.csic.es), an online modular enrichment tool that 1885
DOI: 10.1021/acs.chemrestox.7b00199 Chem. Res. Toxicol. 2017, 30, 1883−1896
Article
Chemical Research in Toxicology assesses combinations of annotations that are significantly enriched.31 For this analysis, the following annotations were selected: GO Biological Process (BP), GO Molecular Function (MF), and KEGG Pathways. Next we utilized STRING v10 (Search Tool for the Retrieval of Interacting Genes/Proteins; http://string.embl.de/) to construct a protein−protein interaction (PPI) network of up- and down-regulated genes. Here, the network is constructed based on the quantitatively integrated interaction data from four sources such as genomic context, high-throughput experiments, (conserved) coexpression, and previous knowledge.32 More specifically, we constructed a conserved coexpression network before Gene Ontology analysis, since it has long been established that coexpression or coregulation is a strong indicator of functional associations. The coexpression scores in STRING v10 are computed using a revised and improved pipeline, making use of all microarray gene expression experiments deposited in NCBI Gene Expression Omnibus.32 3.15. Statistical Analysis. All the experiments were performed in triplicates (n = 3) unless otherwise specified. Data have been presented as mean ± SEM. Data analysis was done using one-way ANOVA test, the Student-t test, and the Tukey post hoc test, as appropriate. Differences among mean were considered significant when P ≤ 0.05.
28 h) was the most sensitive stage for thalidomide teratogenicity with 54% of treated embryos displaying some type of abnormality (Figure 1A,B). Therefore, we treated HH8 staged chick embryo with different doses (10−100 μg) of thalidomide and we found that 60 μg was the most effective dose with respect to abnormal embryos (55%) and optimum survivability (Figure 1C). 4.2. Morphological Abnormalities under Thalidomide Treatment. Chick embryos treated with thalidomide at HH8 stage showed multiple abnormalities as listed in Table 1. In the group of 200 embryos, 69% embryos were found to have abnormalities, while 8% of embryos were dead under thalidomide treatment. The overall deformities included limb (21%), digit (7%), eye (16%), and others (8%); however, the limb deformities were found the major one with the 21% (Table 1). 4.3. Thalidomide Causes Limb Deformities during Embryonic Development. Thalidomide causes several limb deformities during embryonic development. Around 7% of thalidomide treated embryos showed amelia, a type of birth defect where one or more limbs are absent. Ten percent of thalidomide treated embryos showed meromelia, a condition in which a part of limb is absent (Figure 2A2,3). Cartilage staining using Alcian blue showed the irregularity in limb bone cartilage (primitive bone), which leads to limb deformities (Figure 2A4,5). Chicken has tetradactylous pattern of digitization in limbs (Figure 2B1), which is disturbed under thalidomide treatment, and there was decrease in the number of digits formed. Four percent of thalidomide treated embryos showed tridactyly, while 3% of embryos showed didactyly (Figure 2B2,3). 4.4. Other Abnormalities under Thalidomide Treatment. There were several other deformities observed in thalidomide treated embryos such as absence of one or both the eyes, anophthalmia in 4% embryos (Figure 2C1). Reduction in the size of one or both the eyes, that is, microphthalmia, was observed in 12% embryos (Figure 2C2,3). Retarded growth was observed in 17% of the thalidomide treated embryos (Figure 2D1,E), and 3% of embryos had a deformed beak (Figure 2D2). Five percent of embryos had omphalocele, a type of abdominal wall defect in which internal organs like intestine and liver remain outside of the abdomen in a sac because of a defect in the development of the muscles of the abdominal wall (Figure 2D3). Number of lethal embryos was increased from 3% to 8% from vehicle control to thalidomide treated group (Figure 2D4). 4.5. Thalidomide Impaired Vasculogenesis during Early Development. Blood islands start forming at HH8 stage. Primary plexus of blood vessels starts forming at HH 12 and blood vessels begin to remodel and mature from HH15 stage in yolk sac. Figure 3A, C, E, and G show the benzidine staining of extra-embryonic membranes of stages HH10+ (blood island spots), HH12+ (primary plexus), HH14+ (after remodelling of blood vessels). A significant reduction in the number of blood island spots (Figure 3B) and the percentage of primary plexus formation under thalidomide treatment were found (Figure 3D). A significant reduction in the number of junctions was observed under thalidomide treatment in ex-vivo angiogenesis model as well, which further indicates the antiangiogenic property of thalidomide (Figure 3F). Next, we quantified the vasculature in terms of vessel density at HH 19+ stage. Here, we used embryos treated with thalidomide or vehicle control at HH8 stage and yolk sac
4. RESULTS 4.1. Establishment of Model for Thalidomide Teratogenicity. There are several advantages to use chick embryo Table 1. Gross Deformities Observed in Chick Embryo under Thalidomide Treatment in Comparison with Vehicle Control Multiple targets of thalidomide in chick embryos control
p-value
thalidomide
normal embryos 94 ± 4.3% 23 ± 6% dead embryos 3 ± 1.5% 8 ± 2% abnormal embryos 3 ± 2.6% 69 ± 5.5% (a) limb deformities 0% 21% amelia 7% meromelia 10% phocomelia like 4% deformities (b) digit deformities 0% 7% tridactyly 4% didactyly 3% (c) eye deformities 0% 16% anophthalmia 4% mild-microphthalmia 8% acute-microphthalmia 4% (d) retarded growth 2% 17% embryos embryo length (cm) 3.8 ± 0.3 2.8 ± 0.4 (HH stage 32) embryo weight (g) 2.8 ± 0.5 2.3 ± 0.3 (HH stage 32) (e) other deformities 1% 8% omphalocele 1% 5% beak deformities 3% total number of embryos n = 200 for each treatment group
