|Line 291:||Line 291:|
| ||ladder||7249 pDNA||0.5nM DNA/6.25nM Ca control||0.5nM DNA/3.74 nM PO4 control
| ||ladder||7249 pDNA||0.5nM DNA/6.25nM Ca control||0.5nM DNA/3.74 nM PO4 control
| ||0.5nM DNA/6.25nM Ca/
| ||0.5nM DNA/6.25nM Ca/:||0.5nM DNA/6.25nM Ca/:||0.5nM DNA/6.25nM Ca/:
| ||0.5nM DNA/625μM Ca/
| ||0.5nM DNA/625μM Ca/:||0.5nM DNA/625μM Ca/:||0.5nM DNA/625μM Ca/:
| ||0.05nM DNA/625μM Ca/
| ||0.05nM DNA/625μM Ca/:||0.05nM DNA/625μM Ca/:||0.05nM DNA/625μM Ca/:
Revision as of 12:10, 12 June 2013
Will be using links to google drive for files.
Friday May 10
Submitted cadnano design of 66 helices, non-hollow: 66helicev1. Will note that midline there are scaffold crossovers that do not have staples nearby, had incorrectly assumed that staple crossovers must avoid scaffold crossovers. Design approximately 40 nm long and less than 22 nm in diameter.
Sunday May 12
New version edited by Dr. Castro.66 helices, 2-layered, hollow: 66helicev3. Will note that diameter remains the same, is approximately 44 nm long now. Should not be an issue
'Literature review': It seems well accepted that crossing the nucleolus membrane is the greatest challenge facing gene therapy. The calcium phosphate protocol as it stands only takes us to the cytosol, so the origami must be designed to be able to enter the nucleus. A paper in 1988 used a range of gold particles to measure the size able to enter through the nuclear pore complex  and found that the limiting size was 26 nm. This suprised some people as some viral proteins are larger than this, including Hepatitis nucleocapsids. A later paper in 2002 used gold particles with protein complexes to measure a larger size of 32-36nm . We would need to add targeting sequences or nuclear localization signals to the origami. For NLS peptide conjugation to the DNA staples, one would need DNA staples with a free alkylamino group of a thymine. see A_single_nuclear_localization_signal_peptide_is_sufficient_to_carry_DNA_to_the_cell_nucleus
However, peptide conjugation and staple modification is making things more complicated than what is desired or allowable. Perhaps nuclear targeting sequences would be better, though a recent paper in 2011 downplays their importance .
Monday May 13th
Ordered staples for version 3 origami design edited by Castro. The templated had now been decreased to the smaller 7202 template strand. Will be leaving for Puerto Rico while waiting for staples to come in, will be back on the 23rd.
Corresponded with Dr. James Adair at PennU and Dr. Robert 'Bob' Lee of OSU pharma.
Literature Review: In general, some better indications that the origami can be functionalized with cationic peptides to facilitate transport across the nuclear pore complex.
Rosenecker's lab synthesized a tetramer of the PKKKRKV NLSV404 peptide sequnece spaced by glycine residues. (may want to consider finding a lab that has a Applied Biosystems Automatic Synthesizer) . A strong paper, with luciferase transfection assay, in situ flourescent hybridization to detect plasmid dna, and a nuclear import assay with labelled BSA-BODIPY to be transported by the peptide. The SV40 NLS sequence they used is the minimal sequence observed on the T-antigen of SV40 necessary. HeLa cells showed 58.9% transfection within 2 hours, much faster than with the dendrimer alone or with a mutant peptide control. Analysis shows quicker localization of the peptide-plasmid in the nucleus, indicating that the peptides probably helped transport the plasmid across the pore if not through the cytosol as well. Early cell transfection also indicates a lessened need for mitosis to occur. Also did a nuclear transport inhibition protocol by Sebestyen , which had also showed linking of peptides to plasmids by the cyclo-propayrroloindole cross-linker. Remember: Karyopherins are the proteins found to accumulate in the nucleus.
Dr. Davis at Case Western also showed that DNA was binded the nucleolin receptor on the cell membrane, and may indicate a new cellular uptake mechanism with increased transfection efficiency. The DNA nanoparticles were formed by PEGlyated polylysine. We probably don't want to pursue this mechanism, but the effectiveness of the internalization and transport of the nanoparticles is impressive . Davis had also demonstrated using PEGylated=polylisine/lipid/DNA nanoparticles that are 10-20nm in diameter and 100 nm in length that are rapidly uptaked and concentrated into the nucleolus. This confirms the aspect ratio hypothesis, and they also seem to enter by nucleolin as well. Davis also mentions compacted DNA being known to increase efficiency so long as the ellipsoidal diameter of the compacted DNA does not approach 25 nm, and this was shown in non-dividing cells.
Dr. Dean is one of the first to have described the nuclear targeting sequence of the SV40 DNA itself. In his review , Dean describes several methods to nucleari import of nonviral vectors. The classic SV40 NLS PKKKRKV. He also describes the 72 bp sequence that is shown to have binding sites for a number of mammalian transcription factors that are usually brought into the nucleus for RNA transcription. Reich has also idenifitied NFkappaB binding sites that show enhanced expression as well. Using the DNA sequences have an obvious advantage to us in that it eliminates the need for the peptides, but I'm worried that the origami will not allow for the normal binding of the factors even if the binding sequence is spaced and free. Also, we also need to take pore size into account, peptides keep the origami size efficiently small. If we use the peptides, it is important we use a mutant peptide as a negative control. It is also important we check to make sure there is no immunogenic response to the peptide. Dean also mentions the use of the amphiphathic trans-cyclohexane-1,2-diol to disrupt hydrophobic region of central NPC channel. If the DNA origami were to have methyl groups around it, it may help as well.
Side Note: DNA origami would probably be an effective way of DNA labelling, so that only certain regions are exposed for labelling such as biotinylation. Especially if we have a single loop exposed from the origami after it is coated with a protective polymer or such. Then only that loop would be treated with the label.
Friday May 24th
DNA staples arrived two days ago. Pooled the staples into four groups of pre-stock: 96 core staples from plate p086, 41 core staples from plate p085, 13 neighbor staples, and 13 polymer staples. The polymer staples are thus named in the event we want to replace them with staples that will polymerize the origami into filaments.
Created 500nM working stock solution of all staples, and used template strand M13mp18. Final concentrations were 20 nM staples, 10 nM template. The DNA origami was folded under a magnesium screen, ranging from 12, 14, 16, 18, 20, 22, 24, 26 mM. Two groups of 2.5 and 4.5 day ramps were then created. There were 16 groups total.
Monday May 27th
Memorial day, harumph harumph.
- Need to schedule TEM this week, get Nanotech West Access, Order calcium phosphate materials
- Literature notes:
- <300nm is the accepted size for tumor extravasation. keep this in mind for any sort of drug delivery mechanism.
- Zhao Xiaobin makes some good points about FR-targed gene therapy as a target for cancer cells.
- Ward et al. shows that the PEG spacer for the folate is important for efficiency.
- There are groups showing FR-targeting with polyplexes, cationic lipids, PEI. FR-targeted liposomes seem to be more efficient than lipofectim commercial products.
- Concerns: endogenous levels of folic acid and competition, intratumoural diffusion, size, insufficient levels of FR on the tumor cells. Coadministration with anti-estrogen or retinoid receptor ligands which can increase FR expression.
Tuesday May 28th
Found a good protocol for using PEG-bis amine with NHS-activated folate . Scheduled TEM training next week. Called Sigma for their materials, here is a list of what I'm thinking at the moment.
|Igepal CO520||107.00||~500 ml|||
|Calcium chloride||29.90||100 g|||
|Disodium Phosphate||26.10||100 g|||
|Sodium Citrate||23.60||25 g|||
|Sodium Metasilicate||31.00||25 g|||
I don't think the metasilicate is absolutely necessary, it seems that it is mostly served as a nucleation agent to help the precipitation of the CPNP, though I also thought it helped limit the size of the particles. I do know it is in Adair's protocol, and there is literature of CP-silicate nanocomposites that exist elsewhere, so I think it's safe to go ahead and make sure we have it.
Also note: Motskin shows cytotoxicity of hydroxyapatite nanoparticles with concentrations above 250 μg/mL, which I'm not too worried about especially systemically , especially since they were microparticles and they drowned the cells in them, and KEster had shown doped CPNPs had no cytoxicity.
possible flourescence schemes
- Propidium iodide - intercalating flouroescent, excitation 535 nm, emission 617 nm. Good DNA origami marker.
- Dope CPNP with flourescein, lanthanides, or Cy3 amidite (Adair)
- Indocyanine green (ICG) as an infrared flourophore for animal trials (Adair bioconjugation)
- Dr. Castro mentioned YOYO-1 and SYBR Gold. These seem to be ultrasensitive Nucleic acid detection and quantitation, mostly due to their precise proportion of dye to the bases.
Finally, I am still thinking there is a huge potential of origami in DNA engineering, by organizing the DNA origami in such a way that the encoding domains are protected while other non-important regions are exposed to some sort of steric-hindered chemical treatment.
Nongradative intracellular trafficiking of highly compacted polymeric DNA nanoparticles 2011 by Kim and Nanoparticles of Compacted DNA Transfect Postmitotic Cells 2003 by Liu
Both of these papers are highly suggestive that compacted DNA is a good thing for the gene transfer. Both labs use 30-mer lysine polymers linked to PEG. The cationic lysine condense the DNA into compacted shapes. Liu had shown that compacted DNA up to 25 nm showed good nuclear transfection efficiency, which is limited by the size of the DNA (after 9.9 kb). Liu also suggests that the polylysine incorporated into the condensed DNA binds to alpha importin with equal affinitiy as an extended SV40 large T antigen NLS, which he points to two immunosorbent binding assays as proof. That's quite a statement.
In general, it seems that poly Lysine or pLy can form toroids of 80-100 nm with DNA. This structure depends on a lysine/nucleotide ratio of 0.4. Chan's lab had shown that the polylysine works when coupled with T-ag NLS (P101) works for importin binding and transfection efficiency.
Wednesday May 29th
Don't forget Dr. Jim Lee has his general lab meetings on 4pm Sunday afternoons.
Talked to Adair. Sent me a patent that summarizes much of the work. Funny enough, they are using 200 micron glass beads that one would use in road paint to help make them reflect light at night! So that actually makes the stationary phase of the HPLC a lot cheaper, since these are easier to obtain.
Found HPLC at Dr. Poirier and Dr. Ottesen's labs. I've suggested taking 200 micron glass beads from Polysciences Inc., and packing them into a 5x55 mm column. The glass beads are cheap, and they would seem to be effective by Adair's suggested Van Der Waals chromatography in which the calcium phosphate will bind to the glass beads in a nonionic eluent of ethanol. If this is sound theory, then we're looking at a rather simple HPLC protocol. Need to hear back from Ottesen on how she sets up the HPLC and what sort of spectrophotomoter and such I could be using. 280 nm is the absorbance for Igepal CO-520 apparently.
Lab work Storage: The 16 groups using 7249 template, 2.5/4.5 days are stored in the fridge in a box with my name labelled on them. Ran the origami in gel. See OhioMOd2013:Methods/gel electrophoresis for more details. I stored pieces of gel that contain each band of fully folded origami away in the fridge. Uploaded picture of the gel below. Will note that lane 12 has a band that is slightly raised, that is the one I took out of the gel. All other gel pieces were taken from the second band observed. You'll notice that lanes 3-5 and lanes 12-15 have the best looking bands. There is significant agglomeration to the later bands where there was increased Magnesium concentrations.
Lane 1 DNA 1kb ladder. Lane 2 DNA phage 7249 control. Lanes 3-10 2.5 day folding, 12-26 magnesium conc. from left to right. Lanes 11-18 4.5 day folding, 12-26 magnesium conc. from left to right.
Friday May 31st
Not much to do for Thursday or Friday. Simply waiting on others.
- Maleimide-PEG-NH2 from PEG creative works.
- Microemulsion notes: Igepal CO-520 is a nonionic surfactant. Read The colloidal stability of flouresecent calcium phosphosilicate nanoparticles. This paper is the best justification for using Igepal, and using Citrate as the dispersion. Citrate serves as a highly charged small molecule effective for dispersion. Paper discusses whether the particles will irreversibly agglomerate with drying.
- Read the DLVO theory. DIscusses the distribution of charges in ionic solutions. Also discuses double layers forces to explain phenomena found in the colloidal sciences.
- Use this for DNA mass/molar calculator
- So if I use 20 nM template concentration for my folding, with 100 ul final volume, then assuming 14 kb structure size at the end, and assuming 100% folding efficiency, then there is 9.09 ug of origami total, yeech, we can get reporter plasmids for 10 mg. I'd like a 100 μg, but it looks like I'm going to be using a smaller range for loading. Maybe something like 0.1, 0.5, 1, 2 μg for every 1 μg of nanoparticle. I can up the concentration by not bothering with magnesium screen and only use 16 magnesium concentration.
- Since 3x10-3 Molar phosphate concentration and 500x100-3 Molar calcium for the microemulsion method, then phosphate will be a limiting factor by no more than 1:1 molar ratio, unless we get monomer crystallization.
- Should probably first measure mass of calcium phosphate nanoparticles without DNA loading.
Still waiting on others. Here's a picture of the nuclear pore complex taken by Dr. Martin Goldberg from Durham University.
Left image: Cytosol side of the membrane. Composed of the cytoplasmic ring, of eight 50 nm long cytoplasmic filaments, with a gap of 70 nm. Right image: Nucleus side of the membrane. Composed of nuclear filaments joining to create a basket-like ring with a gap of 30-50 nm. The central pore itself has a gap of 45-50 nm. The nucleoporin is composed of around 50 different nucleoporins, which occur in multiples of eight.
It's such a good picture I went ahead and made a page simply for the nuclear pore complex.
Well at least the order was approved, still waiting for delivery for materials.
Composed paper asking donation/collaboration from CLL group. Will probably send to Dr. Byrd with Chris Lucas as cc.
Will continue working on Adobe illustrator and on the website.
Performed crunch-n-squeeze purification on the gels I had cut last week. Most of the lab does their purification soon after the cutting of the gel, but I see no problem with storing the origami in a gel for over a week. Added to protocol in methods.
Will do TEM grid Uranyl stain with Carl today, added protocol in methods. I stained purified lanes 4 and 5. Its stored in the top left cabinet with my name and information taped. 4 carbon grid was significantly damaged, so check hte periphery for that. lane 5 carbon grid was well done, but since it was the second grid performed there may still be more crystallization. Hopefully all goes well Friday. I just need one picture of the folded structure.
Literature Reivew: Went over Maitra's paper  in which she states calcium phosphate can be considered a third generation nanoparticle. Instead of Igepal they used Aerosol OT from sigma, and n-Hexane rather than cyclohexane. A good sign is that she only adds 2.94 μg of pDNA to a 25 ml AOT microemulsion of 70 μL 0.35 M phosphate, which was then added to another 25 ml AOT microemulsion containing 70 μL of 1.36 M calcium solution. They were mixed at 20 drops per minute with continuous stirring at 4 °C for 12 hrs. Then reaction is dissolved in absolute ethanol by vortexing. Solution was pelleted for half an hour at 8000 rpm at 4 °C. The pelleted nanoparticles were washed in ethanol three times, then disperesed in distilled water, and dialyzed overnight using 12 kD dialysis membrane bag.
To determine pDNA encaspulation efficiency, she ran 3 to 9 μg of DNA until distinct DNA bands were observed outside of the nanoparticles. Also ran zeta potential at every pH point. This is a pretty good paper to follow, and I may try to adapt this to Adair's protocols.
Also, I found a paper I hadn't seen before  recently published in Jan 2013. They also use the double reverse microemulsion approach using a different non-ionic surfactant ethylene glycol monobutyl ether. There microemulsion was combined for several minutes as well, and citrate was used as the dispersing agent for 10 minutes. THey then centrifuged at 14000 rpm for 15 minutes, and then washed with ethanol and PBS twice to remove residual ethanol. Then dispersed in PBS. Huh, maybe we can do away with HPLC for the time being.
They also measure loading efficiency by agarose gel elecrophoresis and ethidium bromide staining. They loaded with antisense oligonucleotides, so smaller loading. To make sure citrate is on surface, use characterization by the FTIR spectra. USe X-ray Diffraction to determine the phase.
Here's a dissertation paper on nonviral delivery. Random notes from it: HA1, GALA, and melittin are peptides with well-known ph-sensitive endosome lysis. Luciferase is the most reliable indicator of gene transfer. Can use bioluminescent imaging to quantify the luciferase expression in vivo. BLI uses a camera to detect photon emission at 610 nm. There are little false-positive results. I could simply test to see luciferase in the liver.
John Wolff, Professor of Pediatrics and Medical Genetics, U of W-Madison: "Nuclear targeting is the grand problem facing the non-viral gene therapy community". Jesse Gelsinger first to die from adenovirus gene therapy. Can try using H1 linker histone to help wrap the DNA origami? Useful for delivery possibly. See: Peptide nucleic acid staples. Also, photo-chemical coupling of peptide to DNA.
Oh, and the items were finally shipped yesterday!
Waiting on materials. Will try to further prepare materials prior to their arrival.
Did TEM images with Carl. Will upload quality pictures shortly. Will probably take some dimensioning as well. Structures seem well-folded with a few structures attaching to one another at the ends. This is possibly due to how the structure was edited by Dr. Castro for the possibility of of being filamentous.
I also started another DNA folding for bulk production. Used the same folding excel sheets as last time on Friday two weeks ago. Will run at 16 mM MgCl2 and 2.5 day thermal ramp. Will use 20 nM scaffold and 50 μL total volume for each PCR tube. Will run two of these sets for 16 total, and 800 μL of around 20 nM origami. This suggests a 78.99 μg yield assuming ~100% folding. For my initial reaction runs next week, I can use a small range of 0.1, 0.5, 1, 2 μg of DNA for 4 different reactions, thus using 14.4 μg.
Here are the images taken. 1st picture forgot to add captions to image. There are some structures that are sitting vertically with the hollow end visible. This is somewhat suprising and is rarely seen in origami structures, so a lucky find!
|14 mg conc.|
|16 mg conc.|
Finished bulk production of structures, they're sitting in 4°C currently.
Will be meeting with Dr. Poirier with Dr. Castro today, will be discussing use of the HPLC. Currently I feel that I should explore other purification whether by pelleting and evaporation or by other means before I try to adapt to an expensive and complicated HPLC method.
|Material||Molar Mass||Measure Out||Solution|
|CaCl2||110.98 g/mol||0.5549 g in 50 ml||1x10-1 M|
|HNa2HPO4||141.96 g/mol||0.425 g in 50 ml||6x10-2M|
|Na2SiO||122.06 g/mol||0.05 g in 50 ml||8.2x10-3 M|
|Citrate||294.10 g/mol||0.0147 g in 50 ml||1x10-3 M|
I had the materials arrived today. Did an initial reaction today as a trial. I created the stock solutions using ddH2O and filling up the falcon tubes up to 50 ml. Then sonicated the calcium, phosphate, and metasilicate in the sonicator for 10 minutes. Mixed by vortexing.
- Added 4.06 ml of Igepal in 9.94 ml of cyclohexane to two 50 ml beakers.
- Began stirring, then added the calcium, phosphate, and metasilicate per protocol. No DNA origami was added. ALlowed to equilibrate for 30 minutes. The cyclohexane had evaporated significantly in this time.
- Mixed the microemulsions
- Quenched the reaction after two minutes with the citrate solution.
- Let the mixture continue for ten minutes with the citrate
- Ended microemulsion by adding 50 ml ethanol.
- Saw no particulates, so I transfered to a 50 ml centrifalcon tube(orange) and centrifuged for 10 minutes at 2,000 rpm. No pellet was formed. I do not believe I will be able to perform pellet purificaiton for the nanoparticles, since I believe that the origami and the nanoparticles will not pellet since they are so small.
Need to rethink the CPNP microemulsion protocol. Will start with just aqueous solutions.
- Trial 2 (50 ml falcon tube)
- Same as trial 1, used bulb pipette to mix the two microemulsions. Stored in a centrifalcon tube.
- Trial 3
- Made aq solution particles with no citrate present. Added 65 μL of the stock solutions to a 1.5 tube, and 1 ml of ddH20. Vortex. Will wait to see any visible particulate formation.
- Trial 4
- Same as trial 3 but added citrate after mixing reaction for 8 minutes. Should image.
- Trial 5
- Added 65 μL of the stock solutions, 50 μL of the origami non-purified after folding, 420 μL of ddH2O, and NO citrate was added.
- Trial 6
- To a 15 ml falcon tube with ~9ml of ddH2O, I added 100 μL of each stock solution (0.1M Calcium, 0.06M phosphate, 0.08 M silicate, plus 50 μL origami. After 2 minutes, 250 μL citrate was added.
Trial 5 showed significant aggolomeration of the DNA and calcium phosphate, so much so that it was cloudy and visible. It's somewhat nice to see, but I wonder if it's just the DNA agglomerating due to the high salt concentration or the calcium phosphate forming as well.
Trial 6 shows no visible particulates. Maybe it's working with nanoparticles, maybe it's too dilute.
Therefore, there is a shift in strategy: Focus on Aqueous reactions For tomorrow: Three groups of different stoichometry, three groups of DNA loading.
Will be meeting with Yun tomorrow at Nanotech West. Let's try to make a lot of samples. May want to include 1 ml of the microemulsion in cyclohexane. Let's try 994 ul of cyclohexane, 406 ul of igepal, and add 6.5 ul each of the calcium and 125 ul of the diluted phosphate/silicate.
Sample Group 1, aqueous: Created 9 groups of origami/CPNPs. Added DNA first to tube, then calcium followed by mix of phosphate and silicate for 1 ml total. Mixed for 1 minute at RT by vortex. Quenched with citrate for 1 minute. Sonicated for 15 minutes.
|ratios||1.67:1||16.7:1||167:1||quench 1 min|
|0.5nM origami||6.25 mM Ca, 3.75 mM PO4, 492 μM SiO3||6.25 mM Ca, 375 μM PO4, 49.2 μM SiO3||6.25 mM Ca, 37.5 μM PO4, 4.92 μM SiO3||225μM citrate|
|0.5 nM origami||625 μM Ca, 375 μM PO4, 49.2 μM SiO3||625 μM Ca, 37.5 μM PO4, 4.92 μM SiO3||625 μM Ca, 3.75 μM PO4, 0.492 μM SiO3||22.5μM citrate|
|0.05 nM origami||625 μM Ca, 375 μM PO4, 49.2 μM SiO3||625 μM Ca, 37.5 μM PO4, 4.92 μM SiO3||625 μM Ca, 3.75 μM PO4, 0.492 μM SiO3||22.5μM citrate|
Ran samples of these down the ethidium bromide gel. See lanes below.
|ladder||7249 pDNA||0.5nM DNA/6.25nM Ca control||0.5nM DNA/3.74 nM PO4 control||0.5nM DNA/6.25nM Ca/10:6||0.5nM DNA/6.25nM Ca/100:6||0.5nM DNA/6.25nM Ca/1000:6||0.5nM DNA/625μM Ca/10:6||0.5nM DNA/625μM Ca/100:6||0.5nM DNA/625μM Ca/1000:6||0.05nM DNA/625μM Ca/10:6||0.05nM DNA/625μM Ca/100:6||0.05nM DNA/625μM Ca/1000:6|
As one can see, lanes 14-15 of the 0.05 nM DNA are not visible due to the too dilute DNA. Lanes 7-9 seems to have the least free DNA, and lanes 11-13 have more visible bands and thus more free DNA. I suppose I'll take Lanes 7 and 9 on carbon film and see if I can see them on TEM.
It doesn't seem like this was a good assay. Hopefully DLS/Zetasizer will be more enlightening tomorrow on how the particles are.
- Dworetzky SI, Lanford RE, and Feldherr CM. . pmid:3170630.
- van Gaal EV, Oosting RS, van Eijk R, Bakowska M, Feyen D, Kok RJ, Hennink WE, Crommelin DJ, and Mastrobattista E. . pmid:21424159.
- Ritter W, Plank C, Lausier J, Rudolph C, Zink D, Reinhardt D, and Rosenecker J. . pmid:14574456.
- Sebestyén MG, Ludtke JJ, Bassik MC, Zhang G, Budker V, Lukhtanov EA, Hagstrom JE, and Wolff JA. . pmid:9447599.
- Chen X, Kube DM, Cooper MJ, and Davis PB. . pmid:18059369.
- Liu G, Li D, Pasumarthy MK, Kowalczyk TH, Gedeon CR, Hyatt SL, Payne JM, Miller TJ, Brunovskis P, Fink TL, Muhammad O, Moen RC, Hanson RW, and Cooper MJ. . pmid:12807905.
- van Steenis JH, van Maarseveen EM, Verbaan FJ, Verrijk R, Crommelin DJ, Storm G, and Hennink WE. . pmid:12618033.
- Motskin M, Wright DM, Muller K, Kyle N, Gard TG, Porter AE, and Skepper JN. . pmid:19304317.
- Chan CK, Senden T, and Jans DA. . pmid:11083478.
- Bisht S, Bhakta G, Mitra S, and Maitra A. . pmid:15607268.