{"id":608361,"date":"2024-06-07T20:00:00","date_gmt":"2024-06-08T00:00:00","guid":{"rendered":"https:\/\/platohealth.ai\/?p=608361"},"modified":"2024-06-07T19:44:45","modified_gmt":"2024-06-07T23:44:45","slug":"an-efficient-low-cost-means-of-biophysical-gene-transfection-in-primary-cells-scientific-reports","status":"publish","type":"post","link":"https:\/\/platohealth.ai\/an-efficient-low-cost-means-of-biophysical-gene-transfection-in-primary-cells-scientific-reports\/","title":{"rendered":"An efficient low cost means of biophysical gene transfection in primary cells – Scientific Reports","gt_translate_keys":[{"key":"rendered","format":"text"}]},"content":{"rendered":"

In order to optimize conditions for ME of difficult to transfect cells, we designed a durable, dynamic, low-cost electroporation chamber. As shown in Fig. 1<\/a>A-B, polished 1\u2009\u00d7\u20091 cm sections of 0.024\u2033 thick 316 stainless steel were separated at fixed gap distances ranging from 200\u20131000 \u03bcm, serving as electrodes. Electrodes were then affixed and protected with epoxy on a standard glass slide (Fig. 1<\/a>B, arrows)., While experiments were routinely performed in a class II-A2 biosafety cabinet with cell solutions confined within the electroporation channel, samples could be cover-slipped for real-time analysis on a standard upright or inverted microscope (Fig. 1<\/a>C-D). Due to the reduced electrical requirements resulting from the diminished gap size (700 \u03bcm shown), electronics such as those commonly employed for electrophysiology are capable of supplying sufficient power for electroporation (Fig. 1<\/a>E). In fact, further experimentation demonstrated that the electronics required could be wholly replaced by a battery powered Arduino based system with equal efficiency at a cost of\u2009<\u2009$100 CAD (Fig. 5<\/a>). Using this system, relative cell permeability was assessed by monitoring the relative rate of fluorescence enhancement of cells to the cell-impermeant marker propidium iodide (PI). Under conditions appropriate for electroporation, a rapid increase in PI fluorescence could be observed in a substantial subpopulation of ES cells within 2 min of electroporation (Fig. 1<\/a>F).<\/p>\n

\n
Figure 1<\/b><\/figcaption>
\n
\"figure<\/a><\/div>\n
\n

Microelectroporation cell. (A<\/b>) Schematic overview of unit with most commonly utilized dimensions shown. Cell solution is confined in channel by hydrophobic forces. Direction of electrically induced DNA movement is indicated. (B<\/b>) Polished 1\u2009\u00d7\u20091 cm sections of 316 stainless steel separated by a gap distance of 650 \u03bcm (arrows) with connections soldered and protected using epoxy on a standard glass slide. (C<\/b>) Reusable unit allowed rapid flushing and real-time inspection of electroporation on standard upright microscope. (D<\/b>) Alternative attachment is shown. (E<\/b>) Example of prefab electrophysiology unit which can be utilized for microelectroporation due to reduced gap size and fluidic volume. (F<\/b>) Example of unit used for electroporation of ES cells, demonstrating development of propidium iodide-positive ES cells within well chamber within 2 min following application of pulsed electroporation. Scale bar represents 100 \u03bcm.<\/p>\n<\/div>\n<\/div>\n

Full size image<\/span><\/a><\/div>\n<\/figure>\n<\/div>\n

Microelectroporation optimization<\/h3>\n

We next optimized conditions for electroporation of embryonic stem cells using ME. To determine the true efficiency of the system, attempts were made to vary one critical parameter at a time. Results were assessed as a function of viable puromycin resistant colonies recovered following the electroporation of a 14.8 kb plasmid (Addgene #52961) and subsequent puromycin selection. Doing so allowed not only assessment of faithful plasmid uptake and expression, but also resulting ES cell survival, morphology and growth. As shown in Fig. 2<\/a>A, results were compared between Bio-Rad Gene Pulser with Capacitance Extender and ME using equivalent initial electroporation (5,000 cells\/\u03bcl), and plating (10,000 cells\/well) cell concentrations. Resuspension buffer (EmbryoMax electroporation media) and field strengths (575 versus 571 V\/cm) were also kept comparable. Additional set point conditions for Bio-Rad were 800 \u03bcl of cell solution in a 4 mm cuvette, absolute voltage 230 V, 500 \u03bcF, time constant (\u03c4)\u2009=\u20096.5 ms. For ME values were: 4.3 \u03bcl sample volume, voltage 38 V for a gap distance of 700 \u03bcm. Based on the Bio-Rad time constant (\u03c4), square-wave pulse sequence was 6\u2009\u00d7\u20091 ms. Changes in DNA reporter concentration from 5\u201320 \u03bcg\/ml resulted in consistent and relatively minor alterations in transfection efficiency for the Bio-Rad system (Fig. 2<\/a>A). By comparison, alterations over this range for the ME resulted in significant alteration of the numbers of puromycin resistant colonies observed, with 5 \u03bcg\/ml determined to be optimal compared to all other concentrations tested (Fig. 2<\/a>A). Transfection efficiency as a function of field strength is shown in Fig. 2<\/a>B, demonstrating a maxima at 575 V\/cm for Bio-Rad system and 543 V\/cm (applied voltage: 38 V) for ME among the conditions examined. Additional set point parameters for both systems are as given above for Fig. 2<\/a>A, with the exception that DNA reporter addition was set to 10 \u03bcg\/ml for both units. The results demonstrate that despite similar field strength maxima, ME demonstrated significantly higher transfection rates at its optimum field strength.<\/p>\n

\n
Figure 2<\/b><\/figcaption>
\n
\"figure<\/a><\/div>\n
\n

Optimization of microelectroporation. (A<\/b>) Transfection was examined as a function of DNA reporter concentration; red-Bio-Rad gene pulser, blue-microelectroporator. NEB-no electroporation Bio-Rad, NEM-no electroporation microelectroporator. Experiments were performed in triplicate for each set with n\u2009>\u20094 sets for Bio-Rad conditions, n\u2009>\u20095 sets for microelectroporator. Field strength was set to 575 and 571 V\/cm for Bio-Rad versus microelectroporator respectively. Capacitance for Bio-Rad was set to 500 \u03bcF and (\u03c4)\u2009=\u20096.5 ms. Pulse sequence used for microelectroporator was 6\u2009\u00d7\u20091 ms. *- Denotes significant enhancement at P<\/i>\u2009<\u20090.01 over all Bio-Rad conditions. (B<\/b>) Relative transfection efficiency as a function of applied field strength; red-Bio-Rad gene pulser, blue-microelectroporator. NDNE-no DNA no electroporation. Experiments were performed in triplicate for each set with n\u2009>\u20094 sets for Bio-Rad conditions, n\u2009>\u20095 sets for microelectroporator. DNA reporter set to 10 \u03bcg\/ml for both systems. Capacitance for Bio-Rad was set to 500 \u03bcF and (\u03c4)\u2009=\u20096.5 ms. Pulse sequence used for microelectroporator was 6\u2009\u00d7\u20091 ms. Cells plated at 10 k cells\/well in a 6 well. *- Denotes significant enhancement at P<\/i>\u2009<\u20090.01 over all Bio-Rad conditions. (C<\/b>) Relative transfection efficiency as a function of square-wave pulse; red-Bio-Rad gene pulser, blue -microelectroporator (plating: 10 k cells\/well, 6 well plate), pink-microelectroporator (plating: 50 k cells\/well, 6 well plate). Experiments were performed in triplicate for each set with n\u2009=\u20095 sets for both systems. DNA reporter set to 10 \u03bcg\/ml for both systems, field strength was set to 575 and 571 V\/cm respectively for Bio-Rad versus microelectroporator Capacitance for Bio-Rad was set to 500 \u03bcF and (\u03c4)\u2009=\u20096.5 ms. *- Denotes significant enhancement at P<\/i>\u2009<\u20090.01 over lower plating density (10 k cells\/well).<\/p>\n<\/div>\n<\/div>\n

Full size image<\/span><\/a><\/div>\n<\/figure>\n<\/div>\n

A parameter previously shown to significantly affect transfection efficiency is the nature of the square-wave pulse provided. To this end, the efficiency of different pulse wave parameters were examined, with DNA reporter addition set to 10 \u03bcg\/ml for both systems and field strength set to 575 and 571 V\/cm for Bio-Rad and ME respectively. Modification of pulse length to either longer time periods, or greater pulse number did not enhance overall ME transfection efficiency beyond that seen at 6\u2009\u00d7\u20091 ms (Fig. 2<\/a>C). Thus one millisecond pulses appear capable of generating pores of sufficient size and longevity to allow the uptake of a 14 kb plasmid to occur. However given that embryonic stem cells are known to exhibit significant density dependence with respect to cell survival, it is possible that a portion of transfected cell die subsequently, an effect which might be rescued by enhancing cell density. As such the efficiency of transfection was examined as a function of plating density. As shown in Fig. 2<\/a>C, increasing plating density from 1,042 cells\/cm2<\/sup> (10 k cells\/well of 6-well plate) to 5,208 cells\/cm2<\/sup>, demonstrated significant enhancement in relative transfection at higher plating density compared to lower plating density for all pulse parameters examined (all results presented corrected to 1042 cells\/cm2<\/sup>).<\/p>\n

Despite optimizing transfection efficiency in terms of DNA concentration, field strength and pulse sequence in ME, successful transfection with expression of plasmid-based targets is still only achieved in a minority of cells in difficult to transfect cell lines. One potential source of this inefficiency is the permeabilization characteristics of the targeted population. In order to better understand this process with respect to ME, ES cell morphology was examined as a function of electroporation conditions. Figure 3<\/a>A demonstrates the major morphologic isotypes observed in the presence and absence of electroporation. For these experiments ES cells expressing monomeric cytoplasmic citrine from the Rosa26<\/i> locus were utilized in conjunction with pre-incubation of both the cell permeant marker DNA marker Hoechst 33342 and cell impermeant marker propidium iodide. As shown in Fig. 3<\/a>, under the most optimal ME transfection conditions examined (543 V\/cm, 6\u2009\u00d7\u20091 ms) cells unaffected by electroporation were frequent (PI–<\/sup>, 58\u2009\u00b1\u20094%, n\u2009=\u2009300), exhibiting cytoplasmic citrine fluorescence with only Hoechst staining within the cell nucleus and lacking plasma membrane interruption with propidium iodide entry. By contrast, a minority of cells (PI+<\/sup>, 5\u2009\u00b1\u20092%) exhibited entry of propidium iodide in the period following electroporation (t\u2009=\u200910 min) without cellular collapse or significant disruption of internal structures (yellow arrow). More prevalent (cell death type 1, 14\u2009\u00b1\u20094%) were cells presenting PI permeation with early collapse of internal structures and nuclear blebbing (red arrow). Another common isotype (cell death type 2, 23\u2009\u00b1\u20093%) were cells displaying PI permeation with substantial structural collapse such that cellular volume was\u2009\u2264\u200950% of untransfected cells (brown arrow). Alteration of pulse sequences to longer time periods (5, 10 and 20 ms) increased relative numbers of cell death type 1\/2 cells but did not increase numbers of \u2018PI+\u2019<\/sup> cells, mirroring transfection effects (Fig. 2<\/a>C). Therefore a more extended series of shorter (1 ms) electroporation pulses may be a more optimal approach to enhancing transfection efficiency.<\/p>\n

\n
Figure 3<\/b><\/figcaption>
\n
\"figure<\/a><\/div>\n
\n

Electroporation isotypes. (A<\/b>) Examples of the morphology of major ES cell isotypes observed before and following electroporation. CD1\/2\u2009=\u2009cell death type 1\/2. Scale bar represents a distance of 10 \u03bcm. (B<\/b>\u2013E<\/b>) Examples of pre- (B<\/b>, C<\/b>) and post- (D<\/b>, E<\/b>) electroporated average fields. Scale bar in (E<\/b>) represents a distance of 100 \u03bcm. Where indicated cells are shown 10 min following electroporation (ME: 543 V\/cm, 6\u2009\u00d7\u20091 ms), in which citrine-expressing ES cells were pre-incubated with both the cell permeant marker Hoechst 33342 and cell impermeant marker propidium iodide. Cell types were broadly defined as: PI- cells displaying no membrane disruption or permeation of PI; PI+ cells exhibiting PI permeation without significant disruption of internal structure (yellow arrows); PCD1 cells displaying PI permeation with early collapse of internal structures (red arrows); PCD2 cells displaying PI permeation with complete collapse of cellular volume to\u2009<\u200950% of PI- cells (brown arrows).<\/p>\n<\/div>\n<\/div>\n

Full size image<\/span><\/a><\/div>\n<\/figure>\n<\/div>\n

To further examine the electroporation efficiency of both Bio-Rad and ME electroporators in cell lines recalcitrant to transfection, primary murine fibroblasts and human T lymphocyte Jurkat cells (clone E6-1) were examined at several conditions as shown in Supplemental Figure S1. For purposes of comparison, conditions were set to those previously reported as optimal for the Bio-Rad electroporator. For both cell types and ME and Bio-Rad electroporators, cell and DNA concentration were kept at a constant (3000 cells\/\u03bcl, 20 \u03bcg\/ml mRuby3\/mClover3 fluorescent expression vector-Addgene #74252), despite these parameters being outside the observed optimum for ME electroporation of ES cells (5 versus 20 \u03bcg\/ml, Fig. 2<\/a>A). With respect to primary fibroblasts, Bio-Rad and ME electroporators exhibited similar efficiencies at their observed optimums (Bio-Rad-800 V\/cm, 4 mm cuvette, \u03c4\u2009=\u200914.5 ms; ME- 6\u2009\u00d7\u20091 ms). By contrast, for Jurkat cells (Supplemental Figure S1, S2), ME electroporation exhibited significant enhancement over Bio-Rad at its observed optimum (Bio-Rad-800 V\/cm, 4 mm cuvette, \u03c4 =\u200916.5 ms; ME-6\u2009\u00d7\u20091 ms). In both cell types under these conditions the observed maxima for ME electroporation conditions appeared at lower field strengths compared to that seen with Bio-Rad. Example fields for both primary fibroblasts, and Jurkat cells are shown for ME and Bio-Rad mediated transfection respectively (Supplemental Figure S1). Interestingly for both cell types, ME-transfected cells routinely exhibited greater fluorescence intensity compared to their Bio-Rad transfected counterparts under optimal conditions.<\/p>\n

CRISPR-mediated gene modification using ME<\/h3>\n

Given that a primary motivating factor for designing the ME system was the performance of CRISPR-mediated gene targeting in embryonic stem cells, we assessed the ability of ME to produce targeted mutations for several loci of interest. An example of this is shown in Fig. 4<\/a>A, using multiple CRISPR single guide RNAs (sgRNA) targeting exons 4 and 5 of the Casp3<\/i> gene. To create sgRNA expressing CRISPR plasmids, CRISPR RNA (crRNA) sequences were cloned into a tracrRNA containing Cas9 expression vector (Addgene #52961). CRISPR sgRNA plasmids were then electroporated at a final concentration of 5 \u03bcg\/ml (total) with ES cells (5,000 cells\/\u03bcl) in\u2009~\u20095\u03bc\u03bb at 543 V\/cm (6\u2009\u00d7\u20091 ms) in EmbryoMax electroporation media. Following electroporation, cells were plated at 10,000 cells\/well in 6-well dishes and subjected to puromycin selection on days 2, 3, and 5. Five days following antibiotic removal, puromycin-resistant ES colonies were counted, replica plated and analyzed. As shown in Fig. 4<\/a>B, successful deletion employing both CRISPR sgRNA plasmids would result in a reproducible deletion of a segment of exon 4, exon 5 and the intervening intron, removing a\u2009~\u20091267 nucleotide segment. PCR primers utilized for this analysis are shown (blue arrows, Supplementary Table S1<\/a>).<\/p>\n

\n
Figure 4<\/b><\/figcaption>
\n
\"figure<\/a><\/div>\n
\n

Use of microelectroporation for CRISPR mediated gene editing. (A<\/b>) Example of stratagem employed for modification of Casp3 locus. Combined action of cloned CRISPR single guides (red arrow) results in a deletion of 1267 nucleotide resulting in a frameshift mutation between exons 3 and 4. In such an event PCR primers (blue arrows) identify a band of 149 bp for the mutant allele versus 1416 bp for wildtype (unmodified) locus. (B<\/b>) DNA electrophoresis of 9 puromycin-resistant ES cell clones derived following microelectroporation. Examples of wildtype (Wt), heterozygous (He), and homozygous (Ho) modifications are observed. DNA molecular weight standard (Std.) is shown at left.<\/p>\n<\/div>\n<\/div>\n

Full size image<\/span><\/a><\/div>\n<\/figure>\n<\/div>\n

Using the selection protocol outlined above, we examined the ability of the ME electroporator to produce defined biallelic gene deletions in several target genes of interest via the simultaneous introduction of plasmids encoding CRISPR sgRNAs. The sequence and details of these targeting events are shown in Supplemental Figure S3, together with the genomic deletions induced as determined by direct sequencing of clones. The crRNA sequences used to generate plasmid-encoded CRISPR sgRNAs are shown in Supplemental Table S2<\/a>. Target sequences for genes of interest were obtained from Ensemble and verified via Sanger sequencing of ES cells for the genetic background of interest (129\/Sv). In each case, experimental sequences were identical to those predicted for CRISPR targets despite different genetic backgrounds (C57bl\/6 J versus 129\/Sv). For each gene locus examined, CRISPR sgRNA plasmid pairs were successful in producing biallelic deletions based on Sanger sequencing of the clone populations obtained (Supplemental Figure S3). Conveniently, such method of targeting could be readily identified by standard PCR. The most common deviation observed in this targeting was that several CRISPRs cleaved 1\u20133 nucleotides up\/downstream of the canonical cleavage site (between nucleotides 3 \/ 4 upstream of the PAM site). Consistent with prior observations34<\/a><\/sup>, CRISPRs varied in their ability to produce biallelic deletions at different genetic loci (caspase-3<\/b> WT 48% He 39%, Ho 13%; caspase-8<\/b> WT 37%, He 32%, Ho 31%; MLKL<\/b> WT 23%, He 69%, Ho 8%-of puromycin resistant clones). Despite such differences, none of these frequencies fell below numbers readily obtainable following one ME selection event (>\u200950 puromycin resistant clones).<\/p>\n

Development of microcontroller-based ME wave generator<\/h3>\n

Having validated the operation of the ME chamber (see Methods), we sought to further improve the flexibility, portability and accessibility of ME by re-engineering its electrical aspects. As a reduction in electroporation volume lowers capacitance requirements, standard electrophysiology stimulators such as that shown in Fig. 1<\/a>E (Grass S44) could be utilized to generate the frequency, magnitude and duration of waveforms required. However given recent advances and cost reductions in microprocessers, such requirements could also be met by circuitry supported by Arduino-based microcontrollers. We therefore designed the layout of a small scale Arduino-based electroporator, whose capabilities were well within the waveform, voltage and frequency requirements of ME. Transfection experiments comparing the performance of Grass S44 versus Arduino-based electroporator demonstrated no differences. However the Arduino-based microcontrollers significantly reduced the total cost and size of the electroporation unit while maintaining flexibility of waveform generation. Through provision of a detailed parts list, schematics and instructions (see below and Supplemental Figs. 4<\/a>\u20135<\/a>), even those without detailed electronics experience can readily construct an ME electroporator.<\/p>\n

\n
Figure 5<\/b><\/figcaption>
\n
\"figure<\/a><\/div>\n
\n

Electroporator schematic and practical examples. (A<\/b>) Electrical schematic of Arduino based electroporator. (B<\/b>) Visual example of electroporator assembled on breadboard as indicated, or (C<\/b>) soldered on a protoboard. Connections from wires to other non-visible components are indicated.<\/p>\n<\/div>\n<\/div>\n

Full size image<\/span><\/a><\/div>\n<\/figure>\n<\/div>\n

Construction of microcontroller-based ME wave generator<\/h3>\n

Electroporator parts list. Note: Bolded items<\/b> were obtained from Arduino starter kit (ELEGOO, EL-KIT-003) but may also be acquired separately. Sources indicated for each component. Due to their low cost, extras of individual components are recommended as backup. Individual components can be obtained from local or online electronics distributors such as Mouser Electronics and Digi-key Electronics, or Amazon for more common items.<\/p>\n