As mentioned above PLA-PEG copolymers were synthesized by ring opening polymerization (ROP) of D,L-lactide in the presence of mPEG 2000/5000 (monomethyl ether of polyethylene glycol) using stannous octoate as catalyst . Briefly, mPEG and lactide were taken in two neck flask; sealed and dried at 60Â°C for an hour under vacuum. Vacuum was replaced with nitrogen purging and dried toluene was added with syringe and temperature was raised to 140Â°C, followed by addition of 0.05% w/w stannous octoate. Nitrogen atmosphere was provided to protect from oxygen and moisture. The mixture was refluxed at 140Â°C for 18 h under dry nitrogen atmosphere. The solvent was removed under vacuum; the resulted product was purified by dissolving in DCM (Dichloromethane) and precipitation by cold ethyl ether. The precipitation was done twice to enhance the quality of product. The diblock copolymers were then dried overnight in desiccator.
Results and Discussion
Get your grade
or your money back
using our Essay Writing Service!
Synthesis of PLA-PEG diblock copolymer were attempted according to the method reported by Jain et al 2009 . Products were anlysed by 1H-NMR Spectroscopy. LA/EG ratio in the product was calculated using the intergration value of -CH (Î´ 5.1) of lactide and -CH2 (Î´ 3.7) of PEG. Initial attpemts were to synthesize PLA-PEG block copolymer with 5:1 LA:EG ratio. Due to some unfavourable conditions low degree of polymerisation were obtained; also the end product which is supposed to be completely soluble in dichloromethane (DCM) was found to show some precipitates in DCM solution. These observations indicate that the reaction is not going properly and some undesired products are being formed. To further explore these precipates were dissolved in DMSO-D6 and NMR spectra was recorded (Figure C). NMR spectra was also recored for these precipitates with addition of the batch 3 product (Figure D). As clearly seen in the spectra these precipates are not the desired product and modification in the synthesis enviroment is required to avoid such reactions.
Figure : 1H NMR spectra of synthesized copolymers. A- Block copolymer of batch 2; B- Block copolymer of batch 3; C- DCM precipitates of Batch 3 dissolved in DMSO-D6 ; D- Block copolymer of batch3 + DCM precipitates of Batch 3 dissolved in DMSO-D6.
Specific Aim 2: Formulation of Polymeric Nanoparticles
Polymeric nanoparticles are widely explored and accepted as drug/protein/gene delivery vehicle because of their versatile and adaptive physicochemical characteristics which can be modified according to the requirement [4-7]. Among them PLA/PLGA nanoparticles emerged as most preferred choice by researchers because of their easily mouldable degradation, release and surface properties [8-10].
Materials and methods
Double emulsion solvent evaporation
Method was adapted from Jain et al. 2009 Â with modifications. Briefly, aqueous phase containing protein/DNA was added drop-wise along vortexing into with the organic phase (Dichloromethane, DCM or Ethyl Acetate) having polymer (2-8% w/v). The mixture thus obtained was vortexed for 2 more minutes, followed by
Figure: Schematic presentation of formulation method.
probe sonication (???????) in ice bath for 30Â s. To this water-in-oil (w/o) emulsion, 5/10Â ml of 1-3% (w/v) aqueous polyvinyl alcohol was added, vortexed for 2 minutes, then probe sonicated in ice bath for 3Â min to obtain a w/o/w emulsion. The resultant emulsion was stirred at 500rpm for overnight to evaporate organic phase and to obtain the nanoparticles, which were collected by centrifugation at 22,000-30000Â g washed twice with distilled water to remove PVA and subjected to lyophilization.
Optimization of formulation method
Bovine serum albumin (BSA) was used as model protein to encapsulate within nanoparticles for these studies. Formulation method includes various factors which can be modulated to obtain the desired physicochemical properties of the nanoparticles. These factors can be broadly classified to formulation and process variables. Formulation variables include modulation of concentration and ratio of formula components; whereas, process variables include handling and instrument variables to achieve the optimum conditions. Current study focused on formulation of ~200nm nanoparticles with minimum polydispersity along with substantial protein loading efficiency. For the purpose following variables studied, while all other possible variables were kept constant on the basis literature suggestions.
Choice of organic solvent
Internal aqueous phase volume
PVA concentration/external aqueous phase volume
Protein polymer ratio
Process variables (Probe sonication)
Always on Time
Marked to Standard
Final step of formulation is lyophilisation which produces dried powdered form nanoparticles which is stable on long term storage. Though, lyophilisation poses considerable shear stress on the nanoparticles affecting their stability. Abdelwahed et al 2006  has reviewed stress imposed by lyophilisation steps naming freezing, primary and secondary drying; and how does they affect chemical and physical stability of nanoparticles. A range of cryoprotectants are reported in the literature to sustain these stresses, providing stable, unchanged and ready resipersible formulation [11-13]. Trehalose was used in the current study on the basis of its most promising reports in literature [14-17]. Trehalose forms a glassy matrix around the nanoaprticle which stabilizes them during freezing as well as from aggregation during drying. Lyophilization was performed using Advantage, VirTis, Gardiner, NY, USA freeze dryer. Protocol is adapted from .
Table: Lyophilization protocol
Particle size distribution, zeta potential and surface morphology
Lyophilised nanoparticles were resuspended in deionised water (pH 5.57) for the particle size and zeta potential analysis. Particle size analysis was performed using Zetasizer 3000 (Malvern Instruements Ltd, UK) and zeta potential was measured using Nano ZS (Malvern Instruements Ltd, UK) at 20Â°C. Surface morphology of the formulations were anyslised by scanning electron microscopy (Jeol 6500F) .
High performance liquid chromatography
Method was adapted from Umrethia et al. 2010 . Waters 1525 HPLC system equipped with 996 Photodiode Array Detector, Binary HPLC Pump and Jupiter 5Âµm C5 300A column, 250mm in length x 4.6nm in diameter column (Phenomenex, UK) was used for the analysis. Operating conditions were as follows: Solvent A- 0.1% TFA in acetonitrile, solvent B- 0.1% TFA in water with gradient flow (A:B) from 5:95 to 65:35 in 20 min, run time 33 min, UV detection at 210 nm, flow rate 1.0ml/min, column at room temperature.
Determination of protein encapsulation efficiency
Two methods were evaluated for the determination of encapsulation efficiency of the polymeric formulation: alkaline digestion and acetone extraction method.
Alkaline digestion method
Polymeric nanoparticles were digested with 5% sodium dodecyl sulphate (SDS) in 0.1 M sodium hydroxide solution. Lyophilised nanoparticles were suspended in this solution and kept on rocker overnight at room temperature. Next day samples were centrifuged at 40,000 g for 1 hr and supernatant was taken for HPLC analysis, as mentioned above. Standard curve was prepared in SDS/NaOH in range 5-200 Î¼g/ml.
Acetone extraction method
Lyophilised nanoparticles were dissolved in acetone by vortexing 30 min. Followed by centrifugation at 30,000g for 1 hour to get pallet of BSA. Pellet was re-suspended in acetone and centrifuged again to get purified pellet. Followed by, drying overnight and re-suspension in PBST (0.5% Tween 20 in phosphate buffered saline). Samples were then analysed by HPLC.
Results and Discussion
Validation of HPLC Method
As stated above HPLC method for the determination of BSA is adapted from . Before application method was revalidated for range , precision and accuracy.
Range and Linearity
Calibration curve of Bovine serum albumin was prepared in range of 20-1000Î¼g/ml (n=2). As can be seen in Figure linearity was followed upto 300Î¼g/ml. Thus 200Î¼g/ml was taken as higher limit for further experimentations.
Figure : Calibration curve of BSA in range of 20-1000Î¼g/ml.
To evaluate the precision of the method 3 different calibration curves were prepared (2 intraday and 1 interday). Relative standard deviation for all concentration range was below 3% which is an acceptable limit.
Figure : A, B and C showing three difference calibration curves to check the precision of the method. Y axis is peak area and X axis is concentration in Î¼g/ml.
This Essay is
a Student's Work
This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.Examples of our work
Table: Mean, standard deviation (SD) and relative standard deviation (RSD) of peak area observed in HPLC for three different standard curves.
1.31To validate the accuracy of the method three different stock solutions (A, B and C) were prepared to get the concentration 100 and 200 Î¼g/ml. These samples were analysed by HPLC method, mentioned above, in triplicate. Overall relative standard deviation (RSD) values were found to be ~3.3% which is again well under the acceptable limits ( Â±5%).
Table : Showing the peak area observed in HPLC for the 100 and 200 Î¼g/ml concentrations prepared by three different stock solutions A, B and C.
Optimization of formulation
As all the parameters being considered in current study are independent variables, so rather going for factorial design, serial progressive variation strategy was employed to study the effect of these variables. Particle size and size distribution was kept primary selection parameter followed by percent entrapment efficiency.
Probe sonication is an important step for the formulation of nanoparticles by emulsion solvent evaporation method. It provides the shear force to the globules of emulsion decreasing size to nano range. As process variable effect of wattage supply and percentage amplitude supply to the sample during sonication was studied.
Formulation variables were kept constant to dichloromethane as organic solvent, 4% polymer concentration, 5% PVA concentration, 200 Î¼L internal aqueous phase and 40% amplitude to study the effect of wattage supply. 25 and 50 watt were studied. When 25 watt was used, poor emulsification was observed in case of primary emulsion which was not suitable for further second emulsification, whereas no such problem was witnessed with 50 watt. Thus 50 watt was used for further studies.
Formulation variables were kept constant as stated above. The objective of this study was to find the minimum amplitude required for proper emulsification; in order to get 200nm particles with reduced shear stress imposed on protein/DNA present in internal aqueous phase. Initial set of experiments employed a lower amplification for the 1Â° emulsion compared with 2Â° emulsion. When 10% amplification followed by 20% was used, larger nanoparticles sizing around 700 nm with very high polydispersity index (PDI) i.e. 1 were synthesized. As next step amplification 20% followed by 40% were tested; these experiments also generated large particles sizing this time ~600 nm with high PDI. Considering the fact that a stable and monodispersed primary emulsion is required for smaller and monodispersed nanoparticles, next experiments were performed with same amplification for both of emulsions. In this series 40%, 60% and 80% amplitude were tested. 80% amplitude lead to loss of organic solvent during processing. Thus 40 and 60% amplitude were considered for the further study. 40% amplitude generated 378Â±8 nm particles with 0.415Â±0.134 PDI, whereas 60% amplitude was able to produce 294Â±35 nm particles with 0.154Â±0.045 PDI. On the basis of these results 60% amplitude was used for all further experiments.
Internal aqueous phase volume, polymer concentration and PVA concentration was studied as formulation variable. This particular set of experiments was performed with Ms Judith Martin (Level 4 student). One factor was varied at a time keeping other constant.
Figure : Effect of formulation variables on particle size and polydispersity index.
Internal aqueous phase volume was varied to 100, 150 and 200 Î¼l containing 1%w/v BSA.
Formulation of polymeric nanoparticles with Ethyl acetate
Ethyl acetate was used as organic phase in place of DCM. This organic solvent change exhibited considerable decrease in particle size (from >300nm to <150nm). This may be attributed to the higher miscibility of ethyl acetate in water than DCM. Increased miscibility reduces the surface tension at oil-water interface resulting in lower surface free energy on globules (more thermodynamically stable) that enables to obtain better and stable emulsion. These nano droplets are converted to nanoparticles during solvent evaporation phase when organic solvent is diffused from droplets to external aqueous phase leading to gradual shrinkage of droplets and precipitation of polymer at the interface; this diffused organic solvent is than get evaporated from air-water interface. Contrary to DCM, in case of ethyl acetate this solvent removal is more diffusion controlled than evaporation controlled.
As formulation variable polymer concentration, protein loading and external aqueous phase volume were studied. These experiments were started with the parameters optimised with DCM as organic phase, which included 10ml external aqueous phase (PVA solution). With 10ml PVA solution particles were obtained in micron range, which converted to nano range as soon as the volume was decreased to 5 ml. This effect may be attributed to rapid diffusion of ethyl acetate to large aqueous phase. Thus 5ml PVA solution was used for further experiments. When effect of polymer concentration was studied no significant difference was observed in 2% and 4% formulation. 4% polymer concentration was selected for the further studies as can give higher yield per batch.
Different polymer to protein loadings were studied again no difference in particle size was observed in 0.1-0.3% and 1-5% protein loading to polymer. Mean particle size of all these formulations was found to be 127Â±5.8 nm with 0.135Â±0.051 PDI (n=9), which itself states the consistency of these formulations.
Zeta potential of these formulations was found to be -11.1Â±4.48 eV.
Scanning Electron Microscopy
Scanning electron microscopy (Jeol 6500F) was performed to analyse the surface morphology for formulated polymeric nanoparticles. Images in figure indicate that all polymeric nanoparticles were smooth and spherical. In case of lyophilised nanoparticles, they were placed on metal disc with help of adhesive and were analysed under vacuum. Some samples were prepared by air drying a drop of the nanoparticles suspensions on tin foil.
Figure : Scanning electron microscopy. A&B- Lyophilised NPs prepared with DCM. A shows the flakes of trehalose homing NPs within after lyophilisation; B is magnified image showing NPs embraced in trehalose; C&D are air dried NPs prepared with ethyl acetate.
Efficiency of lyophilisation
Lyophilisation converts the nanoparticles suspension to dried powdered nanoparticles. Trehalose was used as cryoprotectant to stabilise the nanoparticles from collapse and aggregation during lyophilisation. Trehalose forms viscous network around nanoparticles thus prevents the physical stress posed by ice crystals formed during freezing steps. Also, trehalose entangles the nanoparticles in its amorphous network which alleviates the possibility of aggregation during the final drying step. Such facts are also evident in scanning electron microscopy images (figure ).
Reconstitution time and effect on particle size are the primary parameters to evaluate the efficiency of the lyophilisation process. All lyophilized formulations were readily dispersible in water on reconstitution, without need of any external shear force. Further, no considerable difference was observed in size of nanoparticles before and after lyophilisation (Table ).
Table: Effect of lyophilisation on particle size and size distribution of nanoparticles.
Thermogravimetric Analysis (TGA)
Thermogravimetric analysis was performed to check the drying efficiency of lyophilisation protocol. Analysis was performed using Q500 TA instruments Inc. Weighed amount of samples were loaded on aluminium pan and was heated in range of 30-200Â°C with heating rate of 10Â°C per minute.
As shown in figure percent weight loss was found to be 3.2-4.5% and followed by no significant weight loss after 100Â°C upto 200Â°C which indicates absence of bound water in formulation.
Figure: Thermogravimetric analysis of different batches of nanoparticles after lyophilisation. Curves are shifted vertically for better visibility.
Comparison of %EE determination method
BSA was used as model protein to encapsulate within polymeric nanoparticles. These particles are required to rupture to get the encapsulated BSA out, followed by its quantification by HPLC. Two different methods were studied naming alkaline digestion and acetone extraction. In alkaline digestion method the PLGA was digested with 5% SDS in 0.1 N NaOH and supernatant was taken for the HPLC quantification. Whereas, in acetone extraction method polymer was dissolved in acetone which was removed by centrifugation, repeated twice and finally BSA pellet was dissolved in PBST. Results obtained by both of the method are depicted in figure.
For the same samples % EE was found to be higher by alkaline digestion method as compare to acetone extraction method; this may be attributed to involvement of extra steps in later method. Also, protein aggregation was observed in acetone extraction method while resuspending in PBST; this factor might have also attributed to lower efficiency of this method.
*This data is produced with Ms Judith Martin, Level 4 student.
Figure: Showing the difference in the %EE of various formulations determined by alkaline digestion and acetone extraction methods.
When ethyl acetate was used as organic solvent much smaller sized nanoparticles were obtained. Thus more detailed study for the effect of other variable with ethyl acetate is required.
Specific aim 3: Formulation of RALA nanoparticles
Cloning and purification plasmid DNAs
The pE9/GFP (pCArG)[19, 20] and CMV/GFP (pEGFP) were used in this study. As stated earlier pCArG is hypoxia responsive promoter and pEGFP was used as control. These plasmids were cloned in Escherichia coli and extracted to pure form using Maxiprep (Invitrogen Ltd, UK) and Gigaprep (Qiagen, UK) kits. Briefly, starter cultures (10 ml) were grown at 37Â°C in LB media containing 30Î¼g/ml kanamycin using rotary shaking incubator for overnight. This starter culture was inoculated to large volume of LB media according to manufacturer's recommendations. After 24 hrs cell suspension was spun down at 4000g for 20min and pellet was frozen till next steps.
Cell Pellet is resuspended in resuspension buffer (50mM Tris HCL pH8.0, 10mM EDTA, 100Î¼g/ml RNAase A). Followed by lysis of cells by addition of lysis buffer (200mM NaOH, 1% SDS) . The pH of lysate was neutralised by neutralization buffer (3.0 M potassium acetate pH5.5). These lysates were then filtered/centrifuged to remove cell debris. DNA containing supernatant was then purified by column provided in kit. Lysates were poured over column as ran under gravity. Purified DNA was collected by addition of elution buffer (1.25 M NaCl, 50mM TrisHCL). DNA in collected solution was precipitated by addition of isopropanol which was then washed with ethanol and finally pellet was air dried and resuspended in molverular grade water.
Characterisation of plasmid
pCArG is 6.1 Kb and pEGFP is 4.7 Kb long circular DNA. Purified plasmid DNAs were characterised by agarose gel electrophoresis. 0.8% agarose was casted at room temperature for 30 min in ??????????? assembly. Plasmids were digested with enzyme Not1 (????????) at 37Â°C. Digested plasmids were loaded on gel along with non digested plasmids as control. 1Kb ladder was also loaded for the localisation of bands.
Figure: Characterisation of plasmid DNAs by agarose gel electrophoresis. L- Standard DNA ladder (2000 and 5000 indicates the size in Kb of the particular band); CC- pCArG Control; CD- pCArG digested; PC-pEGFP control; PD-pEGFP digested.
Figure depicts the band position in gel after run under UV. L lane is 1Kb ladder, CC is undigested pCArG, CD is digested pCArG, PC is undigested pEGFP and PD is digested pEGFP. Undigested plasmids are circular in shape and supercoiled that's why they run longer in gel; whereas digested plasmids are linear and they run in gel according to their chain length which can be compared with the ladder.
Formulation of RALA nanoparticles
RALA nanoparticles were prepared by electrostatic interaction (method developed and optimized by AleK Zholobenko, unpublished data). Briefly, plasmid DNA and RALA protein were mixed to get N:P ratio (nitrogen to phosphate) 10:1. After mixing the solutions it was kept aside for 30 mins at room temperature. Particle size analysis was performed using zetasiser 5000 and was found to be 24.5Â±4.2 nm (number average diameter).
Stability of DNA plasmid from sonication
One of the major reasons for using RALA nanoparticles for the encapsulation of plasmid DNA in polymeric nanoparticles is to provide the stability to the DNA from harsh shear stress imposed by the probe sonication. Probe sonication is required to decrease the globule size of the emulsion during the formulation of polymeric nanoparticles. A preliminary study was performed to check the efficiency of the RALA nanoparticles to provide stability to plasmid DNA.
II.RALA nanoparticles, have condensed plasmid DNA, suspended in water were subjected to probe sonication for 30 sec at a 60% of amplitude. Two different N:P ratio 5:1 and 10:1 were studied in duplicate. As control, pure DNA was also subjected to sonication in similar conditions. The stability was assessed by agarose gel electrophoresis, as explained earlier. Results are depicted in Figure . Figure I shows the RALA nanoparticles post sonication: Lane C is control plasmid DNA without sonication; Lane N is naked DNA which was subjected to sonication; Lane A are 5:1 RALA NPs and Lane B are 10:1 RALA NPs. Lane N clearly evident the smear, indicating complete degradation of DNA; whereas, nothing can be seen in lanes of A and B. This is attributed to condensation of DNA with RALA protein which restricts plasmid's movement from the well. This indicates possible sustained integrity of the RALA nanoparticles as well as the plasmid DNA after sonication.
Figure : Stability of plasmid DNA from sonication. I- samples post sonication; II- samples digested with SDS post sonication. C- Control non-sonicated plasmid; N- Naked sonicated plasmid; L- 1Kb Ladder; A- 5:1 RALA NPs; B- 10:1 RALA NPs.
To further confirm these results, post sonicated samples were digested with SDS and loaded on gel. SDS being a detergent breaks the electrostatic interaction of RALA protein and DNA. As depicted in Figure II, now A and B lanes show bands similar to the positive control (Lane C) DNA which has not undergone sonication; whereas, no band is visible for the negative control (Lane N) which has undergone sonication; indicating complete degradation of these DNA after sonication. The presence of band in A and B at the similar height as positive control indicate that all plasmid DNA are stable and didn't lost their integrity during sonication, which proves efficiency of RALA nanoparticles against sonication stress. Further, if watched carefully faint bands can be seen in lane A above the common band which is absent in lanes of B. This faint band is because of some plasmid DNA which have felt stress and got converted to uncoiled structure from the stable supercoiled structure. Absence of such bands in 10:1 RALA nanoparticles further confirms the optimization of RALA nanoparticles formulation method.
Methods for the quantification of Plasmid DNA
Picogreen is an ultrasensitive fluorescent nucleic acid stain specific for double strand DNA. Quant-iTâ„¢ PicoGreenÂ® dsDNA reagent (Invitrogen) was used for the current study which is claimed to be detect as small as 25pg/ml double stranded DNA (dsDNA) concentration as compare to traditional 260nm UV spectroscopy method which can detect up to 5Î¼g/ml Ds DNA concentration. This assay is not affected by a range of components and keeps its linearity intact, these include: contaminate nucleic acid preparations, including salts, urea, ethanol, chloroform, detergents, proteins, and agarose.
Picogreen assay was performed to make the standard curve of native pEGFP plasmid and pEGFP plasmid condensed with RALA protein. Different aliquots were prepared by serial dilutions to get 10-1000 ng/ml concentration of DNA for both of the cases. Then equal volume of the reagent was added and incubated for 5 min in room temperature. Fluorescence was measured by fluorescence microplate reader (??????) using standard fluorescein wavelengths (excitation ~480 nm, emission ~520 nm). As can be seen in Figure fluorescence intensity (at 1400 gain) for the RALA nanoparticles was found to be very less as compare to the similar concentration of native pEGFP plasmid. This is attributed to the condensation of plasmid DNA with RALA nanoparticles which prevents DNA to get expose to external environment. These results are in line with the observation of previous study which exhibited the efficiency of RALA nanoparticles to stabilise plasmid DNA from harsh environment such as probe sonication. Thus it is required to break these particles for proper quantification.
Figure Calibration curve prepared by Picogreen assay. Tringle: pEGFP, square: RALA:pEGFP.
Specific Aim 8: Transfection studies
Transfection with Lipofectamine 2000
Preliminary transfection studies were performed with PC-3 cells (human prostate cancer cell line). Cells were grown in RPMI 1640 media (Invitrogen, UK) with 10% foetal calf serum (PAA, UK). 50,000/100,000 cells were seeded in each well of 96-well plate for transfection. Lipofectamine 2000 (Invitrogen, UK) was used as transfection agent; 0.2Î¼g of plasmid was condensed with 0.5Î¼L of lipofectamine. 2 hours before the transfection, culture media was replaced with 100 Î¼l of OptiMem (Invitrogen, UK) and incubated in a CO2 incubator. After 2 hrs lipofectamine condensed plasmid DNA were incubated with the cells for next 4hrs. After incubation the medium was replaced with RPMI1640-FCS media and placed in CO2. Fluorescence microscopy (????????) was performed after 24 and 48 hrs of transfection.
Figure shows the bright field and fluorescent images of 50,000cells per well for control, pCArG-Lipofectamine, pEGFP-Lipofectamine after 24 and 48 hrs of transfection. 24 hrs post transfection images for the wells seeded with 100,000 were also taken. As common trend exhibited in all these images no fluorescence was obeserved in control; pCArg exhibited very small traces of fluorescence and pEGFP having CMV promoter exhibited considerably high amount of fluorescence. This is attributed to the hypoxia selectivity of pCArG which expresses only housekeeping amount in normal conditions whereas expression increases many fold in hypoxic condition.
Transfection Study with RALA nanoparticles
Figure : Fluorescence microscopy images of PC-3 cells. First Column: Control cells with no transfection; Second column: cells transfected with pCArG/lipofectamine; Third column: cells transfected with pEGFP/lipofectamine. A&B: Bright field and fluorescent images (exposure time 9sec) at 200X with 50,000 cell density after 24 hrs; . C&D: Bright field and fluorescent images (exposure time 15sec) at 200X with 100,000 cell density after 24 hrs; E&F: Bright field and fluorescent images (exposure time 9sec) at 400X with 50,000 cell density after 48 hrs.
Formulation of RALA NPs containing polymeric nanoparticles
Transfection studies with hypoxic conditions
Transfection studies with ARPE-19 cell line