|Home | About | Journals | Submit | Contact Us | Français|
We previously reported regulated expression of erythropoietin (EPO) over 4 weeks in the peripheral nerve in vivo using a herpes simplex virus (HSV)-based vector containing a Tet-on regulatable gene expression cassette. In order to create a vector that would be appropriate for the treatment of chronic neuropathy, we constructed a HSV vector with expression of EPO under the control of the Tet-on system in which the HSV latency-associated promoter 2 (LAP2) element was used to drive the expression of the Tet-on transactivator. EPO expression from the vector was tightly controlled by administration of doxycycline (DOX) in vitro. One month after inoculation of the vector to transduce dorsal root ganglion (DRG) in vivo, administration of DOX-containing chow induced expression of EPO. Mice with streptozotocin (STZ)-induced diabetes inoculated with the vector were protected against the development of neuropathy by continuous administration of DOX-containing chow over the course of 3 months. Identical results were achieved when DOX was administered every other week over 3 months of diabetes, but administration of DOX 1 week out of 3 provided only partial protection against the development of neuropathy. Taken together, these results suggest such a vector is well suited for clinical trial for the treatment of chronic or subacutely developing neuropathy.
Extensive preclinical animal studies indicate that systemic delivery of neurotrophic factors can prevent the development of polyneuropathy,1–5 but translation of these findings into clinical treatment has been hampered by off-target effects that preclude systemic administration of neurotrophic factors to patients in doses that are required to produce effects in animal models.6–10 In order to overcome this limitation, we developed a series of non-replicating herpes simplex virus (HSV)-based gene transfer vectors to target and express neurotrophic factors to dorsal root ganglion (DRG) neurons. Transduction of DRG by subcutaneous inoculation of replication-deficient HSV vectors expressing neurotrophin-3 (NT3) or nerve growth factor (NGF) prevents the progression of neuropathy caused by overdose of pyridoxine (PDX) or exposure to cisplatin,11–12 and HSV mediated delivery of EPO, NT3, or vascular endothelial growth factor (VEGF) by subcutaneous inoculation protects against the development of sensory neuropathy in mice rendered diabetic by streptozotocin.13–15
The temporal course of transgene expression in DRG in vivo from HSV-based vectors is dependent on the promoter element utilized. HSV vectors in which gene expression is driven by the human cytomegalovirus immediate early promoter (HCMV IEp) produce relative short-term gene expression that persists for a period of weeks.11 HSV vectors in which transgene expression is under the control of the HSV latency associated promoter 2 (LAP2) element produce at least 6 months of biologically relevant transgene expression in both the central16 and peripheral11, 14 nervous systems.
HSV LAP2-driven prolonged transgene expression would be appropriate for subacutely evolving chronic progressive conditions like sensory polyneuropathy that evolves over a prolonged time frame. But because of the possibility that prolonged expression of a neurotrophic factor might produce unanticipated effects requiring the discontinuation of treatment, it would be desirable to be able to regulate transgene expression. In a proof-of-principle study we previously reported construction and characterization of a non-replicating HSV-based vector in which a Tet-on system expressed under the control of the HCMV IEp was used to regulate the expression of EPO in DRG neurons in vitro and in vivo.17 EPO expression from the vector was strictly regulated by DOX, and intermittent expression of the transgene in vivo achieved by DOX administration 4 days out of 7 was sufficient to protect mice against the development of diabetes-induced neuropathy over a 4-week period.17
In the current study, we constructed a novel HSV vector in which the Tet-on system used to control EPO expression was placed under the control of the HSV LAP2 element and tested the effect of regulated EPO expression from the vector on the development of diabetic neuropathy over a period of 3 months. We found that continuous expression of EPO achieved by continuous administration of DOX to animals inoculated with the vector preserved nerve function in rats with streptozotocin (STZ)-induced diabetes. Intermittent expression of EPO achieved by administration of DOX containing chow every other week (1 week out of 2) provided a similar level of protection against the development of neuropathy. The results have important implications for the development of HSV-based vectors to treat neuropathy.
We started with plasmid pSP27-link-Tet-on that contains the expression cassette of the transactivator of the Tet-on system under the control of the HCMV IEp,17 and replaced the HCMV IEp with the HSV LAP2 to create plasmid pSP72-link-L2-Tet-on. The inducible EPO-HA (hemagglutinin) and GFP-HA fusion expression cassettes were released from plasmid pTRE-Tight-EPO-HA and pTRE-Tight-GFP-HA,17 and ligated into the XhoI site of plasmid pSP72-link-L2-Tet-on to create plasmids pSP72-TRE-Tight-EPO-HA-L2-Tet-on or pSP72-TRE-Tight-GFP-HA-L2-Tet-on. The regualtable gene expression unit was cleaved from plasmid pSP72-TRE-Tight-EPO-HA-L2-Tet-on or pSP72-TRE-Tight-GFP-HA-L2-Tet-on and transferred to a shuttle plasmid SASB3, which contains sequences derived from HSV genome.17 The recombinant vectors were generated by homologous recombination between the end-point plasmid and the non-replicating HSV vector UL41E1G6 (for the EPO vector) or vEPO (for the GFP vector) to produce the vector vL2rtEPO expressing the EPO-HA fusion peptide and the vector vL2rtGFP encoding for the GFP-HA fusion peptide (Figure 1).
We tested the regulated expression of EPO from vL2rtEPO in response to exposure to DOX in complementing 7b cells. Vector DNA replication, which is required for the HSV LAP2 to be active in acute infections, occurs in those cells.18 7b cells were infected with vL2rtEPO or vL2rtGFP at a multiplicity of infection (MOI) of 3 and DOX added 1 hr after infection. EPO expression was assessed by ELISA after 48 hours of DOX treatment. 10 ng/ml DOX resulted in release of 182 ng/ml EPO into the medium from cells infected with vL2rtEPO, which did not increase significantly with increasing concentration of DOX up to 104 ng/ml (Figure 2a). However, there was no detectable release of EPO from cells infected with control vector vL2rtGFP and treated with DOX or from cells infected with vL2rtEPO but not exposed to DOX (Figure 2a). RT-PCR of RNA isolated from infected cells after 48 hrs of 10 ng/ml DOX treatment using EPO-HA specific primers showed an EPO-specific band in vL2rtEPO but not vL2rtGFP infected cells (Figure 2b).
We examined the kinetics of EPO expression by measuring the amount of EPO released into the medium from infected cells using DOX on–off and off–on paradigms. 7b cells were infected with vL2rtEPO or vL2rtGFP at an MOI of 0.01. A substantial release of EPO into the medium was observed when vL2rtEPO-infected cells were exposed to DOX (10 ng/ml) for 2 days. Removal of DOX resulted in a rapid decrease in EPO concentration in the medium from vL2rtEPO-infected cells and by day 4 after DOX removal no EPO was detectable in the medium (Figure 2c). Cells infected with vL2rtEPO and continuously exposed to DOX starting 2 days after infection showed a substantial and statistically significant increase in EPO released into the medium measured on days 4–6. No detectable expression of EPO was seen in cells infected with control vector vL2rtGFP and treated with DOX in a similar fashion on either days 0 and 2 or days 4 through 6 (Figure 2d).
To test the inducible expression of the transgene in vivo, mice were inoculated subcutaneously in both hind feet with vL2rtEPO (108 pfu in 10 μl PBS) and 2 weeks later rendered diabetic by IP injection of STZ. One month after virus inoculation, normal chow was replaced by a DOX-containing diet for 3, 4, 6 or 8 days, and the amount of EPO in DRG determined by RT-PCR and ELISA. Three to 8 days of DOX administration induced the expression of EPO mRNA (insert in Figure 3a) and protein (Figure 3a). No EPO expression was detectable in the animals receiving the vector but not fed DOX-containing diet (day 0). To analyze the kinetics of the shut-off of transgene expression in the absence of DOX in vivo, animals were inoculated with vL2rtEPO and one month after virus inoculation fed DOX-containing chow for 7 days, after which animals were again fed normal chow. EPO mRNA (insert in Figure 3b) and protein (Figure 3b) levels in DRG measured 2 and 4 days after removal of DOX-containing chow fell to undetectable levels by 4 days of normal chow.
To determine whether intermittent expression of EPO from the vector preserves nerve function in diabetic animals over a prolonged time course, we tested the vector in a mouse model of diabetic polyneuropathy. Male Swiss Webster mice were inoculated into the plantar surface of both hind feet with vL2rtEPO (108 pfu in 10 μl PBS) and 2 weeks later rendered diabetic by intraperitoneal injection of STZ. Two weeks after STZ injection (1 month after vector inoculation), diabetic animals infected with vector vL2rtEPO were tested with 4 different regimens: 1) normal food (no DOX); 2) DOX-containing chow 1 week out of 3; 3) DOX-containing chow 1 week out of 2; or 4) Continuous administration of DOX-containing chow (Figure 4a). Diabetic animals with blood glucose concentration higher than 300 mg/dl were included in the study. As a positive control, nondiabetic animals were fed normal chow.
At 14 weeks of diabetes (16 weeks after vector injection, 3 months after initiation of DOX treatment), sensory nerve amplitude and velocity were determined, followed the next day by determination of thermal threshold using the hot plate test. Diabetic animals inoculated with vL2rtEPO and fed normal chow showed significantly reduced sensory nerve amplitude and conduction velocity consistent with the development of neuropathy (Figure 4b,c). Diabetic animals inoculated with vL2rtEPO and fed DOX-containing chow continuously or fed DOX-containing chow on a one-week on and one-week off schedule had amplitudes and conduction velocity indistinguishable from control, non-diabetic mice (Figure 4b,c). Diabetic animals inoculated with vL2rtEPO and fed DOX-containing chow 1 week out of 3 displayed a marked decrease in sensory nerve amplitudes and conduction velocity compared to naïve animals, though statistically significantly higher than that observed in diabetic animals injected with the vector but not fed DOX-containing chow (Figure 4b, c). The behavioral results followed a similar pattern. Diabetic animals infected with vL2rtEPO and fed normal food had a significantly increased latency to withdraw from a painful thermal stimulus, but diabetic mice inoculated with vL2rtEPO and fed DOX-containing food continuously or fed DOX-containing food every other week had a normal withdrawal latency that was identical to that of control, non-diabetic mice (Figure 4d). Animals inoculated with vL2rtEPO and fed DOX-containing food 1 week out of 3 showed increased latency to withdraw from a painful thermal stimulus relative to the control naïve animals, indicating small nerve fibers in those animals were partially damaged.
The principal results of this study are that regulated expression of EPO from an HSV vector using a Tet-on system in which the LAP2 element was used to drive expression of the Tet-responsive transactivator provides prolonged regulated expression in the peripheral nervous system, and that intermittent expression of EPO (every other week) achieved by intermittent administration of DOX over a period of 3 months is sufficient to protect against the development of neuropathy in diabetic animals.
EPO, a hormone that primarily regulates red blood cell production, is a potent neuroprotetive factor that can prevent neuron death either by inhibiting the production of reactive oxygen species or via PI3K/Akt/GSK-3β signaling.15, 19,20 Administration of EPO by intraperitoneal injection prevents the progression of neuropathy resulting from chemotherapy, diabetes or HIV-infection.4, 21–23 In previous work we demonstrated that EPO expression in DRG achieved by subcutaneous inoculation of a non-replicating HSV vector can be used to protect peripheral nerve from damage in diabetic animals as demonstrated by reduced nerve fiber loss in the skin and increased expression of neuropeptide calcitonin gene-related peptide in the dorsal horn.15 We subsequently demonstrated that a non-replicating HSV vector in which EPO expression from a Tet-on based HSV vector was controlled by administration of DOX could be used to achieve similar results when DOX was administered 4 days out of 7 over the course of one month of diabetes.17
In this study, we replaced the HCMV IEp with the HSV LAP2 to drive the expression of the transactivator in the Tet-on system in order to achieve a long-term regulated transgene expression. In vitro and in vivo studies indicated that kinetics of the turn-on and turn-off of the expression of EPO from the vector by DOX was similar to that observed in the non-modified Tet-on based EPO HSV vector in our previous work.17 Moreover, we found intermittent expression of EPO from the vector by feeding animals DOX-containing chow every other week is sufficient to preserve nerve function in diabetic animals for at least 3 months.
We chose to use the Tet-regulated system for regulated gene expression because: 1) the inducer DOX is approved for human use; 2) a safe dose used in humans has been proven to be sufficient to induce transgene expression in rat, mice and nonhuman primates;24–25 3) in vitro and in vivo studies demonstrate that transgene expression can be tightly regulated by DOX;26–28 and 4) there is no the transactivator-specific immunity produced in animals receiving a Tet-on based transgene expression vector.29–30 We chose to use Tet-on system rather than Tet-off system because we reasoned that in clinical use should an adverse off-target effect emerge, the patient would only need to stop taking the activation drug to correct the problem.31
Prolonged regulatable gene expression from other vector platforms has been reported previously.32–35 But gene transfer to the DRG from peripheral inoculation requires the use of HSV vectors, and unlike the other vectors that have been employed in the prior studies of long-term regulated expression, the time course of transgene expression from HSV vectors in vivo is critically dependent on the promoter element. We chose to use HSV LAP2 for the Tet-on system because: 1) HSV LAP2 is naturally active for life in trigeminal and dorsal root ganglion neurons in humans and animals once HSV establishes a latent state;36 and 2) animal studies indicate that LAP2 can drive the long-term biologically relevant expression of therapeutic peptides in the central and peripheral nerve systems.11.14,16 NT3 expressed from a HSV LAP2 based HSV vector prevented the progression of PDX and diabetes-induced neuropathy at least for 6 months.11,14 HSV LAP2-driven expression of EPO and GDNF from a HSV vector protected dopaminergic neurons against MPTP and 6-OHDA-induced degeneration in mouse models of Parkinson’s disease for at least 6 months.16,37
Neurotrophic factors have been demonstrated to preserve peripheral nerve function in animal models of many forms of neuropathy.1–5 Potentially, any one of a number of neurotrophic factors could be used as the transgene for developing a vector for the treatment of neuropathy. In this proof of principle study, we chose to develop and test an EPO vector but it is highly likely that similar results could be achieved with other neurotropic factors.
There are no perfect animal models of human diabetic neuropathy. Therefore, we do not know in patients whether 1 week on followed by 1 week off of EPO expression will prove sufficient to prevent the progression of diabetic neuropathy clinically. However, even if continuous expression of EPO were required to prevent progression of neuropathy in patients, regulated expression would be desirable so that in the event that side effects were observed, EPO expression could be stopped by interrupting the administration of DOX. HSV-mediated gene transfer to the DRG has moved into clinical trial38 and a phase 2 trial of HSV-mediated gene transfer for the treatment of pain is currently underway. The characterization of an HSV vector with prolonged regulated gene expression has important implications for the further development of this therapy to treat chronic problems like polyneuropathy.
7b cells (Joseph Glorioso, University of Pittsburgh), a derivative of Vero cells that expresses HSV ICP27 and ICP4, were maintained and grown in Dulbecco’s modified Eagle’s essential medium supplemented with 10% fetal bovine serum (Atlanta Biologics, Atlanta, GA), 100 units/ml penicillin, 100 μg/ml streptomycin sulfate, 0.03% glutamine, and 0.375% sodium bicarbonate in a 5% CO2 atmosphere. The parental HSV vectors UL41E1G6 (provided by Joseph Glorioso, University of Pittsburgh) and vEPO (Zetang Wu et al., unpublished) used for generation of the recombinant viruses in this study are HSV null mutants deleted for the essential gene ICP27 with both copies of the IE gene ICP4 replaced by GFP or EPO. UL41E1G6 and vEPO were propagated in 7b cells and the virus titer was determined by plaque assay.39
To generate prolonged regulatable transgene expression plasmids, HCMV IEp promoter, which drives the expression of the transactivator in the Tet-on system, was replaced by the HSV LAP2 promoter. 600 bp HSV LAP2 was amplified from replication-deficient HSV mutant UL41E1G6 infected 7b cell DNA by PCR using a upper primer, 5′-CCG CTC GAG GGG TGG TGC GAA AGA CTT TCC-3′ and a lower primer, 5′-CCG GAA TTC GAA GCA GGT GTC TAA CCT ACC TG-3′. The PCR product was ligated into the XhoI and EcoRI sites of plasmid pSP72-Link-Tet-on to replace the HCMV IEp promoter to drive the expression of the transactivator resulting in plasmid pSP 72-link-L2-Tet-on. Plasmid pSP72-Link-Tet-on is a plasmid modified from plasmid Tet-on (Clontech) by changing the MCS to facilitate the downstream manipulation of cloning, which constitutively expresses the transactivaor under the control of HCMV IEp.17 Regulatable EPO-HA and GFP-HA fusion expression cassettes were released from plasmids pTRE-Tight-EPO-HA and pTRE-Tight-GFP-HA, which were constructed in our previous work and express EPO and GFP-HA fusions in the presences of DOX and the transactivator under the control of the tetracycline response element (TRE)-minimal HCMV IEp fusion promoter,17 and ligated into the XhoI site of plasmid pSP72-link-L2-Tet-on. The resultant plasmids pSP72-TRE-Tight-EPO-HA-L2-Tet-on and pSP72-TRE-Tight-GFP-HA-L2-Tet-on express EPO and GFP-HA fusions in the presence of DOX. The whole EPO and GFG-HA fusion-inducible expression units were released from plasmid pSP72-TRE-Tight-EPO-HA-L2-Tet-on and pSP72-TRE-Tight-GFP-HA-L2-Tet-on by BglII cleavage and cloned into the BamHI site of plasmid pSASB3 (Joseph Glorioso, University of Pittsburgh) for facilitating the construction of HSV based EPO and GFP expression vectors. The resulting plasmids pSASB3-LAP2-EPO and pSASB3-LAP2-GFP were used for generation of the corresponding recombinant vectors.
Recombinant HSV vector vL2rtEPO expressing rat EPO-HA was generated based on homologous recombination between plasmid pSASB3-LAP2-EPO and replication-deficient HSV-1 genome DNA (UL41E1G6) in complementing 7b cells and two copies of the regulatable EPO expression unit were inserted into the ICP4 loci of the HSV genome since the regulatable gene expression unit in plasmid pSASB3-LAP2-EPO was flanked by sequences corresponding to the up- and downstream sequences of the ICP4 open reading frame in HSV at the 5′- and 3′-ends, which serve as the basis for homologous recombination between the plasmid DNA and the viral genome. Transfection and infection for generating the recombinant EPO expression vector, the screening, purification and confirmation of the recombinant virus were conducted using the methods previously described.17 Recombinant vector vL2rtGFP which expresses GFP-HA was created using similar methods utilized for the generation of vector vL2rtEPO except that parental virus vEPO (a replication-deficient recombinant HSV virus, in which two copies of EPO with expression under the control of HCMV IEp were inserted into the ICP-4 loci of the HSV genome), was used for generation of the recombinant virus allowing for green-white screening for the recombinant virus under fluorescence microscopy.
Male Swiss Webster mice weighing 20–25 g (Charles River, Wilmington, MA) were used for all the experiments. All animal procedures in this study were performed in compliance with approved institutional animal care and use protocols. For analyzing regulation of EPO expression from the vector by DOX, mice were inoculated with vL2rtEPO in both hind feet [108 plaque forming units (pfu) in 10 μl PBS] and 2 weeks later rendered diabetic by STZ injection twice within 2 days with a dose of 100 mg per kg animal weight. Two weeks after STZ injection, animals with blood glucose concentration higher than 300 mg/dl were fed 625 mg/kg DOX-containing diet (Harlan laboratory, Madison, WI). Six animals were euthanized at day 0, 3, 4, 6 and 8 after DOX treatment and L4-6 DRGs removed. EPO expression in DRG was measured by RT-PCR and ELISA. To examine whether removal of DOX treatment turns off the EPO expression from the vector induced by DOX in animals, animals were fed DOX-containing chow for 7 days one month after virus infection and after which animals fed normal food for 2 or 4 additional days, then euthanized and L4-6 DRGs removed for EPO expression assays. To determine whether regulated expression of EPO from the vector preserves sensory nerve function in diabetic animals, animals were infected by the vector vL2rtEPO and rendered diabetic two weeks after virus injection. Diabetic animals were either fed normal food or fed DOX-containing chow at different schedules (continuously, one week out of two or one week every three weeks) for 3 months (12 weeks). Sensory nerve amplitude and conduction velocity and the sensitivity of animals to a heat stimulus were measured at the end of the 12th weeks after DOX treatment.
Sensory nerve recordings were performed on the right hind foot using a Nicolet Viking III EMG device (Nicolet Biomedical, Madison, WI). Mice were anesthetized with isoflurane for testing, and subcutaneous temperature maintained at 36–37 °C. The hind limbs were secured at an angle of 30° relative to the body and a ground electrode inserted into the tail, electrodes inserted into the sciatic notch as the recording electrode, while the stimulating electrode was placed at the ankle and the reference electrode was positioned at the first digit.17
The sensitivity of animals to heat stimulus was evaluated by hotplate test on a metal plate (UgoBasile, Comeria, Italy). The initiation temperature was set at 47 °C and the temperature programmed to increase at 2 °C per minute. Withdrawal latency was measured in seconds to the time the animals lifted their hind paw from the plate or licked the paw. The cutoff temperature was 49 °C. An animal was placed onto the plate when it was heated to 47 °C, and testing for each animal was conducted in triplicate with a time interval of at least 5 minutes between tests.
EPO ELISA was conducted using an EPO ELISA kit (R&D Systems, Minneapolis, MN) according to the manufacturer’s instructions. The primary antibody in the kit reacts with both rat and mouse EPO. 50 μl of medium or 1–5 μg protein was utilized for EPO assays in medium or cell lysates. All ELISA assays were carried out in triplicates and the results expressed as means ± SD.
Total RNA from tissue and cell pellet were isolated using TRIzol from Invitrogen (Carlsbad, CA) and the RNeasy plus kit from Qiagen (Valencia, CA) respectively according to the manufacturer’s instructions. RNA concentration was measured using a UV spectrophotometer and RNA quality monitored by the ratio of OD260/OD280. Reverse transcription was conducted using the Superscript reverse transcriptase II kit from Invitrogen (Carlsbad, CA) The RNA input was 2 μg and poly (T) used as the primer for reverse transcription. The synthesized complementary DNA was used for quantification of mRNA levels by semi- quantitative PCR.
Polymerase chain reaction (PCR) amplification was carried out in 50 μl using a standard protocol with an initial denaturing step at 94 °C for 5 minutes followed by either 25 cycles for detection of the β-actin mRNA level or 30 cycles for the measurement of the EPO-HA mRNA level at 95 °C, 1 minute/60 °C, 30 seconds/72 °C, 1 minute and the products separated on a 1% agarose gel. The primers used in PCRs are as follows. For identification of vL2rtEPO identity or semi- quantification of exogenous EPO expression, the forward primer is 5′-CCGGAATTCGCCAGGCGCGGAGATG-3′ and the reverse primer that contains EPO and HA sequences is 5′-CGGGATCCTCAAGCGTAATCTGGAACATC-3′. For identification of vL2rtGFP identity, the forward prime is 5′-CGGAATTCCGCCACCATGGCTAGCA AAGGAG-3′ and the reverse primer is 5′-CTAGATCAAGCGTAATCTG GAACATCGTATGGGTATGC-3′. For semi-quantification of β-actin mRNA levels, the forward primer: 5′-CAGTTCGCCATGGATGACGATATC-3′, and the reverse primer: 5′-CACGCTCGGTCAGGATCTTCATG-3′.
The statistical significance of the difference between groups was determined by Student t-test and results expressed as means ± SD.
The authors acknowledge the assistance of Vikram Thakur in vector propagation and Jesssica Myers help in conducting electrophysiolgy testing. The work was supported by NIH grants DK044935 and NS038850, and grants from the Department of Veterans Affairs and the Juvenile Diabetes Research Foundation.
The authors have no competing interests to declare.