Marys Medicine

Single-cell mrna transfection studies: delivery, kinetics and statistics by numbers

Single-cell mRNA transfection studies: Delivery, kinetics and statistics by numbers Carolin Leonhardt a, 1, Gerlinde Schwake a,1, Tobias R. Stögbauer, PhDa, Susanne Rappl a, Jan-Timm Kuhr, PhDb, Thomas S. Ligon, PhDa, Joachim O. Rädler, PhDa,⁎ aFaculty of Physics and Center for NanoScience (CeNS), Ludwig-Maximilians-Universität, München, Germany bInstitut für theoretische Physik, Technische Universität Berlin, Berlin-Charlottenburg, Germany Received 15 March 2013; accepted 18 November 2013 In artificial gene delivery, messenger RNA (mRNA) is an attractive alternative to plasmid DNA (pDNA) since it does not require transfer into the cell nucleus. Here we show that, unlike for pDNA transfection, the delivery statistics and dynamics of mRNA-mediated expression are generic andpredictable in terms of mathematical modeling. We measured the single-cell expression time-courses and levels of enhanced green fluorescent protein(eGFP) using time-lapse microscopy and flow cytometry (FC). The single-cell analysis provides direct access to the distribution of onset times, lifetimes and expression rates of mRNA and eGFP. We introduce a two-step stochastic delivery model that reproduces the number distribution ofsuccessfully delivered and translated mRNA molecules and thereby the dose–response relation. Our results establish a statistical framework formRNA transfection and as such should advance the development of RNA carriers and small interfering/micro RNA-based drugs.
2013 The Authors. Published by Elsevier Inc. All rights reserved.
Key words: mRNA transfection; Non-viral gene delivery; Expression kinetics; Single-cell studies; Pharmacokinetics in gene therapy applications. Firstly, mRNA does not requiretransfer into the nucleus and hence mRNA transfection is also Nucleic acid transfer is widely used in basic research as well as effective in non-dividing cells, which is a major drawback of biomedical applications. In recent years, novel stabilized mRNA pDNA transfectionThis makes mRNA a particularly strong constructs have become more prevalent in therapeutic applications therapeutic agent in dendritic cells which are otherwise hard to showing superior properties compared to plasmid DNThis transfectSecondly, immunogenic response to mRNA progress is mostly due to the discovery of 5′ mRNA anti-reverse activated by Toll-like receptors (specifically TLR3) is less cap analogues (ARCA), to the insertion of additional untranslated pronounced compared to unmethylated CpG motifs of DNA regions, and to poly(A) tails that significantly promote and prolong recognized by TLR9In addition, mRNA transfection efficient translation of foreign mRNA inside cellsIn general, remains transient, preventing the risk of permanently integrating mRNA delivery has considerable advantages over pDNA delivery into the genome. Hence, mRNA delivery is of increasing interestfor future biomedical applications in particular with regards tostrategies that aim to use mRNA as a programmable device for This is an open-access article distributed under the terms of the Creative controlled intracellular mRNA targeting and in situ logic Commons Attribution-NonCommercial-ShareAlike License, which permitsnon-commercial use, distribution, and reproduction in any medium, provided evaluation of disease-related conditions the original author and source are credited.
The major hurdle to clinical trials remains the delivery of Financial support by the Elite Network of Bavaria is gratefully nucleic acid to eukaryotic cells. As a result, an ongoing search is acknowledged by CL. This project was supported by the German Excellence still underway for non-viral delivery methods that are optimized Initiative of the Deutsche Forschungsgemeinschaft (DFG) via the Excellence for efficient and controlled delivery of mRNA. Since the first Cluster "Nanosystems Initiative Munich" (NIM), the Sonderforschungsbereich non-viral delivery of mRNA using cationic lipids by Malone, "Nanoagents" SFB 1032, and by the EU-FP7 project ‘‘NanoTransKinetics".
Felgner and many synthetic delivery systems were E-mail address: (J.O. Rädler).
found to be effective for mRNA delivery, with generally better 1 These authors contributed equally to the work.
efficiency found for liposomes than for It is 1549-9634/$ – see front matter 2013 The Authors. Published by Elsevier Inc. All rights reserved.
Please cite this article as: Leonhardt C., et al., Single-cell mRNA transfection studies: delivery, kinetics and statistics by numbers. Nanomedicine: NBM2014;xx:1-10,

C. Leonhardt et al / Nanomedicine: Nanotechnology, Biology, and Medicine xx (2014) xxx–xxx Figure 1. Comparison of mRNA and pDNA Vectors (both gene vectors encoding for the same eGFP protein) and their respective uptake pathways. (A)Linearized RNA (1192 bases) furnished with a stabilizing CAP sequence, an enhancing UTR sequence, and poly-(A) tail. (B) pDNA (4733 base pairs) under thecontrol of the CMV promoter. The vector transfer under identical transfection protocols differs because mRNA is translated after endosomal escape, whileplasmid DNA must be transferred into the nucleus for the initiation of transcription.
generally accepted that both mRNA as well as pDNA are mRNA vector encoding for eGFP. Single-cell fluorescence time- translocated via endosomal uptake, cytosolic release and - in case courses were fitted based on rate equations for translation and of pDNA - nuclear entry. However, mechanistic insights are mRNA/eGFP degradation yielding the onset time distribution, mostly limited to assessment of changes in the transfection mRNA/eGFP degradation rates, and the expression rate. The efficiency as a function of biochemical or structural variations of mRNA expression model applies to at least three different cell the carrier. A full pharmacokinetic model, which in principle has lines. We interpret the cell-to-cell variability in eGFP levels, i.e.
been established using compartment models and rate the distribution of expression rates, in terms of number of lacks validation due to the multitude of kinetic successfully delivered and translated mRNA. The latter is rates. In comparative studies, it was shown that mRNA estimated using a two-step stochastic delivery model. The model transfection compared to pDNA transfection is faster and yields assumes delivery of mRNA in finite size complexes that are a larger fraction of transfected However, a more taken up stochastically by endosomes and randomly released detailed and quantitative understanding in particular of artificial from endosomes into the cytosol. The model quantitatively mRNA delivery is of increasing importance for gaining a reproduces the dose–response relation and yields the correct systems-level description of the kinetics of RNA-based shape of the distribution function. As such, this work represents deviceThe degree of predictive power describing an advance in predictive modeling of mRNA transfection for synthetic RNA expression level and timing will nevertheless quantitative gene expression studies, which we believe will be depend on the degree of accuracy with which the transfer particularly useful for research on siRNA and miRNA kinetics.
efficiency and transfer kinetics can be described. Moreover,predictive modeling of mRNA transfection will be instrumentalfor the advancement of mRNA based therapies. Yet, any non- viral delivery is inherently stochastic and the expression level pDNA and mRNA-vectors and timing of every single cell is different. Hence, measurementsat the single-cell level and analysis of the corresponding Two different vectors for pDNA and mRNA transfection distribution functions are necessary to acquire the true were designed. The peGFP-N1-Vector (commercially available population response in transfection experiments. Using single- at BD Biosciences Clontech, Germany, 4733 base pairs) is the cell analysis, we recently showed that in the case of pDNA standard eGFP vector. As an mRNA reference construct for in transfection, the distribution of gene expression levels can be vitro transcription, we designed a vector that is based on the reproduced using a stochastic Similarly, a recent pSTI-A120-vector (4746 base pairs, transcript 1192 bases), statistical analysis of nanoparticle dosing exhibited Poisson-type which has previously been described in literaturThe distribution in the number of nanoparticles being taken complete vector map is presented in Figure S1. Both vectors Here, we study gene expression after non-viral delivery of contain the same eGFP gene but differ in their promoter region: synthetic mRNA analyzing single-cell expression traces in terms The peGFP-N1-Vector has a strong CMV-promoter for of numbers of complexes delivered and numbers of proteins expression in vitro. The mRNA is generated with a commercial being expressed. Using single-cell fluorescence time-lapse in vitro transcription kit from the pSTI-A120-vector under the imaging and FC, we monitored expression of a cap-stabilized control of the T7 promoter. The backbone of both vectors is

C. Leonhardt et al / Nanomedicine: Nanotechnology, Biology, and Medicine xx (2014) xxx–xxx Figure 2. Representative FC scatter plots for mRNA- and pDNA- mediated eGFP expression in three different cell lines (arbitrary units). (A-F) Two-dimensionalscatterplots (sideward scatter vs. fluorescence intensity) for HeLa, A549 and MDCKII cells 25 h post-transfection with mRNA and pDNA. (G-I) Averagefluorescence intensity per fluorescent cell (RNA data are shown in blue, DNA data are shown in red); (J-L) Percentage of fluorescent cells (mean ± SD).
based on the pCMV-Script vector. pSTI-A120 has a 120-bp from Invitrogen, Germany. Syto RNAselect was purchased from poly(A) tail and a 3′ untranslated region (UTR) from human β- Life Technologies, Germany. 6-well culture plates (Falcon) were globin enabling in vitro transcription of polyadenylated RNA.
purchased from VWR International GmbH, Germany. Sterile To generate in vitro-transcribed mRNA (IVT RNA), the PBS was prepared in-house. Ham's F-12K, MEM, DMEM and plasmid is linearized downstream of the poly(A) tract by SapI Trypsin-EDTA were purchased from GmbH, Germany.
digestion and purified by phenol/chloroform extraction and sodium acetate precipitation. One μg of the linearized vector isused as a template for the in vitro transcription reaction using the A human alveolar adenocarcinoma cell line (A549, ATCC CCL- Biozym Kit (MessageMAX™ T7 ARCA-Capped Message 185) was grown in Ham's F12K medium supplemented with 10% Transcription Kit). Having an Anti-Reverse Cap Analog FBS. HeLa cells (ATCC CCL-2) were cultured using minimum (ARCA) (m 7, 3′-O G[5′]ppp[5′]G) cap on the 5′ end, ARCA essential medium (MEM) with Earle's salts and L-Glutamine cannot be incorporated in the reverse orientation. Thus, 100% of supplemented with 10% fetal bovine serum (FBS). A Madin-Darby the caps in the produced IVT RNA are in the correct orientation, Canine Kidney epithelial cell line (MDCKII, ATCC CCL-34) was increasing the translation efficiency of the IVT cultured in DMEM with 4,5 g/L glucose and 110 mg/L pyruvate,supplemented with 10% fetal bovine serum. All cell lines weregrown in a humidified atmosphere at 5% CO2 level.
FBS, Leibovitz's L-15 Medium (Gibco), Lipofecta- The cells were transfected with equimolar amounts of pDNA mine™2000, OptiMEM (Gibco) and Sybr Gold were purchased and mRNA for FC measurements and with equal weight amounts

C. Leonhardt et al / Nanomedicine: Nanotechnology, Biology, and Medicine xx (2014) xxx–xxx Figure 3. mRNA- and pDNA-mediated gene expression kinetics. (A, B) Exemplary images of an average transfection of A549 cells 25 h post-transfection(overlay of bright field and eGFP fluorescence image. Scale bars 100 μm). (C, D) Representative fluorescence time-courses of eGFP gene expression aftertransfection with mRNA (C) and pDNA (D). To highlight the characteristic differences, we chose and color-labeled three exemplary time-courses each. mRNAexpression shows early onset and continuous rise in the eGFP level, while pDNA expression exhibits delayed onsets and S-shape expression time-courses.
of pDNA and mRNA for single-cell measurements (see eGFP quantification and calibration Supplementary). The same transfection reagent (Lipofecta-mine2000®) and the same standard transfection protocols were To calculate numbers of eGFP molecules from grey values of used for pDNA and mRNA delivery. For transfection with the recorded time-lapse movies, a calibration-channel system fluorescently labelled mRNA, we followed the standard pro- was developed. Micro channels of known dimensions were filled tocols for labelling mRNA with Sybr Gold/Syto RNAselect and with eGFP solutions of defined concentrations. Images of the prepared lipoplexes with labelled mRNA.
channels were taken under the same experimental conditions asthe monitored expression kinetics data, corrected for backgroundand analysed to get calibration curves. For a detailed description Data acquisition and quantitative image analysis of the calibration method, see Supplementary.
Live-cell imaging was performed on a motorized inverted microscope (Nikon, Eclipse Ti-E) equipped with an objectivelens (CFI PlanFluor DL-10 ×, Phase1, N.A. 0.30; Nikon) and eGFP fluorescence intensity in cells was measured by FC with a temperature-controlled mounting frame for the micro- (Partec, CyFlow space). Flow cytometer settings were adjusted scope stage. To acquire cell images, we used a cooled CCD to discriminate transfected and non-transfected cells. The camera (CLARA-E, Andor). A mercury light source (C-HGFIE Windows™ FloMax® software package was used for data Intensilight, Nikon) was used for illumination and a filter cube analysis. See Supplementary for additional information.
with the filter set 41024 (Chroma Technology Corp., BP450-490, FT510, LP510-565) was used for eGFP detection. Anillumination shutter control was used to prevent bleaching.
Images were taken at 10 fold magnification with a constant mRNA vs. pDNA transfection exposure time of 1300 ms at 10-minute intervals for at least25 hours post-transfection. Fluorescence images were consoli- In a first set of experiments, mRNA-mediated transfection dated into single-image sequence files. Negative control images was quantified using FC and compared to pDNA-mediated were taken to assess lamp threshold values and were subtracted transfection as a reference. As schematically depicted in from corresponding image sequence files to eliminate auto- the design of the mRNA vector A) was chosen for fluorescence effects. Using SINGLECELLTRACKER, an in-house- maximal analogy to the pDNA vector. The pDNA vector is a development software based on fluorescence intensi- commercial eGFP plasmid equipped with a CMV promoter ties were integrated over cell contours and corrected for B). The mRNA construct consists of polyadenylated background noise. The software calculates the cells' fluores- RNAs enabling in vitro transcription under the control of the T7- cence over the entire sequence and connects corresponding promoter and contains 2 sequential human β-globin 3′UTRs as intensities to time-courses of the fluorescence per cell.
well as the anti-reverse cap analog (ARCA) (see also

C. Leonhardt et al / Nanomedicine: Nanotechnology, Biology, and Medicine xx (2014) xxx–xxx the x-axis and the sideward scattering signal on the y-axis showconsistent bimodal populations. Both mRNA and pDNAmediated transfection exhibit eGFP-expressing cells and cellsthat do not express any eGFP. However, for three different celltypes, the fluorescence level of eGFP expressing cells in case ofpDNA mediated expression is more broadly distributed andshifted towards higher values than the eGFP distributionappearing in mRNA transfection. This effect is also seen in theintegrated representation, where the distribution of the averagenumber of eGFP molecules per eGFP expressing cell is shown(, G–I). Here, pDNA transfection is shown in red andmRNA transfection in blue. Note that for pDNA transfection,22% (HeLa), 7% (A549), and 28% (MDCKII) of the cells exhibiteGFP expression levels of 1000 (a.u.) and higher that are notshown for better clarity. In the last row (, J-L), thepercentage of transfected cells are depicted, which is a directmeasure of the transfection efficiencies. We find slightly lowerpercentages of transfected cells for mRNA-transfected cellscompared to pDNA-transfected cells except for MDCKII cells,which feature higher transfection for pDNA vectors.
Single-cell mRNA expression kinetics The most revealing difference between transfection with mRNA and pDNA is seen in the single-cell expression kineticsretrieved from time-lapse studies (). Typically, begin-ning after 1.5 hours of incubation, fluorescence microscopymovies were taken over 25 hours using automated time-lapsemicroscopy. The total fluorescence intensity of each single cellwas followed by image and converted into the numberof eGFP molecules per cell (see Supplementary). showstwo typical microscopy images of transfected cells 25 hourspost-transfection (A and B). Bright field andfluorescence images were overlaid to illustrate the fraction oftransfected cells. C and D show gene expression time-courses of single cells. To highlight the characteristic differencesin the expression kinetics, we picked three representative traceseach and show them in color. While mRNA-transfected cellsshow an early and steady rise to a maximum with a subsequentdecrease, pDNA transfection results in sigmoidal intensity time-courses with a steady-state level of eGFP expression and random Figure 4. Single-cell mRNA translation, analyzed by a kinetic rate model. (A) onset times. In contrast to the ubiquitous early onset of eGFP Time-courses of eGFP expression after mRNA transfection (gray lines). Blue expression with mRNA that mainly occurs within 5 hours after lines are fits according to the rate equation model (shown schematically as transfection, the onset of eGFP expression after transfection with insert in (B)). (B) Shows the same data as (A), normalized to their maximalvalue and shifted by their fitted onset times, t pDNA is spread over the range of 2 hours to 20 hours.
0. (C) Distribution of the onset time t0 (mRNA data shown in blue, pDNA data shown in red). (D) Modeling mRNA expression Distribution of the expression rate kTL · m0. (E) Distribution of the mRNAdegradation rate. The black dashed line shows the Gaussian fit to the Since mRNA transgene expression solely involves transla- experimental data, whereas the red dashed line is the Gaussian fit to simulated tion, quantitative modeling reduces to a simple biochemical data (see Supplementary) (F) Distribution of the eGFP decay rate. Dotted linesrepresent the Gaussian fit to experimental (black) and simulated (red) data.
reaction scheme defined by three kinetic rates as shown in, B. The schematic shows a rate equation model formRNA expression consisting of translation, mRNA, and eGFP-degradation. The model is described by the following set of To collate the outcome of the transfection equations for the changes in the number of eGFP molecules, experiments, identical transfection protocols were followed for G(t), and the number of mRNA molecules, m(t): mRNA and pDNA transfection using the commercial cationiclipid agent Lipofectamine2000®.
The FC data shown in were taken 25 hours post- transfection. The scatterplots with the fluorescence intensity on

C. Leonhardt et al / Nanomedicine: Nanotechnology, Biology, and Medicine xx (2014) xxx–xxx course showing an exponential increase with rate δ-β and a long-term decay with decay rate β (see Supplementary). Each fityields an individual set of parameters. C-F presents thecorresponding distribution of the best-fit parameters, which willbe discussed in the following.
Expression onset time distribution In C, the onset time of mRNA (blue) is shown in comparison to the onset time for pDNA transfection (seeSupplementary). The faster transfer of mRNA is clearlydocumented in this distribution. In the case of A549 cellsshown here, the onset time distribution after transfection withmRNA peaks approximately 3 hours after transfection andhardly shows any delayed expression onset events after 5 hours,whereas the pDNA onset time distribution is spread over theinterval between 2 and 20 hours post-transfection. The time-distribution is an indirect, yet quantitative measure for thetransfer time of delivery. As known from microscopy studies,endosomal uptake already starts 10–30 minutes afterTherefore, the measured delay in case ofmRNA transfer must be limited by endosomal escape rates.
Remarkably, mRNA expression onset ceases after 10 hours,indicating that no more endosomes lyse or (more likely) thatmRNA molecules are degraded in acidic late endosomes. Thebroadly distributed onset times for pDNA are associated withrare nuclear entry events, which are believed to occurpredominately during mitosis.
mRNA degradation rates , E shows the distribution of the mRNA degradation rate retrieved from fitting single-cell time-courses with the Figure 5. Dose–response relation. (A) Percentage of positively transfected described model. The average mRNA degradation rate of 0.062/h A549 cells as a function of increasing amount of mRNA (0.05/0.1/0.5/1/ (corresponding to an mRNA life time of t 2 μg). Squares correspond to FC data. The dashed grey line is a single- rough agreement with the literature value of The value Poisson fit, the black line is a double-Poisson fit according to our stochastic is clearly smaller than the degradation rate of endogenous mRNA delivery model. (B) Corresponding fluorescence intensity distributions as measured by FC (bottom to top with increasing mRNA dose).
δ b which is consistent with the reportedly higher stability of ARCA capped mRNA vectors. The distribution ofmRNA degradation is well described by a Gaussian with half- width 0.024/h. This variability in the degradation rate is on the order of the so-called "extrinsic noise" in Thevalues for the degradation of eGFP (with a mean of 0.056/h) are higher than values that have been reported In TL denotes the translation rate and δ and β the degradation rates of mRNA and eGFP, respectively. With t general, it is noteworthy that the single-cell analysis yields 0 being the time of expression onset and the initial conditions G (t estimates for δ and β with high accuracy. The Gaussian fit yields mean values with less than 6% relative error. Knowing the the following solution for the number of eGFP molecules is obtained: degradation rates is of great value for the improvement of novel vectors and capping sequences. Furthermore, the degradationtimes are a key to predicting the time-course of expression. In mRNAðtÞ ¼ kTL⋅m0 ⋅ fact, analysis of Eq. predicts that the maximum of expressionis reached approximately at tmax = 17 h. The time point of half Applying Eq. to the experimental time-courses, the data are maximum expression value in the declining late phase of indeed well fitted. The blue curves in , A show expression is t1/2 = 45 h. The latter is important because it is a exemplary best fits to single-cell time-courses (from a total of measure for the duration of the transient mRNA expression. Note 281 time-courses). There are four free parameters: the onset time that Eq. also holds for the case δ b β (see Supplementary).
t0, the product of translation (kTL) and initial number of Moreover, the expression rate kTL · m0 and the difference in effectively translated mRNA molecules (m0), as well as mRNA the degradation rates (δ-β) both determining the amplitude and and protein degradation rates (δ and β). Eq. entails a time- hence the maximal expression levels, are uncorrelated (see

C. Leonhardt et al / Nanomedicine: Nanotechnology, Biology, and Medicine xx (2014) xxx–xxx Supplementary, Figure S3C). In E and F, Gaussian fitsto simulated data are additionally shown. For simulation, we usedthe experimentally measured mean degradation rates (seeSupplementary). These fits should represent intrinsic noiseonly, which accounts for about 30% of the total noise. Theadditional width of the experimental data can be attributed toextrinsic sources of noise involved in the gene transfer process.
The kinetics of mRNA proves to be generic because different celltypes show the same mRNA expression curves (seeSupplementary).
A stochastic delivery model by Numbers It is generally understood that mRNA as well as pDNA delivery via artificial, non-viral vectors is stochastic anddominated by rare processes. In the case of mRNA transfection,the limiting steps are endosomal uptake, endosomal lysis, andmRNA release from lipoplexes. Here, we ask the questionwhether the measured distribution of expression levels can bereproduced in a stochastic rate model, where each step isassumed to be described by a random process with definedtransition probability. The fact that a large fraction of cells doesnot express eGFP at all indicates that there is a finite probabilitythat no nucleic acid is successfully transferred. Ashows the dose–response curve in terms of the percentage oftransfected cells versus the concentration of mRNA in μg RNAper ml transfection medium. The corresponding distribution ofeGFP expression levels can be seen in B. Data weretaken 25 h after transfection using FC. The number oftransfected cells monotonically increases with mRNA dosage.
It is instructive to describe the transfection process in terms ofnumber of lipoplexes: Lipoplexes form when cationic lipidliposomes are complexed with nucleic acid. Each lipoplexcontains a large average number of mRNA molecules (asdiscussed below). Hence, the delivery of a single lipoplex resultsin a burst of eGFP expression. If lipoplexes were delivered by Figure 6. Two-step stochastic mRNA delivery model. ( overcoming a single barrier, the dose–response function would A) Schematic drawing of the stochastic uptake of lipoplexes by endosomes, lysis of the endosomes, be described by a Poisson-like process as represented by the and release of the mRNA load by lipoplexes. The model reproduces the dashed line in , A (see Supplementary). In this case, the dose–response relation shown in Figure 6, A. (B) Fluorescence autocorre- average number of effectively delivered lipoplexes would be lation function of lipoplexes showing an average hydrodynamic radius of 〈C〉SP = 0.5. However, as shown in , A, the fraction of Rhydr. = 60 nm. (C) Fluorescence image of fluorescently labeled mRNA transfected cells can be more closely described by a chain of two lipoplexes adsorbed to a petri dish at the concentration that was used for time- successive Poisson processes. In this case, the response does not lapse transfection experiments (dose: 1 μg/ml mRNA). Image analysis led toa typical lipoplex density of order 4000/mm2 corresponding to about 4– rise up to 100% at large mRNA concentration, which is due to 8 lipoplexes per cell (intensity scale inverted for clarity, scale bar 25 μm). (D) the fact that the two Poisson processes are sequential. A physical Typical A549 cell five hours after transfection with fluorescently labeled interpretation of such a chain of events is shown in A: The mRNA-lipoplexes (shown in red, scale bar 25 μm). (E) Predicted distribution scheme shows endosomal uptake of lipoplexes, endosomal lysis, of delivered lipoplexes derived from the dose–response relation. (F) and mRNA release from lipoplexes. It is assumed that N Predicted distribution of delivered mRNA molecules, based on an average endosomes are stochastically loaded with a small number of lipo- of 350 mRNA molecules per lipoplex. (G) Experimental probability distribution of expression rates (kTL · m0, black bars) derived from single- eff, and that subsequently a small fraction of endosomes, cell data. Blue line indicates best fit of mRNA distribution to the expression Neff, undergoes lysis. These two stochastic steps are modeled as distribution, yielding an approximate translation rate of kTL = 170/h.
Poisson processes and determine the number of deliveredlipoplexes, C. If we assume the lipoplex load Leff to be propor-tional to the mRNA concentration, i.e. L eff = λ⋅cmRNA, we obtain a that an average of 〈C〉 = Neff Leff = 2 successfully delivered two-parameter expression for the dose–response function (see complexes is obtained. To demonstrate that such a surprisingly , A and Supplementary). The best fit yields Neff = 0.9 and small number of effectively delivered lipoplexes is realistic, we λ = 1.1 μg-1, meaning that at the highest dose of 2 μg, an effective assessed the average number of lipoplexes resting on a single cell in number of Leff = 2.2 lipoplexes are contained per endosome and an experiment. At a dose of 1 μg mRNA and after one hour C. Leonhardt et al / Nanomedicine: Nanotechnology, Biology, and Medicine xx (2014) xxx–xxx incubation time, we found a lipoplex surface density of about can predict the transient course of therapeutic efficacy of mRNA 4000/mm2, corresponding to an average of 4–8 lipoplexes per therapeutics in preclinical studies. For example, the development cell (C). This number is strongly dependent on of improved capping sequences of mRNA vectors can be carried incubation time due to the diffusion limited transport of the out using destabilized eGFP variants. In this case, the protein lipoplexes. After five hours of incubation, the number of level decreases substantially faster and long observation times lipoplexes doubles as seen in , D. We can safely assume causing experimental difficulties can be circumvented (see that almost all lipoplexes that hit the cell surface will be taken Figure S7, Supplementary). Based on kinetic rates obtained in up by endocytosis over time as reported by How- such studies, the time-course of arbitrary gene products with ever, not every endosome releases its lipoplex cargo into the longer half-life times can be inferred. In this context, it should be cytosol. We find that a lysis rate of about 25–50% leads to noted that the half-life of about 12 hours for eGFP determined accordance of the experimental dose–response relation with the from single-cell tracks is shorter than previously reported in above theoretical estimate.
ensemble measurements, which necessarily average over the A single lipoplex contains an average of 〈m〉 = 350 mRNA somewhat heterogeneous timing of whole We molecules. This number is derived knowing the size and packing also showed that the cell-to-cell variability in the expression density of lipoplexes (see Supplementary). The mRNA lipo- levels is well described by a two-step Poisson process. The two- plexes used here exhibit an average hydrodynamic radius of step stochastic model is capable of reproducing the measured 60 nm as measured by fluorescence correlation spectroscopy dose–response curve consistently with the statistical distribution (FCS) , B). The structure and packing density have of expression rates. However, it is limited to transfection in vitro been measured previously using small angle X-ray scattering and provides only an approximate description of the underlying , E shows the theoretical distribution of delivery cascade. The most important element provided by our delivered lipoplexes based on the double-Poisson model and the model is the account of quantal delivery of mRNA in form of mRNA dose that was used for these experiments (1 μg). If this lipoplexes, which is in quantitative agreement with the measured distribution is multiplied with the number of mRNA molecules distribution functions. The small number of successfully per lipoplex, we obtain the theoretical distribution of mRNA per delivered lipoplexes per cell is the key to understanding the cell as shown in , F. It is noteworthy that the theoretical stochastic outcome of transfection experiments that inherently distribution (, G, blue curve) is in very satisfying allow a finite number of non-transfected cells. More refined agreement with the shape of the experimental distribution modeling has to be done to picture the dynamics of transfection (G, black bars) of expression rates. Comparing the and to reproduce the onset time distribution. Here, computational theoretical mRNA distribution with the actually measured representation of size-dependent uptake rates, the nature of distribution of expression rates, kTL · m0, we find kTL = 170/h.
endosome lysis, and intracellular diffusion need to be solved.
This translation rate, which emerges from the analysis of single- Furthermore, computational modeling of extracellular delivery, cell expression rates, is in the range of independently published mimicking in vivo situations, needs to be advanced to gain values of translation impact on translational medicine.
In our experiments, the single-cell time-courses of mRNA- mediated transfection showed excellent agreement with the standard biochemical rate model of translation. Hence, single-cell analysis enables direct determination of expression rates as We studied the expression kinetics of eGFP following well as decay rates for both mRNA and eGFP with great transfection mediated by mRNA and pDNA. While pDNA accuracy and provides a quantitative foundation for kinetic complexes have to enter the nucleus, mRNA molecules released studies on mRNA translational regulation as for example RNA from mRNA lipoplexes can be translated immediately after interference. The fact that mRNA transfection exhibits a narrow endosomal escape. Consequently, mRNA-induced expression is time window of delivery is beneficial for kinetic studies. This profoundly earlier and more homogeneously timed than pDNA- advantage should be of practical importance for future time- induced expression. This behavior is generic and similar onset resolved studies on siRNA knockdown and RNA constructs for time distributions are observed e.g. for HeLa and MDCKII cells programmed gene regulatory operations.
(data not shown). The high transfection efficiencies for pDNAtransfected cells as compared to mRNA transfected cells might be a result of size-dependent lipoplex uptake that has beenreported We determined the pDNA-lipoplexes to We thank Carsten Rudolph for the friendly gift of the vector be about 230 nm in diameter (data not shown), as opposed to pSTI-A120, Svenja Lippok for FCS measurements, David Smith 120 nm for mRNA-lipoplexes. The narrow timing of mRNA for proof-reading of the manuscript, and Maria P. Dobay for expression onset at approximately 3 hours post-transfection is in agreement with the observed timing found for endosomal uptakeand release in single-particle tracking Therefore, themRNA expression onset distribution might serve as a valuable Appendix A. Supplementary data indicator for the endosomal release time distribution and couldbe useful for the advancement of artificial endosomolytic agents.
Supplementary data to this article can be found online at Furthermore, our data imply that mRNA expression modeling C. Leonhardt et al / Nanomedicine: Nanotechnology, Biology, and Medicine xx (2014) xxx–xxx 20. Rinaudo K, Bleris L, Maddamsetti R, Subramanian S, Weiss R, Benenson Y. A universal RNAi-based logic evaluator that operates in 1. Tavernier G, Andries O, Demeester J, Sanders NN, De Smedt SC, mammalian cells. Nat Biotechnol 2007;25(7):795-801.
Rejman J. mRNA as gene therapeutic: how to control protein expression.
21. Win MN, Smolke CD. Higher-order cellular information processing with J Control Release 2011;150(3):238-47.
synthetic RNA devices. Science 2008;322(5900):456-60.
2. Yamamoto A, Kormann M, Rosenecker J, Rudolph C. Current 22. Leisner M, Bleris L, Lohmueller J, Xie Z, Benenson Y. Rationally prospects for mRNA gene delivery. Eur J Pharm Biopharm 2009; designed logic integration of regulatory signals in mammalian cells. Nat 3. Andries O, De Filette M, Rejman J, De Smedt SC, Demeester J, Van 23. Xie Z, Wroblewska L, Prochazka L, Weiss R, Benenson Y. Multi-input Poucke M, et al. Comparison of the gene transfer efficiency of mRNA/ RNAi-based logic circuit for identification of specific cancer cells.
GL67 and pDNA/GL67 complexes in respiratory cells. Mol Pharm 24. Carothers JM, Goler JA, Juminaga D, Keasling JD. Model-driven 4. Stepinski J, Waddell C, Stolarski R, Darzynkiewicz E, Rhoads RE.
engineering of RNA devices to quantitatively program gene expression.
Synthesis and properties of mRNAs containing the novel "anti-reverse" cap analogs 7-methyl(3′-O-methyl)GpppG and 7-methyl (3′-deoxy) 25. Malone RW, Felgner PL, Verma IM. Cationic liposome-mediated RNA GpppG. RNA (New York, NY) 2001;7(10):1486-95.
transfection. Proc Natl Acad Sci U S A 1989;86(16):6077-81.
5. Jemielity J, Stepinski J, Jaremko M, Haber D, Stolarski R, Rhoads RE, 26. Bettinger T, Carlisle RC, Read ML, Ogris M, Seymour LW. Peptide- et al. Synthesis of novel mRNA 5¢ cap-analogues: dinucleoside P1, mediated RNA delivery: a novel approach for enhanced transfection of P3-Tri-, P1, P4-Tetra-, and P1, P5-Pentaphosphates. Nucleosides primary and post-mitotic cells. Nucleic Acids Res 2001;29(18):3882-91.
Nucleotides Nucleic Acids 2003;22(5–8):691-4.
27. Rejman J, Tavernier G, Bavarsad N, Demeester J, De Smedt SC. mRNA 6. Zohra FT, Chowdhury EH, Tada S, Hoshiba T, Akaike T. Effective transfection of cervical carcinoma and mesenchymal stem cells mediated delivery with enhanced translational activity synergistically accelerates by cationic carriers. J Control Release 2010;147(3):385-91.
mRNA-based transfection. Biochem Biophys Res Commun 2007;358(1): 28. Debus H, Baumhof P, Probst J, Kissel T. Delivery of messenger RNA using poly(ethylene imine)-poly(ethylene glycol)-copolymer blends for 7. Kuhn AN, Diken M, Kreiter S, Selmi A, Kowalska J, Jemielity J, et al.
polyplex formation: biophysical characterization and in vitro transfection Phosphorothioate cap analogs increase stability and translational properties. J Control Release 2010;148(3):334-43.
efficiency of RNA vaccines in immature dendritic cells and induce 29. Lin AJ, Slack NL, Ahmad A, George CX, Samuel CE, Safinya CR.
superior immune responses in vivo. Gene Ther 2010;17(8):961-71.
Three-dimensional imaging of lipid gene-carriers: membrane charge 8. Pesole G, Grillo G, Larizza A, Liuni S. The untranslated regions of density controls universal transfection behavior in lamellar cationic eukaryotic mRNAs: structure, function, evolution and bioinformatic liposome-DNA complexes. Biophys J 2003;84(5):3307-16.
tools for their analysis. Brief Bioinform 2000;1(3):236-49.
30. Chan C-L, Majzoub RN, Shirazi RS, Ewert KK, Chen Y-J, Liang KS, et 9. Kormann MSD, Hasenpusch G, Aneja MK, Nica G, Flemmer AW, al. Endosomal escape and transfection efficiency of PEGylated cationic Herber-Jonat S, et al. Expression of therapeutic proteins after delivery of liposome–DNA complexes prepared with an acid-labile PEG-lipid.
chemically modified mRNA in mice. Nat Biotech 2011;29(2):154-7.
10. Holtkamp S, Kreiter S, Selmi A, Simon P, Koslowski M, Huber C, et al.
31. Safinya CR, Ewert K, Ahmad A, Evans HM, Raviv U, Needleman DJ, et al.
Modification of antigen-encoding RNA increases stability, translational Cationic liposome–DNA complexes: from liquid crystal science to gene efficacy, and T-cell stimulatory capacity of dendritic cells. Blood delivery applications. Philos Transact R Soc A, Math Phys Eng Sci 11. Zabner J, Fasbender AJ, Moninger T, Poellinger KA, Welsh MJ. Cellular 32. Varga CM, Hong K, Lauffenburger DA. Quantitative analysis of synthetic and molecular barriers to gene transfer by a cationic lipid. J Biol Chem gene delivery vector design properties. Mol Ther 2001;4(5):438-46.
33. Dinh AT, Pangarkar C, Theofanous T, Mitragotri S. Understanding 12. Wilke M, Fortunati E, vandenBroek M, Hoogeveen AT, Scholte BJ.
intracellular transport processes pertinent to synthetic gene delivery via Efficacy of a peptide-based gene delivery system depends on mitotic stochastic simulations and sensitivity analyses. Biophys J 2007;92(3):831-46.
activity. Gene Ther 1996;3(12):1133-42.
34. Kamiya H, Akita H, Harashima H. Pharmacokinetic and pharmacody- 13. Brunner S, Sauer T, Carotta S, Cotten M, Saltik M, Wagner E. Cell cycle namic considerations in gene therapy. Drug Discov Today 2003;8(21): dependence of gene transfer by lipoplex polyplex and recombinant adenovirus. Gene Ther 2000;7(5):401-7.
35. Zou S, Scarfo K, Nantz MH, Hecker JG. Lipid-mediated delivery of 14. Mockey M, Goncalves C, Dupuy FP, Lemoine FM, Pichon C, Midoux P.
RNA is more efficient than delivery of DNA in non-dividing cells. Int J mRNA transfection of dendritic cells: synergistic effect of ARCA mRNA capping with Poly(A) chains in cis and in trans for a high protein 36. Guo P. The emerging field of RNA nanotechnology. Nat Nanotechnol expression level. Biochem Biophys Res Commun 2006;340(4):1062-8.
15. Kuhn AN, Diken M, Kreiter S, Vallazza B, Türeci Ö, Sahin U.
37. Schwake G, Youssef S, Kuhr JT, Gude S, David MP, Mendoza E, et al.
Determinants of intracellular RNA pharmacokinetics: implications for Predictive modeling of non-viral gene transfer. Biotechnol Bioeng RNA-based immunotherapeutics. RNA Biol 2011;8(1):35-43.
16. De Haes W, Van Mol G, Merlin C, De Smedt SC, Vanham G, Rejman J.
38. Summers HD, Rees P, Holton MD, Rowan Brown M, Chappell SC, Internalization of mRNA lipoplexes by dendritic cells. Mol Pharm Smith PJ, et al. Statistical analysis of nanoparticle dosing in a dynamic cellular system. Nat Nano 2011;6(3):170-4.
17. Arthur JF, Butterfield LH, Roth MD, Bui LA, Kiertscher SM, Lau R, 39. Grudzien E, Stepinski J, Jankowska-Anyszka M, Stolarski R, Darzyn- et al. A comparison of gene transfer methods in human dendritic cells.
kiewicz E, Rhoads RE. Novel cap analogs for in vitro synthesis of Cancer Gene Ther 1997;4(1):17-25.
mRNAs with high translational efficiency. RNA (New York, NY.
18. Strobel I, Berchtold S, Gotze A, Schulze U, Schuler G, Steinkasserer A.
Human dendritic cells transfected with either RNA or DNA encoding 40. Youssef S, Gude S, Rädler JO. Automated tracking in live-cell time- influenza matrix protein M1 differ in their ability to stimulate cytotoxic T lapse movies. Integr Biol 2011;3(11):1095-101.
lymphocytes. Gene Ther 2000;7(23):2028-35.
41. de Bruin K, Ruthardt N, von Gersdorff K, Bausinger R, Wagner E, Ogris 19. Seelig G, Soloveichik D, Zhang DY, Winfree E. Enzyme-free nucleic M, et al. Cellular dynamics of EGF receptor-targeted synthetic viruses.
acid logic circuits. Science 2006;314(5805):1585-8.
Mol Ther 2007;15(7):1297-305.
C. Leonhardt et al / Nanomedicine: Nanotechnology, Biology, and Medicine xx (2014) xxx–xxx 42. Sacchetti A, El Sewedy T, Nasr AF, Alberti S. Efficient GFP mutations 45. Rädler JO, Koltover I, Salditt T, Safinya CR. Structure of DNA-cationic profoundly affect mRNA transcription and translation rates. FEBS Lett liposome complexes: DNA intercalation in multilamellar membranes in distinct interhelical packing regimes. Science 1997;275(5301):810-4.
43. Blake WJ, Kaern M, Cantor CR, Collins JJ. Noise in eukaryotic gene 46. Rädler JO, Koltover I, Jamieson A, Salditt T, Safinya CR. Structure and expression. Nature 2003;422(6932):633-7.
interfacial aspects of self-assembled cationic lipid−DNA gene carrier 44. Li X, Zhao X, Fang Y, Jiang X, Duong T, Fan C, et al. Generation of complexes§. Langmuir 1998;14(15):4272-83.
destabilized green fluorescent protein as a transcription reporter. J Biol 47. Ross PC, Hui SW. Lipoplex size is a major determinant of in vitro lipofection efficiency. Gene Ther 1999;6(4):651-9.


Rapid detection of genetically modified organisms on a continuous-flow polymerase chain reaction microfluidics

Analytical Biochemistry 385 (2009) 42–49 Contents lists available at Analytical Biochemistry Rapid detection of genetically modified organisms on a continuous-flowpolymerase chain reaction microfluidics Yuyuan Li, Da Xing *, Chunsun Zhang MOE Key Laboratory of Laser Life Science and Institute of Laser Life Science, South China Normal University, No. 55, Zhongshan Avenue West, Tianhe District,Guangzhou 510631, People's Republic of China

TAKORADI POLYTECHNIC JOURNAL OF TECHNOLOGY AN OFFICIAL JOURNAL OF TAKORADI POLYTECHNIC Volume 3, Number 1, April 2014 Published by TAKORADI POLYTECHNIC TAKORADI, GHANA, WEST AFRICA © Takoradi Polytechnic Journal of Technology All rights reserved; no part of this publication may be reproduced, stored, in a retrieval system, transmitted