Hepatocyte Based Insulin Gene Therapy for Diabetes

Abstract

A method and vectors for controlling blood glucose levels in a mammal are disclosed. In one embodiment, the method comprises the steps of: treating the hepatocyte cells of a patient with a first, second or third vector, wherein the first vector comprises a promoter enhancer, glucose inducible regulatory elements, a liver-specific promoter, a gene encoding human insulin with modified peptidase and an albumin 3′UTR and lacks an HGH intron, wherein the second vector comprises an HGH intron, glucose inducible regulatory elements, a liver-specific promoter, a gene encoding human insulin with modified peptidase site and an albumin 3′UTR and lacks a promoter enhancer, wherein the third vector comprises an HGH intron, glucose inducible regulatory elements, a liver-specific promoter, a gene encoding human insulin with modified peptidase site, an albumin 3′UTR and a promoter enhancer and observing the patient's insulin levels, wherein the patient's insulin levels are controlled.

Claims

1. A method for obtaining glucose-regulated expression of insulin ex vivo in mammalian hepatocytes, wherein the method comprises delivering a first, second or third genetic vector for glucose-regulated synthesis of insulin into an isolated mammalian hepatocyte and wherein the insulin levels in the blood of the mammal stays within 0.5 μU-100 μU/ml and the blood glucose concentration of the mammal stays within 80-150 mg/dl for at least 10 days after transplant of the mammalian hepatocytes into a mammal, wherein the first vector comprises a promoter enhancer, one to five glucose inducible regulatory elements, a liver-specific promoter, a gene encoding insulin with modified peptidase sites and an albumin 3′UTR and lacks a human growth hormone (HGH) intron, wherein the second vector comprises a HGH intron, glucose inducible regulatory elements, a liver-specific promoter, a gene encoding insulin with modified peptidase sites and an albumin 3′UTR and lacks a promoter enhancer, wherein the third vector comprises a HGH intron, glucose inducible regulatory elements, a liver-specific promoter, a gene encoding insulin with modified peptidase sites, an albumin 3′UTR and a promoter enhancer, and wherein glucose-regulated expression of insulin occurs.

2. The method of claim 1 additionally comprising the step of transplanting the hepatocytes back into a mammal.

3. The method of claim 1, wherein the genetic vector is delivered by exposing the cells to a virus infective for the cells, wherein the virus comprises the genetic construct, and whereby at least a portion of the cells are infected by the virus under suitable conditions and at a sufficient multiplicity.

4. The method of claim 1 wherein the mammal is human.

5. The method of claim 1 wherein the insulin is human insulin.

6. A vector suitable for controlling blood glucose levels in a mammal, wherein the vector comprises a promoter enhancer, glucose inducible regulatory elements, a liver-specific promoter, a gene encoding insulin with modified peptidase sites and an albumin 3′UTR and lacks a HGH intron or wherein the vector comprises an HGH intron, glucose inducible regulatory elements, a liver-specific promoter, a gene encoding insulin with modified peptidase sites and an albumin 3′UTR and lacks a promoter enhancer or wherein the vector comprises an HGH intron, glucose inducible regulatory elements, a liver-specific promoter, a gene encoding human insulin with modified peptidase sites, an albumin 3′UTR and a promoter enhancer.

7. The vector of claim 6 wherein additionally comprising a VEGF-Enhancer.

8. The vector of claim 6 wherein the promoter is an albumin promoter.

9. The vector of claim 6 wherein the insulin is human insulin.

10. A method of controlling blood glucose levels in a mammal, comprising the steps of: treating a mammal with a first, second or third vector, wherein the first vector comprises a promoter enhancer, 1-5 glucose inducible regulatory elements, a liver-specific promoter, a gene encoding insulin with modified peptidase sites and an albumin 3′UTR and lacks a HGH intron and wherein the second vector comprises an HGH intron, glucose inducible regulatory elements, a liver-specific promoter, a gene encoding insulin with modified peptidase sites and an albumin 3′UTR and lacks a promoter enhancer, wherein the third vector comprises an HGH intron, glucose inducible regulatory elements, a liver-specific promoter, a gene encoding insulin with modified peptidase sites and an albumin 3′UTR and a promoter enhancer, and observing the mammal's insulin levels, wherein the insulin levels in the blood of the mammal stays within 0.5 μU-100 μU/ml and the blood glucose concentration of the mammal stays within 80-150 mg/dl for at least 10 days after transplant of the mammalian hepatocytes into a mammal.

11. The method of claim 1 wherein the vector is in a minicircle format.

12. The method of claim 10 wherein the vector is in a minicircle format.

13. The method of claim 10 wherein the mammal is human.

14. The method of claim 13 wherein the insulin is human insulin.

15. The method of claim 10 wherein the mammal's cholesterol level decreases after treatment.

16. The method of claim 10 wherein the mammal's triglyceride level decreases after treatment.

17. The method of claim 10 wherein the mammal is a cat.

18. The method of claim 10 wherein the mammal is a dog.

19. The method of claim 10 wherein the mammal is selected for the group consisting of hamsters, gerbils, rats, mice, rabbits, guinea pigs, chinchillas and ferrets.

20. The method of claim 10 wherein the mammal is a non-human mammal.

21. The method of claim 10 wherein the patient has a decrease in the plasma level of a compound selected from the group of AST, ALT, and alkaline phosphatase after treatment.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0023] FIG. 1 is a diagram comparing fasting blood glucose versus time in STZ-treated diabetic rats treated with various levels of TA-1M.

[0024] FIG. 2 is a graph comparing postprandial blood glucose levels versus time for TA-1 minicircle DNA treated STZ-treated diabetic rats.

[0025] FIG. 3 is a graph comparing fasting body weight versus time in TA-1 minicircle DNA treated STZ-treated diabetic rats.

[0026] FIG. 4 is a diagram of the TA1 expression cassette.

[0027] FIG. 5 is a diagram of the TA4 expression cassette.

[0028] FIG. 6 is the DNA sequence of the TA1 expression cassette.

[0029] FIG. 7 is the DNA sequence of the TA4 expression cassette.

[0030] FIG. 8 is a diagram of glucose levels versus time versus insulin levels during an intraperitoneal glucose tolerance test of TA1-treated diabetic rats.

[0031] FIG. 9 is a bar chart comparing ex vivo insulin production in hepatocytes treated with various insulin gene constructs.

[0032] FIG. 10 is a diagram of the TA2 expression cassette.

[0033] FIG. 11 is the DNA sequence of the TA2 expression cassette.

[0034] FIG. 12 is a diagram of blood glucose versus days in rats, fasted overnight and treated with TA1 and Ta4 constructs.

[0035] FIG. 13 is a diagram of body weight versus days in rats treated with TA1 gene construct.

[0036] FIG. 14 is a graph of body weight versus days on rats treated with TA1 and TA4 gene construct.

[0037] FIG. 15 is a table of human insulin levels in serum of diabetic rats treated with plasmid or minicircle DNA.

[0038] FIG. 16 is a graph of blood glucose levels versus time for rats experiencing a second treatment of TA1 minicircle DNA.

[0039] FIG. 17 is a table of glucose-dependent insulin production from human stem cells derived from hepatocytes.

[0040] FIG. 18 is a graph of body weight versus days for diabetic rats treated with TA1M, TA2M and TA3M.

DESCRIPTION OF THE INVENTION

In General

[0041] Due to a shortage of donor pancreata and the limited long-term success of islet transplants, alternatives for treating Type I diabetes (T1D) are needed. We have developed a gene therapy-based glucose regulated hepatic insulin production therapy that demonstrates great promise in treating T1D in experimental animals. Our approach is to apply insulin gene therapy to autologous native hepatocytes or stem cell-derived hepatocytes in an attempt to overcome the two critical shortcomings in treating T1D, which are the shortage of donor organs and the need for life-long use of immunosuppression in transplantation patients.

[0042] As the Examples below demonstrate, we examined novel DNA constructs for the ability to improve insulin production. For example, a novel insulin construct (TA1, described below) which contains the human growth hormone (HGH) intron, a translational enhancer, glucose inducible regulatory elements, albumin promoter, human insulin with modified peptidase sites, and the albumin 3′-UTR improved insulin production in cultured hepatocytes and diabetic rats and mice. TA1 resulted in a ˜25-fold increase in insulin production from isolated rat hepatocytes compared to our previously published insulin construct [Alam & Sollinger, Transplantation. 2002 Dec. 27; 74(12):1781-7].

[0043] In one aspect, the present invention is a novel DNA construct designed to improve insulin production in hepatocytes. Another aspect of the present invention is the creation of hepatocytes with improved insulin production. Another aspect of the present invention is a method of relieving the symptoms of Type I diabetes in a mammalian patient by modulating the production of insulin.

Constructs of the Present Invention

[0044] The Examples below demonstrate four insulin constructions containing various elements. The Examples demonstrate that three constructs (TA1, TA2 and TA4) were successful in providing an insulin gene therapy that provides tight control of insulin production. Therefore, the present invention encompasses three types of vector. The first vector type comprises a transcriptional enhancer, glucose inducible regulatory elements, a gene promoter, a translational enhancer, a gene encoding insulin with modified peptidase site and an albumin 3′ UTR and lacks an HGH intron. The second vector type comprises an HGH intron, glucose inducible regulatory elements, a gene promoter, a translational enhancer, a gene encoding insulin with modified peptidase site and an albumin 3′ UTR and lacks a transcriptional enhancer. The third vector type comprises all of the listed elements.

[0045] In another embodiment of the present invention, the constructs of the present invention consist essentially of the elements listed above. By “consist essentially of” we mean that a vector of the present invention will consist of the element described above and possibly other regulatory elements necessary for vector function. For example, plasmids and minicircle vectors may include sequences to facilitate the addition or removal of functional elements, such as restriction sites, or sequences necessary for the replication of the vector itself.

[0046] Applicants note that SEQ ID NO: 1, SEQ ID NO: 2 and SEQ ID NO: 3 (FIGS. 6, 7 and 11) are the entire nucleotide sequence of the TA1, TA2 and TA4 expression constructs, respectively. These listings do not include sequences that correspond to minicircle recombination sites, etc. For example, SEQ ID NOs: 1 and 2 do not include specific sequences used for recombination that are found within the commercial minicircle parental plasmid that flank the expression cassettes. Because all these sequences are part of commercial plasmids and readily available, the sequences are not included in the provided information.

[0047] SEQ ID NO: 1, SEQ ID NO: 2 and SEQ ID NO:3 include elements that are necessary for the present invention (for example, the elements listed below) and linking and nonessential sequences that are useful for cloning but may be substituted by many other sequences with similar functions.

TABLE-US-00001 Residue Residue Residue Location Location Location in SEQ ID in SEQ ID in SEQ ID NO: 1 NO: 2 NO: 3 (TA1) (TA4) (TA2) Alpha fetoprotein (AFP)  15-265 —  15-265 Enhancer Glucose Inducible 279-369  46-136 279-369 Regulatory Elements Albumin Promoter 380-706 147-474 380-707 Sequence HGH Intron — 475-754 708-987 VEGF Translational 707-870 755-918  988-1151 Enhancer Gene Encoding Human  871-1203  919-1251 1152-1484 Insulin with Modified B-C and C-A Peptide Junctions for Furin Compatibility Albumin 3′ UTR 1204-2077 1252-2125 1485-2358

[0048] More specifically, the vectors of the present invention comprise the following elements:

Promoter Enhancer

[0049] By “promoter enhancer,” we preferably mean the alpha-fetoprotein enhancer. The Examples below disclose the use of the alpha-fetoprotein enhancer. This element is designed to enhance transcription of the functionally linked gene sequence encoding a protein in liver cells. Alpha-fetoprotein enhancer increases the effectiveness of albumin promoter and increases the binding of RNA polymerase complex, thereby producing more mRNA, ultimately leading to an increase in protein production. The endogenous factors present in liver cells interact with alpha-fetoprotein enhancer region which activates the albumin and alpha-fetoprotein promoters during liver development and in fully developed liver. Because the effect of AFP enhancer is extinguished in fully developed liver cells through repression of its activity, the region associated with repression is not included in our AFP enhancer sequence, which allows the enhancer activity to persist in fully developed liver cells.

[0050] A suitable form of the AFP enhancer of the present invention is disclosed in Jin et al., Developmental Biology 336 (2009) 294-300. A specific sequence of the AFP enhancer can be found at residues 15-265 of SEQ ID NO: 1 (TA1 expression cassette).

[0051] In another embodiment of the invention, one would use other promoter enhancers suitable for use with a liver-specific promoter. Many normal promoters are quite large in size and contain multiple regions that modulate transcriptional activity as required for the existing physiological needs at a given time. Therefore, selection of an appropriate promoter enhancer is context dependent. It must work with the promoter in question. If empirical determinations validate functional efficacy of enhancers from other promoters, in conjunction with the liver-specific promoter used, appropriate modifications in insulin expression cassettes can be made to achieve the desired results. Currently, according to the Cold Spring Harbor Laboratory database, there are approximately 400 known regulatory regions and elements that function in liver cells.

Translational Enhancer

[0052] By “translational enhancer,” we preferably mean the VEGF translational enhancer. The Examples below disclose the use of the VEGF translational enhancer. This element is designed to enhance translation of the functionally linked protein encoding sequence. The VEGF translational enhancer acts as a ribosomal entry site; it increases the effectiveness of the translation process. Thus, its presence causes a larger amount of insulin protein production from a given amount of insulin mRNA.

[0053] A specific sequence of the VEGF translational enhancer can be found at residues 707-870 of SEQ ID NO: 1 (TA1 expression cassette) and residues 755-918 of SEQ ID NO: 2 (TA4 expression cassette).

Glucose Inducible Regulatory Elements (GIREs)

[0054] The vector of the present invention requires 1-5 GIREs, preferably 2-4 GIREs, most preferably 3 GIRES. One may find the sequence for suitable GIREs at residues 279-369 in SEQ ID NO: 1 and 46-136 in SEQ ID NO: 2. Suitable GIREs are also described below in the Examples and may also be found at U.S. Pat. Nos. 7,425,443 and 6,933,133, both incorporated by reference.

[0055] As used herein, a “glucose inducible regulatory element” (GIRE) refers to a polynucleotide sequence containing at least one pair of perfect CACGTG motifs, each member of the pair separated from the other by a sequence of five base pairs. A “glucose responsive regulatory module” contains one or more GIREs. In one example, the regulatory elements were inserted 5′ of the 5′ untranslated region of the human proinsulin gene and then cloned into an adenovirus vector which was used to transfect hepatocytes. As the Examples below demonstrate, the GIREs provide transcriptional regulation of insulin mRNA in hepatocytes in response to physiologically relevant glucose concentrations.

Promoter Sequence

[0056] The constructs of the present invention also involve the use of a gene promoter, preferably an albumin promoter. The albumin promoter is a hepatocyte (liver) specific promoter and is used to ensure that production of insulin is restricted only to liver cells. Therefore, if some of the insulin gene construct ends up in organs other than liver, the construct will not be expressed. Additionally, various components and mechanisms necessary to confer glucose responsiveness to insulin expression using gene constructs of the present invention are endogenous to liver cells. As illustrated in Examples herein, the rat albumin promoter (184 bp), (Heard et al., Determinant of rat albumin promoter tissue specificity analyzed by an improved transient expression system. Mol Cell Biol 1987; 7: 2425) was generated by PCR using rat genomic DNA template, as described previously (Alam et al., Glucose-regulated insulin production in hepatocytes. Transplantation 2002; 74:1781). The use of the rat albumin promoter sequence in the example is provided for illustrative purposes only. Constructs containing an albumin promoter from other species, such as humans, are expected to confer similar properties to the constructs.

[0057] One may obtain an albumin promoter by use of primers and PCR amplification after examination of SEQ ID NOs: 1 and 2. The promoter sequence is found at residues 380-706 of SEQ ID NO: 1 and 147-474 of SEQ ID NO: 2 and 380-707 of SEQ ID NO:3.

[0058] In principle, any constitutively active liver cell specific promoter capable of sustained moderate to high level transcription can be substituted for albumin promoter. An example of such a promoter is alpha 1-antitrypsin inhibitor (Hafenrichter D G et al. Blood 1994; 84, 3394-404). Currently, according to the Cold Spring Harbor Laboratory database there are approximately 300 known liver specific promoters.

Gene Encoding Insulin with Modified Peptidase Site

[0059] The vectors of the present invention comprise a gene encoding insulin, preferably human insulin or non-human mammalian insulin, with a modified peptidase site. The insulin genes of the present invention are also disclosed in SEQ ID NO: 1 at residue 871-1203 and SEQ ID NO: 2 at residue 919-1251.

[0060] In one aspect of the invention, one may wish to treat non-human animals. To ensure that no immune reaction to insulin occurs when diabetic animals, such as cats or dogs, are treated using insulin gene therapy, one would use species-specific insulin in minicircle DNA for treating animals. For example, one would use published sequences of insulin for cats and dogs (Kwok et al. 1983 J Biol Chem 258 2357-2363) to generate 3′ and 5′ primers to amplify the coding sequence of insulin from cDNA preparations made from isolated pancreatic RNA from respective species by standard molecular biology techniques. Alternatively, the coding sequence can also be chemically synthesized. Similarly one may wish to substitute insulin sequences from other animals when treating those animals. These sequences are readily available.

[0061] Human insulin cDNA was modified at two junctions of proinsulin where proteolytic processing and maturation of insulin occurs by specific enzymes residing in beta cells but absent in liver cells. The modification at the B and C peptide of human insulin from KTRR to RTKR and at the C and A peptide from LQKR to RQKR makes the two insulin junctions compatible with cleavage specificity of endogenous protease, furin, of liver cells. These modifications are described in the following publications: Simonson G D, Groskreutz C M, Gorman C M, et al. Synthesis and processing of genetically modified human proinsulin by rat myoblast primary cultures. Hum Gene Ther 1996; 7: 71; Groskreutz D J, Sliwkowski M X, Gorman C M. Genetically engineered pro-insulin constitutively processed and secreted as mature, active insulin. J Biol Chem 1994; 269: 6241.

[0062] Regarding the use of non-human insulin, all modifications to the sequence of preproinsulin will be similar in nature to that described for human insulin, wherein the recognition/processing sites for peptidases found in ß-cells (and neuroendocrine tissue) will be changed to sites that can be processed by commonly found proteases in liver (such as furin) and other cells. There are some minor sequence differences in insulin from various species but the key point is to retain the authentic sequence of mature insulin for the given species after processing.

[0063] The purpose of the specific mutation is to change the amino acid sequence in such a way that proteolytic processing is possible by commonly found furin. There are multiple codons for several of the amino acids. Theoretically, one can alter the DNA sequence by using an alternative codon but still produce the same polypeptide.

[0064] These modifications have been successfully used by Applicants in a published report (Alam T, Sollinger H W. Glucose-regulated insulin production in hepatocytes. Transplantation 2002; 74:1781). An unmodified insulin gene will produce unprocessed proinsulin because the specific enzymes necessary for the maturation by proteolytic processing are absent in liver cells. Proinsulin has minimal biological activity of approximately 100 fold less than the mature insulin.

[0065] One may obtain a modified insulin gene by use of primers and PCR amplification with knowledge of the insulin gene in SEQ ID NOs: 1, 2 and 3.

Albumin 3′ UTR

[0066] The albumin 3′ UTR is known to contribute to longevity of the albumin mRNA in hepatocytes. This sequence was obtained from an expression vector plasmid from Mirus (pMIR0375) but this sequence can also be amplified by PCR using reverse transcribed mRNA from liver. The albumin 3′ UTR sequence is disclosed in SEQ ID NO: 1 at 1204-2077 and at SEQ ID NO: 2 at residues 1252-2125.

HGH Intron

[0067] Two of the constructs of the present invention, TA4 and TA2, comprise the HGH intron. The HGH intron is known to add to the efficiency of mRNA processing and helps in yielding quantitatively more mRNA. There are several other introns, such as beta-globin, that serve similar function to a varying degree. However, the HGH intron is known to function well and is preferred. The HGH intron may be amplified by PCR from the commercially available plasmid pAAV-LacZ [Stratagene, La Jolla, Calif.]. The sequence can also be readily amplified by PCR using genomic DNA as the template. The sequence of the HGH intron is disclosed at residues 475-754 of SEQ ID NO: 2 and 708-987 of SEQ ID NO:3.

Minicircle Embodiment

[0068] Optionally, the vector of the present invention is in the “minicircle DNA” format. This is a vector that is virtually devoid of all DNA sequences that are unrelated to expression of insulin. The original minicircle DNA production vector was obtained from the laboratory of Mark Kay, described in the following publications: Chen, Z Y, He, C Y, Ehrhardt, A and Kay, M A (2003). Minicircle DNA vectors devoid of bacterial DNA result in persistent and high-level transgene expression in vivo. Mol Ther 8: 495-500, and Chen Z Y, He C Y, Kay M A (2005) Improved production and purification of minicircle DNA vector free of plasmid bacterial sequence and capable of persistent transgene expression in vivo. Human Gene Ther, 16:126. A newly revised method to easily produce minicircle DNA was published recently (Kay M A, He C, Chen Z. A robust system for production of minicircle DNA vectors. 2010. Nature Biotech 28, 1287). The vector and the E. coli needed to produce the minicircle are commercially available from System Biosciences (SBI), Mountain View, Calif. (systembio.com), and are currently used by us.

[0069] The present invention of insulin expression constructs conforms to the generally accepted and proven placement scheme of various elements in relation to each other. Thus, the gene expression constructs of the present invention are comprised of AFP enhancer-conditional inducer-promoter-intron 1-gene-intron 2-termination/5′ UTR. In our Examples, the GIREs are the conditional inducers and there is a translational enhancer inserted after the HGH intron and before the modified insulin gene. After the insulin gene, the second intron for efficient mRNA processing is from albumin followed by the 3′ UTR of albumin.

[0070] FIGS. 4, 5 and 10 disclose preferred placement of elements in gene constructs of the present invention.

Method of the Present Invention

[0071] In one aspect, the present invention is a method of controlling blood glucose levels in a mammalian patient (preferably a human or non-human mammal), comprising the steps of treating a mammal with hepatocytes that have been modified with a first, second or third vector, as described above. The first vector comprises promoter enhancer, glucose inducible regulatory elements, a liver-specific promoter, a gene encoding insulin with modified peptidase site and an albumin 3′ UTR and lacks an HGH intron. The second vector comprises an HGH intron, glucose inducible regulatory elements, a liver-specific promoter, a gene encoding insulin with modified peptidase site and the albumin 3′ UTR and lacks promoter enhancer. The third vector comprises an HGH intron, glucose inducible regulatory elements, a liver-specific promoter, a gene encoding insulin with modified peptidase site, the albumin 3′ UTR and promoter enhancer.

[0072] One would observe the mammal's blood glucose and insulin levels after vector treatment and note that the mammal's blood glucose or insulin levels are controlled and normal.

[0073] To test our insulin expression constructs ex vivo in hepatocytes, TA1, TA2 and TA4 insulin constructs were cloned in adenovirus vector, as described earlier (Alam T, Sollinger H W. Glucose-regulated insulin production in hepatocytes. Transplantation 2002; 74:1781.) Freshly isolated normal rat hepatocytes were plated on collagen coated cell culture plates and transfected with adenovirus containing the insulin gene construct. These cells were then exposed to low (3.5 mM), normal (5.6 mM) and high (27.5 mM) concentrations of glucose. Aliquots of medium were drawn at various time intervals and insulin present in the culture medium was quantitated by ELISA. Results showed that hepatocytes transfected with each insulin construct produced insulin in a glucose concentration dependent manner (FIG. 9). At the high concentration of glucose, the amount of insulin production was 4-10× higher than at the low concentration of glucose.

[0074] By “controlled,” we mean that the method of the present invention is preferably characterized by tight control of glucose regulation. The tight control refers to the empirical observation of glucose regulation itself. In non-diabetic individuals, the blood glucose returns to normal at 2 hr post meal. Before the present invention, one would have anticipated that following the correction of hyperglycemia in a mammal in response to elevation in blood glucose levels, the preformed insulin mRNA would remain for a while and continue producing insulin. Depending on how long such a condition persists, one would expect that the mammal would then become hypoglycemic. However, our results showed that the insulin levels in serum increased soon after the increase in blood glucose levels, as we had anticipated, but the insulin levels did not stay high for too long and followed the blood glucose level curve with a delay of about 15-30 min (See FIG. 8).

[0075] Typically, the present invention provides that insulin levels will stay within 0.5 μU-100 μU/ml. (This is comparable to the maximum amount of insulin that is released from normal islets under hyperglycemic challenge of approximately 100 μU/ml.).

[0076] Typically, blood glucose concentration will stay within 80-150 mg/dl after treatment. The high end relates to a temporary rise soon after having a meal. If glucose concentration does rise above 150, the level does not stay at that level for beyond a short period (30-60 min).

[0077] Poorly controlled diabetes causes hyperlipidemia and the severity of hyperlipidemia is dependent on the degree of hyperglycemia. Liver function tests are performed for two reasons. Severe diabetes is associated with a degree of systemic inflammatory responses, including elevation in serum levels of some liver enzymes. Our data provide evidence that following insulin gene therapy, a correction in serum levels of liver enzymes is apparent. Secondly, the hydrodynamic deliver procedure is known to cause a transient stress to liver but this damage is short-lived. Our data support these findings and assert that there is no long-term risk associated with gene-therapy in the context of liver function. In fact, the therapy normalizes the liver function, as evidenced by the albumin production.

[0078] In one embodiment of the present invention, the lipid and/or liver enzyme profile of the treated animal is corrected. In one embodiment of the present invention, the animal will have a lipid/enzymes panel wherein the plasma lipid or liver enzymes concentration is equivalent to or less than a normal control. By “normal control” we mean an animal who is not diabetic. By “lipid/enzymes panel” we mean either a plasma triglycerides measurement, an alanine transaminase (AST), an aspartate transaminase (ALT) or plasma albumin measurement. In another embodiment of the invention, we would expect to see a drop in cholesterol (mg/dl) of at least 20% compared to a diabetic control. Based on a conservative estimate, approximately 1 week may be sufficient for a substantial correction or normalization of hyperlipidemia.

[0079] The method of the present invention involves the treatment of a mammal, preferably a human patient or non-human mammal, with the vectors of the present invention.

[0080] Introduction of our insulin expression constructs into liver cells can be achieved either in vivo or ex vivo. In the first method, the gene construct will be introduced either using a minicircle DNA by hydrodynamic method described herein, injecting the condensed minicircle DNA nanoparticles as such, or after coating nanoparticles with compounds that are known to target liver cells. Alternatively, liver cells will be harvested from the patient through a biopsy, expanded in cell culture, and transfected with a construct of the present invention using minicircle or safe viral vectors such as adeno-associated virus (AAV) (already used in many clinical trials).

[0081] The transfected cells will be tested for their ability to produce insulin and modulate the quantity of production of insulin in response to changes in concentrations of glucose. Appropriate number of cells to provide necessary amount of insulin (assisted with information from ex vivo measurements) will be transplanted into the liver of the mammal via radiological and ultrasound guidance. The ex vivo method allows for lower vector load as well as for targeted delivery of gene.

[0082] Incorporating scaffold/matrix-attachment regions (S/MARs) that serve as constitutively active anchors for the nuclear scaffold (Wong S P et al. Gene Ther. 2011; 18(1):82-7; Heng H H et al. J Cell Sci. 2004; 117(Pt 7):999-1008; Lufino M M et al., Nucleic Acids Res. 2007; 35(15):e98. PMCID: 1976449) will likely increase the survival of TA1minicircle (TA1m)-containing S/MARs (TA1m-S/MAR) or other minicircles due to the association of SARs with the nuclear scaffold and may also reduce gene silencing (Wong S P et al. Gene Ther. 2011; 18(1):82-7).

[0083] The Examples below describe the delivery of DNA into rat liver by hydrodynamic procedures (Zhang G. et al., Methods Mol Biol 245:251-264, 2004; Zhang G. et al., Hum Gene Ther 8:1763-1772, 1997). The delivery of the DNA into mammalian liver will typically not be by the hydrodynamic method described in the Examples. Other alternatives, such as nano-particles of condensed DNA that do not require a large volume and high pressure, will be employed. In a preferred embodiment, suitable glucose regulation will last at least 2-4 weeks after treatment with the constructs.

[0084] Hydrodynamic venous delivery of naked plasmid DNA has led to successful gene uptake and transduction of liver cells (Sebestyen M G et al. Hum Gene Ther. 2007; 18(3):269-85. PMCID: 2268901; Wooddell C I et al. J Gene Med. 2008; 10(5):551-63; Wooddell C I et al. Hum Gene Ther. 2011; 22(7):889-903. PMCID: 3135275). This method relies on a rapid, high-volume intravenous injection (˜10% of body weight). However, with hydrodynamic venous delivery, much of the DNA is absorbed in the venous system, particularly the lungs, prior to its delivery to the liver. As an alternative, one may inject the vectors into the arterial system, such as the femoral artery, which has a systolic blood pressure of approximately 120 mm/Hg. The remaining vector will reach the liver via the portal vein at 10-12 cm.Math.H.sub.2O. (Blood pressure is usually presented in mm of Mercury (mm/Hg). Venous blood pressure is lower and sometimes a unit based on cm of water (cm/H.sub.2O) is used.) Overall, intra-arterial injection should greatly enhance TA1m constructs uptake compared to hydrodynamic venous delivery.

[0085] Injection into the femoral artery should dramatically increase vector liver uptake and transduction of hepatocytes. We expect that treated animals will maintain normoglycemia for more than 1 month, even when fed ad libitum.

EXAMPLES

Example 1: Generation of Constructs

[0086] Human insulin based gene constructs containing various elements to modulate expression were generated with the aim of producing biologically active insulin in response to changes in glucose levels.

Four insulin constructs designated TA1, TA2, TA3, and TA4 contain various elements in order shown below:

TABLE-US-00002 AFP- Alb- HGH- Human Alb 3′- Enhancer 3GIREs Promoter intron Insulin UTR TA1 Y Y Y N Y Y TA2 Y Y Y Y Y Y TA3 N Y Y N Y Y TA4 N Y Y Y Y Y
AFP-Enhancer: Alphafetoprotein enhancer was used from Mirus vector pMIR0375.
3GIREs: 3 units of glucose inducible regulatory elements are connected in tandem; the sequences are based on S14. S14 is a glucose responsive transcriptional enhancer. The elements responsible for glucose-dependent transcriptional enhancement have been identified in published work: Shih H, Towle H C. Definition of the carbohydrate response element of the rat S14 gene: Context of the CACGTG motif determines the specificity of carbohydrate regulation. J Biol Chem 1994; 269: 9380.
Alb-Promoter: Albumin promoter (Albumin promoter was amplified by PCR from rat genomic DNA, as described in U.S. Pat. Nos. 6,352,857, 6,933,133, 7,425,443 and the following publication: Alam T, Sollinger H W. Glucose-regulated insulin production in hepatocytes. Transplantation 2002; 74:1781)
HGH-intron: Human growth hormone intron was amplified by PCR from a commercially available plasmid pAAV-LacZ (Stratagene).
Human Insulin: Human insulin cDNA sequence was modified at B-C and C-A junction for furin cleavage compatibility so that liver cells are able to process preproinsulin to functional insulin.
Alb 3′-UTR: Albumin 3′-untranslated region was used from Mirus vector pMIR0375. It can also be amplified by PCR using reverse transcribed mRNA from liver.

[0087] Insulin expression constructs described above were incorporated into replication-defective adenovirus for transient expression for initial testing purposes. Insulin expression in rat hepatocytes, ex vivo, was glucose responsive and each construct yielded significantly higher amount of insulin (4-12 fold improvement over the previously described constructs), Alam & Sollinger, Transplantation. 2002 Dec. 27; 74(12):1781-7.

[0088] TA2 and TA3 were also tested in ex vivo insulin production. (See FIG. 9).

[0089] Initially we incorporated TA1 in an adenovirus vector to test its ability to control hyperglycemia in rats that were rendered diabetic by streptozotocin (STZ) treatment. Results showed that such a treatment fully corrected fasting hyperglycemia, restored the weight loss caused by diabetes to normal rate of weight gain and significantly reduced postprandial hyperglycemia.

[0090] The adenovirus vectors containing TA1-TA4 were tested individually for their ability to correct diabetic hyperglycemia in STZ-rats. Results showed a full correction of fasting hyperglycemia and a partial correction of postprandial hyperglycemia. The benefit of a single gene therapy treatment on overall metabolism and preventing body weight loss lasted well beyond the time of full correction of fasting hyperglycemia. During the period of full correction of fasting hyperglycemia, the rate of weight gain in diabetic rats treated with our insulin gene constructs was indistinguishable from the normal controls.

[0091] To improve efficacy of gene therapy through better gene expression by increasing levels and duration of insulin expression, minicircles of DNA containing only the gene expression constructs were produced. All of the above insulin gene expression constructs were cloned into a plasmid vector p2ϕC31 as previously described (Chen et al., 2003, Mol Ther, 8, 495-500; Chen et al., 2005, Human Gene Ther, 16, 126-131). This published method of Chen et al. was substantially modified to improve purity of minicircle DNA as described below.

[0092] Encouraged by results from transient expression afforded by adenovirus vector, we generated a plasmid vector, known as “minicircle DNA,” that is virtually devoid of all DNA sequences that are unrelated to expression of insulin. TA1 minicircle was introduced into the livers of rats via an established hydrodynamic procedure (Zhang G. et al., Methods Mol Biol 245:251-264, 2004; Zhang G. et al., Hum Gene Ther 8:1763-1772, 1997). Results obtained from STZ-diabetic rats treated with TA1 minicircle DNA show a full correction of hyperglycemia among ad lib fed animals (FIG. 1) in addition to restoration of rate of weight gain to normal (FIG. 2) and correction of fasting hyperglycemia.

[0093] The TA1-minicircle-treated diabetic rats were subjected to a glucose tolerance test by intraperitoneal injections of 4 gm/kg glucose. Results from these experiments bore a marked similarity to observations from normal control rats. The peak of elevated blood glucose levels appeared at 15 min post injection and hyperglycemia dissipated in about 60 min (FIG. 8). The time to correct hyperglycemia induced by 4 gm/kg glucose IP injection is similar to normal animals. An increased insulin output in response to elevated levels of glucose closely follows the rise in glucose levels and insulin production declines as the level of glucose progressively reduces.

[0094] To confirm that insulin production was glucose-dependent, we measured human insulin levels in plasma at 30 min time intervals and found that the human insulin levels peaked at 30 min and declined relatively quickly, essentially following the blood glucose profile with a 15 min delay. Thus, there was an approximately 15 min lag between the profiles of blood glucose and insulin levels.

[0095] Given the nature of glucose-induced transcription of insulin mRNA that gives rise to circulating insulin, after achieving euglycemia, continued presence of insulin mRNA could have caused a sustained secretion of high levels of insulin until the mRNA was degraded. Reduction in insulin levels in only ˜60 min to the levels observed in fasting animals prior to glucose injections, is an unexpected, albeit very desirable, result.

[0096] Modifications to the published minicircle DNA production method were useful and necessary to obtain pure minicircle DNA that was free of detectable unprocessed or partially processed minicircle DNA. These modifications involved elimination of a 2 hr incubation step, claimed to be necessary for in vivo digestion of the DNA circle that consists of the unneeded sequences from the parental plasmid that were eliminated from the minicircle containing the gene of interest. In practice, this step was only partially effective.

[0097] In our procedure, elimination of this 2 hr incubation step caused no perceptible change in quality or quantity of recovered DNA and the final product was comprised of a mixture of minicircle DNA and the parental unprocessed plasmid DNA as well as partially processed plasmid DNA. The mixture of DNA thus produced was treated, ex vivo, with a restriction enzyme that could cut the parental plasmid but not the minicircle DNA containing our insulin gene constructs. The product of this reaction was purified by CsCl equilibrium density gradient to separate the circular DNA from linear DNA.

[0098] The TA1 insulin minicircle DNA was tested for its ability to correct diabetic hyperglycemia in STZ-treated diabetic rats. Groups of rats were rendered diabetic by intravenous streptozotocin injections (100 mg/kg). The TA1 insulin minicircle DNA was injected via tail vein into diabetic rats according to a previously published method (hydrodynamic delivery method described by J. Wolff group, Zhang G. et al., Methods Mol Biol 245:251-264, 2004; Zhang G. et al., Hum Gene Ther 8:1763-1772, 1997). Four groups of diabetic rats were injected with indicated amounts of TA1 minicircle DNA (1.0 μg, 0.75 μg, 0.5 μg, and 0.025 μg per gm body weight). Results are shown in FIGS. 1, 2, and 3.

[0099] This is the first time we have been able to fully correct blood glucose levels in diabetic rats fed ad libitum (FIG. 2) by insulin gene therapy. This treatment fully restored rate of weight gain in diabetic rats (FIG. 3).

Example 2: Creation of Adenovirus Constructs

[0100] Referring to FIG. 12, insulin constructs TA1 and TA4 were created in adenovirus and equal plaque forming units injected into rat livers, as indicated. TA1 and TA4 were used under identical conditions. Both were able to fully correct fasting hyperglycemia, as shown in FIG. 12. However, the vectors did not fully correct blood glucose levels in rats fed ad libitum.

[0101] Referring to FIG. 13, TA1 in adenovirus was injected into rat livers as indicated. Note that the TA1 treated rats regained the weight initially lost due to diabetes and maintained a higher weight than the STZ diabetic rats.

[0102] Referring to FIG. 14, TA1 or TA4 in adenovirus was injected into livers of diabetic rat groups, as indicated. Note that both TA1 treated and TA4 treated rats had maintained higher body weight than STZ diabetic rats. FIG. 14 shows a subset of data points from FIG. 13, representing 0-48 days. The information on rate of weight gain in various groups of rats (shown in the inset box on the top left corner) is derived from data obtained 24 days post treatment (d8-d31). The rate of weight gain in diabetic rats treated with TA1 in adenovirus is equal to normal control rats, whereas the diabetic untreated rats experienced a net loss of weight on a daily basis. FIG. 13 has many more data points, and the early days occupy only a small portion of the area, and therefore the degree of correction in weight gain may be somewhat difficult to appreciate. FIG. 14 shows this aspect more clearly.

Example 3: Evaluation of Human Insulin in Serum of Diabetic Rats

[0103] Referring to FIG. 15, a comparison of human insulin in rat serum shows that animals treated with TA1 minicircle DNA produce larger amounts of insulin compared to other plasmid vectors (pTED110, TA1/pENTR). The plasmid pENTR contains an ampicillin resistance gene which has been replaced by a Kanamycin resistance gene in pTED110, and S/MAR has also been added to it; both have the TA1 insulin construct. The TA1 minicircle containing S/MAR produced the most insulin in vivo. The Ultrasensitive Human Insulin ELISA (Mercodia, Inc) has a detection limit of 0.15-20 mU/L, as advertised. The table in FIG. 15 shows the relative in vivo effectiveness of various vectors used for insulin gene therapy. All insulin vectors contained the TA1 expression cassette. pENTR is a commercial plasmid containing an ampicillin resistance gene. The pTED is our modified plasmid where we replaced the ampicillin resistance gene with kanamycin resistance gene to increase the in vivo survival of the vector in non-dividing cells.

[0104] We also added S/MAR to pTED110 to increase survival of vector in dividing cells and to some degree, increase the overall expression. The data are in agreement with our vector design expectations.

[0105] Finally, when extraneous sequences were eliminated and insulin gene constructs were used as minicircle DNA molecules, the expression levels of insulin were significantly increased, more so when S/MAR was included in the minicircle. In all four sets of experiments, the molar equivalence of TA1 was maintained at a constant level.

[0106] The table (FIG. 15) in conjunction with data obtained from adenovirus mediated transduction ex vivo (FIG. 9) and in vivo (FIG. 12 and FIG. 13, FIG. 14, and FIG. 17) provides the proof that the insulin expression cassettes are able to produce glucose-dependent insulin, as intended. However, the magnitude and longevity of insulin expression does depend on the vector employed to deliver the insulin expression cassette(s). We expect that these insulin expression cassettes will be readily adaptable to take advantage of new vectors developed to have better characteristics for gene therapy, such as ease of delivery, expression level, and longevity of expression

Example 4: Second Treatment with TA1M

[0107] Referring to FIG. 16, a second treatment with TA1 minicircle DNA corrects the gradually elevating fasting blood glucose to normal level.

[0108] FIG. 1 discloses data that are relevant to the information presented in FIG. 16. As the information in two figures comes from different studies, individual data points are not identical, but the trend is the same. One should refer to the curve corresponding to use of 1.0 μg/gm TA1 minicircle DNA in FIG. 1 for comparison. FIG. 16 shows that it appears possible to correct hyperglycemia by a second treatment as the effect of the first treatment diminishes.

Example 5: Hepatocytes Derived from Human Stem Cells

[0109] Referring to FIG. 17, hepatocytes were derived from human embryonic stem (hES) cells and human adult induced pluripotent (iPS) cells as described in Si-Tayeb et al 2010. Hepatology 51(1):297-305: PMCID: 2946078 and were transduced with TA1 in adenovirus, ex vivo. Three cell culture plates were kept in medium with low glucose (3.5 mM), and three plates were kept in high glucose (27.5 mM). Freshly prepared normal rat hepatocytes were treated similarly. Flasks containing stem cell derived hepatocytes were confluent with cells, whereas primary rat hepatocyte plates were ˜60% confluent.

[0110] Results (FIG. 17) show a robust production of glucose-dependent insulin in embryonic and induced pluripotent-cell-derived hepatocytes.

Example 6: Examination of Rat Weight after Treatment with Minicircle Vectors

[0111] Referring to FIG. 18, groups of diabetic rats (minimum number of rats in a group=5) were injected with the indicated insulin minicircle DNA, TA1, TA2, or TA3, via tail veins. Fasting body weights (Mean±S.D) of rats are shown. Normalization in rate of weight gain is similar when TA1m or TA2m was used, whereas TA3m was less effective.

Example 7: Evaluation of Lipid Profiles

[0112] Two groups of rats were rendered diabetic by intravenous streptozotocin treatment. One group of diabetic rats (n=5) was treated with 1 μg TA1 minicircle DNA/gm body weight of animal. The second group of diabetic rats was used as an untreated control. A third group of normal rats was included as age matched healthy controls. Blood was drawn from each experimental animal after 10 days, and plasma was analyzed for lipid contents and various markers of liver damage and hepatic function, as shown in the table below:

TABLE-US-00003 Alka- Tri- Aspartate Alanine line Plasma glyce- Choles- Transa- Transa- Pos- Albu- Animal rides terol minase minase phatase min Groups (mg/dl) (mg/dl) (U/L) (U/L) (U/L) (g/dl) TA1m  53 ± 34 141 ± 15 302 ± 33  77 ± 40 210 ± 95 3.4 ± 0.2 Treated Diabetic 704 ± 313 191 ± 36 617 ± 349 152 ± 75 423 ± 73 2.5 ± 0.2 Normal 100 ± 14 129 ± 6 504 ± 100 106 ± 16 172 ± 55 3.4 ± 0.1 Control

[0113] TA1m treatment corrected all deficiencies caused by the uncontrolled diabetes. Thus, the levels of cholesterol and triglyceride in plasma of treated rats were reversed to normal levels. Likewise liver function markers showed an improvement and reduced levels of albumin in diabetic rats returned to normal.