WO2006032952A1 - Therapeutic random oligonucleotides from genomic dna - Google Patents

Abstract

This invention provides methods to produce bulk oligonucleotides of random sequence over a wide range of sizes from biologic genomic DNA. The invention further provides how to monitor and regulate the degree of degradation of a DNA sample. It is further described how degraded genomic DNA can be used to suppress proliferation of cultured human cells, and can additionally induce apoptosis in sensitive cells, potentially making it an effective antiproliferative therapy.

Description

Description

Therapeutic Random Oligonucleotides from Genomic DNA

Technical Field

[1] This invention relates generally to the field of oligonucleotide-based medical therapies. Specifically, it is a reduction to practice of concepts presented in PCT/ US03/20696 which demonstrates that synthetic random sequence oligonucleotides might be used to treat cancer and other proliferative disorders in humans. This invention describes different methods to produce from a biologic source of DNA random sequence oligonucleotides that also can inhibit cellular proliferation. This invention further creates random sequence oligonucleotides much more economically and as random sizes that appear to be more effective at growth suppression and in addition to induce apoptosis in proliferating cells.

Background Art

[2] Two hallmarks of cancer cell behavior are proliferation and unregulated expression of a plethora of normally quiescent genes ( Singh H. Sekinger EA. Gross DS. Chromatin and cancer: causes and consequences.Jo«r«α/ of Cellular Biochemistry - Suppl 35:61-8, 2000 , Hanash S. Integrated global profiling of cancer .Nature Reviews. Cancer. 4(8):638-44, 2004 Aug , Leonhardt H. Cardoso MC. DNA methylation, nuclear structure, gene expression and cancer. Journal of Cellular Biochemistry - Suppl 35:78-83, 2000 ). Both attributes are controlled at the level of DNA by DNA replication and transcription. We tested random-sequence single-stranded DNA oligonucleotides that could inhibit excessive cell proliferation by several different mechanisms. One mechanism of action that would affect DNA replication is to substitute for RNA primers that are essential for formation of Okazaki fragments along the 5'->3' strand of replicating DNA and stabilization of the replication fork ( Lehninger Principles of Biochemistry, By Albert L. Lehninger / W H Freeman & Co / April 2004 / 0716743396). Another mechanism of action would be to bind com¬ petitively to replication protein A (RPA, or eukaryotic single-stranded DNA binding protein), which could inhibit DNA repair or induce apoptosis ( Thoma BS. Vasquez KM. Critical DNA damage recognition functions of XPC-hHR23B and XPA-RPA in nucleotide excision repair.Mo/ecw/αr Carcinogenesis. 38(1): 1-13, 2003 Sep ). Synthetic random sequence oligomers of varying lengths have been tested both in in vitro and in vivo models of cancer (Figl & 2), and smaller oligos appear to be more effective with greater intracellular accumulation than larger oligos. The use of random sequence oligonucleotides as an invention for therapy of proliferative disorders including cancer is protected under PCT/US03/20696.

[3] This present invention is both a reduction to practice of the first invention and an improvement of the original concept and involves the production of random-sequence AND random-size DNA fragments from a biologic source of DNA. (DNA fragments are herein defined as constituents or fragments of synthetic or biologic source DNA that are the result of chemical or mechanical degradation.) This degraded DNA (dDNA) prepared and aliquoted as described below has been demonstrated to inhibit growth of various cancer cell lines in vitro to varying degrees. In sensitive cell lines, the dDNA was more potent and had a greater maximal effect on growth than the random sequence 7 mer oligonucleotides that have been presented under patent PCT/ US03/20696. Disclosure of Invention

Technical Problem.

[4] Major limitations of the therapeutic random-sequence oligonucleotides described and protected under PCT/US03/20696 are the time, complexity, and cost of manufacture and the narrow range of oligonucleotide sizes obtainable. The purified reagents necessary to synthesize oligonucleotides are expensive with a limited shelf- life. In addition, even a dedicated large-scale oligonucleotide synthesizer would be limited to producing uniform length oligonucleotides. As discussed earlier, a variety of different length oligonucleotides are expected to affect more DNA binding targets and exert a greater inhibition of abnormal DNA transcription and replication. In order to synthesize random sequence oligonucleotides with random lengths over a continuous range of 4 - 100 bases, the synthetic protocol would need to be changed repeatedly or utilize scores of different machines, later mixing the synthesized oligonucleotides together. Even then, the efficiency of synthesizing oligonucleotides would pro¬ gressively drop as larger sequences are being made, and synthesis of sequences larger than 100 bases would not be practical.

Technical Solution

[5] One solution to these problems is to mechanically, enzymatically, and/or chemically degrade biologic source genomic DNA (gDNA) into degraded DNA (dDNA). Various methods for this degradation process, such as sonication, au- toclaving, boiling with or without acid or alkaline hydrolysis or DNAse digestion are discussed within this application. Methods for monitoring the degradation process to quantitate and qualify the size distribution and yield of dDNA using gel elec¬ trophoresis and mass spectroscopy are also discussed. The dDNA can be further processed to obtain rapidly different size fractions that might be used for different therapeutic effects.

Advantageous Effects

[6] This present invention is both a reduction to practice of the first invention and an improvement of the original concept and involves the production of random-sequence AND random-size DNA fragments from a biologic souxce of DNA. (DNA fragments are herein defined as constituents or fragments of synthetic or biologic source DNA that are the result of chemical or mechanical degradation.) The rationales for using a biologic source are four-fold. First, this would allow rapid production of a large amount of random sequence oligonucleotide at a substantial cost savings over what could be synthesized de novo. Second, the representative mixture of DlSfA sequences found in vivo, including those from non-coding repeat sequences, may have biologic effects that are unattainable from synthetic oligos. Third, unlike synthetic random sequence oligos, oligos and DNA fragments produced from a biologic source, as described in this patent, could generate a wide size range from 2 to thousands of nu¬ cleotides. Although smaller oligonucleotides are expected to enter cells more ef¬ ficiently, they may not be able to bind efficiently to many single-stranded DNA binding proteins, including RPA, and may lack many additional biologic effects. Fourth, biologic source oligonucleotides and DNA fragments may contain trace amounts of unspecified DNA-binding proteins, such as histones, chapexone proteins, and RNAase-resistant RNA (e.g., double-stranded RNA or RNA complexed to other molecules) that may also exert favorable biologic effects.

[7] DNA fragments (dDNA) degraded from biologic source DNA shares the potential therapeutic effects of synthetic random sequence oligonucletoide combinations described in patent PCT/US03/20696, but in addition dDNA appears to exert additional favorable biologic effects. Like the random sequence 7 mer oligonu¬ cleotides, the dDNA was able to reduce the incorporation of BrdU into cancer cells (A549), suggesting that its biologic effect at least in part is due to inhibition of DNA replication (Fig 3). Differential mRNA expression in A549 cells treated with dDNA versus the synthetic 7 mer and 25 mer oligonucleotide combination (described in patent PCT7US03/20696) were compared by Affymetrix microarray analysis. Transcripts from several genes involved in malignant invasion and cell cycle progression were reduced in common for both the synthetic oligonucleotide-treated and dDNA-treated cells. The dDNA treatment, however, inhibited transcription of more genes than the synthetic oligonucleotides, even at less than half the DNA dose by weight. Additional transcripts suppressed included CDC42 (GTPase involved in en¬ dothelial morphogenesis during angiogenesis) and XRCC5 (implicated in DNA repair).

[8] The dDNA was able to induce apoptosis and inhibition of Gl -> S phase cell cycling in susceptible cell lines, as suggested by in vitro TUNEL assay (NU417, human small cell lung cancer) and cell cycle analysis (HL60, human erythroleukemia), effects that were difficult to appreciate with synthetic oligonucleotides (Fig 4). Clearly any enhanced activity against cancer cells could be the result of a much, wider array of different sized random sequence oligonucleotides. However, dDNA incubation with trypsin was as effective as incubation with DNAse in reducing the inhibitory effect of dDNA on proliferating cell lines, suggesting that an unspecified trace protein in dDNA may responsible for the added biologic effect it has over synthetic oligonucleotides (Fig 5).

[9] Preliminary in vivo experiments demonstrate no overt toxicity to mice given daily intravenous injections of dDNA for up to 2 weeks at doses similar to those used for the therapeutic random sequence 7 mer oligonucleotides previously. Further in vivo ex¬ periments to compare the tumor growth inhibitory effect of dDNA compared to the random sequence 7 mer oligonucleotides are pending. Extrapolating from in vivo results seen with synthetic random sequence oligonucleotides, we would anticipate that dDNA would be best utilized to inhibit the growth of a minimal cancer cell burden or for prevention of cancer recurrence. The lack of toxicity from synthetic DNA oligonu¬ cleotides and their known short half life suggests that similar fragments of native DNA could also be well tolerated with daily dosing for more than one month at a time.

Description of Drawings

[10] Figure 1. Growth suppression of SKBR3 Breast cancer cell line after treatment for

24, 48, and 72 hours with micromolar range concentrations of random sequence 25 mer oligonucleotides. Cell Density measured by MTT cell viability assay.

[11] Figure 2. Combination Treatment with 25mer palindromic ODN and random sequence ODN in vivo, (a) Improved Survival in B16 Murine Melanoma Model. Mice were inoculated subcutaneously with Bl 6 Fl cells and given subcutaneous injections of oligos or vehicle controls daily for 4 weeks into the cancer inoculation site and later intratumorally. The random sequence ODN used were 25mers. Animals were sacrificed when tumors reached a maximal dimension of 2 cm. In reference to this endpoint, median survival for oligo-treated mice was doubled relative to control mice, (b) Complete response in p388 Murine Leukemia Ascites Model. In this model, the random sequence ODN were changed from 25mers to 7mers. Mice were inoculated in- traperitoneally with p388 cells, and all treatments were intravenously administered daily for 4 weeks. A 100% cure rate was seen in mice treated with the combination of random 7mer ODN and 25mer palindromic ODN at 300 mg/kg, each. 60%, 90%, and 0% cure rates were seen in mice treated with random 7mer ODN alone at 600 mg/kg, positive control treatment cyclophosphamide, and PBS, respectively.

[12] Figure 3. BrdU incorporation by N417Lu small cell lung cancer cells after treatment with a. nothing, b. random sequence 7mer oligos (R), c. palindromic 25mer oligo (P), and d. both random 7mers and palindromic 25mer (R & P). BrdU in¬ corporation is diminished by random 7mer treatment alone, but not by palindromic 25mer treatment, and only minimally by the combination of oligos. Reduced effect of the oligo combination may be explained by the cell internalization studies, which demonstrate that in combination, the 25mer inhibits internalization of the random 7mers.

[13] Figure 4. Effect of 7mer random oligo DNA in vitro. Human monocytic leukemia

HL60 cells treated at 10 -6 M for 96 hours demonstrated a reduction in cell cycle S phase, and increase in Gl, as well as a nearly 4-fold increase in subGl DNA suggestive of apoptosis. Percentage of diploid cells in Gl, G2, and S phase were calculated using ModFit cell cycle analysis software. SubGl is calculated as a percentage of total events.

[ 14] Figure 5. MTT Cell Viability Assay of dDNA-sensitive cell line N417 after 48 hour exposure to dDNA. Cell Viability is measured as an arbitrary optical density at 562 nm. On the far left is viability of control cells not treated with dDNA. On the far right is a greater than 80% reduction in cell viability of cells exposed to dDNA. Pre- treatment of dDNA with DNAse I modestly inhibited the reduction in viability resulting from dDNA exposure. Pretreatment of dDNA with trypsin, however, almost completely abrogated the reduction in viability from dDNA exposure. DNAse I followed by trypsin pretreatment of dDNA did not significantly enhance the effect of trypsin alone, suggesting that an unknown protein component of dDNA may be more important in inhibiting the growth of N417 than the oligonucleotide constituents themselves.

[15] Figure 6. Q-Tof Mass Spectrometry of a dDNA sample shows molecular weight peaks that are regular multiples of single nucleoside size. The distribution of sizes is relatively uniform over the lower molecular weight range, and this technique is able to resolve single nucleotide fragments of dDNA.

Best Mode

[16] One of the most abundant and least expensive sources of high purity DNA is from salmon testes. These preparations contain a lyophilized mixture of genomic DNA (gDNA) homogenized, extracted, and precipitated to sizes less than 2 kB (e.g., Sigma Chemicals Company, catalog # D- 1626). This lyophilized gDNA was resuspended in double-distilled, carbon-filtered, UV-irradiated (polymerase chain reaction-grade, or PCR- grade) water to a concentration of 10 mg/ml. This gDNA was then used directly or further purified by twice extracting with phenol/chloroform and precipitating in absolute ethanol with standard Sodium Perchlorate (NaClO4) or Ammonium Chloride (NH4C1) solution, followed by washing the DNA pellet in 75% Ethanol. The resulting DNA pellet was dried and resuspended again in PCR-grade water at 10 mg/ml. Both of these solutions of gDNA were subjected to several different degradation methods described below, but the simplest, most reproducible method for small scale production of dDNA was repetitive autoclaving.

[17] The salmon testis gDNA in 100 ml volumes of 10 mg/ml aqueous solution were repetitively autoclaved in standard liquid sterilization cycles. After each autoclave cycle, the DNA solution was cooled and PCR-grade water back to reestablish the 100 ml volume, and a sample of DNA was run on a 2% agarose gel containing Ethidium bromide for staining. Autoclaving was continued until the cumulative optical density (OD) of the DNA sample detectable on electrophoresis was reduced to 5% (2-10%) of the original. Despite, the reduction in visibility by ethidium bromide staining, the 260 nm OD for DNA samples before and after degradation remained stable, demonstrating that nucleotide content was preserved, and that decreased ethidium bromide staining was due to inability of this agent to intercalate into the smaller sized DNA fragments in the processed dDNA samples.

Mode for Invention

[18] Another method of gDNA degradation was sonication. Ten mg/ml aqueous solutions of salmon testis gDNA were then processed in 10 ml aliquots contained in polypropylene tubes held on ice. The samples were sonicated intermittently for 4 hours using an ultrasonic probe with an operating frequency of 20 kHz and amplitude of 50- 100 ?m (XL 2000 - MicrosonTM - Part No. XL 2000, Misonix, Inc.). Sonication was interrupted intermittently for several minutes to recool the DNA solution below 10° C. Continued sonication further reduced the detectability of the processed DNA by ethidium bromide staining on electrophoresis, while preserving the 260 nm OD of the DNA solutions. Repeated sonication resulted in DNA samples with reduction in ethidium bromide staining (to 2-10% of original) similar to that achieved by processing the salmon testis gDNA by repeated autoclaving.

[19] Sonication of salmon testis gDNA resuspended in 0.1 N hydrochloric acid (HCl) solution was also tested. The addition of acid enhanced DNA degradation by the sonication process allowing shorter processing times. After sonication, the resulting dDNA samples required neutralization using sodium hydroxide (NaOH), generally doubling the solution volumes.

[20] Various combinations of autoclaving and sonication to degrade gDNA into dDNA were tested, with or without HCl acid. AU samples of dDNA were tested in vitro directly or after one isopropanol precipitation and ethanol wash to desalt the specimen (as described above) followed by resuspending the residual dDNA in a minimal volume of PCR-grade water. All these preparations had similar effects on cancer cell growth inhibition in vitro as long as the extent of DNA degradation to dDNA was similar as gauged by the reduction in ethidium bromide staining on electrophoresis. A final alcohol precipitation step after processing gDNA into dDNA did appear to reduce the growth inhibitory activity of the dDNA preparation, presumably by washing away the smallest fragments of DNA that are unable to precipitate.

Industrial Applicability

[21 ] Other methods for mechanical degradation of genomic DNA (gDNA) might be more cost effective and safer for large-scale commercial manufacture of dDNA than the methods utilized by us to date. Some of these methods are already employed to manufacture commercial preparations of 'sheared' DNA, which is a common reagent used to reduce nonspecific interactions in nucleic acid hybridization experiments. Such alternative methods could include, but are not limited to, the following.

[22] First, aerosolization, nebulization, or ultrasonic nebulization of a solution of gDNA has also been shown to result in DNA shearing. A gas flowing through a minute orifice into the solution of DNA at a rate sufficient to create an aerosol could be used. Typical flow rates are greater than 10 liters per minute through a pores that permit extrusion of droplets in the range of one to ten microns in diameter. Ultrasonic transducers enhance the efficiency of the nebulization and can more precisely regulate the size of the droplets being formed based on characteristics like viscosity of the solution, and hence regulate the degree of DNA shearing within the droplets. Any gas could be used, including but not limited to ambient air, oxygen, carbon dioxide, or nitrogen. Carbon dioxide may facilitate the degradation of DNA by lowering the pH of the DNA solution and enhancing hydrolysis of the DNA deoxyriobose backbone. Inert gases, such as nitrogen, may be favorable if later gDNA might be modified to a chemically labile intermediate form prior to degradation. Likewise, gases containing specific con¬ centrations of oxygen may be favorable for differently modified chemical variants of gDNA depending on the optimal redox conditions for it.

[23] Second, glass or metallic beads of varying size can be mixed in with an aqueous solution of genomic DNA and agitated within a closed container, in a process that can be referred to as bead-shearing. After a brief centrifugation of the mixture to sediment any particulate matter from the beads or precipitated components of the DNA solution, supernatant collected from the mixture would contain sheared DNA.

[24] Third, gDNA solutions can be injected into or extruded through a small orifice to shear it into smaller fragments. Examples of such methods include, but are not limited to, a French press or a syringe with a narrow (>/= 20 gauge) needle. If a needle is used, the DNA solution can be aspirated in and out of the needle with a syringe under sufficient pressure to aerosolize the solution. As the DNA solution becomes pro¬ gressively more fragmented and less viscous, it can be aspirated through successively more narrow needles to increase DNA shearing.

[25] Fourth, gDNA solutions could be subjected to ultrasonic energy (without neb¬ ulization). Tubes or vials of gDNA solutions can be immersed in sonicating water baths. The ultrasonic waves could be partially transmitted through the container into the gDNA solution. High-intensity focused ultrasound (HIFU) techniques also could be used to transduce ultrasound energy directly into a solution of gDNA. The gDNA solution could be continuously mixed or stirred during the ultrasonication process to ensure homogeneity of the DNA shearing.

[26] Fifth, vacillating temperature extremes could aid mechanical degradation of gDNA.

Such procedures could involve alternation of boiling and freezing, repeated as frequently as necessary. This method alone is relatively inefficient, but could facilitate other methods described already.

[27] Any conceivable DNA degradation method including the ones discussed above could be repeated sequentially on the same gDNA specimen to produce the smaller size therapeutic DNA fragments described in this patent. The degree of DNA degradation and the size of DNA fragments generated during any of these methods could be monitored crudely using DNA-binding fluorescent dyes. This method is effective because smaller fragments of DNA below a threshold size are intercalated by these dyes less efficiently. The dDNA binding fluorescence could then be correlated to 260 nm absorbance of the dDNA sample by spectroscopy (representing total nucleoside content) and used as a semi-quantitative index of gDNA degradation. Any DNA-intercalating fluorescent dye might be used for quantification, including, but not limited to, propidium iodide, acridine orange, 7-amino Actinomycin D (7- AAD), and ethidium bromide. Flourescence of dDNA solutions containing dye could be measured directly, subtracting background flourescence from dye solution alone. Densitometry measurements of ethidium bromide-stained dDNA at specified size bands on agarose gel electrophoresis would give even more detailed information on the degree of dDNA degradation and the distribution of resulting sizes.

[28] Fluorescent measurements of dDNA while simple, however, are limited because they yield non-linear semiquantitative measurements and because very small fragments below an undetermined threshold may no longer detectably bind in¬ tercalating dyes. To compensate for these limitations, Q-Tof mass spectrometry could be adapted to measure relative quantities of different DNA sizes within the degraded gDNA samples. Preliminary results demonstrate that the molecular weight species present in dDNA after sonication or repetitive autoclaving differ by regular increments, roughly the mass of single nucleosides (Fig 6). Q-Tof and other mass spectrometry methods thus can quantify the size and distribution of DNA fragments in a dDNA sample.

[29] The source of gDNA may also influence the biologic activity of the processed dDNA. If random sequence DNA fragments of greater than 16 bases have therapeutic activity, they may be acting by sequence-specific hybridization to or other interactions with genes in cells under treatment. Such target sequences might be protein-coding regions of genes or repetitive sequences in non-translated regulatory areas of the genome. This would suggest that the ideal source of gDNA from which to derive therapeutic dDNA would be the target tumor or organ itself, or at least cells of human origin. Likewise, therapeutic activity that might be the result of trace proteins (e.g., histones) or other molecules associated with the gDNA also might be greater if the gDNA is derived from the same cancer, patient, or species under treatment. It is conceivable that dDNA might be later derived from gDNA of cultured immortalized human cells (such as epithelial cells transformed with oncogenes or cancer cell lines) or resected tumors from individual patients who could be then treated with dDNA derived from their own tumors. A potential advantage of using salmon testes DNA is that it contains both haploid as well as diploid gDNA, since larger oligonucleotides derived from haploid DNA may tend to stay more readily denatured.

[30] Other solutions also could be used to dissolve gDNA for degradation. An alkaline solution can facilitate hydrolysis of gDNA subjected to mechanical degradation, as the acidic solutions we used earlier did. Solutions of gDNA containing differing con¬ centrations of various DNAses with optimized buffers and magnesium concentrations incubated at optimized temperatures could enzymatically degrade the gDNA to desired sizes. DNAse can then be inactivated by heating the gDNA solution enough to denature the enzyme. This enzymatic degradation could also be combined sequentially with mechanical degradation methods.

[31] The same degradation techniques could be used to varying degrees on mixtures of synthetic oligonucleotides of discrete lengths to create a range of different sequence lengths. If situations exist where synthetic random sequence oligonucleotides might be more effective therapy than dDNA, the same degradation methods used for gDNA could be adopted for mixtures of larger random sequence synthetic oligonucleotides to obtain a range of different lengths

[32] Any given solution of dDNA or partially degraded synthetic random sequence oligonucleotides could also be passed through size exclusion columns or membranes to isolate DNA fragments within a specific size range. By isolating and testing different size ranges of dDNA, we would hope to identify optimal size fractions for selected biologic activities. For example, the optimal size range of dDNA to inhibit DNA pro¬ liferation may be smaller than that to bind RPA and inhibit DNA repair or tran¬ scription. Size exclusion columns or membranes could be applied sequentially to DNA samples to isolate different size ranges of DNA that may have different therapeutic potentials.

[33] Biologic source DNA either before or after degradation might be chemically modified to improve biologic half-life, pharmacokinetics, and other pharmacologic properties. Such modifications could include, but are not limited to, oxidation or reduction, addition of ester, ether, or amide groups to random or specific free hydroxyl groups on the DNA backbone or on nucleotide bases.

[34] After processing dDNA into optimal therapeutic molecules, they may or may not be further added to solvents, stabilizers, flavorings, or fragrances to facilitate phar¬ maceutical administration. The resulting therapeutic dDNA preparation may be given by different routes including, but not limited to, intravenous, intramuscular, sub¬ cutaneous, transmucosal, inhalational, oral, topical, gargle, or enema administration.

Sequence List Text

[35]

Claims

Claims

[1] Chemically and/or mechanically degraded purified biologic source gDNA

(genomic deoxyribonucleic acid) for the treatment of proliferative disorders in humans and animals. Treatment of a proliferative disorder refers to the im¬ provement, stabilization, or prevention of the disorder.

[2] Repetitive use of procedures for mechanical degradation of gDNA to produce degraded DNA (dDNA), including autoclaving, sonication, high intensity focused ultrasonication, aerosolization, nebulization, extrusion, boiling, freeze- thawing, and bead-shearing.

[3] Genomic DNA (gDNA) from any biologic source might be utilized to produce dDNA, as long as the source DNA can be purified by standard phenol/ chloroform extraction and alcohol precipitation, and different gDNA sources may have different therapeutic advantages.

[4] dDNA may be prepared with solvents, stabilizers, flavorings, or fragrances to facilitate different routes of administration, including, but not limited to, in¬ travenous, intramuscular, subcutaneous, transmucosal, inhalational, oral, topical, gargle, or enema administration.

[5] Quantitation of DNA-binding dye fluorescence with or without 260 nm wavelength absorbance as a semi- quantitative index to monitor gDNA degradation.

[6] Q-Tof and other mass spectrometry methods for semi-quantitating and monitoring the size distribution of gDNA degradation.