Efficient gene transduction of neurons by lentivirus with ... - Nature

Gene Therapy (2007) 14, 872–882 & 2007 Nature Publishing Group All rights reserved 0969-7128/07 $30.00 www.nature.com/gt

ORIGINAL ARTICLE

Efficient gene transduction of neurons by lentivirus with enhanced neuron-specific promoters H Hioki1, H Kameda1, H Nakamura1, T Okunomiya1, K Ohira1, K Nakamura1,2, M Kuroda1, T Furuta1 and T Kaneko1,2 1

Department of Morphological Brain Science, Graduate School of Medicine, Kyoto University, Kyoto, Japan and 2Core Research for Evolutional Science and Technology (CREST), Japan Science and Technology Agency, Kawaguchi, Japan

In the field of basic and clinical neurosciences, it is important to develop a method for easy delivery and persistent expression of transgene in central neurons. We firstly generated lentiviral vectors with five kinds of neuron-specific promoters, such as synapsin I (SYN), calcium/calmodulindependent protein kinase II, tubulin alpha I, neuron-specific enolase and platelet-derived growth factor beta chain promoters and then novel hybrid promoters by fusing cytomegalovirus enhancer (E) to those neuron-specific promoters. Neuron-specific expression of green fluorescent protein (GFP) with those promoters was examined in vivo by injecting the lentiviral vectors into the rat neostriatum, thalamus and neocortex. Among all the promoters, SYN

promoter displayed the highest specificity for neuronal expression in all the regions examined (more than 96%). Although GFP production by the hybrid promoters was about 2–4 times larger than the non-enhanced promoters, the neuronal specificity was significantly decreased in most cases. However, the neuronal specificity of E/SYN hybrid promoter exhibited the least decrease only in the thalamus. Furthermore, the transcriptional activity and neuronal specificity of E/SYN promoter were sustained for up to 8 weeks. Thus, lentivirus with E/SYN promoter is the best vector for strong persistent expression in neurons. Gene Therapy (2007) 14, 872–882. doi:10.1038/sj.gt.3302924; published online 15 March 2007

Keywords: lentivirus; neuron specific; cytomegalovirus; enhancer; promoter; hybrid

Introduction Gene delivery to the central nervous system with recombinant viral vectors have aroused interest in the field of basic and clinical neurosciences, and many kinds of viral vectors have been developed for investigation of neuronal morphology, gene function, modeling of human diseases and therapeutic applications.1,2 Among available viral vectors, lentiviral vectors offer unique advantages of stably integrating transgene into the genome of dividing and non-dividing cells, and providing the basis for sustained gene expression without any toxicity and immune response. Lentiviral vectors are, therefore, considered to be the most appropriate ones for experiments requiring long-term gene expression and gene therapeutic applications. Self-inactivating and replication-defective lentiviral vectors, derived from human immunodeficiency virus type 1 (HIV-1), have been developed for transduction of mammalian cells.3–6 These vectors are pseudotyped with vesicular stomatitis virus G-protein (VSV-G), and VSV-G achieves a broad transduction spectrum.7 Many gene delivery vectors have used viral promoters, particularly Correspondence: Dr T Kaneko, Department of Morphological Brain Science, Graduate School of Medicine, Kyoto University, Kyoto 6068501, Japan. E-mail: [email protected] Received 24 October 2006; revised 22 December 2006; accepted 25 December 2006; published online 15 March 2007

cytomegalovirus (CMV) or rous sarcoma virus promoter, because they have high transcription activities in all the infected cells whatever their cell types are. Taken together, VSV-G-pseudotyped lentiviral vectors with CMV promoter express transgene in not only neuronal but also glial cells in the central nervous system3–5,8 Thus, it is indispensable to utilize neuron-specific promoters for stable neuron-specific expression of transgene with VSV-G-pseudotyped lentiviral vectors. Many kinds of neuron-specific promoters have been developed and applied to transgenic animals and viral vectors.9 As lentiviral vectors cannot accept large genetic sequences (up to 10 kb between the long terminal repeats), we selected relatively small promoters. In the present study, we used five kinds of neuron-specific promoters; human synapsin I (SYN),10 mouse calcium/ calmodulin-dependent protein kinase II (CaMKII),11 rat tubulin alpha I (Ta1),12 rat neuron-specific enolase (NSE)13 and human platelet-derived growth factor-beta chain (PDGF) promoters.14 Although these promoters are known to be specific for neuronal expression, the transcriptional activities are much weaker than those of viral promoters such as CMV promoter. In the present study, we generated novel hybrid promoters by a combination of CMV enhancer and the neuron-specific promoters listed above to enhance promoter activities, and then characterized their neuronal specificities and transcriptional activities in the rat neostriatum, thalamic nuclei and cerebral cortex, using enhanced green fluorescent protein (GFP) as a reporter protein. The purpose of

Lentivirus with enhanced neuron-specific promoters H Hioki et al

873

the present study is to explore the best promoter for strong, persistent and specific expression of transgene in neurons with VSV-G-pseudotyped lentiviral vectors.

Results GFP expression by neuron-specific and enhanced neuron-specific promoters in the rat neostriatum, thalamus and neocortex We generated 11 kinds of VSV-G-pseudotyped lentiviral vectors, which express GFP under CMV, neuron-specific and enhanced neuron-specific promoters (Figure 1, Table 1). At 1 week after injection of the lentiviral vectors into the rat neostriatum, ventrobasal thalamic nuclei and somatosensory cortex, GFP was visualized by immunoperoxidase staining (diaminobenzidine-4HCl (DAB) reaction). GFP expression driven by CMV promoter was observed in not only neuronal but also glial cells (arrowheads and insets in Figure 2a1–a3). The immunoreactivity was very intense and detected in cell bodies as well as distal dendrites (Figure 2a1–a3). Under SYN promoter, GFP expression was found only in neurons, but the immunoreactivity was conspicuously weak, especially in the thalamus (Figure 2b1–b3). Using E/SYN promoter, intense immunoreactivity was observed predominantly in neurons, and distal dendrites were also well visualized (Figure 2c1–c3). These results indicate that the neuronal specificities of SYN and E/SYN promoters are higher than that of CMV promoter, and the expression levels of CMV and E/SYN promoters are stronger than that of SYN promoter. In the other neuronspecific promoters, the immunoreactivity for GFP also became more intense but neuronal specificity was

Figure 1 Construction of pLenti6-Pro-GFP-WPRE. We firstly introduced 11 kinds of promoters (CMV, neuron-specific and enhanced neuron-specific promoters), GFP and WPRE into pBSII SK (a), and then KpnI-to-NotI fragment was inserted into the KpnI/ NotI sites of pENTRIA (b). The insert (Pro-GFP-WPRE) was transferred pLenti6 with LR recombination reaction, and resulting to pLenti6-Pro-GFP-WPRE (c). CMV-E, the enhancer region of human cytomegalovirus promoter; GFP, enhanced green fluorescence protein; LTR, long-terminal repeat; f, HIV-1 packaging signal; Pro, promoter, RPE, HIV-1, Rev response element; WPRE, woodchuck hepatitis virus post-transcriptional element. See text for further detail.

Table 1 Neuron-specific promoters used in the present study Promoter

CMV enhancer

+

Human synapsin I (1889–2289: M55301, GenBank) (F: 50 -CTGCAGAGGGCCCTGCGTAT-30 , R: 50 -CGCCGCAGCGCAGATGGTCG-30 )

SYN (401 bp)

E/SYN (769 bp)

Mouse calcium/calmodulin-dependent protein kinase II (7625–7988: AJ222796) (F: 50 -ACTTGTGGACTAAGTTTGTT-30 , R: 50 -GCTGCCCCCAGAACTAGGGG-30 )

CaMKII (364 bp)

E/CaMKII (732 bp)

Rat tubulin alpha I (Gloster et al.12) (F: 50 -CCGTATTAGAAGGGATGGCT-30 , R: 50 -TTCTTACAGCGCGACTCTTA-30 )

Ta1 (1034 bp)

E/Ta1 (1402 bp)

Rat neuron-specific enolase (Forss-Petter et al.13) (F: 50 -GAGCTCCTCCTCTGCTCGCC-30 , R: 50 -CTCGAGGACTGCAGACTCAG-30 )

NSE (1807 bp)

E/NSE (2175 bp)

Human platelet-derived growth factor-beta chain (Sasahara et al.14) (F: 50 -GGATCCACAGTCTCCTGAGT-30 , R: 50 -CTAGGGAGGCAGCGGGGGAG-30 )

PDGF (1468 bp)

E/PDGF (1836 bp)

Abbreviations: CaMKII, calcium/calmodulin-dependent protein kinase II; CMV, cytomegalovirus; NSE, neuron-specific enolase; PDGF, platelet-derived growth factor; SYN, synapsin; TaI, tubulin alpha I. We amplified CMV promoter and five kinds of neuron-specific promoters by polymerase chain reaction with the primer sets indicated above, and inserted them into plasmid vectors. We further developed novel hybrid promoters by fusing CMV enhancer to those neuron-specific promoters. See text for further details. Gene Therapy


874

Figure 2 Immunoperoxidase staining for GFP at 1 week after injection of lentiviral vectors with CMV, SYN or E/SYN promoter into the rat neostriatum, ventrobasal thalamic nuclei and somatosensory cortex. GFP expression under CMV promoter was observed in both neuronal and glial cells (a1–a3), whereas the expression by SYN or E/SYN was neuron-specific at all the brain regions (b1–b3, c1–c3). GFP immunoreactivity was strong under CMV or E/SYN promoter (a1–a3, c1–c3), but weak under SYN promoter (b1–b3). Arrowheads indicate glial cells, and they were shown in insets (a1–a3). Scale bar in inset (a3) ¼ 40 mm.

decreased to some extent after addition of CMV enhancer. Transduction efficiencies of neuron-specific and enhanced neuron-specific promoters appeared somewhat lower than that of CMV promoter, this is probably because most of glial cells did not express GFP under those promoters even though transgene was integrated into their genomes. In addition, we observed no remarkable differences of GFP-expressing cell numbers between neuron-specific and enhanced neuron-specific promoters. In the following experiments, we quantitatively analyzed neuronal specificities and expression strength of all promoters listed in Table 1.

Neuronal specificity At 1 week after injection of lentiviral vectors with all promoters listed in Table 1, we randomly selected 100–200 GFP-expressing cells around the injection sites in the rat neostriatum, ventrobasal thalamic nuclei and somatosensory cortex, and then examined whether or not these cells might show immunoreactivity for NeuN (Figure 3; in the neostriatum). We compared neuronal specificities between promoters with or without CMV enhancer by using the two-tailed Student’s t-test (Figure 4a). After addition of CMV enhancer, the neuronal Gene Therapy

specificities of CaMKII, Ta1 and NSE promoters exhibited a statistically significant decrease at all the brain regions examined, whereas that of PDGF promoter did not display any significant change. SYN promoter showed the highest specificity for neuronal expression among all the promoters (more than 96%), whereas E/SYN promoter exhibited the least decrease in neuronal specificity only in the thalamic nuclei (P ¼ 0.042). However, the specificity of E/SYN promoter remained relatively high at all the brain regions compared with the other enhanced neuron-specific promoters. We further examined time course of neuronal specificities from 3 days to 8 weeks after injection of lentiviral vectors with CMV, SYN or E/SYN promoter into the neostriatum (Figure 4b). The neuronal specificities of these promoters showed no remarkable changes and seemed constant throughout 8 weeks in the neostriatum; the specificities of SYN and E/SYN promoters stayed around 98%, whereas that of CMV promoter remained about 50%.

Time course of GFP expression levels To determine time course of GFP expression levels, we injected lentiviral vectors with SYN or E/SYN promoter into the rat neostriatum, allowed the animals to survive


875

Figure 3 Confocal laser-scanning microscopic images of double-immunofluorescence staining for GFP and NeuN in the neostriatum. At 1 week after injection of lentiviral vectors with CMV, SYN or E/SYN promoter, GFP and NeuN were labeled with AlexaFluor 488 and AlexaFluor 647, respectively. The, digital images were captured at about 200-mm far from injection sites. GFP was expressed in both neuronal and glial cells under CMV promoter (a–a’’), whereas the expression by SYN or E/SYN promoter was neuron specific (b–b’’, c–c’’). Arrowheads point to the colocalization of GFP and NeuN.

from 3 days to 8 weeks, and observed GFP-immunofluorescence (GFP-IF) or GFP-native-fluorescence (GFPNF) under confocal laser-scanning microscope. It should be noted that GFP-IF and GFP-NF were observed mainly in somata, whereas the fluorescence signals in dendrites were obviously weak (Figure 5a, Figure 6b, b0 and d, d0 ), as compared with immunoperoxidase staining (Figure 2b1–c3). We measured average intensity per pixel of GFP-IF or GFP-NF in soma (IIF or INF (U)) with software ImageJ, and then estimated GFP-IF or GFP-NF in somatic volume (IIFV or INFV (106 U mm3)), presuming the neuronal cell bodies to be spherical (Figure 5a; somatic volume V ¼ 4A3/2/3p1/2 (mm3), where A is sectional area (mm2)). We found that most GFP-IF was actually accumulated in somata, and that IIFV reflected about 80% of GFP produced by infected neurons (see Supplementary Figure 1 online). The curves of I and IV were, when normalized, similar to each other in both GFP-IF and GFP-NF (Figure 5b,c and d,e), probably because the vast majority of neurons in the neostriatum is composed of one principal neuronal cell type (medium-sized spiny neuron; X95%) and the somatic volume of cell body was rather homogeneous in the present study (7.271.4 (103 mm3), mean7s.d., n ¼ 270).

The increase of IIF and IIFV had considerably slowed along the time course and seemed saturated by 8 weeks (Figure 5b and c). On the other hand, INF and INFV were almost linearly increased from 1 to 8 weeks after the injection (Figure 5d and e). GFP-NF at 3 days after the injection was so weak that we could not detect the fluorescence signals. To investigate the discrepancy between GFP-IF and GFP-NF curves (Figure 5b,d and c,e), we examined the effect of immunofluorescence staining for GFP. At 1 and 2 weeks after injection of lentiviral vectors with SYN or E/SYN promoter into the neostriatum, INF was measured from 15 neurons for each promoter. After immunofluorescence staining for GFP, IIF was obtained from the identical neurons (Figure 6b, b0 and d, d0 ). A correlation between INF and IIF was examined with linear or logistic regression analysis. At 1 week after the injection, INF was linearly amplified by immunofluorescence staining (Figure 6a), but not at 2 weeks (Figure 6c). This result indicates that when the amount of GFP surpassed certain range, immunofluorescence staining did not linearly amplify the fluorescence signals anymore. Thus, we determined to compare expression strength of all promoters listed in Table 1 at 1 week after the injection. Gene Therapy


876

brain regions examined (two-tailed Student’s t-test, n ¼ 15, 7.9 1016ppp5.0 103). We further compared IIF between CMV and all the hybrid promoters by one-way analysis of variance (ANOVA) followed by Dunnett’s post hoc test. E/Ta1 promoter displayed higher intensity than CMV promoter in the neostriatum, whereas E/CaMKII and E/PDGF promoters exhibited lower intensities than CMV promoter in the thalamus and neocortex. No significant differences were detected between CMV and the other enhanced promoters in the other regions examined. IIF was generally weaker in the thalamus than in the neostriatum and neocortex. However, this does not mean that the promoter activity was weak in the thalamus, because IIF did not reflect the total amount of GFP in single neurons especially when compared with small and large neurons, such as somatic volumes of neostriatal and thalamic neurons (7.771.6 (103 mm3) in the neostriatum and 2175.3 (103 mm3) in the thalamus, mean7s.d., n ¼ 165). Thus, to estimate promoter activity in neurons of different sizes, we calculated IIFV from average IIF and somatic volume (V). IIFV produced by all the hybrid promoters was 1.6–3.8 times higher than non-enhanced promoters at all the brain regions examined (Figure 7b, two-tailed Student’s t-test, n ¼ 15, 4.0 1011 pPp1.6 104). Multiple comparison tests between CMV and all the hybrid promoters were also performed by one-way ANOVA followed by Dennett’s post hoc test. E/Ta1 promoter was stronger than CMV promoter in the neostriatum, whereas E/CaMKII promoter in the thalamus and E/PDGF promoter in the thalamus and neocortex were weaker than CMV promoter. No significant differences were detected between CMV and the other enhanced promoters in the other regions examined. IIFV in the neocortex was larger than those in the neostriatum and thalamus, suggesting that the transcriptional activity was generally stronger in the neocortex.

Discussion Figure 4 Neuronal specificities in the rat neostriatum, ventrobasal thalamic nuclei and somatosensory cortex. (a) We randomly selected 100–200 GFP-expressing cells, and examined whether or not they might show immunoreactivity for NeuN at 1 week after viral injection. Each column represents the mean7s.d. (n ¼ 3) and P-values were calculated by using the two-tailed Student’s t-test. (b) Time course of neuronal specificities of CMV, SYN or E/SYN promoter in the neostriatum. We randomly selected 100–200 GFPexpressing cells, and examined whether or not they might show immunoreactivity for NeuN from 3 days to 8 weeks after injection of lentiviral vectors. Each symbol represents the mean7s.d. (n ¼ 3).

Comparisons of GFP expression levels by neuronspecific and enhanced neuron-specific promoters At 1 week after injection of lentiviral vectors with all promoters listed in Table 1, IIF and IIFV were measured in the infected neurons of the rat neostriatum, ventrobasal thalamic nuclei and layer II/III of somatosensory cortex (Figure 7a and b), as described before (Figure 5a). We first compared IIF between neuron-specific and enhanced neuron-specific promoters (Figure 7a). IIF was increased by 1.4–3.6-fold in all the enhanced promoters at all the Gene Therapy

We developed novel hybrid promoters by fusing CMV enhancer to neuron-specific promoters, and quantitatively examined their characteristics in vivo with VSV-Gpseudotyped lentiviral vectors derived from HIV-1. After addition of CMV enhancer to neuron-specific promoters, the neuronal specificities of most promoters were significantly decreased, although the expression levels were increased by about two- to four-fold in all the hybrid promoters. Among all the hybrid promoters, the neuronal specificity of E/SYN promoter remained comparatively high, and the expression levels of E/SYN promoter was almost the same as that of CMV promoter at all the brain regions examined. These results lead to the conclusion that E/SYN promoter might be the most appropriate one to express a gene of interest in neurons efficiently and specifically.

Technical consideration: quantification of promoter activity Transcriptional activity of promoter has been analyzed using several reporter enzymes; chloramphenicol acetyltransferase, luciferase, alkaline phosphatase and b-galactosidase.15 Recently, it was demonstrated that GFP also had all the essential properties of a quantitative


877

Figure 5 Time course of GFP expression levels by CMV, SYN or E/SYN promoter in the rat neostriatum. We observed GFP-IF or GFP-NF by confocal laser-scanning microscope from 3 days to 8 weeks after injection of lentiviral vectors. Average intensity per pixel of GFP-IF or GFPNF in soma (IIF or INF (U)) and sectional area (A (mm2)) of GFP-expressing neurons were measured by using software ImageJ (a; SYN, GFP-IF, 4 weeks). Presuming the neuronal cell bodies to be spherical, we estimated somatic volume (V (mm3)) of GFP-expressing neurons and GFP-IF or GFP-NF in somatic volume (IIFV or INFV (106 U mm3)) (a). IIF and IIFV had become saturated by 8 weeks (b, c), whereas INF and INFV increased linearly along the time course up to 8 weeks after viral injection (d, e). Each symbol represents the mean7s.d. (n ¼ 15).

reporter protein and concluded that GFP is reliable and useful one for analysis of promoter activity by measuring fluorescence intensity.16 As GFP has the advantage that promoter activity can be easily monitored in single cells in sections, we selected GFP as a reporter protein in the present study. To analyze promoter activity precisely, total amount of GFP should be wholly calculated in single cells. Morphology of neuronal cells, however, is so complicated that it is impractical to completely estimate total amount of GFP in single cells in sections. In the present study, we estimated transcriptional activities by calculating IIFV from average immunofluorescence intensity per pixel (IIF) and somatic volume (V). Although this estimation did not include GFP in dendrites and axons, IIF was obviously weak in those processes (Figure 5a, Figure 6b, b0 and d, d0 ). Thus, the underestimation of GFP amount was considered to be not so high in the present

study. Actually, GFP-IF was not detected in axonal structures, and most GFP-IF was localized in somata. Furthermore, GFP-IF in somatic volume reflected about 80% of GFP produced by neurons, suggesting that the total GFP produced by neurons might be predicted by calculating IV (see Supplementary Figure 1 online). Under SYN and E/SYN promoters, INF and INFV were linearly increased from 1 to 8 weeks, whereas IIF and IIFV became saturated by 8 weeks after viral injection into the neostriatum (Figure 5). To investigate this discrepancy, we compared INF and IIF from identical neurons, and analyzed their correlation at 1 or 2 weeks after viral injection into the neostriatum (Figure 6). When INF was about within 60 (U) in the present study, GFP-NF was linearly amplified by immunofluorescence staining but not at over 60 (U). This indicates that immunofluorescence staining linearly reflects the amount of antigen within at most 1 week in the present study. Gene Therapy


878

Figure 6 A correlation between average intensities per pixel of GFP-NF and GFP-IF in soma (INF and IIF). At 1 and 2 weeks after injection of lentiviral vectors with SYN or E/SYN promoter into the rat neostriatum, INF was measured from 15 neurons for each promoter, as in the legend of Figure 5a. GFP was subsequently immunolabeled with AlexaFluor 488, and IIF was obtained from the identical neurons (b,b0 and d,d0 ). A correlation between INF and IIF at 1 or 2 weeks after viral injection was analyzed by linear or logistic regression, respectively (a, c).

Figure 7 Average intensity per pixel of GFP-IF in soma (IIF) and GFP-IF in somatic volume (IIFV) in the rat neostriatum, ventrobasal thalamic nuclei and somatosensory cortex. One week after viral injection, IIF (a) and IIFV (b) were calculated, as in the legend of Figure 5a. After addition of CMV enhancer, IIF and IIFV displayed a statistically significant increase in all the neuron-specific promoters at all the brain regions examined (two-tailed Student’s t-test; *, 4.0 1011 pPp5.0 103). We further performed multiple comparison tests between CMV and all the hybrid promoters by one-way ANOVA followed by Dennett’s post hoc test (wPo 0.05; wwPo 0.01).

The effect of CMV enhancer After addition of CMV enhancer to neuron-specific promoters, their characteristics were markedly changed; neuronal specificities were decreased and transcriptional Gene Therapy

activities were drastically enhanced. Although, it has not yet been answered how the enhancer altered characteristics of these neuron-specific promoters, a considerable number of studies have been made on the functional


mechanisms of CMV enhancer over the past few decades. Several distinct sets of cellular transcription factors bind to and stimulate the enhancer, so that it functions as a strong transcriptional enhancer.17 Thus, it can be assumed that transcriptional activities were augmented but the neuronal specificities were loosened in the hybrid promoters owing to such a functional feature of CMV enhancer. The potential of CMV enhancer for hybrid approach has been demonstrated by several reports. CMV enhancer/chicken beta-actin promoter is a very strong one and has been widely applied to transgenic animals and viral vectors.18 A hybrid promoter by fusing CMV enhancer to PDGF promoter was also reported, and this hybrid promoter significantly augmented transgene expression in neurons in vitro and in vivo, using plasmid vectors, recombinant adeno-associated virus vectors and baculovirus-derived vectors.19–21 Moreover, the transcriptional activity of glial fibrillary acidic protein promoter was strikingly increased after addition of CMV enhancer, whereas the astrocytic specificity of the promoter remained high.22 These reports and the present results suggest that a hybrid approach might be effective in improvement on transcriptional activities of cellular promoters.

Time course of GFP expression We observed time course of transcriptional activities and neuronal specificities of SYN and E/SYN promoters from 3 days to 8 weeks after viral injection into the neostriatum. Surprisingly, GFP-NF linearly increased even up to 8 weeks in both the promoters, indicating that the transcriptional activities of these promoters can be maintained at least during 8 weeks. Furthermore, the neuronal specificities of both the promoters displayed no remarkable changes throughout 8 weeks, suggesting that neuronal specificity is independent of time course and amount of GFP products for at least 8 weeks. It is noteworthy that although the transcriptional activity of E/SYN promoter was almost as strong as that of CMV promoter, the transcriptional activity and neuronal specificity remained very high for up to 8 weeks. As persistent transcriptional activity and neuronal specificity are crucial issues for in vivo modeling and gene therapy of human neurodegenerative diseases, a combination of lentiviral vectors and E/SYN promoter is suitable for strong persistent gene expression in central neurons. Prospects of VSV-G-pseudotyped lentiviral vectors with E/SYN promoter Lentiviral vectors are now considered to be a most promising tool for experiments requiring long-term gene expression and gene therapeutic applications in the central nervous system23–25 by the following reasons: lentiviral vectors (1) integrate transgene into the genome of dividing and non-dividing cells; (2) enable us to express transgene permanently in the infected cells without immune response; (3) can be directly delivered to specific brain region by local injection; and (4) can be applied to mammalian species other than rats at anytime during the lifespan of the animals. Actually, lentiviral vectors have provided new strategies for in vivo modeling and treatments of human neurodegenerative diseases such as Parkinson’s disease (PD).26,27

In the modeling of PD, for example, mutated alphasynuclein was overexpressed in the rat substantia nigra by local injection of VSV-G-pseudotyped lentiviral vector with phosphoglycerate kinase (PGK) promoter.28 After the viral injection, most of nigral dopaminergic neurons were selectively abolished and the immunoreactivity for tyrosine hydroxylase (TH) was significantly decreased, indicating that an adult onset model of PD can be generated by direct stereotaxic injection of viral vectors expressing a gene linked to PD. However, as PGK promoter is one of the ubiquitous promoters, transgene is generally expressed in both neuronal and glial cells. Thus, it cannot be denied that the genetic model of PD was affected by overexpression of mutated alphasynuclein in glial cells. As alpha-synuclein is produced only in neurons, the expression should be restricted in neurons for establishing more sophisticated animal models and analyzing pathogenetic mechanisms more precisely. It was further reported that VSV-G-pseudotyped lentivirus can achieve functional improvement in a rat model of PD.29 In the experiment, nigral dopaminergic neurons were damaged by administration of 6-hydroxydopamine, and then TH, aromatic amino acid dopa decarboxylase and GTP cyclohydrolase 1 were delivered into the dopamine-denervated striatum, using lentivirus with CMV promoter. After the viral injection, sustained transgene expression, dopamine production and functional improvement were observed, suggesting that this method has the potential for gene therapy of late-stage PD patients. Although VSV-G-pseudotyped lentiviral vectors have been also developed for gene therapy experiments of the other neurodegenerative diseases, ubiquitous promoters were employed to express transgene in almost all the cases. For avoidance of unexpected side effects caused by ectopic gene expression, it is desirable to use tissue-specific promoters. In the present study, we demonstrated that E/SYN hybrid promoter is highly neuron-specific and has strong transcriptional activity. Thus, E/SYN promoter may replace ubiquitous promoters for gene transductions of neurons, and VSV-G-pseudotyped lentivirus with E/SYN promoter may accelerate the experiments for in vivo modeling and treatments of human neurodegenerative diseases.

879

Concluding remarks In summary, E/SYN promoter was proved to be an excellent one for gene expression in neuronal tissues by the following reasons; (1) the neuronal specificity was relatively high compared with the other hybrid promoters, and kept around 98% in the neostriatum during 8 weeks; (2) the transcriptional activity was almost as strong as that of CMV promoter, and kept for at least 8 weeks; and (3) the length of E/SYN promoter is short (769 bp), which enable us to apply the E/SYN promoter to many kinds of viral vectors. We expect that this novel E/SYN promoter will be a useful tool in the field of basic and clinical neurosciences.

Materials and methods The experiments were conducted in accordance with the Committees for Animal Care and Use of the Graduate Gene Therapy


880

School of Medicine at Kyoto University, and for Recombinant DNA Study in Kyoto University. All efforts were made to minimize animal suffering and the number of animals used.

Plasmids construction The lentiviral vector was derived from HIV-1 (Invitrogen, Carlsbad, CA, USA), and constructed as follows. We used human CMV (nucleotides 1–589 of gb: U57609, GenBank) promoter and five kinds of neuron-specific promoters for human SYN, mouse CaMKII, rat Ta1 (a gift from Dr Miller), rat NSE (a gift from Dr Forss-Petter and Dr Sutcliffe) and human PDGF (a gift from Dr Sasahara and Dr Collins) promoters (Table 1). These six promoters, enhanced GFP (Clontech, Palo Alto, CA, USA) and woodchuck hepatitis virus posttranscriptional regulatory element (WPRE; nucleotides 1093–1684 of gb: U57609; a gift from Dr Hope30) were amplified by polymerase chain reaction (PCR), and inserted into HincII, EcoRV and SmaI sites of pBluescript II SK (+) (pBSII SK; Stratagene, La Jolla, CA, USA), respectively (Figure 1a). To generate novel hybrid promoters by a combination of CMV enhancer and neuron-specific promoters (Table 1), the enhancer region of CMV promoter was firstly amplified by PCR with the following primers: 50 -TAGTTATTAA TAGTAATCAA-30 and 50 -TTTTCACGTGCCATGGTAA TAGCGATGACT-30 (PmaCI site underlined), and subcloned into HincII site of pBSII SK. Then, five kinds of neuron-specific promoters (Table 1), GFP and WPRE were inserted into the PmaCI, EcoRV and SmaI sites, respectively (Figure 1a). These totally eleven constructs were confirmed by sequencing and named as pBSII-SKPro-GFP-WPRE. To generate a Gateway entry vector, a KpnI-to-NotI fragment from pBSII-SK-Pro-GFP-WPRE was inserted into the KpnI/NotI sites of pENTR1A (Invitrogen, Figure 1b). Then, the insert from the entry vector (Pro-GFP-WPRE) was transferred to the destination vector pLenti6/BLOCK-iT-DEST by homologous recombination with LR clonase (Invitrogen), resulting in pLenti6-Pro-GFP-WPRE (Figure 1c). Production and concentration of VSV-G-pseudotyped lentivirus The production of VSV-G-pseudotyped lentivirus was performed according to the manufacturer’s instructions, ViraPower Lentiviral Expression System (Invitrogen). The destination plasmid pLenti6-Pro-GFP-WPRE was cotransfected with the mixture of the packaging plasmids (pLP1, pLP2 and pLP/VSVG; Invitrogen) into the 293FT producer cell line (Invitrogen), using Lipofectamine 2000 (Invitrogen). The medium was replaced at 8 h after transfection with Dulbecco’s modified Eagle’s medium containing 10% fetal bovine serum. After 60 h from transfection, the viral particles in the culture supernatant were collected, filtered through 0.45-mm filters (Millipore, Corning, NY, USA) following low speed centrifugation (3000 g, 15 min), and then concentrated with Centricon Plus-20 (Millipore). Viral titers (transducing units/ml) were determined by transduction of 293FT cells with serial dilutions of the viral solution and colony counting after blasticidin selection, and adjusted to 1.0 106 TU/ml. The virus solution was stored in aliquots at 801C until use for delivery to brain tissues. This viral vector was replication deficient and Gene Therapy

had the least chance for production of parent viral particles in the infected cells.

Injection of viruses, fixation and immunoperoxidase staining Forty-one adult male Wistar rats (250–300 g; Japan SLC, Shizuoka, Japan) were deeply anesthetized with chloral hydrate (35 mg/100 g body weight). The virus (1.0 ml of 1.0 106 TU/ml) was stereotaxically injected by pressure through a glass micropipette attached to Picospritzer III (General Valve Corporation, East Hanover, NJ, USA) into the rat neostriatum, ventrobasal thalamic nuclei and somatosensory cortex. The rats were allowed to survive from 3 days to 8 weeks after the injection. The rats were deeply anesthetized again with chloral hydrate (70 mg/100 g body weight), and perfused transcardially with 200 ml of 5 mM phosphate-buffered 0.9% (w/v) saline (PBS; pH 7.4). The rats were further perfused for 30 min with 200 ml of 3% (w/v) formaldehyde, 75%-saturated picric acid and 0.1 M Na2HPO4 (pH 7.0; adjusted with NaOH). The brains were removed, cut into several blocks and post-fixed with the same fixative above for 8 h at 41C. After cryoprotection with 30% (w/w) sucrose in PBS, the blocks were cut into 30-mm-thick sections on a freezing microtome. Some sections were mounted onto gelatinized glass slides without immunostaining, and coverslipped with 50% (v/v) glycerol and 2.5% (w/v) triethylenediamine (antifading reagent) in PBS. Some sections were incubated overnight with 0.2 mg/ ml affinity-purified rabbit antibody to GFP,31 and then for 1 h with 10 mg/ml biotinylated anti-(rabbit immunoglobulin (Ig)G) donkey antibody (Chemicon, Temecula, CA, USA). The incubation was carried out at room temperature in PBS containing 0.3% (v/v) Triton X-100, 0.25% (w/v) l-carrageenan and 1% (v/v) donkey serum (PBS-XCD) and followed by a rinse with PBS containing 0.3% (v/v) Triton X-100 (PBS-X). The sections were further incubated for 1 h with avidin-biotinylated peroxidase complex (ABC-Elite; Vector Laboratories, Burlingame, CA, USA) in PBS-X. After a rinse with PBS-X, the sections were reacted for 20–40 min with 0.02% (w/v) DAB and 0.001% (v/v) H2O2 in 50 mM Tris-HCl (pH 7.6), mounted onto gelatinized glass slides, dehydrated in ethanol series, cleared in xylene and coverslipped. Photographs were taken by the digital camera QICAM (QIMAGING, Burnaby, BC, Canada), modified (720% contrast enhancement) in software Canvas 8 (ACD Systems, Saanichton, BC, Canada) and saved as 8-bit TIFF files. Double immunofluorescence labeling The brain sections, which were obtained as described above, were incubated overnight in PBS-XCD with a mixture of 1 mg/ml anti-GFP rabbit antibody31 and 1 mg/ ml anti-NeuN mouse antibody (Chemicon). After a rinse with PBS-X, the sections were incubated for 2 h with 5 mg/ml AlexaFluor 488-conjugated anti-(rabbit IgG) goat antibody and 5 mg/ml AlexaFluor 647-conjugated anti-(mouse IgG) goat antibody (Molecular Probes, Eugene, OR, USA). The sections were mounted onto gelatinized glass slides and coverslipped with 50% (v/v) glycerol and 2.5% (w/v) triethylenediamine (antifading reagent) in PBS. Digital pseudocolor images were


captured by confocal laser-scanning microscope LSM 5 Pascal (Carl Zeiss, Oberkochen, Germany) with optical slice thickness (Pinhole corresponding to 1 airy unit), using a 40 objective lens (Plan-NEOFLUAR, NA ¼ 0.75, Carl Zeiss). AlexaFluor 488 and 647 were excited with 488- and 633-nm laser beams and observed through 510–530- and X650-nm emission filters, respectively. The images were modified (720% contrast enhancement) in software Canvas 8 and saved as 8-bit TIFF files.

Measuring fluorescence intensity GFP-IF or GFP-NF was observed under confocal laserscanning microscope LSM 5 Pascal as described above. The digital images were captured in the rat neostriatum, ventrobasal thalamic nuclei and layer II/III of somatosensory cortex and saved as 12-bit TIFF files (without contrast enhancement). We measured average intensity of florescence per pixel (I (U)), sectional area (A (mm2)) with software ImageJ (http://rsb.info.nih.gov/ij). To keep the condition for taking digital images constant, we adjusted laser power, gain and offset each time by monitoring intensity of fluorescence beads (#F-14791, Molecular Probes). Statistics Two-tailed Student’s t-test, one-way ANOVA followed by Dennett’s post hoc test, and linear or logistic regression analysis were performed by using software Excel (Microsoft Corporation, Redmond, WA, USA), Prism (Graphpad Software Inc., San Diego, CA, USA) and DeltaGraph (RockWare Inc., Golden, CO, USA), respectively.

Acknowledgements This research was supported by Grant-in-Aid for Scientific Research 18700341, 16200025, 17650100 and Grant-in-Aid for Scientific Research on Priority Areas System study on higher-order brain functions 17022020 from The Ministry of Education, Culture, Sports, Science and Technology of Japan (MEXT).

References 1 Davidson BL, Breakefield XO. Viral vectors for gene delivery to the nervous system. Nat Rev Neurosci 2003; 4: 353–364. 2 Callaway EM. A molecular and genetic arsenal for systems neuroscience. Trends Neurosci 2005; 28: 196–201. 3 Naldini L, Blomer U, Gallay P, Ory D, Mulligan R, Gage FH et al. In vivo gene delivery and stable transduction of nondividing cells by a lentiviral vector. Science 1996; 272: 263–267. 4 Miyoshi H, Blomer U, Takahashi M, Gage FH, Verma IM. Development of a self-inactivating lentivirus vector. J Virol 1998; 72: 8150–8157. 5 Zufferey R, Nagy D, Mandel RJ, Naldini L, Trono D. Multiply attenuated lentiviral vector achieves efficient gene delivery in vivo. Nat Biotechnol 1997; 15: 871–875. 6 Amendola M, Venneri MA, Biffi A, Vigna E, Naldini L. Coordinate dual-gene transgenesis by lentiviral vectors carrying synthetic bidirectional promoters. Nat Biotechnol 2005; 23: 108–116. 7 Cronin J, Zhang XY, Reiser J. Altering the tropism of lentiviral vectors through pseudotyping. Curr Gene Ther 2005; 5: 387–398.

881 8 Dittgen T, Nimmerjahn A, Komai S, Licznerski P, Waters J, Margrie TW et al. Lentivirus-based genetic manipulations of cortical neurons and their optical and electrophysiological monitoring in vivo. Proc Natl Acad Sci USA 2004; 101: 18206–18211. 9 Wells T, Carter DA. Genetic engineering of neural function in transgenic rodents: towards a comprehensive strategy? J Neurosci Methods 2001; 108: 111–130. 10 Li L, Suzuki T, Mori N, Greengard P. Identification of a functional silencer element involved in neuron-specific expression of the synapsin I gene. Proc Natl Acad Sci USA 1993; 90: 1460–1464. 11 Mayford M, Baranes D, Podsypanina K, Kandel ER. The 30 -untranslated region of CaMKII alpha is a cis-acting signal for the localization and translation of mRNA in dendrites. Proc Natl Acad Sci USA 1996; 93: 13250–13255. 12 Gloster A, Wu W, Speelman A, Weiss S, Causing C, Pozniak C et al. The T alpha 1 alpha-tubulin promoter specifies gene expression as a function of neuronal growth and regeneration in transgenic mice. J Neurosci 1994; 14: 7319–7330. 13 Forss-Petter S, Danielson PE, Catsicas S, Battenberg E, Price J, Nerenberg M et al. Transgenic mice expressing beta-galactosidase in mature neurons under neuron-specific enolase promoter control. Neuron 1990; 5: 187–197. 14 Sasahara M, Fries JW, Raines EW, Gown AM, Westrum LE, Frosch MP et al. PDGF B-chain in neurons of the central nervous system, posterior pituitary, and in a transgenic model. Cell 1991; 64: 217–227. 15 Alam J, Cook JL. Reporter genes: application to the study of mammalian gene transcription. Anal Biochem 1990; 188: 245–254. 16 Soboleski MR, Oaks J, Halford WP. Green fluorescent protein is a quantitative reporter of gene expression in individual eukaryotic cells. FASEB J 2005; 19: 440–442. 17 Mocarski ES. Cytomegaloviruses and their replication. In: Fields BN, Knipe DM, Howley PM (eds). Fields Virology. LippincottRaven: Philadelphia, PA, 1996, pp 2447–2492. 18 Niwa H, Yamamura K, Miyazaki J. Efficient selection for highexpression transfectants with a novel eukaryotic vector. Gene 1991; 108: 193–199. 19 Wang CY, Guo HY, Lim TM, Ng YK, Neo HP, Hwang PY et al. Improved neuronal transgene expression from an AAV-2 vector with a hybrid CMV enhancer/PDGF-beta promoter. J Gene Med 2005; 7: 945–955. 20 Liu BH, Wang X, Ma YX, Wang S. CMV enhancer/human PDGFbeta promoter for neuron-specific transgene expression. Gene Therapy 2004; 11: 52–60. 21 Li Y, Yang Y, Wang S. Neuronal gene transfer by baculovirusderived vectors accommodating a neurone-specific promoter. Exp Physiol 2005; 90: 39–44. 22 Wang CY, Wang S. Astrocytic expression of transgene in the rat brain mediated by baculovirus vectors containing an astrocytespecific promoter. Gene Therapy 2006; 13: 1447–1456. 23 Wong LF, Goodhead L, Prat C, Mitrophanous KA, Kingsman SM, Mazarakis ND. Lentivirus-mediated gene transfer to the central nervous system: therapeutic and research applications. Hum Gene Ther 2006; 17: 1–9. 24 Kay MA, Glorioso JC, Naldini L. Viral vectors for gene therapy: the art of turning infectious agents into vehicles of therapeutics. Nat Med 2001; 7: 33–40. 25 Azzouz M, Kingsman SM, Mazarakis ND. Lentiviral vectors for treating and modeling human CNS disorders. J Gene Med 2004; 6: 951–962. 26 Maries E, Dass B, Collier TJ, Kordower JH, Steece-Collier K. The role of alpha-synuclein in Parkinson’s disease: insights from animal models. Nat Rev Neurosci 2003; 4: 727–738. 27 Dauer W, Przedborski S. Parkinson’s disease: mechanisms and models. Neuron 2003; 39: 889–909. Gene Therapy


882 28 Lo Bianco C, Ridet JL, Schneider BL, Deglon N, Aebischer P. alpha -Synucleinopathy and selective dopaminergic neuron loss in a rat lentiviral-based model of Parkinson’s disease. Proc Natl Acad Sci USA 2002; 99: 10813–10818. 29 Azzouz M, Martin-Rendon E, Barber RD, Mitrophanous KA, Carter EE, Rohll JB et al. Multicistronic lentiviral vectormediated striatal gene transfer of aromatic L-amino acid decarboxylase, tyrosine hydroxylase, and GTP cyclohydrolase I induces sustained transgene expression, dopamine production,

and functional improvement in a rat model of Parkinson’s disease. J Neurosci 2002; 22: 10302–10312. 30 Zufferey R, Donello JE, Trono D, Hope TJ. Woodchuck hepatitis virus posttranscriptional regulatory element enhances expression of transgenes delivered by retroviral vectors. J Virol 1999; 73: 2886–2892. 31 Tamamaki N, Nakamura K, Furuta T, Asamoto K, Kaneko T. Neurons in Golgi-stain-like images revealed by GFP-adenovirus infection in vivo. Neurosci Res 2000; 38: 231–236.

Supplementary Information accompanies the paper on Gene Therapy website (http://www.nature.com/gt)

Gene Therapy