WO2002089855A1

WO2002089855A1 - Peroral transduction of hepatocytes in the treatment of disease

Info

Publication number: WO2002089855A1
Application number: PCT/NZ2002/000076
Authority: WO
Inventors: Ruian Xu; Hsu Ma; Matthew John During
Original assignee: Auckland Uniservices Limited
Priority date: 2001-04-27
Filing date: 2002-04-23
Publication date: 2002-11-14

Abstract

There is disclosed a method of expressing a therapeutically effective amount of a biologically active protein or peptide in a hepatocyte comprising oral administration of a vector, or of a therapeutic agent comprising the vector, wherein the vector comprises a nucleotide sequence encoding the protein or peptide operably linked to a regulatory sequence capable of driving expression of the nucleotide sequence in the hepatocyte, and wherein the hepatocyte harbouring the vector secretes the expressed protein or peptide into the systemic circulation locally.

Description

PERORAL TRANSDUCTION OF HEPATOCYTES IN THE TREATMENT OF DISEASE

TECHNICAL FIELD

The present invention includes improved methods and composition for transduction of hepatocytes in vivo and in particular to in vivo hepatocyte expression of therapeutic proteins and peptides in the treatment of diabetes and other glucose metabolism disorders.

BACKGROUND ART

A primary aim of gene therapy for DM is achievement of euglycemia and elimination of the need for repeated injections or tissue transplantation requiring immunosupression. Direct in vivo gene transfer using an "off the shelf vector, which is not associated with problems of tissue availability and harvesting, may offer a significant alternative to the promising developments in islet transplantation. A number of gene therapy approaches have

- previously been investigated. These range from transfer of cell survival genes directly to pancreatic islet cells (1) to the generation of artificial β cells (2,3).

Although in general physiological regulation of insulin secretion has not been obtained in these early studies, and some of the ex vivo cell transplant approaches may prove to be unstable or potentially unsafe for clinical practice (4). A recent paper suggests that direct in vivo gene transfer of an insulin analogue can lead to persistent euglycemia in diabetic rodents, and be associated with some transcriptional regulation of the transgene (5). Recombinant adeno-associated virus (rAAV) possesses a number of characteristics that make it a particularly attractive candidate for clinical gene therapy (6). Recently we have demonstrated that a perorally administered AAV vector leads to persistent (> 6 months) expression of a transgene in both gut epithelial cells and lamina propria, resulting in long-term phenotypic recovery in an animal model of lactose intolerance (7). This approach avoids many problems such as limited tissue supply and adverse effects associated with more invasive access by intravenous, intraportal, direct intraparenchymal (liver or muscle) injection or transplantation of genetically altered cells with uncertain growth characteristics.

A major impediment to successful insulin gene therapy has been the difficulty in coupling the synthesis and release of the transgene insulin to serum glucose concentrations. It was recently shown to direct sufficient gene expression to restore euglycemia in diabetic rodents following intraportal delivery (5). These studies demonstrated for the first time the potential therapeutic value of insulin gene therapy. However, although glucose induced the transcription of the transgene, release of the insulin analogue was substantially delayed. Unless transcriptional control is coupled to appropriate processing to mature insulin, concentration in storage granules, and tied to glucose responsive release, it is unlikely to fully restore euglycemia and avoid hypoglycemia under the rapidly changing demands met by intermittent feeding. An alternative approach may be the use of the insulin promoter itself, and particularly because of its small size, a 412 bp rat insulin I promoter (RIP) fragment which retains the elements necessary for transcriptional regulation by glucose is of interest (8). Our hypothesis was that insulin derived from rAAV under control of the RIP fragment, and administered perorally, might result in sufficient expression in diffusion neuroendocrine cells (DNES) to obtain both constitutive secretion as well as a component of appropriately regulated insulin release in the absence of functional islet cells. This regulation of release would not simply be based on transcriptional control, but rather ectopic expression of insulin in DNES cells, that possess the machinery for both effective processing as well as peptide storage, would lead to the acute release of both the endogenous gut peptide as well as the appropriately-processed mature insulin from these cells in response to a meal. The DNES L cells release glucagon-like peptide with kinetics very similar to that of insulin itself following carbohydrate ingestion (9), therefore the transgenic insulin in these cells might be appropriately released peri- and postprandially. Moreover, mature insulin would be released into the portal circulation - the most physiological route of insulin delivery (10). However, the secretion of insulin from transduced DNES cells and other cells in the gut decreased sharply 3 days after oral administration, resulting in hyperglycemia of these diabetic rats, probably due to the loss of transduced cells through naturally sloughing process. It is therefore an object of the present invention to overcome or at least ameliorate one or more of the disadvantages of the prior art or to provide a useful alternative.

SUMMARY OF THE INVENTION

According to a first aspect there is provided a method of expressing a therapeutically effective amount of a biologically active protein or peptide in a hepatocyte comprising oral administration of a vector, or of a therapeutic agent comprising said vector, wherein the vector comprises a nucleotide sequence encoding the protein or peptide operably linked to a regulatory sequence capable of driving expression of the nucleotide sequence in the hepatocyte, and wherein the hepatocyte harbouring said vector secretes the expressed protein or peptide into the systemic circulation or locally. According to a second aspect there is provided a method of expressing a therapeutically effective amount of insulin in a hepatocyte comprising oral administration of a vector, or of a therapeutic agent comprising said vector, wherein the vector comprises a nucleotide sequence encoding insulin operably linked to a regulatory sequence capable of driving expression of the nucleotide sequence in the hepatocyte, and wherein the hepatocyte harbouring said vector secretes the expressed insulin into the systemic circulation or locally. According to a third aspect there is provided a method for improved expression of a therapeutically effective amount of a biologically active protein or peptide in a hepatocyte comprising oral administration of a vector, or of a therapeutic agent comprising said vector, to a subject in which such expression is desirable, wherein the vector comprises a nucleotide sequence encoding the protein or peptide operably linked to a regulatory sequence capable of driving expression of the nucleotide sequence in the hepatocyte, whereby the hepatocyte harbouring said vector secretes the expressed protein or peptide into the systemic circulation or locally and wherein the subject is fasted prior to oral administration for a time sufficient to achieve complete or near complete emptying of gastric and/or intestinal contents.

According to a fourth aspect there is provided a method of expressing a therapeutically effective amount of insulin in a hepatocyte comprising oral administration of a vector, or of a therapeutic agent comprising said vector, to a subject in which such expression is desirable, wherein the vector comprises a nucleotide sequence encoding insulin operably linked to a regulatory sequence capable of driving expression of the nucleotide sequence in the hepatocyte, whereby the hepatocyte harbouring said vector secretes the expressed insulin into the systemic circulation or locally and wherein the subject is fasted prior to oral administration for a time sufficient to achieve complete or near complete emptying of gastric and/or intestinal contents. According to a fifth aspect there is provided a method of improving in vivo hepatocyte expression of an exogenous nucleotide sequence following oral administration to a subject of a vector comprising the sequence, said method comprising the step of fasting the subject before oral administration of said vector for a time sufficient to achieve complete or near complete emptying of gastric and/or intestinal contents. According to a sixth aspect there is provided a method of treating diabetes comprising oral administration to a subject requiring such treatment of a vector, or of a therapeutic agent comprising said vector, wherein the vector comprises a nucleotide sequence encoding insulin operably linked to a regulatory sequence capable of driving expression of the nucleotide sequence in the hepatocyte, and wherein the hepatocyte harbouring said vector secretes the expressed insulin into the systemic circulation in the a mount sufficient to treat diabetes.

According to a seventh aspect there is provided a method of regulating blood glucose level in a diabetic subject comprising oral administration to a subject requiring such treatment of a vector, or of a therapeutic agent comprising said viral vector, wherein the vector comprises a nucleotide sequence encoding insulin operably linked to a regulatory sequence capable of driving expression of the nucleotide sequence in the hepatocyte, and wherein the hepatocyte harbouring said vector secretes the expressed insulin into the systemic circulation in the a mount sufficient to regulate blood glucose levels.

In preferred embodiments the vector is a viral vector and can be preferably selected from the group consisting of adeno-assocaited vector, an adenovirus vector, a lentivirus vector, a parvovirus vector and a herpes virus vector.

Preferred regulatory sequence is a promoter preferably selected from the group consisting of insulin promoter, rat insulin promoter, glucagon promoter, glucokinase promoter and L-pyruvate kinase promoter

In a preferred embodiment the protein or peptide such as insulin is expressed in, and secreted from, the hepatocyte for a period of 9 months. Where the protein or peptide expressed in and secreted from the hepatocyte is insulin it is preferably secreted from the hepatocyte in response to elevated blood glucose levels. Also preferred is the secretion from the hepatocyte of insulin in the amount and for a time sufficient to normalise blood glucose levels.

The vector is preferably administered orally, but other known means by which the vector can access the gastric and intestinal cavities may also be used. Also preferred is administration of the vector in the amount of from about 0.5x10^π to about 5x10ⁿ particles/dose. Even more preferred is the amount of about 2x10^π particles/dose.

BRIEF DESCRIPTION OF THE FIGURES

Figure 1: (A) rAAV/EF-human insulin vector diagram. (B) rAAV/RIP- furin insulin vector diagram. (C) C-peptide and (D) rat insulin levels in the blood of the experimental rats (STZ rats, N=101, Normal, N=6, mean+SD) prior to oral administration diagram (ELISA, Mercodia AB, Sweden). (E)

RT-PCR analysis of total RNA extracted from the proximal intestine from rat at one month post oral dosing of rAAV/RIP-furlns vector. Lane 1 DNA ladder (size indicated along left margin); Lane 2 positive control.

Amplification of the insulin cDNA with upper primer and lower primer

(lower arrow). Lane 3, 4, 5, 6, 7 and 8 mRNA extracted from the gut, liver, spleen, stomach, brain and testes. Lane 9 (no reverse transcriptase) and 10 (no template) negative control. ?-actin cDNA served as an internal control for the PCR reactions (upper arrow). In Situ hybridization to the gut sections by DIG immunological detection kit (Boehringer) (F) control rats and (G) RlP-furlns rats one month post oral dosing. Immunostaining of proinsulin / insulin protein expression in the gut of the rats administered with 2x 10^π EF- hlns and 2x 10¹¹ RIP-furlns insulin particles/animal for one and 3 months respectively. Immunofluorescent labelling using monoclonal antibody to proinsulin (1:250; dilution, Biodesign, E8609M) with propidium iodide counterstaining (H. control, I. 1 month), and an antibody to insulin (1 :250 dilution, Linco H145P; L. 1 month, M. 3 months). Arrows indicate proinsulin/insulin granules. Sections were incubated with primary antibodies and detected with Cy-3 or Cy-5 conjugated secondary antibodies prior to confocal imaging. Double immunofluorescent labelling of sections obtained from rats 3 months following vector administration using a polyclonal antibody to PCl/3 (J and N) (1 :250 dilution, Chemicon, AB1260), and monoclonal antibody to proinsulin (K, 1:250 dilution, Biodesign, E86209M) and ami primary antibody to insulin (O, 1:250 dilution, Linco, H145P). Arrows indicate DNES cells. Figure 2: Transgenic insulin expression in the hepatocytes following administration with 2x 10ⁿ RIP-furIn particles for (A) control, and at (B) 1 month, (C and D) 3 months, (E) 6 months, and (F) 9 months. Sections were incubated with unlabelled primary antibody (1:250 dilution, Linco, H145P), and were detected with Cy-3 conjugated antibodies and propidium iodide counterstaining prior to confocal imaging. Arrows indicate insulin granules. Scale bar: (A, D and F) 20 μm, (B and E) 24 μm and (C) 60 μm.

Figure 3: Time course of blood glucose and insulin levels in individual STZ rats treated with 2x 10¹¹ RIP-furlns particles/animal (A and C), lx 10¹¹ RIP-furlns particles/animal (B). There was an inverse relationship between the insulin and the blood glucose in the treated STZ rats during the experimental period (D) first 3 days (N=20, mean ±SEM) and (E) 6 months (N=8, mean ±SEM). Comparison of levels of transgenic insulin (IMX, Abbott, Tokyo) and blood glucose in the euglycemic (F) and hyperglycemic (G) STZ treated rats post oral dosing. Red open circles - human insulin level, black triangles - blood glucose level.

Figure 4: (A) STZ-rats response to a glucose challenge. At 4 months following peroral AAV administration, RIP-furlns STZ-rats, normal rats and untreated STZ-rats were fasted (ad libitum access to water only) for 12 hours. They were then perorally administered 2.5ml of a 20% glucose solution. Blood samples were taken before glucose administration and at 30, 120, 240 and 360 minutes (N=6, mean+SEM) following the challenge. Triangles - untreated STZ rats, open circles - treated STZ rats, squares - normal rats. (B) Dynamic balance and relationship betwee blood glucose levels and insulin release.

STZ-rats, 4 months following dosing, were fasted for 12 hours, and were then perorally administered 2.5ml of a 20% glucose solution. Blood samples were taken before and 30, 120, 180 and 360 min (N=-6, mean+SEM) after glucose administration. Squares — blood glucose level, diamonds - transgenic insulin level (IMX, Abbot lab, Tokyo).

(C) Ketone levels in normal rats, STZ rats at 14 days following administration of RIP-furlns or control virus (rAAV/RIP-luciferase) (N=6, mean ± SEM). Serum ketones were measured by spotting serum on a ketone test strip using a Bioscanner Test system (Polymer Technology Systems, Indiana).

DESCRIPTION OF THE PREFERRED EMBODIMENTS

The present invention is based at least in part on the observation that increasing fasting time, before oral administration of a vector comprising the nucleotide sequence of interest, significantly improved transduction of hepatocytes can be achieved with concomitant increase in expression and secretion of the product of the nucleotide sequence. This approach, with the assistance of appropriate dosage, formulation and handling of the vector, provides improved methods for oral therapy of disorders which rely on maintaining sufficiently high levels of biologically active proteins or peptides in the circulation, such as for example insulin in the treatment of diabetes or other glucose metabolic disorders. In the initial experiments, STZ rats (N=16) received 2xlO^π particles of rAAV/EF-human insulin (EF-hlns), a vector expressing an unmodified human proinsulin cDNA under control of the constitutive elongation factor l (EF) promoter (Fig.lA). The EF-hlns vector was administered via an orogastric tube into the stomach of STZ-treated diabetic rats as defined by high glucose level (>20mmol/L) and low circulating C-peptide (Fig.lC, < 20 pmol/L) and rat insulin levels (Fig.lD, <3.8 μU/ml). Plasma glucose fell from about 30mM to less than 3 mM within 12 hours. Acute hypoglycemia was corrected using 1-2 ml of 20% glucose solution administered intraperitoneally and 10% glucose provided in the drinking water for the first 48 hours. Additional control rats received identical doses of rAAV expressing β-galactosidase (AAVlac, n-6) or luciferase (AAVluc, n=6). To determine human proinsulin gene expression in target intestinal cells, rats were sacrificed at day 3, 17, 30 and 90 days (n=4 for each timepoint) following oral administration. Transgene expression (using a monoclonal proinsulin antibody) was observed primarily in the lamina propria (Fig.ll) as previously described for AAVlac-treated animals (7, 11) with a greater number of proinsulin-expressing cells at 30 and 90 days compared to earlier time points, whereas at no time was insulin-immunoreactivity detected in naϊve, AAVlac or AAVluc-treated animals (Fig.lH). Transduced DNES cells showed persistent expression at 3 months (Fig.lJ and IK), however, only about 5% of these cells (defined by immunoreactivity to prohormone convertase) were transduced. Proinsulin in blood was detectable in this group using a sensitive and specific ELISA (12 - 14 pmol/L, Mercodia AB, Sweden). However mature human insulin levels were below the detectable range (<0.5 μU/ml) using the IMX system (Abbott Lab, Tokyo). Blood glucose levels in rats treated with EF-hlns showed only a transient drop with baseline hyperglycemia returning within two weeks and remaining at high levels for the duration of the study (90 days). This suggests that peroral administration of a vector that contains an unmodified proinsulin gene under a strong constitutive promoter leads to insufficient conversion of proinsulin into mature insulin to result in euglycemia. Higher doses of this vector, which contained a constitutive promoter, were not used in view of the profound acute hypoglycemia in this animal model.

In preferred embodiments of the present invention a furin- consensus site and BIO modified human proinsulin cDNA was generated to enable more efficient processing of proinsulin and secretion of mature insulin from potential target cells. This cDNA was inserted into an expression cassette, containing a 412 bp rat insulin promoter and a tripartite postregulatory element to facilitate mRNA transport and stability, flanked by 145 bp AAV terminal repeats. This AAV cis plasmid (AAVfurlns) was used to generate recombinant AAV vectors, together with control vectors expressing luciferase (AAVluc) and galactosidase (AAVlac). High titer (genomic particles ~10el2/ml) rAAV was administered perorally to streptozotocin diabetic rats (blood glucose >20mM; undetectable C-peptide). In all AAVfurlns treated rats, ketones were absent and in approximately 40% of animals which received a dose of 2x1 Oel 1 viral particles, after a delay period of 5-7 weeks, blood glucose levels reached euglycemia which were maintained for the 6 month study period.

These 'responders' had higher levels of insulin and proinsulin. Specificity of the transgene human insulin was determined by using an IMX assay that did not crossreact with rat insulin. The 412 bp insulin promoter provided some transcriptional control over insulin expression and facilitated expression in gut neuroendocrine cells. As a result, ectopic insulin release increased 2-3 fold in response to a glucosechallenge. Apart from a very transient drop in blood glucose between 6 and 18 hours after oral administration, the long-term euglycemia was not punctuated with significant hypoglycemic or hyperglycemic episodes.

Unexpectedly, there was transduction of hepatocytes following peroral administration. The delayed restoration of metabolic control was associated with liver transduction, suggesting that it was important for the development of long-term euglycemia.

The invention will now be described more particularly with reference to non-limiting examples. EXAMPLES

Example 1: Rat model of diabetes

Young Wistar adult male rats, 250-280 grams, maintained in individual cages under conditions of controlled lighting, humidity and temperature, were used in all studies under a research protocol (#N798 to Dr Xu) approved by the university of Auckland Animal Ethics Committee. The rats were kept in their individual cages. These animals were injected twice with streptozotocin (lOOmg/kg i.p.) on Day 0 then again on Day 5. Two dose schedule gave a more consistent diabetic phenotype with over 80% of the rats having a fasting blood glucose of greater than 20mM. Blood glucose levels of STZ-treated rats were measured both prior to and following STZ administration at 3 -day intervals by a Glucocard Test Strip II (KDK Co. Kyoto, Japan). In order to keep animal fasting at identical condition, just prior to fasting, all cases were replaced with those never being used previously, and animals were supplied with drinking water only, food was not provided until blood sampling was completed. The rats were fasted overnight for 12 hours and tail bleeds performed to obtain 50 μl of blood at 8:30-9:30 am. Rats whose blood glucose level exceeded 20mmol/L (~360 mg/dl) with a very low level of C peptide (<20 pmol/L, ELISA, Mercodia, Sweden) and rat insulin (<3.75 μU/ml, ELISA, Mercodia) were regarded as STZ-induced diabetic (Fig. 1C and ID) and subsequently randomised for further study. Example 2: Generation of rAAV vector

For the purposes of examplifying the present invention, second generation rAAV/RIP -furin insulin (RIP-furlns) vector was constructed. Human proinsulin was subcloned into two rAAV packaging plasmids: one with a 2.5kb human elongation factor (EF) 1 alpha promoter (FiglA), and the second, a 412 bp rat insulin promoter (RIP). A woodchuck hepatitis B virus post-transcriptional regulatory element (WPRE) was also added to selected constructs to boost expression levels. (Z.M., Huang, and T.S. Yen, J. Virol. 68, 3193, 1994). Recombinant AAV vectors expressing proinsulin, insulin, luciferase and β-galactosidase were packaged as previously described using the pDG helper plas id (12). To determine the quantity of the packaging virus, aliquots of the media from HEK 293 cells infected (MOI 1000) with the EF promoter and RIP promoter virus were assayed respectively. EF promoter was 3-5 fold stronger than the RIP promoter in driving gene expression based on HEK 293 cell with ELISA assay (Mercodia). To examine if virus preparation contains proinsulin/insulin after vector extraction and processing, aliquots of vectors were analysed by ELISA (Mercodia AB, Sweden). The results showed that the level of proinsulin insulin in virus preparation was undetectable. All viral vectors were stored at -80° C prior to oral dosing. Minimal modification of the human insulin cDNA allows translation of a proinsulin gene product whose B-C and C-A junctions encode recognition sites for a ubiquitous endopeptidase, furin (Fig. IB). Groskreutz and colleages further engineered the insulin gene with a B10 mutation, leading to improve secrection of mature insulin (13). To facilitate transgene expression in DNES and hepatocytes which retain the ability to more efficiently process human insulin cDNA that contains both furin consensus and the B10 mutation were inserted into 412 bp RIP promoter (Fig. IB). Example 3 : RT-PCR amplification of rAAV mRNA

After peroral delivery of rAAV, rats were killed and the guts, liver, spleen, stomach, brain and testes were removed and placed on dry ice immediately. The samples were stored at -80° C before analysis. RNA from 100 mg of the proximal intestine and other organs was extracted using Trizol (Life Tech.). First-stand cDNA was synthesised using 5.0ug of total RNA, which was primed with Oligo dt (0.5μg, Promega), then reverse-transcribed using Superscript II RNase H reverse transcritase (150U; Life Tech.) at 42° C for 90 min. Duplicate reactions without Superscript II were negative controls. Insulin oligonucleotide primers In- 1 5'-CAGCCTTTG TGAACCAACAC-3' and In-2 5'-GCGTCTAGTTGCAGTAGTTC-3' were used to generate product. Analysis of β-actin cDNA was an internal control for the PCR reactions. Primers for β-actin PCR were βA- 1 (5'-CTCTTCCA GCCTTCCTTCC-3') and βA-2 5'-GTCACCTTCACCGTTCCAG-3'). The cycling parameters were 5 min at 94° C, followed by 40 cycles of 1 min of 60 ⁰ C 1 min at 72 ° C. After amplification, 5 μl of PCR products was electrophoresised on a 2% agarose gel containing ethidium bromide solution (Life Tech) and visualized with UV light. Example 4: In situ hybridization

Intestine slices were fixed for 7 min in 4% formaldehyde and washed in PBS for 3 min, 2xSSC for 10 min. The sections were hybridized at 37° C for 24 hr in a mixture containing 4xSSC, 10% dextran sulfate, lx Denhardt's solution, 2mM EDTA, 50% deionised formamide, 500 μg/ml herring sperm DNA. The slices hybridized with DIG-labelled antisense cRNA. The labeling procedure was according to the DIG RNA labelling kid (Boehringer). The negative controls hybridized with DID-labeled sense cRNA. Following high stringency post hybridization washes in 60% formamide in 0.2 x SSC at 37 C for 15 min and in 2xSSC room temperature for 10 min. Hybridization was detected by DIG immunological detection kid (Boehringer). Example 5: Peroral delivery of vector constructs In preliminary experiments, STZ rats (n=4 per group) were dosed perorally at 0.5, 1, 2 and 5 xl0^π particles/animal. Blood glucose fell from a stable hyperglycemic level of ~30mM to 2-6 mM within 12 hours in the first three groups. Rats that received the highest dose of 5x 10^π particles suffered severe hypoglycemia with blood glucose levels <3mM. To protect against hypoglycemia, 10% glucose was provided in the drinking water for the initial 48 hours immediately following vector dosing, with only the animals receiving the highest dose of RIP-furlns requiring further supplemental glucose treatment. Plasma glucose returned to the baseline hyperglycemic levels over 2-3 days and remained elevated with only two exceptions. Both of these rats had received a dose of 2x10* ^x virions and glucose levels fell from ~30mM to 12 mM at 4 weeks following vector administration, dropping even lower at 5 weeks and remaining within the physiological range (5-10 mM) for 6 months.

An additional group of rats (n=6) were administered 2x10ⁿ RIP-furlns particles perorally to evaluate mRNA expression in target cells. Transgene expression was analyzed at one and three months (n=3 at each timepoint) following oral dosing using RT-PCR analysis. Positive amplification was obtained in mRNA isolated from the gut and the liver (Fig.lE), with no expression detectable in the spleen, testes, lung, stomach or brain (Fig.lE). The absence of detectable insulin mRNA in tissues other than the liver and the gut, suggested that the gut and the liver are the sources for insulin production. Location of insulin gene expression in the gut was further confirmed by in situ hybridization (Fig.lG). We tested whether mRNA in the gut and the liver was effectively translated into protein using immunocytochemistry with a specific anti-human insulin antibody (Lot HI45P, Linco, MO). Insulin expression was observed as described above for proinsulin expression at one and three months post oral dosing (Fig.lL and 1M). Only a relatively few scattered epithelial cells within the intestine showed transgene expression at later time points, with the bulk of expression being in the lamina propria and scattered DNES cells (Fig.lN-O). Nevertheless, insulin expression with the RIP promoter was not as strong nor as widely distributed as the proinsulin expression obtained with the EF vector with the exception of DNES cells, with approximately 5-10 % of prohormone convertase immunoreactive cells transduced. Transgenic insulin mRNA was present in the liver (Fig.lE) of the RIP-furlns treated rats one month after oral dosing, although transgenic insulin protein content in hepatocytes was low (Fig.2B). Over ensuing weeks, this ectopic hepatocyte expression increased significantly. This delay in transgene expression in the liver is consistent with the timecourse of AAV vector genome conversion to a duplex replicating form, integration and improved access to the transcriptional machinery. Miao et al. similarly demonstrated that the single- stranded intraportally-delivered AAV vector genomes were progressively converted into double-stranded, head-to-tail concatamers in hepatocytes of mice over a period of five weeks (14). In our study, transgenic insulin in hepatocytes increased gradually over 3 months (Fig.2C and D) and then plateaued.

Additional STZ diabetic rats were divided into 3 groups. A control group (n=6) was given AAVlac. Ten animals were treated with 10¹¹ RIP- furlns particles and an additional ten STZ rats received 2 x 10^π RIP-furlns particles. Higher doses were not administered in view of the transient but severe hypoglycemia in the first 24 hours, necessitating glucose supplementation. In this experiment, glucose was not added to the diet or drinking water and blood glucose levels were closely monitored for the first 3 days. No animal died from hypoglycemia, nor did glucose drop below 2 mM. We used an IRMA method (IMX, Abbott Lab, Tokyo) specific for human mature insulin (claimed cross reactivity with proinsulin <0.01%) to quantitate circulating transgenic mature insulin in the blood of these rats. To further demonstrate the specificity of this assay, non-diabetic control rats (n=6), were administered a glucose load (2.5ml of 20% dextrose) with peripheral blood samples taken at baseline and at 30 minutes when endogenous rat insulin levels would be significantly elevated. In none of these samples was human insulin detected by the IMX assay (<0.5 μU/ml). In RIP-furlns treated STZ rats, as predicted, we found an inverse relationship between the human insulin concentration and glucose levels (Fig.3D and 3E). In both the low and high dose groups, vector administration led to significant gene expression and insulin release occurring within 6-12 hours following peroral dosing (Fig.3D). This was associated with a drop in plasma glucose in diabetic animals to 2-5 mM at 10-14 hours. Glucose levels subsequently climbed back to baseline hyperglycemic levels by 3 days. This pattern is consistent with an early transduction of gut epithelial cells, a rapidly proliferating cell population that sheds every 2-3 days. Moreover, abluminal protein secretion has previously been seen with intestinal epithelial cells transduced using adenoviral vectors (15). Over the course of 1-3 days, intestinal epithelial cells migrate from the crypts and are sloughed off the tips of the villi. The drop in plasma transgenic insulin levels after 24 hours is consistent with the shedding of these cells.

Glucose levels remained high for the first few weeks. But 5-10 weeks following vector administration, the glucose level in 4 out of 10 rats in the 2x10¹¹ virion group fell to the physiological range and remained at an euglycemic plateau for over 6 months (Fig.3A). C-peptide levels were at or under the detection limit (= 15 pmol/L, Mercodia AB, Sweden), when measured at one, three and six months following vector dosing (N=10 for each time point), values considerably below those seen in normal rats (66+ 1, mean + SEM). Similarly, four out of 10 rats in the low dose similarly reached euglycemia with a similar time course. These animals had slightly greater variability in plasma glucose than the high dose group and were unable to maintain precise euglycemia (Fig.3B). Human insulin levels in these animals were <0.7 μU/ml for the first 5-10 weeks, then climbed associated with a corresponding drop in plasma glucose levels. Glucose concentrations remained within the near normal physiological range throughout the rest of the experimental period. Non-endocrine cells including fibroblasts, hepatocytes and epithelial cells express furin, a distinct Kex2 family endopeptidase (16). Furin recognises a specific consensus sequence in the prohormone for its cleavage. Thule et al. (17) recently demonstrated that a mutated proinsulin was cleavable by the ubiquitous convertase furin in vitro by hepatocytes. The altered cleavage sites in our RIP-furlns construct (Fig.lB) allow the insulin to be processed by the ubiquitous endoprotease furin. Moreover, a histidine BIO to aspartic acid point mutation creates a more stable form of insulin leading to an increase in its accumulation and release (13). Hepatocyte insulin expression, upon reaching high level expression in the liver at ~3 months, remained stably elevated at 6 and 9 months (Fig.2C-F). Furthermore on high power confocal microscopy, insulin showed a punctate expression pattern (Fig.2B-F). This is consistent with an earlier report that the introduction of insulin cDNA into a human hepatoma cell line resulted in synthesis, storage and acutely regulated insulin release (3). The secretory dynamics of the transgenic insulin is likely to be a function of the type and number of transduced cells at any given timepoint. The timecourse of the increase in plasma insulin concentrations paralleled the ectopic insulin content in hepatocytes whereas gut lamina propria expression appeared stable over this time as previously reported (7), suggesting that liver transduction is most likely responsible for the observed delayed, and potentially permanent, therapeutic effect. They found in transgenic mice in which an insulin promoter drove oncogene expression, tumor formation occurred in β cells as expected, but they also found tumors in the liver and the DNES, including gut L and K cells (18). In the setting of rAAV, heterologous promoters show even greater leakiness as the retained AAV ITRs have intrinsic promoter/enhancer activity (19). Moreover, apart from β cells and related neuroendocrine cells, transduced hepatocytes appear to be one of the few cell-types that possess the potential for glucose- dependent proinsulin gene transcription, processing and insulin secretion (20). Example 6: Effect of fasting on hepatocyte transduction

To assess why only about 40% of the treated-STZ rats became euglycemic, we studied an additional nine STZ rats administered the well- tolerated dose of 2x 10¹¹ RIP-furlns particles. The results were again similar (Fig.3C), with only 3 out of 9 rats reaching euglycemia after a significant lag period (6 + 2 weeks). Furthermore, the time course (one representative rat, Fig.3F) of circulating insulin in the blood of these rats corresponded to the pattern we had previously observed (Fig.3D and E). In STZ rats that did not become euglycemic following vector administration, plasma insulin was measurable in the first 3 days following dosing (consistent with epithelial cell transduction), but fell subsequently to either non-detectable or very low levels (<0.5 μU/ml) throughout the monitoring period (Fig. 3G). Examination of the gut and liver of all animals by immunostaining (7), showed that the approximate number of transduced cells (>10⁷) in the gut and liver of euglycemic rats, as determined by cell counts on selected tissue sections, was much higher than those in the hyperglycemic rats (~10⁵). This result suggested that transduction efficiency was the critical factor for obtaining euglycemia, with a threshold effect likely. Consistent with this hypothesis, (human) proinsulin measurements (Mercodia AB, Sweden) in plasma samples taken 3 months following furlns vector administration showed levels of 14 ± 1.1 pmol/1 in the euglycemic rats, whereas the equivalently treated "non-responders" had levels of 4.1± 1.8 pmol/1. To investigate the cause for the poor transduction efficiency in significant cohorts of hyperglycemic rats, we then selected 10 rats weighing 280-300 grams, fasted them for 12 hours, and then analyzed the proximal intestine. The presence of food in the stomach and proximal intestine in ~60% of rats, consistent with the percentage of "non-responders", indicated that fasting for 12 hours was insufficient to empty the proximal intestine in many animals. The presence of intestinal contents and/or active digestion may have significantly reduced transduction efficiency, however this is only one mechanism to account for the inefficiency of transduction in the non- responders. Alternative explanations could relate to gut motility, neutralizing immunological responses, and AAV receptor polymorphisms in target cells in these outbred rats preventing effective uptake and transduction.

Example 7: Effect of peroral transduction of hepatocytes on the blood glucose levels in diabetic rats

In non-diabetic animals, pancreatic β cells are exposed to elevated glucose levels for minutes rather than hours after food uptake. We therefore determined whether transduction with RIP-furlns would lead to insulin expression in target cells that is capable of rapidly responding to an acute glucose challenge. In naive rats, plasma glucose increased transiently and rapidly dropped to normal after oral glucose dosing, while those of untreated control STZ diabetic rats increased markedly and remained at a plateau for the period ofthe experiment (5.5 hours). In contrast, in the STZ rats which had received the RIP-furlns and subsequently shown fasting euglycemia, plasma glucose increased sharply within 30 min ofthe oral administration of glucose solution, but dropped to normal by the end ofthe experiment (Fig.4A and B). Although, this glycemic curve is clearly not normal, the acute glucose tolerance is improved significantly compared to the STZ controls (p<0.0001 at 0, 2 and 6hour time point). Moreover, there was a significant increase (2.4 fold, p<0.05) in plasma human insulin in RIP-furlns treated STZ rats following this glucose challenge, with a peak reached at 30 minutes, and levels subsequently falling towards baseline values as euglycemia was reached at 2 hours (Fig.4B). These data suggest that the 412 bp rat insulin I promoter fragment retains some ofthe elements necessary for transcriptional regulation by glucose and also leads to transduction of cells which have some capacity for coupling peptide release to glucose intake. Recently, Leibiger and colleagues (8) demonstrated that insulin transcriptional regulation, which had previously been thought to require exposure to high glucose levels for many hours, was extremely rapid. Using the 412 bp RIP fragment coupled to a GFP transgene, these investigators showed that exposure to high glucose for an interval as brief as 15 minutes resulted in a significant increase in GFP fluorescence. Furthermore, some degree of both negative and positive regulation appeared likely in our study, as blood glucose levels in this group of rats remained within the range of 2.6- 12.5mM (Fig.3A-C). Example 8: Effect of peroral transduction of hepatocytes on the ketone levels in diabetic rats

To further characterize the therapeutic effect ofthe insulin expressing vectors, blood was taken from normal rats, untreated STZ rats, STZ rats at 3 days following administration of a luciferase-expressing vector (rAAV/RIP- luc), and from STZ rats at 3 days, 3 and 6 months following dosing with the RIP-furlns vector. Serum ketones were measured by spotting serum on a ketone test strip using a Bioscanner Test System (Polymer Technology Systems, Indiana, USA). The level of ketones in untreated diabetic rats (2 weeks post STZ) and in the RIP -luc STZ rats was double that of normal rats (P<0. 000001, ANOVA) (Fig.4C). However, in all RlP-furlns-treated STZ rats, ketone levels dropped to normal at each timepoint measured in both "responders", i.e. those rats that became euglycemic, as well as "non- responders", who remained hyperglycemic, consistent with the hypothesis that transduction efficiency was sufficient to prevent ketosis, but not to restore a euglycemic phenotype in these rats.

In summary, transgenic insulin gene expression in the liver and the gut of STZ rats treated with 1 or 2 xlO¹¹ RIP-furlns particles restored near normal glucose levels for at least 6 months. When treated rats were fed with a 20%) dextrose solution, circulating transgenic insulin levels increased and glucose levels dropped to normal within 6 hours. There was an inverse relationship between blood glucose and human insulin levels in the insulin vector-treated rats, and blood glucose and ketone levels remained at near normal physiological levels for the duration ofthe study. These data suggest that the transgenic insulin in the blood was at least to some extent processed and had biological activity. Moreover, there was sustained synthesis and the transgenic insulin had release kinetics approximating normal physiology. The levels of mature insulin obtained were much lower than that typically found in normal rats, this is consistent with only partial processing with some biological effect mediated through proinsulin in addition to the mature insulin.

The genetic constructs and the methods ofthe present invention, as described herein, have uses in gene therapy by way of oral administration of said constructs. In particular, the embodiments describing constructs comprising nucleotide sequences encoding insulin are useful in diabetes gene therapy or in management of blood glucose levels in a variety of glucose metabolic disorders.

Although the present invention has been described with reference to certain preferred embodiments and examples it will be understood by those skilled in the art that variations in keeping with the principles ofthe invention and the inventive concept are also within the scope ofthe present invention.

References

1. S. Deng, KL.Brayman, Mol. Med. Today 5, 400 (1999). 2. H.P. Moore, M.D. Walker, F. Lee, R.B. Kelly, Cell 35, 531 (1983).

3. L. Gros, L. Montoliu, E. Riu, L. Lebrigand, F. Bosch, Hum. Gene Ther. 8, 2249 (1997). 4. N. Welsh, Gene Ther. 7, 181 (2000).

5. H.C. Lee, S.J. Kim, K.S. Kim, H.C. Shin, J.W. Yoon, Nature 408, 483 (2000). 6. T.R. Flotte and B.J. Carter, Gene Ther. 2, 357 (1996).

7. M.J. During, et al., Nat. Med. 4, 1131 (1998).

8. B. Leibiger, et al., Proc. Natl. Acad. Sci. USA 95, 9307 (1998).

9. C. Herrmann, et al., Digestion, 56, 117 (1995).

10. T.M. Kolodka, M. Finegold, L. Moss, S.L., Woo, Proc. Natl. Acad. Sci. USA 92, 3293 (1995).

11. M.J. During et al., Science 287, 1553 (2000).

12. Grimm, D., et al., Human Gene Therapy, 9, 2745-2760 (1998).

13. D.J. Groskreutz, M.X. Sliwkowski, CM. Gorman, J. Biol. Chem. 269, 6241 (1994).

14. CH. Miao et al., Nat Genet. 19, 13 (1998)

15. J.N. Lozier, et al., Hum. Gene Ther. 8, 1481 (1997).

16. R.S. Fuller, A.J. Brake, J. Thorner, Science 246, 482 (1989). 17. P.M. Thule, J.Liu, L.S. Phillips, Gene Ther. 7, 205 (2000).

18. G. Rindi, et al, Am. J. Pαthol. 136, 1349 (1990).

19. R.P. Haberman, T.J. McCown, R.J. Samulski, J. Virol. 74, 8732 (2000).

20. A. Khan, Nature medicine, 553, 6 (2000)

Claims

THE CLAIMS DEFINING THE INVENTION ARE AS FOLLOWS:

1. Method of expressing a therapeutically effective amount of a biologically active protein or peptide in a hepatocyte comprising oral administration of a vector, or of a therapeutic agent comprising said vector, wherein the vector comprises a nucleotide sequence encoding the protein or peptide operably linked to a regulatory sequence capable of driving expression ofthe nucleotide sequence in the hepatocyte, and wherein the hepatocyte harbouring said vector secretes the expressed protein or peptide into the systemic circulation or locally.

2. Method of expressing a therapeutically effective amount of insulin in a hepatocyte comprising oral administration of a vector, or of a therapeutic agent comprising said vector, wherein the vector comprises a nucleotide sequence encoding insulin operably linked to a regulatory sequence capable of driving expression ofthe nucleotide sequence in the hepatocyte, and wherein the hepatocyte harbouring said vector secretes the expressed insulin into the systemic circulation or locally.

3. Method for improved expression of a therapeutically effective amount of a biologically active protein or peptide in a hepatocyte comprising oral administration of a vector, or of a therapeutic agent comprising said vector, to a subject in which such expression is desirable, wherein the vector comprises a nucleotide sequence encoding the protein or peptide operably linked to a regulatory sequence capable of driving expression ofthe nucleotide sequence in the hepatocyte, whereby the hepatocyte harbouring said vector secretes the expressed protein or peptide into the systemic circulation or locally and wherein the subject is fasted prior to oral administration for a time sufficient to achieve complete or near complete emptying of gastric and/or intestinal contents.

4. Method of expressing a therapeutically effective amount of insulin in a hepatocyte comprising oral administration of a vector, or of a therapeutic agent comprising said vector, to a subject in which such expression is desirable, wherein the vector comprises a nucleotide sequence encoding insulin operably linked to a regulatory sequence capable of driving expression ofthe nucleotide sequence in the hepatocyte, whereby the hepatocyte harbouring said vector secretes the expressed insulin into the systemic circulation or locally and wherein the subject is fasted prior to oral administration for a time sufficient to achieve complete or near complete emptying of gastric and/or intestinal contents.

5. Method of improving in vivo hepatocyte expression of an exogenous nucleotide sequence following oral administration to a subject of a vector comprising the sequence, said method comprising the step of fasting the subject before oral administration of said vector for a time sufficient to achieve complete or near complete emptying of gastric and/or intestinal contents.

6. Method of treating diabetes comprising oral administration to a subject requiring such treatment of a vector, or of a therapeutic agent comprising said vector, wherein the vector comprises a nucleotide sequence encoding insulin operably linked to a regulatory sequence capable of driving expression ofthe nucleotide sequence in the hepatocyte, and wherein the hepatocyte harbouring said vector , secretes the expressed insulin into the systemic circulation in the a mount sufficient to treat diabetes.

7. Method of regulating blood glucose level in a diabetic subject comprising oral administration to a subject requiring such treatment of a vector, or of a therapeutic agent comprising said viral vector, wherein the vector comprises a nucleotide sequence encoding insulin operably linked to a regulatory sequence capable of driving expression ofthe nucleotide sequence in the hepatocyte, and wherein the hepatocyte harbouring said vector secretes the expressed insulin into the systemic circulation in the a mount sufficient to regulate blood glucose levels.

8. Method according to any one ofthe preceding claims wherein the vector is a viral vector and the vector is selected from the group consisting of adeno-assocaited vector, an adenovirus vector, a lentivirus vector, a parvovirus vector and a herpes virus vector.

9. Method according to any one ofthe preceding claims wherein the regulatory sequence is a promoter and said promoter is selected from the group consisting of insulin promoter, rat insulin promoter, elongation factor 1 alpha promoter, glucagon promoter, glucokinase promoter and L-pyruvate kinase promoter

10. Method according to any one of the preceding claims wherein the protein or peptide is insulin.

11. Method according to claim 10, wherein insulin is expressed in, and secreted from, the hepatocyte for a period of 9 months.

12. Method according to any one of claims 2, 4, 7, 10 or 11 , wherein insulin is expressed in and secreted from the hepatocyte in response to elevated blood glucose levels.

13. Method according to claim 13, wherein insulin is secreted from the hepatocyte in the amount and for a time sufficient to normalise blood glucose levels.

14. Method according to any one ofthe preceding claims wherein the vector is administered in the amount of from about 0.5x10^U to about 5xlO^π .

15. Method according to claim 14, wherein the vector is administered in the amount of about 2x10¹¹.

16. Method substantially as herein described and with reference to any one or more ofthe Examples.