Biomod/2014/OhioMOD/results

From OpenWetWare

Jump to: navigation, search

Project and Results

Novel microRNA antisense therapeutic delivery using DNA origami nanostructures

Table of Contents
1. Structures
2. miR 21 Sequestering
3. Cellular Uptaking
4. PTEN Expression
5. Cellular Viability
6. Discussion
7. Conclusion and Future Work
8. References

1. Structures

Hypothesis: Combination of designed staples and scaffold DNA, in an optimized cationic solution will result in the synthesis of nanostructures with specific geometry.

Structures were designed to emulate and verify previous reports of ideal nanoparticle size for cellular uptake. It has been demonstrated that rod-shaped nanostructures, with sizes ranging from 20 to 120 nm, are uptaken more effectively than structures with a more cube-like shape[1]. Using this information, we designed two structures at opposite ends of the spectrum. The Block-O is essentially a torus, with outer dimensions of 30x30x24 nm, while the Buckeye Branch is a 90 nm long rod-like shape (Figure 5). Both structures were designed using caDNAno[2] and are able to incorporate ssDNA overhangs complementary to miR-21 (56 overhangs on The Block-O and 42 on the Branch). Three versions of each structure were designed; one with miR-21 complementary overhangs, a control with scrambled-sequence overhangs, and a structure without overhangs. Importantly, the scrambled overhangs were randomized so as to not create any epigenetic effects in the target.

Figure 5: Block O and Branch Structures

To overcome the electrostatic repulsion of DNA oligomers, the addition of MgCl2 to the folding reaction (a solution containing ssDNA staple strands in a 10 fold molar excess relative to ssDNA scaffold strands, as well as a buffer containing EDTA and Tris) provides divalent cations which facilitates binding interactions. The concentration of these cations has been shown to have a pronounced effect on the yield and quality of folded structures[3]. To optimize this concentration, structures were folded with MgCl2 concentrations ranging from 12mM to 26mM. Subsequent analysis by means of both agarose gel electrophoresis and spectrophotometry showed that a concentration of 20mM MgCl2 provided the highest yields with the lowest concentration of misfolded structures.

The structures were further validated by direct imaging using Transmission Electron Microscopy (TEM), which showed branch structures with dimensions of 10 nm by 12 nm by 90 nm, and Block O structures with dimensions of 30 nm by 30 nm by 24 nm (Figure 6). This closely matches the designed dimensions. A small degree of global twist is evident in the branch, a result of inter-helical strain induced by the square helix pattern used in the design[4].

Figure 6: TEM Images of Block O and Branch

2. miR 21 Sequestering

Hypothesis: Structures are able to remove miR-21 from solution via complementary base pairing with miR-21 complementary overhangs.

In order to demonstrate the ability of the DNA nanostructures to sequester miR-21, structures functionalized with ssDNA overhangs complementary to miR-21 were incubated in solution at 37℃ with fluorescently-tagged miR-21 (Figure 7), purified via agarose gel electrophoresis, and subsequently analyzed by means of bulk fluorescence imaging. Sequestration of miR-21 is evident in sample bands when the fluorescence of the labeled miR-21 is present in the band containing our structures and fluorescence is diminished in the band containing free miR-21. A titration of miR-21 allowed us to determine the ability of the structures to sequester an increasing amount of free miR-21 before becoming saturated (Figure 8). As expected, when the control structure with scrambled overhangs was incubated with fluorescent miR there was no binding observed.

Figure 7: miR Sequestration Schematic

Figure 8: Bulk fluorescence image of a scrambled overhang control structure (c) incubated with a 3x excess of miR, functional structure incubated with a titration of miR

3. Cellular Uptaking


Hypothesis: Our structures utilize the endolysosomal pathway to infiltrate cells.In addtion, the difference in size and aspect ratio of the structures will affect how they are uptaken

Once the structures were characterized and functionalized with miR-21 complementary overhangs, we sought to test whether they would be successfully uptaken into cells. One possible mode of uptake would be through endocytosis, where the structures would be enclosed in an endosome and allowed into the cell. To test this, structures labeled with a fluorescent intercalating dye (TOPRO3) were incubated with cells whose lysosomes were labeled with Lysotracker Green at a final concentration of 0.1 nM. After a 4 hour incubation period, the cells were imaged with a fluorescent microscope to observe colocalization between the signals originating from the structures and the lysosomes.

Figure 9: Fluorescent microscopy images exhibiting endosome (labeled in green) and DNA nanostructure (labeled in red) colocalization. Overlay of fluorescence signals with a DIC image is indicative of an endosomal cellular uptake mechanism.

As shown in Figure 9, fluorescence from the Lysotracker (shown in green) and fluorescence from TOPRO3 (shown in red) occupy the same coordinates within a cell, suggesting that the structures and endosomes occupy the same space within the cell. In addition to suggesting a possible mechanism for uptake, the Lysotracker also aids in affirming the position of the structures within the cell. Since these images are two-dimensional, without the Lysotracker, there would be no way to determine whether the signal from the structures originated from within the cell, or from above or below the cell.

One advantage of generating structures with very different aspect ratios was the ability to test which structures were more easily uptaken by cells. Although both structures fall within the ideal dimensional parameters for cell uptake[5], the different geometries of the elongated branch versus the compact Block O may have an impact on cell uptake efficiency[6]. While epifluorescent microscopy images do not necessarily provide quantitative data on the uptaking efficiency of structures, some preliminary assessments may be made nonetheless. Firstly, more instances of co-localization could be observed in the presence of branch structures relative to Block O structures. Perhaps more intriguingly, in the presence of branch, all instances of fluorescence in the 640 nm channel occurred either within viable cells or within dead cell debris following an incubation time of 4 hrs. However, a few instances of fluorescence were found outside of cells in the presence of Block O, as shown in Figure 10.

Figure 10: Fluorescent microscopy images exhibiting endosomes (green) and DNA nanostructure (red) colocalization. The fluorescence signal outside the cell in the Block O sample is indicative of slower uptake.

The fluorescence signal, with an intensity comparable to the intensity of the TOPRO3 dye, originated at the very edge of the cell, on the cell membrane. This may suggest that while branch structures are quickly uptaken once they come in contact with cells, the Block O structures experience a delay in uptake, and thus tend to congregate outside the cell. While this introduces an interesting possibility in terms of the mechanism of uptake, more experiments must be conducted both using qualitative methods such as fluorescent imaging, and quantitative methods such as qPCR to detect intracellular levels of DNA nanostructures to form a conclusion on this matter.

4. PTEN Expression

4.1 Protein Expression


Hypothesis: An increase in the relative expression of PTEN, a target protein for miR-21, will be observed after incubation of cells with our structures.

The proof of successful structure uptake into cells allowed for the focus to shift to elucidating a possible mechanism through which the structures affect cell viability. As stated earlier, PTEN is a well established target of miR-21, and so it was hypothesized that any changes in miR-21 levels brought about by the structures will have an observable impact on relative PTEN protein levels within cells.

Cells were incubated with structures with complementary overhangs and scrambled overhangs at a final concentration of 0.1 nM for either 24 or 48 hrs. After the incubation time, total protein was extracted from the cells and a western blot was performed to compare the levels of PTEN between the treatment and the control. The levels of PTEN were normalized to GAPDH, a housekeeping protein, to control for sample and loading variability.

Figure 11: (a) Western blot image showing PTEN protein band (top) and GAPDH protein band (bottom) for different samples after a 24 hour (left) or 48 hour (right) incubation period. (b) Densitometric analysis showing normalized intensity relative to scrambled control for different samples at 24 hours (left) or 48 hours (right). The intensities of both PTEN and GAPDH bands were measured using ImageJ. The PTEN intensities were then divided by the corresponding GAPDH intensities to obtain a normalized value. The normalized intensities for each treatment (Branch miR, Block O miR) were normalized to its corresponding control (Branch scr, Block O scr) and plotted on a bar graph.

The intensity of the bands on the blot, as shown in Figure 11, correspond to the level of protein expression in the sample. In samples incubated for 24 hours, densitometric analysis showed increased PTEN expression relative to the control for branch only. However, after 48 hours of incubation, an increase in PTEN expression was seen in both branch and Block O experimental samples.

This data reinforces our hypothesis by correlating the presence of miR-21 complementary structures to increased PTEN levels. It also reaffirms characteristics of the structures studies earlier, such as their ability to sequester miR-21, even within the cell environment. Of particular interest is the discrepancies between the behavior of the two structures. The samples incubated with Block O showed a delayed response to the treatment compared to the branch. If the Block O structures are indeed uptaken at a slower rate, it follows that they will reach the critical concentration necessary to facilitate change at a slower rate, resulting in the delayed response observed in this experiment. However, once uptaken, it is possible that the Block O structures are more effective at raising PTEN than branch is, since the Block O treatment resulted in a higher expression of PTEN compared to the branch treatment. This may be because the Block O is functionalized with more complementary overhangs than the branch, allowing it to induce a greater change per structure than the branch.

4.2 mRNA Expression


Hypothesis: Since miRNA functions by suppressing mRNA, a reduction in miRNA concentration by our structures should result in an increase in mRNA concentrations of PTEN, a miR-21 target gene

Another line of evidence for establishing a mechanism can be generated by studying changes in the relative level of the mRNA that codes for the target protein. mRNA quantification is usually carried out through a quantitative polymerase chain reaction (qPCR), which can detect the presence of minute quantities of target DNA with high specificity. The higher level of specificity allows for the detection of miniscule changes in mRNA levels that may not translate over to the protein level.

Since mRNAs are the direct targets of miRNA, it was believed that changes in the mRNA level would be observed sooner than the resultant changes in the protein level. Therefore, incubation periods of 12 and 24 hours were chosen for analysis. As with protein expression experiments, the cells were incubated with structures at 0.1 nM final concentration for the selected time, and then lysed. However, instead of protein the total RNA from the cell was isolated, and subsequently converted to cDNA via reverse transcriptase PCR. qPCR was then performed on the samples with primers for PTEN and Actin, a housekeeping gene used to normalize the data.

Figure 12: Comparative analysis of normalized PTEN mRNA expression in treatment to expression in control 12 hours after incubation (left) and 24 hours after incubation (right). All samples were first normalized by their corresponding Actin expression levels. The expression in experimental samples were then normalized to control, and the data was plotted on a graph.


The above figure shows the levels of PTEN gene expression measured from the experiment. The 12 hour experiment showed no significant difference in mRNA expression between the complementary miR treatment and the scrambled control. At 24 hours, however, both the branch and the Block O miR treatments resulted in higher levels of PTEN mRNA, suggesting that the structures do have a direct impact on the mRNA levels.

Even though an increase in PTEN mRNA was observed, the increase was significantly lower than the levels observed in PTEN protein. Since mRNA is necessary for the translation of protein and is soon degraded afterwards, a spike in mRNA concentration would cause a delayed increase in protein concentration, by which time the mRNA concentration would fall again. Therefore, one obstacle with analyzing mRNA levels is the possibility that the time point of measurement fall within the downturn of mRNA concentration. Therefore, to obtain a clearer picture mRNA levels, experiments at a wide array of time points must be conducted.

5. Cellular Viability


Hypothesis: Cells exposed to miR-21 complement structures will show decreased viability because of an increase in apoptosis-inducing proteins. Cells exposed to scrambled structures should show no change in viability when compared to a cells only control.

Our ultimate goal in creating complementary miR-21 overhang functionalized DNA nanostructures is to show its effectiveness in reducing the viability of OSU CLL cells. To test for cell viability, branch structures, both with scrambled and miR overhangs, were incubated with cells for 24 hours at 0.5 nM final structure concentration. After incubation, the exclusion dye propidium iodide was added to stain dead cells, and the samples were imaged using fluorescence microscopy.

Figure 13: Characteristic composite image of cells in the different samples stained with propidium iodide(PI) (shown in green). The PI stains DNA escaping the nucleus of cells undergoing apoptosis.


The images shown in Figure 13 are representative of the images in general. The morphology of the miR complement cells showed that the cells were generally unhealthy, showing irregular shape and granularity of the cytoplasm. These characteristics were not nearly as common in the under the other conditions. The histogram for signal intensity of each set of images also suggests that there was a much higher signal in the miR complement samples. The 488 signal peaked at a higher intensity and continued to higher intensities in the miR samples, while the other samples showed lower peaks and no signal at higher intensities.

Figure 14: Comparison of cell viability after 24 hour incubation normalized to cells only control


Quantitative analysis of the number of fluorescent cells in the sample population confirmed the qualitative assessments made previously. The data, presented in Figure 14 illustrate the marked decrease in the percentage of viable cells after treatment with miR-21 complementary overhangs for 24 hours.

6. Discussion

Our goal to create a new approach for treating CLL patients using DNA nanotechnology gives us a promising technology. While others have used conjugation based methods, liposomes, polymer nanoparticles, or antibodies as a means to deliver antagomiR[7], this constitutes the first successful attempt at adding miR-21 complementary strands to a DNA nanostructure. We used TEM and agarose gel electrophoresis to test for proper structure folding, and used fluorescence imaging to confirm that our structures were capable of sequestering miR-21. Using fluorescence microscopy, initial experiments demonstrating cellular uptake in OSU-CLL cells were performed. Colocalization of structures and lysosomes suggests that structures are uptaken via the endolysosomal pathway. Increased expression of PTEN, a protein targeted by miR-21, was demonstrated in the protein level using western blotting, and in the mRNA level using pPCR. Finally, OSU-CLL cells showed a marked decrease in viability after incubation with miR-21 complementary structures when compared to scrambled controls. Thus we were able to characterize the efficacy of our structure in all facets of its function.

DNA origami provides an efficient route to use antisense delivery because of the precise structural control that can be specified when creating the nanostructures. In addition, a DNA based carrier provides the advantages of high biocompatibility, low cytotoxicity, and failure to induce immunogenicity[8]. Finally, it is possible to add targeting moieties to DNA origami structure such as antibodies[9] and folates[10] in a spatially selective manner. Two structures were chosen specifically because one is rod-shaped and one is roughly spherical. There is currently a lack of concrete information about how shape affects the endocytosis of nanostructures, but shape has been shown to affect uptake. Therefore, we wanted to chose structures at two extremes of the shape spectrum to compare uptake results. The higher success rate for the Branch suggests that cells may take up rod like structures of around 100 nm in length more efficiently. Another possible explanation for this outcome is that the Block O has a higher tendency to degrade in the lysosomal environment due to acidity. Typically, lysosomes have a pH of about 5, and DNA nanostructures rapidly degrade at a pH of about 4[11]. Their degradation is depended upon structure, however, so the Block O might be more sensitive to this environment.

While fluorescence microscopy showed instances of uptake for both branch and Block O, there was evidence to suggest that the Block O was not as efficiently uptaken as branch. This was further corroborated by the results from the PTEN protein experiment, where the Block O structures exhibited a delayed response in the cell compared to branch. While the specific mechanism for the uptake of nanoparticles remains unclear, research shows that the rate of uptake is size dependent. While Gratton et al showed that rod like structures are more preferentially uptaken than other geometries[1], Chithrani et al showed that, when the size of the particles falls below 100 nm, the trend changes and spheres are more easily uptaken[1]. Thus, our structures provide a conundrum since the branch is rod like and close to the 100 nm threshold, whereas the Block O is more compact and also well below the 100 nm range. This provides a unique opportunity to compare cellular uptake efficacies of nanostructures with ideal geometries on both ends of the spectrum. At first glance, the data suggests that rod-shaped structures at 100 nm are more efficient at being uptaken than compact shapes with diameters less than 100 nm. However, there are a few confounding variables to consider. Firstly, the Block O contains a void in the center, which may reduce the effective surface area in contact with the cell. None of the literature reviewed tested a structure with a central void, so their conclusions do not extend to this case. Secondly,while there have been studies conducted on the uptake potency of spherical and rod-shaped structures, the little research done with nonspherical structures indicates that their interactions with cells may be much more complex[1]. Therefore, more research must be conducted to determine the factors resulting in the lower uptake efficacy for Block O.

Studies have previously shown that decreasing miR-21 levels using anti-miR-21 resulted in an increase in PTEN levels in human hepatocellular cancer cells[15] and colorectal cancer cells[16]. However, no studies have attempted to regulate PTEN levels in CLL using antisense methods. Here we have shown that treating OSU CLL cells with DNA nanostructures functionalized with miR-21 complements induces an increase in PTEN expression both in the protein level and in the mRNA level. However, certain discrepancies were observed in the data. Most prominent is the fact that the Block O had a delayed effect on PTEN protein levels compared to branch. This could be due to reduced levels of uptake observed in the cellular uptake experiments. If the structures must reach a critical concentration before they can facilitate change, then it would be logical for the Block O to show a slowed reaction, since it also has a slower rate of uptake.

However, in both the protein and mRNA experiments, the eventual increase in PTEN brought about by the Block O was slightly higher than the increase brought about by the branch. This illustrates that, once successfully uptaken by the cell, the Block O may be slightly more potent than the branch. The most feasible explanation for this phenomenon is the difference in the number of overhangs between Block O and branch. Once enough time has passed for the concentrations of both structures within the cell to equilibrate to a similar value, the transient effect from the reduced uptake efficiency of the Block O can be ignored, and the impact on gene expression is dependent only on the intensive properties of the structures. Since the Block O has 12 more overhangs per structure, it is logical to believe that the Block O structures would cause a greater effect on expression compared to the branch structures in the same concentration.

One final point of interest arises with the comparison between PTEN protein expressions and PTEN mRNA expressions observed in our experiments. The increases in PTEN mRNA observed is miniscule compared to the increases in PTEN protein, even at the same incubation time and structure concentration. Since mRNA which code for proteins involved in regulating the cell cycle are short lived[17], a spike in mRNA concentration would be transient and cause a delayed increase in protein concentration. Therefore, one obstacle with analyzing mRNA levels is the possibility that the time point of measurement fall within the downturn of mRNA concentration. Performing a qPCR analysis at multiple time points could resolve this issue. Another possible explanation could be that a single PTEN mRNA molecule could produce a large number of proteins before degrading. A large-scale study quantifying cellular mRNA and protein expression levels in parallel found that the median translation rate constant of 40 protein molecules per mRNA per hour, and found the maximal rate to be as high as 180 proteins per mRNA per hour[18]. While the specific rate constant for PTEN specifically has yet to be measured, the data suggests that even a small increase in mRNA levels can have a significant impact in the protein expression observed.

Although fluorescence microscopy shows that our structures induce a decrease in viability, more questions remain. The use of flow cytometry could better quantify cell death, and could yield other information about the nature of the cell death. We hypothesize that our structures cause cell death by reactivating the pathways that induce apoptosis, but imaging doesn’t distinguish between apoptotic cell death and necrotic cell death. There are flow cytometry assays that can make that distinction, and their use could be vital to confirming that our structures operate using the hypothesized mechanism. It’s also possible that further study reveals that our structures do not alter cell viability in a significant way. If so, they might still alter how the cells grow and proliferate. To investigate, we would compare the growth of cell populations when exposed to our structures.

Here, we reported on the design and validation of two novel DNA origami nanostructures capable of sequestering miR-21 in cells, thereby altering PTEN expression levels and inducing apoptosis in OSU-CLL cells. We used TEM to show that structures were properly folded. Fluorescent gel electrophoresis was used to ensure that the structures could sequester miR-21 in solution. Fluorescence microscopy showed colocalization of our structures with lysosomes in the cell, suggesting that our structures are absorbed via the endolysosomal pathway. The expression of PTEN, a target of miR-21, was analyzed through western blotting and it was found that our structures are capable of changing PTEN levels. Viability was measured by propidium iodide analysis with fluorescence microscopy, and it was also found that our structures were capable of decreasing the viability of OSU-CLL cells. Our results suggest that DNA nanostructures with miR complement overhangs could be a promising new anti-sense therapeutic method for combatting cancer and other human diseases. Our structures are simple and flexible, capable of supporting targeting and acting as vehicles for other types of therapeutics. The structures are not limited to treatment of a specific type of cells or a specific RNA, thus the method we used may be applied in studies of various cells and microRNA profiles. .

7. Conclusion and Future Work

While preliminary results are quite promising, experimental trials need to be repeated to establish statistical significance. Furthermore, although our initial data suggest that our nanostructures exhibit a marked effect on PTEN, a more comprehensive study would involve the measurement of a manifold assortment of potential genes and proteins which may be affected. Additionally, therapeutic efficiency and uptake may be improved by the addition of targeting moieties to our DNA nanostructures. Previous reports have demonstrated that targeting significantly increases the rate of endocytosis of nanoparticles[10]. Additionally, the versatility and exacting addressability of DNA origami nanostructures allows for the combination of multiple therapeutic approaches. For instance, the intercalating chemotherapy drug Doxorubicin (dox) can be loaded into DNA origami nanostructures and delivered to cells[12] and functionalized overhangs could then be tailored to change protein expression in a way that makes cancer cells more vulnerable to dox.

It’s also important to note that there is nothing specific about the use of CLL or the targeting of miR-21 in this project. Both were chosen due to the depth of research available, however it is a simple process to apply our nanostructures to other miRNA targets or cell types. MiRNA dysregulation is symptomatic of many cancers, and is implicated in other diseases such as diabetes[10] and heart disease[13], which suggests that our structures may have significant implications beyond cancer. Our structures also have valuable applications to the study of fundamental cellular gene expression pathways which may provide insight into how diseases such as cancer develop.

Here, we reported on the design and validation of two novel DNA origami nanostructures capable of sequestering miRNA in cells, thereby altering protein expression levels.

8. References


[1] Albanese, A., Tang, P. S., & Chan, W. C. (2012). The Effect of Nanoparticle Size, Shape, and Surface Chemistry on Biological Systems. Annual Review of Biomedical Engineering, 14(1), 1-16. doi: 10.1146/annurev-bioeng-071811-150124
[2] Douglas, S. M., Marblestone, A. H., Teerapittayanon, S., Vazquez, A., Church, G. M., & Shih, W. M. (2009). Rapid prototyping of 3D DNA-origami shapes with caDNAno. Nucleic Acids Research, 37(15), 5001-5006. doi: 10.1093/nar/gkp436
[3] Sungwook Woo & Paul W. K. Rothemund (2014). Self-assembly of two-dimensional DNA origami lattices using cation-controlled surface diffusion. Nature Communications 5, 4889
[4] Ke, Y., Douglas, S. M., Liu, M., Sharma, J., Cheng, A., Leung, A., ... Yan, H. (2009). Multilayer DNA Origami Packed on a Square Lattice. Journal of the American Chemical Society, 131(43), 15903-15908. doi: 10.1021/ja906381y
[5] Le e, H., Lytton-Jean, A. K., Chen, Y., Love, K. T., Park, A. I., Karagiannis, E. D., ... Anderson, D. G. (2012). Molecularly self-assembled nucleic acid nanoparticles for targeted in vivo siRNA delivery. Nature Nanotechnology, 7(6), 389-393. doi: 10.1038/nnano.2012.73
[6] Shi, X., von dem Bussche, A., Hurt, R., Kane, A., Gao, H (2011). Cell entry of one-dimensional nanomaterials occurs by tip recognition and rotation. Nature, 6, 714-719. doi: 10.1038nnano.2011.151
[7] Li, Zhonghan, and Tariq M. Rana. "Therapeutic Targeting of MicroRNAs: Current Status and Future Challenges." Nature Reviews Drug Discovery 13.8 (2014): 622-38. Web.
[8] Zhang, Q., Jiang, Q., Li, N., Dai, L., Liu, Q., Song, L., Wang, J., Li, Y., Tian, J., Ding, B., Du, Y (2014). DNA Origami as an In Vivo Drug Delivery Vehicle for Cancer Therapy. ACS Nano, 8(7): 6633-6643. doi: 10.1021/nn502058j
[9] Douglas, S. M., Bachelet, I., & Church, G. M. (2012). A Logic-Gated Nanorobot for Targeted Transport of Molecular Payloads. Science, 335(6070), 831-834. doi: 10.1126/science.1214081
[10] Liu, X., Song, W., Sun, T., Zhang, P., & Wang, J. (2011). Targeted Delivery of Antisense Inhibitor of miRNA for Antiangiogenesis Therapy Using cRGD-Functionalized Nanoparticles. Molecular Pharmaceutics, 8(1), 250-259. doi: 10.1021/mp100315q
[11] H Padh, J Ha, M Lavasa and T L Steck. (1993). A post-lysosomal compartment in Dictyostelium discoideum. The Journal of Biological Chemistry, 268, 6742-6747
[12] Zhao, Y., Shaw, A., Zeng, X., Benson, E., Nyström, A. M., & Högberg, B. (2012). DNA Origami Delivery System for Cancer Therapy with Tunable Release Properties. ACS Nano, 6(10), 8684-8691. doi: 10.1021/nn3022662
[13] Ikeda, S., Kong, S. W., Lu, J., Bisping, E., Zhang, H., Allen, P. D., ... Pu, W. T. (2007). Altered microRNA expression in human heart disease.Physiological Genomics, 31(3), 367-373. doi: 10.1152/physiolgenomics.00144.200
[15] Meng, Fanyin, Roger Henson, Hania Wehbe–Janek, Kalpana Ghoshal, Samson T. Jacob, and Tushar Patel. "MicroRNA-21 Regulates Expression of the PTEN Tumor Suppressor Gene in Human Hepatocellular Cancer." Gastroenterology: 647-58. Print.
[16] Xiong, Binghong. "MiR-21 Regulates Biological Behavior through the PTEN/PI-3 K/Akt Signaling Pathway in Human Colorectal Cancer Cells." International Journal of Oncology(2012). Print.
[17] Ross, Jeff. "MRNA Stability in Mammalian Cells." Microbiological Reviews 59.3 (1995): 423-50. Print.
[18] Schwanhäusser, Björn, Dorothea Busse, Na Li, Gunnar Dittmar, Johannes Schuchhardt, Jana Wolf, Wei Chen, and Matthias Selbach. "Global Quantification of Mammalian Gene Expression Control." Nature (2011): 337-42. Print.
Edit Project and Results

Views
Personal tools