Expression of β-galactosidase in mouse brain: utilization of a novel nonreplicative Sindbis virus vector as a neuronal gene delivery system

Gene Therapy (1997) 4, 815–822  1997 Stockton Press All rights reserved 0969-7128/97 $12.00

Expression of b-galactosidase in mouse brain: utilization of a novel nonreplicative Sindbis virus vector as a neuronal gene delivery system S Altman-Hamamdzic1, C Groseclose1, J-X Ma2, D Hamamdzic3, NS Vrindavanam 1, LD Middaugh1, NP Parratto1 and FR Sallee Departments of 1Psychiatry and Behavioral Medicine, 3Microbiology and Immunology and 2 Ophthalmology, Medical University of South Carolina, Charleston, SC, USA

Sindbis virus expression has been used for in vitro investigations of antigen processing, presentation and epitope mapping. The recent development of a replication-deficient recombinant Sindbis virus expression vector has made in vivo expression possible with minimal pathogenic risk. Advantages of Sindbis virus over other available viral systems include a comparatively smaller genome size making it possible to clone larger inserts, the ability to infect a wide range of host cell types with reduced pathogenicity for humans. These features suggest the possible utility of Sindbis virus for the in vivo delivery of genes to neural cells. We used the recombinant Sindbis viral expression system to target delivery of the lacZ gene to neuronal cells of mice by stereotactic surgery. Sindbis viral mRNA obtained by in vitro transcription was used to transfect baby hamster kidney (BHK) cells. As shown by histochemistry, b-galactosidase (b-gal) was expressed in approximately

50% of transfected cells. Cells were then cotransfected with DH-BB helper sequences that enabled the recombinant Sindbis virus RNA packaging. Nonreplicative Sindbis viral stock was collected 24 h after transfection. BHK cells were then infected with viral stock and histochemistry analysis was performed. Again, approximately 50% of the cells expressed b-gal. The same viral stock was infused into the nucleus caudatus/putamen and nucleus accumbens septi and histochemical analysis on frozen sections from the relevant brain areas confirmed that b-gal was expressed in neurons in a time-dependent manner. b-Gal was detected at 24 h after inoculation and was present for at least 14 days, with maximum expression at 48 h. These results suggest that a nonreplicative Sindbis virus expression system may be useful for delivery of foreign genes into the central nervous system (CNS).

Keywords: Sindbis virus; lacZ; gene delivery; neuronal cells

Introduction The potential of gene therapy for brain dysfunction is complicated by factors which include the blood–brain barrier, the post-mitotic status of neural cells, the pathogenic risk of viral homologous recombination, and the stability of foreign gene expression. Viral vectors commonly used for therapeutic DNA delivery include replication-deficient forms of adenovirus (AD), retrovirus, and herpes simplex virus (HSV).1–7 This report describes the expression of b-galactosidase (b-gal) in targeted brain areas using a replication-defective Sindbis virus expression system. The Sindbis virus expression system differs from currently available viral delivery systems in that it is an RNA virus, known to generate high levels of protein expression in vitro.8 The potential advantages of this system include a high infection efficiency attributed to a nonreceptor-mediated cellular internalization and the ability to infect post-mitotic neural cells. The Sindbis Correspondence: FR Sallee, Department of Psychiatry and Behavioral Medicine, Youth Division, 67 President Street, Charleston, SC 29403, USA J-X Ma and D Hamamdzic contributed equally to this work Received 29 October 1996; accepted 24 March 1997

virus is less pathogenic8 to humans and infections usually manifest themselves as headaches and rashes. 9 In contrast, recent reports associate encephalitis with other viral systems.10 Expression of foreign gene products in vivo has already been successfully demonstrated using a recombinant Sindbis viral vector to sensitize cytotoxic T lymphocytes to heterologously expressed MHC class I-restricted antigens.11 The present system utilizes a recombinant Sindbis vector, pSinRep5, which is approximately 10 kb in size and contains the Sindbis virus nonstructural protein genes 1– 4 (nsP1–4), the promoter for subgenomic RNA and a multiple cloning site (MCS). Recombinant RNA is synthesized in vitro using this vector into which the appropriate gene has been cloned.12–14 A helper for the defective virus (DH-BB template) contains the genes for the four structural proteins required for Sindbis virus packaging. It is also transcribed with the resultant RNA cotransfected into BHK cells to produce Sindbis virus vector particles containing the recombinant RNA. To investigate this expression system for in vivo delivery to the brain, we have generated nonreplicative Sindbis virus expressing the lacZ (b-gal) gene and stereotactically placed it into the nucleus caudatus/putamen and nucleus accumbens septi. These areas are a part of

Sindbis virus as a tool for in vivo gene delivery S Altman Hamamdzic et al

816

Figure 1 (a) Rep/lacZ is a 13104 bp linearized DNA template containing the lacZ gene. The construct was linearized at the XhoI site and contains only nonstructural genes. (b) DH-BB (defective helper) is a 6729 bp linearized DNA template and contains the genes for the four structural proteins required for packaging of the Sindbis viral genome. This template does not contain a packaging signal, so it will not form a defective interfering particle.

the dopaminergic mesolimbic pathway and would provide for behavioral and motor monitoring in future studies. In this report we aim to demonstrate that the nonreplicative Sindbis virus expression system can be used to express b-gal in brain tissue.

Results Generation of viral stocks To obtain viral stocks, in vitro transcription reactions were performed on linearized SinRep/lacZ and DH-BB helper constructs (Invitrogen, San Diego, CA, USA). Figure 1a represents SinRep/lacZ linearized DNA template containing the lacZ gene. This template contains only nonstructural genes. The DH-BB helper virus contains the structural genes (Figure 1b) necessary for viral packaging. Figure 2 represents polyA RNA generated by in vitro transcription using SP6 polymerase and labeled with a32P-CTP. From Figure 2 it can be concluded that on agarose gel electrophoresis, the RNA transcribed from the Sindbis virus construct containing the lacZ sequence (31 335 bp) migrates slower than one without insert. BHK cells were transfected (5 mg of SinRep/lacZ mRNA and

5 mg of DH-BB mRNA) as described in Materials and methods. Nontransfected control cells were negative for the production of b-gal (Figure 3a) and transfected cells, incubated 24 h before performing b-gal histochemistry, demonstrated that the transfection was successful. BHK cells produced active b-gal 24 h following transfection (Figure 3b). Medium from transfected cells was collected as viral stock and stored in aliquots at −70°C until used for delivery to brain tissue.

Infection of BHK cells and expression of b-galactosidase BHK cells were seeded on six-well plates at a density of 3 × 105 cells per well and infected with 250 ml of the viral stock. b-Galactosidase histochemistry was performed 24 h following infection. Approximately 50% of infected cells expressed b-gal (Figure 4b), while noninfected cells failed to show b-gal expression (Figure 4a). Western blot analysis using a monoclonal antibody against b-gal (Gibco/BRL, Gaithersburg, MD, USA) confirmed the presence of b-gal in whole cell lysate (117 kDa) (Figure 4c). Cells which were not infected with virus were used as a control. Delivery of virus to mouse brain and investigation of b-galactosidase expression Mice were infused with 2.5 ml of either viral stock or sterile vehicle as described in the Materials and methods. Observation of the animals during the 14-day period before death revealed no differences between the two groups with regard to body weight or normal daily activities. Brain tissue of control mice showed no X-gal staining (Figure 5a), in contrast to that of mice infused with virus stocks (Figure 5b–f). The expression of b-gal was detected 1 day after infusion and was sustained for up to 14 days (Figure 5). Densitometric analysis performed on X-gal-stained brain tissue over this time course (Figure 6), indicates that maximum expression of b-gal occurred 2 days after infusion and then slowly declined over the time course. Fourteen days after infection, b-gal was still detectable, although at lower levels (Figure 6).

Discussion

Figure 2 In vitro transcription: after in vitro transcription reaction in the presence of the tracer (32P-CTP) mRNA samples were run on 2% agarose gel. This gel was dried at room temperature using a vacuum dryer and exposed to radiographic film for 2 min. 1, SinRep5; 2, SinRep/lacZ; 3, DH-BB helper; 4, control vector (pTRI-XefI).

For gene therapy of the CNS, efficient and minimally pathogenic vehicles are required. Although naked DNA or liposomal delivery systems are comparatively safe, they are less efficient than viral systems.51 Considerations when using a viral system for in vivo genetic expression include the safety of the recombinant virus, the host cell range, the overall level of expression, the cell cycle characteristics and the technical ease of use. Available DNA viral vector systems like AD, adeno-associated viruses (AAV) and HSV systems have serious potentially pathogenic consequences.10 Even their nonreplicative forms have a high probability of reversion to potentially virulent replicative viruses by homologous recombination. Among RNA-based systems in use today, the retroviral expression system is problematic with the possibility of germline modification due to its ability to be incorporated into the host genome. Utility of the retroviral delivery system in the CNS is limited, as retroviruses cannot infect postmitotic cells. The use of alpha viruses as a delivery/expression system in the CNS has

Sindbis virus as a tool for in vivo gene delivery S Altman Hamamdzic et al

817

Figure 3 Expression of b-galactosidase in BHK cells 24 h after transfection with SinRep/lacZ. (a) Control cells transfected with SinRep5; (b) cells transfected with SinRep/lacZ.

Sindbis virus as a tool for in vivo gene delivery S Altman Hamamdzic et al

818

Figure 4 Expression of b-galactosidase in BHK cells 24 h after infection with SinRep/lacZ. (a) Control cells (infected with pSinRep5); (b) BHK cells infected with SinRep/lacZ. (c) Western blot analysis: 24 h after infection cell lysate was prepared and 25 ml of total lysate was run on 4–15% trisglycine gel; lane 1, marker; lane 2, noninfected cell lysates; lane 3, cells infected with pSindRep5; lane 4, cells infected with SinRep/lacZ; lane 5, control (0.5 mg/ml b-galactosidase; Gibco).

Sindbis virus as a tool for in vivo gene delivery S Altman Hamamdzic et al

Materials and methods Cell culture BHK cells were maintained in minimal essential medium (aMEM), GIBCO BRL) supplemented with l-glutamine (GIBCO/BRL) 1 × penicillin, streptomycin, fungizone (PSF) antibiotic mixture (GIBCO/BRL) and 5% fetal bovine serum (FBS) (JRH Biosciences, Lenexa, KS, USA). Cells were grown in 175 cm2 flasks (Baxter, McGraw Park, IL, USA) until the monolayer was 80–90% confluent and then harvested using trypsin-EDTA (GIBCO/BRL) and transferred into 35-mm (six-well) plates, at a density of 2 × 105 per well. Cells were incubated overnight at 37°C in a humidified atmosphere with 5% CO2. Figure 4 (Continued).

not been studied extensively but seems to have potential as an efficient and minimally pathogenic system.8,16 This system also has the ability to infect a wide range of cells including nonmammalian cells and has been found to express significant quantities of heterologous proteins in vitro,17–19 and in vivo.8 Sindbis virus is an infectious virus whose life cycle ends in cytolysis in its normal hosts. Genetic engineering of this virus has currently provided us with a nonreplicative, recombinant, mature and infectious viral form capable of expressing any gene of interest.20 The Sindbis virus expression system has been used for the efficient expression of heterologous proteins in quantities sufficient for structure/function analysis. It has a moderately small genome (approximately 10 kb) making the in vitro manipulation of the vector and cellular transfection easier, strong viral promoters that drive expression of both the viral and foreign genes under investigation,8 and it infects a wide range of cell types. In the present study, recombinant Sindbis virus as a vector, efficiently delivered the lacZ gene into mouse brain. No significant morbidity was observed in comparison to sham-infected controls as feeding or drinking behavior was not impaired. To examine the efficiency of transfection and infection by the viral stocks, histochemistry was performed in the isolated cell culture system. Results indicate that the efficiency of transfection was sufficient to produce high numbers of viral particles (Figures 3 and 4). Histochemistry performed on frozen brain sections of mice infused with the viral stock indicated that the expression of b-gal was colocalized to neurons identified by Nissl body staining. Delivery of the viral stock to the brain tissue was carefully conducted and monitored to allow the reproducible administration of virus to a localized area. If surgical or viral damage had occurred, the effects would have been identified in comparison to the contralateral nontreated brain area analyzed and in the post-injection activity of the animals. These results suggest that this expression system may be of particular interest for gene delivery to the CNS and could be instrumental in the development of animal models of neurologic diseases. The time frame of expression (1–14 days) of a foreign gene in brain tissue is sufficient to allow investigations of the effect of the expressed genes on behavioral patterns. The Sindbis virus expression system is a candidate for further studies of the potential therapeutic applications.

In vitro transcription Plasmid vectors pSinRep5, SinRep/lacZ and DH-BB virus helper were obtained from Invitrogen Sindbis Expression System (Invitrogen). Template DNA, pSinRep5 was amplified using Qiagen’s Plasmid Maxi Kit (Qiagen, Chatsworth, CA, USA) in a final concentration of 1 mg/ml and linearized using XhoI restriction endonuclease (Boehringer Mannheim, Indianapolis, IN, USA), SinRep/lacZ and DH-BB helper virus were used as provided. In vitro transcription reactions were performed by using MEGAscript Sp6 kit (Ambion, Austin, TX, USA) as per the manufacturer’s instructions. The total amount of DNA per reaction was 2 mg/ml. Transcription reactions were incubated for 4 h at 37°C in the presence of a32PCTP (Amersham, Arlington Heights, IL, USA) and 1 ml of the mRNA obtained was run on a 1% agarose gel. The gel was dried and exposed to radiographic film for 2 min. Transfection of BHK cells and X-gal staining Transfection was done by electroporation (Electroporator II; Invitrogen). 1 × 107 Cells in a volume of 0.5 ml were transfected with 5 mg of SinRep/lacZ and DH-BB helper virus RNA respectively at a ratio of 1:1, mixed with 9.5 ml of complete aMEM after electroporation and 10 ml of cell suspension was plated in six-well plates (2 ml per well). Twenty-four hours after transfection, the supernatant containing the Sindbis virus vector particles was harvested and this virus stock was frozen in liquid nitrogen and stored at −70°C. To check the transfection efficiency BHK cells were stained uzing X-gal (Boehringer Mannheim). To confirm b-gal expression, cells were stained 24 h after transfection.17 Virus titer One hundred microliters of the viral stock was mixed with 400 ml of aMEM containing 1% FCS and 2 × 104 cells per tissue culture treated 1-chamber slide (Nunc, Naperville, IL, USA) were infected. Cells were then incubated for 8 h at 37°C under 5% CO2 and additional 1.5 ml of aMEM containing 1% FCS was added. Cells were incubated under the same conditions for an additional 16 h and then stained with X-gal. The same experiment was done by using 100 ml of the viral stock per chamber slide and 2 h after incubation, the virus was removed and fresh complete medium was added to the cells. The cells were assayed after 8, 16, 24 and 48 h, respectively. X-gal at a concentration of 20 mg/ml was used for the assay. There was a gradual increase in the number of cells expressing b-gal up to 24 h and then it was constant. Infectivity was

819

Sindbis virus as a tool for in vivo gene delivery S Altman Hamamdzic et al

820

Figure 5 Time course of the expression of b-galactosidase in mouse brain (a) Uninfected control; (b) 24 h after infusion; (c) 48 h after infusion; (d) 7 days after infusion; (e) 10 days after infusion and (f) 14 days after infusion.

calculated by comparing the number of cells expressing b-gal per field (1 cm2 ) to the total number of cells present and extrapolating it to derive an infectious particle count of approximately 1 × 104 particles per milliliter of the viral stock. Cell count was done using a CK2 Olympus microscope (Lake Success, NY, USA) under × 40 magnification.

Infection of BHK cells with Sindbis virus vector particles 3 × 105 Cells were washed twice with 1 × PBS and infected with 250 ml of viral stock for 1 h at room temperature. After the infection reaction, 2 ml of aMEM + 1% FBS per plate was added and the cells were incubated at 37°C for 24 h. To determine the presence of b-gal, cells

were stained using the X-gal staining method following the manufacturer’s protocol (Invitrogen).

Western blot analysis Twenty-four hours following infection, BHK cells were rinsed twice with cold 1 × PBS, removed from the dish by scraping in 1 × PBS, pelleted by centrifugation and lysed in 200 ml of lysis buffer (50 mm HEPES, 1% NP-40, 1 mg/ml aprotinin and 100 mg/ml PMSF) (equivalent of 3 × 105 /200 ml). Twenty microliters of cell lysate was run on a pre-made 4–15% Tris-glycine denaturing acrylamide gel (Bio-Rad, Hercules, CA, USA). b-Gal was detected by Coomassie blue staining and by Western blot analysis. The proteins were transferred to transblot 0.45 mm nitro-

Sindbis virus as a tool for in vivo gene delivery S Altman Hamamdzic et al

Figure 6 Densitometric analysis: images were analyzed by the NIH image program and results expressed in arbitrary units of optical density (OD). (a) Uninfected control; (b) 24 h after infusion; (c) 48 h after infusion; (d) 7 days after infusion; (e) 10 days after infusion and (f) 14 days after infusion.

cellulose membranes (Bio-Rad) using a semi-dry transfer system (milliblot-SDE transfer system; Millipore, Bedford, MA, USA) for 1 h at 100 mA. As a positive control 1.0 mg of purified b-gal (GIBCO/BRL) was run on a polyacrylamide gel.

Animal manipulations and surgical procedures C57BL/6 experimentally naive mice (26–32 g) obtained from the Jackson Laboratories (Bar Harbor, ME, USA) were used. The animals were housed in single cages (one mouse per cage). The colony room was maintained at 22°C and was on a 12 h light/dark cycle. Water and food were available ad libitum. The animals were approximately 12 weeks old at the time of surgery and were randomly assigned to groups on the basis of post-operative brain extraction intervals. Intracranial surgery began 10 min following intraperitoneal injections of an anesthetic cocktail (ketamine hydrochloride: 100 mg/kg body weight, Sigma, St Louis, MO, USA; xylazine: 3.0 mg/kg body weight). A small rodent stereotaxic apparatus (Kopf model 900 small animal stereotax) modified for mice (Kopf model 921 mouse adaptor/921-F&G Ear Bar Assembly, David Kopf, Tugunga, CA, USA) was used to guide cannula placement. The mouse was placed on an elevated platform (4.1 cm) and its head secured to the stereotax via the tooth bar and ear bar assembly. Its eyes were coated with neosporine and the skull area was shaved and swabbed with betadine. An incision was made along the midline skull area and the skull was exposed deviating the skin laterally with hemostats and scraping the periosteum. The head was secured on a horizontal plane via a series of skull surface measurements at equidistant points bilateral to the bregma. Rostral–caudal leveling was accomplished by measurements on the mid-line sagittal suture caudally where it intersected with the frontal suture. After leveling the skull, a No. 8 dental burr was used to drill a hole in the skull at the coordinates to allow

Figure 7 Brain sections stained for nuclei and Nissl bodies in Darrow red (× 40).

821

Sindbis virus as a tool for in vivo gene delivery S Altman Hamamdzic et al

822

the insertion of a canulae (10 ml syringe needle) to access the nucleus caudatus/putamen and nucleus accumbens septi. The coordinates were adjusted from those obtained from the atlas of Slotnick and Leonard:21 0.4 mm anterior to the bregma; 2.1 mm lateral to the midline sagittal suture; and 3.5 mm ventral to the skull surface. Following surgery, the skull was filled with bone wax. The skull was washed with saline, swabbed with neosporin, and the wound closed with surgical staples. Mice were injected with 2.5 ml of the viral stock at a rate of 1.25 ml per min. The brains were extracted postoperatively at day 1, 2, 7, 10 and day 14 after injection. Six control mice were infused with a sterile PBS vehicle (1.25 ml per min) and two untreated mouse brains were used as negative control.

Tissue preparation and histology The mice brains were frozen over liquid nitrogen and stored at −70°C until use. The tissue was mounted on to cryostat chucks using Histo-Prep embedding matrix (Fisher, Fair Lawn, NJ, USA) and coronal sections were cut to 5–7 mm on a cryostat (JUNG Frogcut 2800N; Leica Instruments, Nussloch, Germany) at −18°C. Sections were thaw-mounted on to cold gelatin/chrome alum-coated slides and dried for 5 min at 37°C. Between four and six sections were mounted on each slide and stored at −70°C, or they were analyzed for b-gal expression. Sections were fixed with 0.25% glutaraldehyde (Sigma, St Louis, MO, USA) in 1 × PBS at room temperature for 30 min, washed with 1 × PBS twice, and incubated overnight at 37°C in X-gal staining solution (5 mm K-ferrocyanide; 5 mm Kferricyanide; 2 mm MgCl2; 1 mm spermidine; 0.02% NP40; 0.01% Na deoxycholate and 20 mg/ml of X-gal in 1 × PBS). Sections were then washed twice with 1 × PBS and rinsed with distilled water. The sections were then airdried at room temperature overnight, covered with coverslides using Cytoseal 60 mounting medium-low viscosity (Stephens, Riverdale, NJ, USA) and stored at room temperature. For the detection of Nissl bodies, sections first stained for X-gal were incubated for 30 min in Darrow red stain (50 mg of Darrow red in 200 ml 0.2 m glacial acetic acid; pH 2.7) at room temperature (Figure 7). The sections were then washed in distilled water, differentiated and dehydrated through 50, 70 and 95% alcohol, fixed in 4% formaldehyde for 5 min at room temperature, washed with 1 × PBS and air-dried. Microscopy was performed using a Nikon Optiphot-2 microscope (Nikon, Melville, NY, USA).

Acknowledgements The authors would like to thank Dr Phillipe Arnaud for reviewing the manuscript and helpful suggestions and Dr Steven D London for helpful suggestions. This research was supported by grant DA06881.

References 1 Akli S et al. Transfer of a foreign gene into the brain using adenovirus vectors. Nat Genet 1993; 3: 224–228.

2 Bajocchi G, Feldman SH, Crystal RG, Mastrangeli A. Direct in vivo gene transfer to ependymal cells in the central nervous system using recombinant adenovirus vectors. Nat Genet 1993; 3: 229–234. 3 Culver WC et al. In vivo gene transfer with retroviral vector– producer cells for treatment of experimental brain tumors. Science 1992; 256: 1550–1552. 4 Fink JD et al. In vivo expression of b-galactosidase in hippocampal neurons by HSV-mediated gene transfer. Hum Gene Ther 1992; 3: 11–19. 5 Wolfe JH, Deshmane SL, Frasier NW. Herpes virus vector gene transfer and expression of b-glucuronidase in the central nervous system of MPS VII mice. Nat Genet 1992; 1: 379–384. 6 Tenenbaum L, Darling JL, Hooghe-Peters E. Adeno-associated virus (AAV) as a vector for gene transfer into glial cells of the human central nervous system. Gene Therapy 1994; 1 (Suppl. 1): S80. 7 Austin CP, Cepko CL. Retrovirus mediated gene transduction into the vertebrate CNS. Gene Therapy 1994; 1 (Suppl. 1): S6–S9. 8 Piper RC et al. Recombinant Sindbis virus as an expression system for cell biology. Meth Cell Biol 1994; 43: 55–78. 9 Griffin DE. Alphavirus pathogenesis and immunity. In: Schlesinger S, Schlesinger MJ (ed). The Togaviridae and Flaviviridae. Plenum Press: New York, 1990, pp 209–251. 10 Wood MJ et al. Inflammatory effects of gene transfer into the CNS with defective HSV-1 vectors. Gene Therapy 1994; 1: 283– 291. 11 Hahn CS, Hahn YS, Braciale TJ, Rice CM. Infectious Sindbis virus transient expression vectors for studying antigen processing and presentation. Proc Natl Acad Sci USA 1992; 89: 2679–2683. 12 Bredenbeek PJ. Sindbis virus expression vectors: packaging of RNA replicons by using defective helper RNAs. J Virol 1993; 67: 6439–6446. 13 Johanning FW. A Sindbis virus mRNA polynucleotide vector achieves prolonged and high level heterologous gene expression in vivo. Nucleic Acids Res 1995; 23: 1495–1501. 14 Raju R, Subramaniam SV, Hajjou M. Genesis of Sindbis virus by in vivo recombination of nonreplicative RNA precursors. J Virol 1995; 69: 7391–7401. 15 Imaoka T et al. Preliminary results of gene transfer to central nervous system by continuous injection of DNA–liposome complex. Cell Transplant 1995; 4: 23–26. 16 Dubensky TW Jr et al. Sindbis virus DNA-based expression vectors: utility for in vitro and in vivo gene transfer. J Virol 1996; 70: 508–519. 17 Rice CM, Levis R, Strauss JH, Huang HV. Production of infectious RNA transcripts from Sindbis virus clones: mapping of lethal mutations, rescue of a temperature-sensitive marker and in vitro mutagenesis to generate defined mutants. J Virol 1987; 61: 3809–3819. 18 Levis R, Huang H, Schlesinger S. Engineered defective interfering RNAs of Sindbis virus express bacterial chloramphenicol acetyltransferase in avian cells. Proc Natl Acad Sci USA 1987; 84: 4811–4815. 19 Levis R, Schlesinger S, Huang HV. Promoter for Sindbis virus RNA-dependent RNA transcription. J Virol 1990; 64: 1726–1733. 20 Geigenmuller-Gnirke U et al. Complementation between Sindbis viral RNAs produces infectious particles with a bipartite genome. Proc Natl Acad Sci USA 1991; 88: 3253–3257. 21 Slotnick BM, Leonard CM. A Stereotaxic Atlas of the Albino Mouse Forebrain. US Dept of Health, Education and Welfare: Rockville, MD, 1975.