Polymeric cations are widely used as nonviral transfection agents. A fundamental understanding of the structure-internalization relationship that underlie the cytotoxicity of these molecules, however, is still lacking to date and mostly based on empirical observations. We have assessed the cytotoxicity profile of a polymer library synthesized from (built from?) L-lysine monomer units containing 17 members in order to systematically investigate the influence of all the structural parameters on cellular toxicity. The library contained linear, hyperbranched and dendritic L-lysine analogues in a broad range of molecular weights 400-140 000 g/mol. Hyperbranched poly-L-lysine, which is newly designed alternative to polyethylenimine in gene delivery, was compared to linear and dendritic poly-L-lysine for its toxicity. We investigated the effect of the molecular weight (Mn), degree of branching and polydispersity on the mechanism of short and long term cell toxicity in vitro in two different cell lines. The mechanisms underlying cell death at various stages of cell exposure to polycation were identified. Acute cell death was triggered via lethal membrane injury and increased osmotic pressure. In contrast delayed cell death was related to the intracellular accumulation of the polycation resulting eventually to mitochondrial collapse. The onset and extent of these specific modes of cell death were shown to be dependent on the molecular weight and degree of branching of the poly-L-lysine analogues. The number of cells displaying signs of early apoptosis was related to the degree of branching and molecular weight in similar fashion as for the acute cell death. Most importantly the novel L-lysine analogue, hyperbranched poly-L-lysine, elicits the same stress induced response in cells as its structural analogue dendritic poly-L-lysine of comparable molecular weight and polyethyleneimine.
Keywords: linear poly-L-lysine, L-lysine dendrimer, hyperbranched poly-L-lysine, degree of branching, apoptosis, toxicity, acute cell death.
Introduction
Increasing number of studies (investigations) focuses on the application of newly designed polycations for delivery of nucleic acids for transient gene expression. Polyethylenimine (PEI), linear Poly-L-lysine (LPL) or polyamidoamine (PAMAM) represent examples of widely used polycations in biotechnological or medicinal research for the gene delivery to mammalian cells in vitro and in vivo.1 The native, unmodified form of these polymers is directly investigated for this purpose, however structural derivatization of PAMAM, PEI and LPL lead in many cases to polycations with improved properties like increased degradability2, enhanced transfection efficiency3 or targetability4. These results provide motivation to understand the effect of the polymer architecture on the basic biological properties and to advance the rational design and synthesis of more efficient carriers. The most important criteria in focus for gene delivery are delivery efficiency5, internalization properties6-8, proper intracellular trafficking9, 10 and toxicity/degradability11, 12. These properties are directly determined by structural characteristics of the polymer and can be fine-tuned for example by modulating the polymer sequence, its final molecular weight or by optimizing the (co)monomer(s) type13, 14.
The polycation induces in mammalian cells two types of stress related responses which are separated in time - immediate and delayed cell death. These two mechanisms induce distinct morphological and molecular changes in the cells that can be differentiated.
The relationship between immediate toxicity and molecular weight of a polycation was documented in the past and it was demonstrated that the extent of the induced acute cell death is clearly concentration and molecular weight dependent for polycations.15 The positive charge of the polycations is effective in disrupting plasma membranes in living cells or model membranes16-18. Membrane lysis induced by electrostatic interactions thus has been proposed as the main cause of immediate toxicity induced by polycations. This observation correlates with the extensive hemolysis of red blood cells or lactate dehydrogenase release in vitro caused by polycations19, 11. Suppression of the lytic activity of polycations can be achieved by end group modification aimed to reduce the overall charge density. In this perspective PEGylation as well as esterification with acetic acid or long chain fatty acids lead to 10 fold increase in cell viability (biocompatibility). 20 21, 22
Long term impact of the polycation exposure to the cells has been recently analyzed in vitro 23-26. These studies demonstrated that at delayed phases after the polycation treatment the delayed cell death program in the cells is lanced via a cascade of predefined events. It has been proposed that once internalized into the cells polycations may elicit the apoptotic response by distinct yet simultaneous mechanisms27 and that, based on their structure polymers have variation in their mechanism of cell death24. In general polycations have been shown to induce apoptosis through alteration of mitochondrial activity, where depolarization of mitochondrial membrane potential and Cyt C release occurs with activation of caspase-9 and caspase-3 27, 28. However, most of the newly designed and synthesized polycations are only very rarely tested for their pro-apoptotic activity; consequently their long term cell compatibility is generally unknown. Similarly systematic studies on the apoptosis as a function of polymer architecture are missing. Apoptosis, as a long term mode of cell death, has been investigated to this time for poly-L-lysine 24 PAMAM dendrimer 29 and linear or branched PEI 23, 27, however these studies did not directly focus on the concrete role of the architecture of the given polycation.
In this manuscript we present a detailed investigation focusing on the cytotoxicity profile of a polymer library based on a single trifunctional monomer unit L-lyisine. Our library consists of linear- (LPL), dendritic- (DPL) and hyperbranched poly-L-lysine (HBPL) polymers of varying size, and contains altogether 17 members. LPL and DPL are monodisperse and well characterized polymers and widely studied as vehicles for drug and gene delivery applications .HBPL, a recently developed polycation, is a branched analogue of linear poly-L-lysine, that can be cost effective synthesized on a large scale via a one step polycondensation process30, 31. HBPL is a promising candidate for applications in gene delivery (ZK manuscript in preparation). Due to its efficiency in gene delivery, low production costs, and improved biodegradability, HBPL may be especially suitable for in-vivo applications and large scale production of recombinant proteins via transient gene expression. Thus the additional aim of this study was to characterize the cytotoxic properties of HBPL in comparison to LPL and DPL. For this purpose we used industrially relevant cell lines CHO DG4432. Our systematic investigation of the influence of degree of branching, molecular weight and architecture of polycation on the execution and extent of apoptotis may be of help to further optimize the design of the polycation vectors used for large scale gene delivery purposes.
Experimental Section
Materials
Supporting Information
Techniques
See the Supporting Information for Techniques section
Methods-Synthetic procedures
See the Supporting Information for Methods-Synthetic procedures section
Cell culture
CHO DG44. Suspension cultures of CHO DG44 cells were routinely maintained in square-shaped glass flasks (250mL) in serum free ProCHO5 medium (Lonza AG, Verviers, Belgium) supplemented with 13.6 mg/l hypoxanthine, 3.84 mg/l thymidine, and 4 mM glutamine (SAFC Biosciences, St. Louis, MO) as described by Muller et al 33. Cultures were maintained according to standard protocols at 37 °C in a 95% humidified incubator with 5% CO2.
Fluorescence Microscopy. Zeiss Axiovert 200M inverted microscope was used for fluorescence light microscopy, equipped with Zeiss AxioCam MRm camera. Images were analyzed with Axiovision software.
Flow cytometry. Flow cytometric analysis was performed using a Cyan ADP (Beckman Coulter, Fullerton, CA) equipped with three lasers (405nm / 488nm / 635nm) using Summit 4.2 software (BeckmanCoulter, Fullerton, CA). Data were further processed with Flow Jo Software.
Osmolarity measurement. The values of osmolarity were measured with a freezing point osmometer Multi Osmette Osmometer (Precision SystemsInc, Natick, MA) calibrated with a range of standards 100, 500, 1000, 1500 and 2000 mOsm/kg. Each sample was measured in three separate experiments with three separate solution preparation.
In vitro assessment of acute cytotoxicity. In vitro assay applying Viacount Reagent (Guava Technologies) was performed as following. For typical cell sample at a concentration of 2*106 CHO cells per mL 20uL of cell suspension was mixed with 180 uL Viacount Reagent in a 5mL round bottom polypropylene tube. Cells were resuspended and incubated for 5 minutes. The total volume was diluted with 1mL PBS. Samples were filtered prior the flow cytometry analysis. Experiment was carried out in three independent replicate. All polymer samples were tested in the concentration range of 10-2 M to 10-10M,
Membrane permeabilization assessment. Cells were treated with the polycation with defined range of concentration for 1hour. Cells were afterwards washed with PBS and centrifuged at low speed. Dextran-rhodamine solution in PBS was added to the cells and incubated for 15 min at 4°C with shaking. Cells with permeabilized cell membrane presented diffusive cytoplasmic staining of cytosol. Experiments were performed in triplicate. Standard errors were within 15 % of the mean values.
Mitochondrial Membrane Potential Measurement. The change in the mitochondrial membrane potential (ψm) was analyzed by fluorescence wide field microscopy and flow cytometry using the ψm-sensitive dye 5,5',6,6'-tetrachloro-1,1',3,3'tetraethylbenzimidazolylcarbocyanine iodide (JC-1) (Invitrogen AG, Basel, Switzerland). Briefly, cells were treated with polycation for up to 4 hours. Cell aliquots harvested at defined time point (0.5h, 1h, 2h, and 4h), were extensively washed in PBS and resuspended in culture medium containing 1 mM JC-1. Followed incubation at 37 °C for 20min, cells were washed with PBS and analyzed by wide field microscopy and flow cytometry. The emission maxima of JC-1 monomers and aggregates are 527 and 590 nm respectively. . Camptothecin at a concentration of 10 μM was used as a positive control for this experiment.
Annexin V-FITC/PI assay for apoptosis. Apoptotic cells and dead cells were detected by Annexin-V binding and nuclear PI incorporation, respectively. Approximately 2*105 cells were sampled and washed in cold PBS. Binding buffer was prepared beforehand (Hepes buffer: 10 mM HEPES/NaOH, pH 7.4, 150 mM NaCl, 5 mM KCl, 1 mM MgCl2, 1.8 mM CaCl2). FITC-Annexin V was diluted at a concentration of 1 mg/mL in binding buffer and cells were resuspended in this solution. 0.1 mL of PI solution was added (final concentration,3 μM) and incubated in the dark at room temperature. Samples were diluted with PBS to the final volume of 2 mL and quantification of the apoptotic / dead cell population was carried out with flow cytometry.
Immunocytochemistry - Detection of caspase 3, caspasae 9, cytochrome C. To evaluate the presence of caspase 3, caspase 9, and cytochrome C within the cells immunocytochemistry analysis were performed on PC12 cells. Morphological features were examined by growing them on 35 mm round coverlips. Before treatment, cells were seeded into 6-weel plates with coverslips inside, at density of 1.5X105/well, left overnight to attach to the substratum and to acquire the normal flattened morphology, and finally challenged with veichles (they were treated for 12h with different polymers). After the treatment period cells were washed twice with 1X PBS and then fixed with 4% paraformaldehyde for 30 min on ice. Fixed cells were washed again with 1X PBS and permeabilized with 1X PBS containing 0.1% Tween, for 15 minutes at room temperature. Non-specific binding was blocked by 30 minutes incubation at 37°C with PBS containing 5% secondary antibody serum (goat for caspases 3 and 9, rabbit for cytochrome C). For proteins visualization coverslips were stained with specific primary antibody (from Sigma-Aldrich, Milan, Italy) 1:200 in 1X PBS containing 1% secondary serum overnight at 4°. Then they were washed three times with 1X PBS and subsequently incubated with secondary antibody FITC conjugated (from Sigma-Aldrich, Milan, Italy) 1:500 in 1X PBS, for 45 minutes in the dark at room temperature. Cells were washed three times with 1X PBS and subsequently were incubated for 20 minutes in the dark at room temperature with 1 mg/ml 4,6-diamidino-2-phenylindole (DAPI, Sigma), which selectively stains DNA and allows for the evaluation of nuclear morphology. Slides were mounted with DABCO glycerol. Cells were examined under a fluorescent microscope equipped with the appropriate filters (Nikon Italy).
Results and Discussion
Generation of a polymer library base on L-Lysine monomers
In order to determine how the architecture of polymers may contribute to their potential cytotoxycity we have synthesized a polymer library based on L-lysine. Our library consisted of three structural types of polymers, i.e. linear, branched, and dendrimeric (Fig. 1), and contained samples with varying molecular weight, degree of branching and polydispersity. The type of poly-Lysine counterion, was homologous for all the samples, i.e. HCl, in order to limit the crosstalk between additional parameter 34, other than the Mn, DB and PDI. The final aim is evaluation of the role of the polymer architecture on cell cytotoxicity we initially characterized all the synthesized polymers for their degree of branching, molecular weight, polydispersity, number of accessible amino groups and buffering capacity.
Insert figure 1 here
Synthesis of linear Poly-L-lysine. Linear poly-L-lysine was synthesized via polymerization of N-carboxy anhydride monomer. This strategy allowed us to successfully synthesize linear poly-L-lysine molecules with a molecular weight of up to 20 000 g/mol (Table 1). For the different analogues, the theoretical molecular weight, determined by the ratio of initiator to monomer was in agreement with the values obtained by NMR analysis. On the other hand GPC measurements (Figure S4) lead to a certain overestimation of the molecular weight. GPC analysis also revealed a relatively low polydispersity for all analyzed samples.
Insert table 1 here
Synthesis of dendrimer of L-lysine. Dendrimers were synthesized up to the 6th generation via standard peptide coupling strategy, employing HOBT and HBTU as coupling agents (Table 2). The structure of dendrimers was verified by 1H NMR (Figure 1), integrals of the structural unit correspond with theoretical values and the deviation does not exceed 8 % (table S1). Dendrimer displayed relatively low polydispersity as documented by GPC chromatograms (Figure S3). However, further MS (ES+) analysis revealed a certain heterogeneity especially for those dendrimers of generation higher than the forth. The divergent strategy used here for the construction of the dendrimers is prone to induce structural errors, mainly due to incomplete coupling occurring during the sequential steps of polymerization. Similar types of irregularities were previously reported for the synthesis of poly (propyleneimine) and PAMAM dendrimers35.
Insert table 2 here.
Synthesis of hyperbranched poly-L-lysine. Polycondensation of the AB2 monomer L-lysine, are known to lead to inherently heterogenous structures with relatively high polydispersity. In addition, the thermal degradation of the monomer units or of the growing polyamide chain can generate additional structural imperfections 36. We attempted to limit these side reactions and thus reduce heterogeneity, by carrying out the polymerization reaction under a continuous stream of nitrogen. When determined experimentally by ???,compounds obtained that way, exhibited a slightly lower number of free amino groups than the theoretically predicted value. In order to distinguish and to understand the influence of the structural irregularities of HBPL on the toxicity, all the in vitro studies were carried out simultaneously with structurally faultless analogues LPL and DPL.
As evident from the table 3 the polycondensation and subsequent fractionation, leads to multibranched peptidic structure of very high molecular weight. Contrary to the conventional peptide coupling strategy or NCA polymerization used for LPL and D respectively, polycondensation is easily scalable and cost effective synthesis method. Furthermore, the reaction conditions are free of any potentially toxic catalyst or reagent making the purification procedure extremely easy.
Insert table 3 here
IMMEDIATE CELL DEATH UPON POLYCATION ADDITION (EXPOSURE)
Acute toxicity has been traditionally defined as the ability of a substance to cause severe biological harm, or death, upon short time or low doses exposure (Hagan, 1959). In our studies we determined acute toxicity induced by the different polymers belonging to our library, by quantifying cell viability after cell exposure to polymers for a time period of 1 hour. In order to determine the half maximal effective concentration (EC50) values, relative to the different polymers, we tested increasing concentrations of the single compounds, ranging from 10e-8 M up to 10e-1 M. Our data show a direct correlation between EC50 values and the size and structure of the different polymers (Fig. 2). In particular we observed a direct correlation between enhanced cytotoxycity and increasing molecular weight, independently from the architecture of the compounds. Moreover at similar molecular weights the EC50 values for the branched poly-L-lysine forms (i.e. Dendrimer and HBPL) were up to xy fold lower than those measured for the linear compounds, suggesting a toxic effect induced by the free xy groups present in the branched polymers. . The EC50 values measured for the L-lysine dendrimers ranged from 0.009+/- XY to 7x10e-5 +/- XY M, for molecular weights of 900 to 20000 g/mol respectively. These data are in agreement with previous studies, where the MTT assay was used to determine dendrimer cytotoxicity on various cell types.7, 37, 38. Likewise the observed evolution of the toxicity of L-lysine dendrimer with increasing generation type was similar to the one reported for the dendritic PAMAM polymer 8. At equivalent sizes dendritic and hyperbranched poly-L-lysine induced acute in vitro toxicity at similar levels, especially for polymer sizes ranging from 3000 up to 20000 g/mol. (25 kDa) PEI is one of the most effective vectors for gene delivery, consequently it is extensively used in both biotechnology and gene therapy research (add ref). Interestingly despite both HBPL and dendrimer polymer types, were overall showing the highest cell toxicity, their EC50 values were only slightly lower than the one measured for PEI when size analogous were tested. It should also be noticed that whereas the PEI sample was expected to have a Mn of 25'000 g/mol as suggested by the supplier, its molecular weight assessed by GPC (?), was only of 10'000 g/mol.
. The slope of the function f(Mn HBPL)= EC50 decreases for high molecular weight samples and the gradual change of EC 50 with Mn is slower than for PL and dendrimer. This observation indicates either of the role of polydispersity and structural imperfections of hyperbranched poly-L-lysine or of aggregation of high Mn samples of HBPL due to lower solubility.
The impact of high polydispersity on EC 50 value was confirmed by measuring the EC 50 value of crude (yet dialyzed) HBPL 7. This sample of nonfractionated HBPL contains contaminants of very high molecular weight. These "impurities" are extremely effective in triggering acute cell death at low doses. Indeed, the EC 50 value of HBPL 7 is approximately equal to the EC 50 value of high moleculer weight HBPL 3 or the EC 50 of the 6th generation dendrimer. The variability of the EC 50 value of this sample was notably high, although the testing conditions were identical as for all the previous samples. We observed that linear poly-L-lysine displayed the lowest toxicity of all the tested analogues.
Insert Figure 2 here
We further investigated the potential causes of the immediate cell death observed upon polymer exposure. Osmolarity is one of the most important variables affecting the cell viability, indeed increased osmolarity of the extracellular fluid has been proposed among the possible causes of immediate loss of cell viability 39-41. In their protonated form, polycations can contribute to the osmotic stress by every single charged amino group and as well by the corresponding chloride counterion and thus exerting a cytotoxic effect by increasing the osmolarity of the cell culture medium.., Optimal osmotic values of cell culturing media lie in the range of 280-320 mOsm/kg11, and changes in morphology and cell viability can be already observed at an osmotic pressure of 450 mOsm/kg for suspension cell culture39. In order to investigate cytotoxic effect due to changes in osmolarity, the osmotic pressure of cell culturing medium supplemented with the tested polycations at concentrations corresponding to their EC50 values was measured (Table S2). Interestingly the low molecular weight polycations LPL 1 and HBPL 1 showed the highest osmotic values of 630 and 570 mOsm/kg, respectively, whereas In a model experiment aimed to test whether these osmotic values are indeed detrimental to cells, cell populations were incubated in a NaCl solution having comparable osmolarity (Table S2). This abrupt change of the osmotic pressure, induced necrosis of cells as evidenced by trypan blue inclusion in to the cytoplasm of the cells (Figure S5-3B). These data suggests that the cytotoxicity observed for low weight polycations may directly result from changes in osmolarity of the polymer solution compared to cell culture medium. On the other hand, when high Mn polycations were added to cell cultures at their EC50 concentrations, only a limited increase of the osmotic pressure was observed (Table S2). Furthermore such levels of osmolarity were not detrimental to cells as verified by equi-osmotic NaCl solutions (Figure S5-2A and 2B). We can thus suggest that for high molecular weight polycations osmolarity may not be the predominant factor inducing cell death.
Apart from the hypertonicity, an additional event that may contribute to early cell death upon exposure to polycations, is the potential ability of the polycation itself to promote cell membrane pore formation42-44. This process has been evidenced and analyzed experimentally in both, model biomembranes and in living cells (add ref). The ability of the polycation to disrupt cell membranes is commonly followed by the spectrophotometrical quantification of lactate dehydrogenase release (add ref). Alternatively, plasma membrane permeability can be evaluated by the cellular uptake of tracer molecules that are normally excluded from the cell, like for example high molecular weight dextran-fluorescent conjugates 45.. This approach is more relevant to the application of the polycations as gene delivery compounds (explain way?). On the basis of the previously published data showing enhanced LDH release upon cell incubation with polymers of higher Mw (REF?) we hypothesize that the high molecular weight polycations trigger the cell death by compromising the cell membrane integrity, and the extent of this incident is charge density dependent. In an attempt to investigate such hypothesis and determine cell membrane disruption induced by polycation exposure, cells were incubated with the various polycations and analyzed for membrane permeability to 70 kDa rhodamine-dextran entry. For each polymer three different concentrations, below their respective EC50 values were tested. Following polymer exposure, cells were washed and incubated in media containing rhodamine-dextran. Finally, dextran uptake into the cytoplasm of damaged cells was visualized by confocal scanning microscopy (Figure S6) and the percentage of cells presenting rhodamin-dextran in their cytoplasm cells was quantified by flow cytometry (Figure 3). From. the heat map in Figure 3, representing the percentage of rhodamin-dextran positive obtained by flow cytometry, we can observe that cell membrane permeability increases with increasing polymer molecular weight and DB. These data suggest a direct dependency of the degree cell membrane permeabilization upon the polymer charge density. We can therefore hypotize that the e cell membrane damaging is a key factor during induction of acute cell death especially for the branched poly-L-Lysine types.
Insert figure 3 here
LONG TERM TOXICITY
Activation of pro-apoptotic proteins (ref), direct mitochondrial permeabilization (ref) and perturbation of the membranes of intracellular organelles with the subsequent release of apoptosis inducing factors29, are among the proposed mechanisms by which polycations may trigger the apoptotic response.
The pioneering work of Moghimi et col23 defined the important events in late stage cytotoxicity of polycations. Their study focused on PEI and poly-L-lysine and identified them as apoptotic inducers. Authors pointed out the importance of the molecular weight of linear poly-L-lysine24 and PEI and its role on the mechanism of apoptosis. On the basis of their results the authors hypothesized that various polycationic macromolecules may initiate the apoptotic signaling pathway differently and with different efficiency. In relation to this hypothesis we investigate the extent of apoptosis induction as a function of polycation architecture. Distinct features characteristic for early apoptotic progression were monitored following cell exposure to polycations. These events involve translocation of phosphatidylserine from the inner to the outer side of the cell membrane, mitochonodrial depolarization and Cytochrome C (Cyt C) release. Further on we followed induction of the apoptotic pathway to its later stages via detection of caspase 3 and caspase 9 activation. Our final aim was to understand the temporal pattern of intracellular changes in response to polycation exposure, and how such events may correlate to the architectural structure of the different polycations.
Detection of the phosphatidylserine translocation belongs to the well established method of identifying population of early apoptotic cells (add ref). The plasma membrane of the early apoptotic cells loses its asymmetry and the phosphatidylserine is exposed to the outer leaflet as an external signal of cell demise.46 We have quantified the percentage of the apoptotic cells in CHO DG44 culture exposed to polycations via flow cytometry, employing the phosphatidylserine binding protein, Annexin V-FITC. The concomitant propidium iodide staining was used as indicator of cell necrosis. The histograms presented in Figure 4 represent Annexin V-FITC binding occurring 3 hours after the challenge of the cells with the polycation at a concentration of 20 μg/mL. Since this concentration is largely below the EC50 values previously measured, we assumed that under these experimental conditions, apoptosis was the primary mode of cell death. This was confirmed by the very low number of necrotic cells detected in all samples tested, as assessed by propidium iodide inclusion measurements (Figure 4B). As shown in Figure 4 for equi-size polycations, apoptotic induction detected by phosphatidylserine cell membrane translocation, reveals following tendency LPL< HBPL < DPL.
Insert figure 4 here
The next signaling event in the apoptosis cascade implicates mitochondria which membrane depolarization finally results in the release of Cyt C into the cell cytoplasm (add ref maybe :Susin et al., 1998). Cyt C plays a central role in the regulation of apoptosis by acting as an activator of a series of caspase cascades (add ref maybe Cai et al., 1998; Liu et al., 1996). The JC-1 reagent was used to assess the dissipation of cell mitochondrial potential as response to polycation exposure. JC-1 is a lipophilic probe bearing a delocalized positive charge wich depending on mitochondrial transmembrane potential (Δψm) accumulates either as red emitting aggregates in functional hyperpolarized mitochondria or, following mitochondrial membrane collapse, as a green monomer in the cytoplasm of apoptotic cells. (add ref Cossarizza A et. al Biochem Biophys Res Commun. 1993 Nov 30;197(1):40-5). Cells, per-exposed to polycations, were stained with the JC-1 compound and immediately analyzed either via flow cytometry (Figuer 5) or by fluorescence microscopy (Figure S7). In all the experiments, any cell fixation procedure was omitted in order to preserve mitocondria integrity. In accordance to the previous results on phosphatidylserine translocation, the highest degree of apoptotic induction, determined by the maximal number of the cells presenting green diffusive JC-1 fluorescence, occurred 2-3 hours after addition of the polymers to the cell cultures. Flow cytometric analyses demonstrate that the induction of mitocondria depolarization is directly dependent on the molecular size of the polycation. Indeed only in case of exposure to high weight polycation, cells presenting depolarized mitocondria were detectable upon JC-1 staining (Figure 5). On the other hand cells exposed to low molecular weight polycations did not show signs of dysfunctional mitochondria even after extended periods of incubation with the polymer samples (Figure 5). Since equi-size polymers showed similar result the structure of the polymer itself seems not to have influence on the onset of mitochondria collapse.
Insert Figure 5 here
It is generally accepted that the release of Cyt C from mitochondria is one of the very early events in the development of an apoptotic program that is a prerequisite for subsequent engagement of caspase cascades (add ref). Cyt C belongs to a family of mitochondrial intermembrane proteins47 and its cytoplasmatic release is often defined as a "point of no return" in the course of apoptosis (add ref). According to the work of Mather and Rottenberg 48 dual mechanism is responsible for the release of Cyt C from the mitochondria after the treatment with polycations: in a first putative mechanism Cyt C release might be a direct consequence of the mitochondrial depolarizarion induced during apoptotic cell death, a second hypothetical mechanism rely on the direct destabilization of the outer mitochondrial membrane through the electrostatic interactions of polycations with the anionic phospholipids of the mitochondrial membrane. Nevertheless both of these mechanisms contribute to the apoptotic program. The presence of the Cyt C in the cytoplasm of PC 12 cells was detected by indirect immunocytochemistry 12 h after addition of the polycation to cell cultures. For each polymer sample increasing concentrations were tested ranging from 10-8 to 10-4 M. Confirmatory to the trend observed for the depolarization of the mitochondrial membrane, cytoplasmatic Cyt C was detectable in cells incubated with high molecular weight polymers already when using concentrations as low as 10-7 M. On the other hand for low molecular weight samples, Cyt C release was evident only when high concentrations of polycations were added to the cell cultures (Figure 6). Such a trend was generally consistent with the tree types of polycation analogous, i.e. LPL, HBPL and DPL. Although the branched types of polymers induced a stronger response.
Insert Figure 6 here
Finally, the last stage, the activation of the caspase zymogens was detected and its presence was compared. Interestingly both apoptotic signals - Cyt C release and Caspase 9 activation - are detectable at a concentration range were apoptosis was already detected for PAMAM dendrimers.29 This may suggest …???. Once activated caspases cleave host of cellular substrates, leading to morphological hallmarks of apoptosis including DNA fragmentation and condensation of cellular organelles. The fragmentation of chromatin after 12 hour treatment with polymer was observed for all the samples except LPL samples in CHO cells (Figure S8).
All the check points in the intrinsic apoptosis pathway were observed if the cells (CHO and PC 12) were challenged with low doses of branched structures of high molecular weight. The extrinsic apoptotic pathway (caspase independent) was ruled out following the absence of caspase 8.
Insert Figure 7 here
Conclusion.
The polycations which are the subject of this study are the main tools in transient gene expression for delivering the required genetic material for recombinant proteins expression in mammalian cells either for biotechnology purposes or in the view of medicinal therapy. For this reason the exposure of the cell to the polycation must not lead to lethal injury or must not become a triggering event for apoptosis. If a (sub)lethal injury occurs the cell may enter into the commitment step of the death program without any chance to recover. Such a cell is ineffective producer of recombinant protein either in "bioreactor" 49-51 or in vivo23, 52, 53.
All four types of polycations exhibited similar trends in the commitment of the cell death, although branched analogues were more effective in causing stress related response of the cells. The differential cytotoxicity of the polycations results most probably from the different degradability pattern and degradation rate of these analogues. Dendritic and hyperbranched poly-L-lysine contains ε amide bonds, therefore partially resists to the cleavage by proteases. The high molecular weight of hyperbranched poly-L-lysine HBPL 5 and HBPL 6 probably contributes as well to the structural hindrance between a protease (i.e. trypsin) and the bond to be cleaved. These two samples were not well sustained by the cells. The degradation behavious is going to be further studied for our library… (These results will be contained in the next paper on gene delivery)
In summary we have shown that the novel L-lysine analogue, hyperbranched poly-L-lysine, elicits the same response in the cells as PEI and dendritic poly-L-lysine of the same molecular weight. In agreement with previous studies we have identified two separate mechanism of the cell death caused by polymers: acute cell death & apoptotis. The immediate cytotoxicity causing the cell death by the polycations demonstrates the ability of polycations to affect the cell morphology, metabolism and cell membrane permeability. The acute cell death can be caused by lethal cell injury caused via fatal membrane permeabilization or high osmolytic shock. The extent of the acute cell death and apoptosis is directly linked to molecular weight and DB of the polycation. Low molecular weight polycations trigger the acute cell death at high concentration by osmolytic shock. Polycations of low Mn do not initiate apoptosis even at high concentration, irrespectively of their architecture. Our data indicate that the molecular mechanisms of the apoptosis is mediated via mitochondria, we have detected phosphatidyl serine translocation, mitochondrial depolarization, cytochrome C release and caspase 9 activation.
To conclude the objective of this study was to clarify the relationship between architecture and Mn of the polycation with the degree of their in vitro toxicity. The mechanism of the acute cell toxicity and delayed cell death was investigated together with the underlying mechanism. Additionaly, a novel L-lysine analogue, hyperbranched poly-L-lysine, was investigated for its biocompatibility and the data proves that it is a suitable candidate for gene delivery applications based on toxicity criterion.
Figures & Tables
Figure 1. Schematic representation of the architecture of the poly-L-lysine analogues. A) dendritic poly-L-lysine, B) hyperbranched poly-L-lysine, C) linear poly-L-lysine.
Table 1. Ring opening polymerization of N-carboxy-(N-ε-(tertbutoxycarbonyl) L-lysine anhydride.
Sample
[I0]/ [M0]a
t/[h]b
Mtheorc [g*mol-1]
M1H NMRd [g*mol-1]
Mn e [g*mol-1]
Mwf [g*mol-1]
PDIg
LPL1
1:40
72
5120
800
2400
3300
1.4
LPL2
1:40
72
5120
2500
7000
11300
1.6
LPL3
1:50
72
6400
4400
11000
14300
1.3
LPL4
1:150
78
19200
14500
18700
26100
1.4
LPL5
1:200
65
25600
19600
16600
43100
2.6
PEI
-
-
-
-
9100
28600
3.2
a Ratio of initiator to monomer; molar concentration in the polymerization feed, b polymerization time c Theoretical Mn determined from initial [M]:[I] ratio and conversion. d Determined by 1H NMR spectroscopy by comparing the initiator-derived resonance to 2H attached to epsilon carbon e, f,g Experimentally determined by GPC.
Table 2. Synthesis of dendrimer of L-lysine
Sample
Mtheora[g*mol-1]
ESI MSb
ESI MSc
Mn d [g*mol-1]
Mwe
[ g*mol-1]
PDIf
DBg
DPL1
372
372.55
([M + H]+) 372.55
-
-
-
1
DPL2
885
443.34
([M + 2H]2+) 443.35
-
-
-
1
DPL3
1909
478.18
([M + 4H]4+) 478.37
1200
1300
1.1
1
DPL4
3959
660.64¯
([M + 6H]6+) 660.83
3100
4000
1.3
1
DPL5
8058
1009.01¯
([M + 8H]8+)1008.25
5600
7200
1.3
1
DPL6
16266
2323.19¯
([M + 7H]7+)2323.30
12000
18000
1.5
1
a theoretical molecular weight, b Experimental value, c m/z theoretical, d,e,f Experimentally determined by GPC, ¯ multiple peaks present in the spectra
Table 3. Synthesis of hyperbranched poly-L-lysine
Sample
Mn h [g*mol-1]
Mw i [g*mol-1]
PDIj
NNH k
DB l
HBPL_1
1400
2700
1.9
7±3
0.68
HBPL_2
8100
14500
1.8
78±14
0.68
HBPL_3
21000
44300
2.1
148±28
0.68
HBPL_4
46200
101200
2.2
280±41
0.68
HBPL_5
83900
232200
2.8
562±57
0.68
HBPL_6
146800
352600
2.4
756±24
0.68
HBPL_7
19800
105200
5
326±35
0.68
h, i, j Experimentally determined by GPC. k number of amino groups experimentally determined by acid base titration. l Average number of branches were calculated as described in SI, Section Techniques.
Figure 2. Quantification of acute cytotoxicity. CHO cells were incubated for 1h with various polymers as indicated. Acute cell death was then assessed by flow cytometry. EC50 values were extrapolated by assessing cell viability after incubation with increasing polymer concentrations. For each sample at least 10 concentrations were tested ranging from 10-8 to 10-1 M.
Figure 3. The cell membrane integrity was assessed by its permeability to rhodamine-dextran. The diffusion of rhodamine-dextran into the cytosol of CHO cells after 1h incubation with a polymer at a concentration range of 10, 20 and 30 μg/mL. Color scale of the heatmap (a) was obtained by measuring the sample of the cells treated with 0.1% solution of triton (positive control) and PEG 5000 at a concentration of 100ug/mL (negative control). The heatmap represents the percentage of the fluorescent cells after the treatment with HBPL (b), DPL (c), and LPL (d) the percentage of fluorescent cells was determined by flow cytometry. .
Figure 4. Phosphatidyl serine translocation upon polycation treatment. CHO cells displaying phosphatidyl serine were indetified via Annexin V binding 3h after incubation with LPL, DPL or HBPL polymers. Non viable cells were excluded by PI staining.
Figure 5. ΔψM as a function of the incubation time and the Mn of polycation. The percentage of cells displaying depolarized mitochondria after the exposure of the cells to LPLs (a), DPLs (b) or HBPLs (c) was determined. The ΔψM of mitochondria was analyzed at 30min (â- ), 60min (â-), 120min (â-²), 240min (â-¼) post incubation with polycation. Analysis was performed by flow cytometry using the JC-1 stain. Wide field fluorescent microscopy of cells with depolarized mitochondria (green staining of the cytoplasm) and with intact mitochondria (punctual red staining). Cells treated with negative control (PEG 5000) at a time 0 and 4h (d and e respectively). As a positive control, cells were treated with camptothecin and images were acquired at the time 0(f) and 4h (g). Images h and i shows mitochondrial depolarization after the treatment of the cells with HBPL 6 (0h and 4h).
Not done yet
Not done yet
Figure 6. Detection of Cyt C cytoplasmatic release in PC12 cells after 12h incubation with different polymers. The release of Cyt C into the cytosol is preceded by the decrease in mitochondrial transmembrane potential after the stress inducing factor. Cyt C release is evident in almost all samples, for lower molecular weight samples the release is detectable only when high concentration is applied. The presence of Cyt C is more pronounced for the dendritic and hyperbranched samples and samples of higher molecular weight. The Cyt C was not detected for the low molecular weight HBPL and LPL.
Not done yet
Figure 7. Caspases 9 activation in PC 12 cells after 12h treatment with polycation. Caspase 9 is a crucial upstream activator of the Cyt C-dependant apoptotic pathway., and its presence (green) was detected mainly with high molecular weight samples -DPL 6, HBPL 6, LPL 5.
