Minigenes for heterologous expression of human and mouse cationic trypsinogen

TL;DR

This study explores how minigenes can improve the expression of human and mouse trypsinogen proteins in cells, with mixed results on protein secretion.

Contribution

The novel contribution is demonstrating that minigene constructs increase mRNA levels but not always protein secretion for trypsinogen.

Findings

Minigene constructs increased PRSS1 and Prss3b mRNA levels by 2.5-fold and 4.5-fold respectively.

Secretion of mouse cationic trypsinogen increased 2.9-fold, but human trypsinogen secretion remained unchanged.

Increased mRNA levels may not always lead to higher protein secretion due to translation or folding limitations.

Abstract

Inborn mutations in the PRSS1 gene encoding human cationic trypsinogen cause hereditary pancreatitis. In mouse models, PRSS1 mutations are often studied in the context of the Prss3b gene that codes for mouse cationic trypsinogen. To characterize the cellular and biochemical effects of trypsinogen mutations, heterologous expression in transfected cell lines is often employed. Recent studies with the human and mouse trypsin inhibitor SPINK1 indicated that minigene expression constructs carrying a short intron yield markedly higher recombinant protein levels than cDNA constructs. Here, we investigated whether the minigene approach would increase the expression of human and mouse cationic trypsinogen in transfected HEK 293T cells. We found that compared with the cDNA, minigene constructs increased PRSS1 and Prss3b mRNA levels by 2.5-fold and 4.5-fold on average, respectively. Surprisingly, however, the amount of secreted human cationic trypsinogen remained unchanged while secretion of mouse cationic trypsinogen was increased 2.9-fold. The observations indicate that minigene expression constructs are effective in boosting mRNA levels in transfected cells, however, this may not always translate to elevated protein secretion. In these cases, inefficient protein translation and/or folding may be rate limiting.

Figures

Schematic representation of human (PRSS1) and mouse (Prss3b) cationic trypsinogen expression constructs used for transfections.Exons numbered with Roman numerals are illustrated by rounded rectangles and the introns are indicated by horizontal lines.

Expression of human cationic trypsinogen in HEK 293T cells transfected with PRSS1 cDNA and minigene constructs.A, Secreted trypsinogen protein (arrow) in the conditioned medium 24 hours and 48 hours after transfection was assessed by SDS-PAGE and Coomassie Blue staining. Representative gels from 3 independent transfections are shown. B, Trypsinogen levels in the conditioned medium were determined by trypsin activity measurement after activation with enteropeptidase. Individual values from 4 transfections with duplicates (n = 8) are shown, with the mean and SD indicated. C, Western blot analysis of trypsinogen (arrow) levels in cell lysates 48 hours post-transfection. Alpha-tubulin was measured as loading control. Representative blots are shown. D, PRSS1 mRNA levels were measured 48 hours after transfection by reverse-transcription quantitative PCR and expressed as fold change relative to the average value of the cDNA construct. Individual values from 3 transfections with duplicates (n = 6) are shown, with the mean and SD indicated. E, Splicing of PRSS1 mRNA expressed from cDNA and minigene constructs was analyzed 48 hours after transfection by reverse-transcription PCR and agarose gel electrophoresis. The arrow indicates the correctly spliced PRSS1 band. The smaller faint band in the minigene samples indicated by the asterisk corresponds to an aberrant splice product in which nucleotide c.40 was spliced to c.114 resulting in the deletion of the mini-intron and an extra 74 nucleotides from exon 2.

Expression of mouse (T7) cationic trypsinogen in HEK 293T cells transfected with Prss3b cDNA and minigene constructs.A, Secreted trypsinogen protein (arrow) in the conditioned medium 24 hours and 48 hours after transfection was assessed by SDS-PAGE and Coomassie Blue staining. Representative gels from 3 independent transfections are shown. The faint bands in the 12-14 kDa region are due to autolysis of trypsinogen at Arg123. B, Trypsinogen levels in the conditioned medium were determined by trypsin activity measurement after activation with enteropeptidase. Individual values from 4 transfections with duplicates (n = 8) are shown, with the mean and SD indicated. C, Prss3b mRNA levels were measured 48 hours after transfection by reverse-transcription quantitative PCR and expressed as fold change relative to the average value of the cDNA construct. Individual values from 3 transfections with duplicates (n = 6) are shown, with the mean and SD indicated. D, Splicing of Prss3b mRNA expressed from cDNA and minigene constructs was analyzed 48 hours post-transfection by reverse-transcription PCR and agarose gel electrophoresis. The arrow indicates the correctly spliced Prss3b band.

Full text

Trypsinogen is the most abundant digestive proteolytic proenzyme in mammals. The protease precursor is synthesized, stored, and secreted by the pancreatic acinar cells. As a component of the pancreatic juice, trypsinogen travels to the duodenum via the ductal tree and becomes activated to trypsin by the intestinal protease enteropeptidase [1]. Trypsinogen is often coded by multiple genes generating highly homologous isoforms. In humans, three isoforms are produced in the pancreas, encoded by the PRSS1, PRRS2, and PRSS3 genes, and commonly referred to as cationic trypsinogen, anionic trypsinogen, and mesotrypsinogen [2]. Their relative abundance in the normal human pancreas is 55–60%, 30–35%, and 5–10%, respectively [3–6]. The mouse pancreas secretes 4 major isoforms, cationic trypsinogen, also known as isoform T7 (encoded by the mouse Prss3b gene), and 3 anionic trypsinogen species, named isoforms T8, T9, and T20 [7]. In C57BL/6 mice, cationic trypsinogen represents about 60% of the total pancreatic trypsinogen content [8].

Trypsinogens play an important role not only in physiological protein digestion but also in the inflammatory diseases of the pancreas, called pancreatitis [9–11]. Inborn mutations in human cationic trypsinogen cause autosomal dominant hereditary pancreatitis or sporadic chronic pancreatitis. The mutations stimulate trypsinogen autoactivation or prevent its protective degradation and thereby increase intrapancreatic trypsin activity to pathological levels. Copy number variations can increase trypsin activity through gene-dosage effect, whereas promoter variants can be protective by reducing trypsinogen expression [12,13]. To study hereditary pancreatitis in animal models, mice carrying mutations in the mouse cationic trypsinogen were developed [14–18]. These mutant versions of mouse cationic trypsinogen autoactivate rapidly and cause progressive spontaneous pancreatitis or increase susceptibility to experimentally induced pancreatitis.

To understand the mechanistic basis of hereditary pancreatitis, recombinant human cationic trypsinogen mutants have been studied extensively [9, 10 and references therein]. Several mutations were also analyzed biochemically in the context of mouse cationic trypsinogen [7,14,15,17,18]. In these experiments, trypsinogen was typically expressed in E. coli as inclusion bodies, refolded in vitro, and purified by ecotin affinity chromatography. Bacterial expression systems, however, do not recapitulate mammalian post-translational modifications, such as sulfation of Tyr154 in human cationic trypsinogen [19]. Furthermore, the N termini of recombinant trypsinogens produced in E. coli may not be correctly processed resulting in functional differences from their native counterpart [20]. To address these potential shortcomings, heterologous expression in mammalian cells, such as the commonly used laboratory cell line HEK 293T, is a viable approach. Unfortunately, both human [21,22] and mouse (unpublished) cationic trypsinogen are secreted relatively poorly from transiently transfected HEK 293T cells, due to reasons that are not readily apparent. We recently found that minigenes containing 100 or 200 nucleotide long mini-introns from the human serine protease inhibitor Kazal type 1 (SPINK1) gene markedly increased human and mouse SPINK1 expression both at the mRNA and protein levels, in multiple cell types [23,24]. In this paper, we investigated whether the same minigene approach would improve expression of human and mouse cationic trypsinogen in transfected cells.

Methods

Gene accession numbers. NM_002769.5, NCBI reference sequence for Homo sapiens serine protease 1 (PRSS1) mRNA, encoding human cationic trypsinogen. NM_023333.4, NCBI reference sequence for Mus musculus serine protease 3B (Prss3b) mRNA sequence, formerly known as RIKEN cDNA 2210010C04 (2210010C04Rik) mRNA, encoding the mouse cationic (T7) trypsinogen.

Plasmid constructs. Construction of the pcDNA3.1(-) expression plasmid harboring the coding DNA (cDNA) for human cationic trypsinogen (PRSS1) was described previously [25]. The cDNA for mouse cationic trypsinogen (Prss3b) was PCR-amplified from IMAGE clone #30306963 (catalog number 10088751, ATCC, Manassas, VA) using the following primers. NheI forward primer, 5’- AAA TTT GCT AGC CCA CAG TGA GCA ACC ATG AAG ACC TTA ATC TTC CTT GCC 3’ and BamHI reverse primer, 5’- AAA TTT GGA TCC GGT AAA AAC AAA ATG TTT TTC TCT TGA TAG – 3.’ The PCR product was digested with NheI and BamHI and cloned into the pcDNA3.1(-) vector. Minigene constructs were generated by gene synthesis (GenScript, Piscataway, NJ) and cloned into the pcDNA3.1(-) vector using the XhoI – BamHI (PRSS1 minigene) or NheI – BamHI (Prss3b minigenes) sites. The PRSS1 minigene construct contained a 100 nucleotide-long mini-intron from the human SPINK1 gene [23,24] placed between exons 1 and 2. The Prss3b minigene constructs carried a 200 nucleotide-long mini-intron from human SPINK1 [24] placed between exons 1 and 2 (minigene v1) or exons 2 and 3 (minigene v2). Fig 1 shows the schematic representation of the expression constructs, and the Supporting Information file contains the DNA sequences.

Schematic representation of human (PRSS1) and mouse (Prss3b) cationic trypsinogen expression constructs used for transfections.Exons numbered with Roman numerals are illustrated by rounded rectangles and the introns are indicated by horizontal lines.

Cell culture and transfection. Cell culture reagents were purchased from Thermo Fisher Scientific. HEK 293T cells were cultured at 37^o^C in DMEM growth medium (catalog number 10313039) supplemented with 10% fetal bovine serum, 4 mM glutamine and 1% penicillin/streptomycin. Cells were seeded in six-well tissue culture plates at a density of 1.5 × 10^6^ cells per well. Transfections were carried out by adding 0.5 mL Opti-MEM medium (catalog number 11058021) containing 4 µg plasmid DNA and 5 µL Lipofectamine 2000 (catalog number 11668019) to 1.5 mL DMEM medium. After overnight incubation, cells were rinsed with 1 mL phosphate-buffered saline (pH 7.4) and covered with 1.5 mL OptiMEM containing 1 mM benzamidine trypsin inhibitor. Conditioned medium and cells were harvested 24 hours and 48 hours after the addition of OptiMEM.

Measurement of trypsinogen secretion. Conditioned media (175 µL) were precipitated with 10% trichloroacetic acid (final concentration). The precipitate was collected by centrifugation (10 min, 13,200 rpm, 4^o^C) and dissolved in 25 µL 2 × Laemmli Sample Buffer (catalog number 1610737, Bio-Rad, Hercules, CA) supplemented with 100 mM dithiothreitol and 150 mM NaOH. The samples were heat-denatured at 95ºC for 5 min, electrophoresed on 15% SDS-polyacrylamide gels, and stained with Brilliant Blue R-250 (Coomassie Blue).

Measurement of trypsin activity from conditioned medium. Trypsinogen in the conditioned medium was activated with human enteropeptidase for 1 hour at 37^o^C. The activation mix (110 µL) contained 5.5 µL medium, 0.1 M Tris-HCl (pH 8.0), 1 mM CaCl_2_, 0.05% Tween 20, and 2.2 µL of 100-fold diluted human enteropeptidase (catalog number 1585SE010, R&D Systems, Minneapolis, MN). Trypsin activity was measured in duplicate by mixing 50 µL activated medium with 150 µL of 200 µM Z-Gly-Pro-Arg-p-nitroanilide substrate. Release of the yellow p-nitroaniline was followed for 1 minute at 405 nm in a microplate reader at 22^o^C. The rate of substrate cleavage was determined from the linear portion of the curves in milliOD/min units. The activity was converted to trypsin concentration in nM units using purified, active site-titrated human and mouse cationic trypsin standards.

Preparation of cell lysates. The conditioned medium was removed from the tissue-culture plate wells, and the cells were rinsed with 1 mL phosphate-buffered saline (pH 7.4). The cells were then suspended in 1 mL phosphate-buffered saline, and centrifuged at 850 g for 10 minutes at 4^o^C. The pellets were resuspended in 500 µL ice-cold RIPA Buffer (catalog number R0278-50ML, Sigma), supplemented with protease inhibitors (Halt Protease and Phosphatase Inhibitor Cocktail, catalog number 78444, Thermo Fisher Scientific, 100 × factory stock diluted to 1×). After a 15-minute incubation on ice, lysates were sonicated three times for 10 seconds.

Western blotting. Cell lysates (10 µL per lane) were electrophoresed on 15% mini-gels and transferred to Immobilon-P membrane (catalog number IPVH00010, Sigma). The anti-PRSS1 sheep polyclonal antibody (catalog number AF3848, R&D Systems) was used at 1:5000 dilution. As secondary antibody, horseradish peroxidase (HRP)-conjugated donkey polyclonal anti-sheep IgG (catalog number HAF016, R&D Systems) was used at 1:2000 dilution. Human α-tubulin was measured as loading control using a mouse monoclonal antibody (DM1A) against α-tubulin (CP06, MilliporeSigma) at 1:2000 dilution followed by HRP-conjugated goat polyclonal anti-mouse IgG (HAF007, R&D Systems) at 1:2500 dilution. Incubation with the antibodies was carried out at 22^o^C. The primary antibodies were applied overnight, while the secondary antibodies were used for 1 hour. Protein bands were detected using the SuperSignal West Pico PLUS Chemiluminescent Substrate (catalog number 34580, Thermo Fisher Scientific). Densitometric evaluation of human cationic trypsinogen band intensities was performed with the Bio-Rad Image Lab 6.1 software.

Reverse-transcription quantitative PCR. Total RNA was extracted from transfected cells using the RNeasy Plus Mini Kit (catalog number 74136, Qiagen, Valencia, CA). Two µg RNA was reverse-transcribed using the High Capacity cDNA Reverse Transcription Kit (catalog number 4368814, Thermo Fisher Scientific). TaqMan Gene Expression Assays Hs00605631_g1, Mm00481058_m1, Hs00607129_gH, and Hs00358796_g1 were used with TaqMan Universal PCR Master Mix (catalog number 4304437, Thermo Fisher Scientific) to determine mRNA expression levels for human PRSS1, mouse Prss3b, human HSPA5 (BiP) and DDIT3 (CHOP), respectively. As housekeeping reference gene, GAPDH (Hs02758991_g1) was measured. Expression was quantitated using the comparative cycle threshold method, and results were expressed as fold-change relative to the average value of the cDNA samples (PRSS1, Prss3b) or empty vector (HSPA5, DDIT3) within each transfection.

Analysis of mRNA splicing. The PRSS1 and Prss3b mRNA were amplified using reverse-transcription PCR with the following primers. PRSS1 sense primer, 5′ – ACA CTC TAC CAC CAT GAA TCC – 3′, PRSS1 antisense primer, 5′ – GGT CAC TTT ATT GGT ATA GAG ACT G – 3′. Prss3b sense primer: 5′ – CCA CAG TGA GCA ACC ATG – 3′, Prss3b antisense primer: 5′ – GAG AGG GCT TAT GAA GAG C – 3′. The Supporting Information file indicates the position of the primers in the PRSS1 and Prss3b sequences. The amplicons were resolved on 2.5% agarose gels and stained with GreenGlo Safe DNA Dye (catalog number C788T73, Thomas Scientific, Swedesboro, NJ). The bands were excised from the agarose gel, purified, and analyzed by Sanger sequencing.

Statistical analysis. The difference of means between two groups and multiple groups was analyzed by unpaired t test and one-way ANOVA with Tukey’s post hoc test, respectively, using the Prism 10 software (GraphPad Software, Boston, MA). P < 0.05 was considered statistically significant.

Results

Study design. Our aim was to test the effect of previously characterized mini-introns from the human SPINK1 gene [23,24] on the expression of human and mouse cationic trypsinogen. The question of whether trypsinogen-derived intronic sequences might boost expression will be addressed in a separate study. Since prior experiments demonstrated that the optimal mini-intron length was 100 or 200 nt, we utilized both sizes in our construct design, using the shorter version for PRSS1 and the longer one for Prss3b. We also investigated how placement of the mini-intron at different locations within Prss3b affects expression.

Human cationic trypsinogen (PRSS1) minigene construct. We placed a 100 nucleotide long mini-intron from the human SPINK1 gene [23,24] in the cDNA of human PRSS1 between exons 1 and 2 (Fig 1, Supporting Information). To study the effect of the minigene on trypsinogen expression, HEK 293T cells were transiently transfected with the cDNA and minigene PRSS1 constructs. Conditioned media were collected 24 hours and 48 hours after transfection and analyzed by SDS-PAGE and Coomassie Blue staining (Fig 2A, Supporting Information). Furthermore, trypsinogen in the conditioned media was activated with human enteropeptidase and trypsin activity was assayed (Fig 2B). As expected, relative to the 24-hour time point, there was about twice as much secreted trypsinogen in the growth medium at 48 hours. However, at both times, there was no difference in the trypsinogen content of media from cells transfected with cDNA or minigene constructs. Western blot analysis of cell lysates 48 hours after transfection indicated a small (1.2-fold) increase in the intracellular trypsinogen content of cells transfected with PRSS1 minigene (band density 120 ± 12%, mean ± SD, n = 4) when compared to those with cDNA (100 ± 5%) (Fig 2C, Supporting Information). Reverse-transcription (RT) quantitative PCR analysis 48 hours post-transfection demonstrated a 2.5-fold increase in PRSS1 mRNA levels in cells transfected with the minigene construct relative to those with the cDNA construct (Fig 2D). Splicing fidelity was verified by RT-PCR, agarose gel electrophoresis, and DNA sequencing of the bands visible on the gels (Fig 2E, Supporting Information). As expected, the cDNA construct gave rise to a single band, which had the expected PRSS1 sequence. The minigene construct, however, yielded 2 bands. The larger, more intense band was the correctly spliced PRSS1 sequence while the smaller faint band was a mis-spliced form with an extra 74 nucleotides deleted from exon 2. We conclude that the presence of the mini-intron increased PRSS1 mRNA expression levels, however, this did not translate to increased cationic trypsinogen secretion. A likely explanation is that protein translation and/or folding represents a bottleneck in the heterologous expression of human cationic trypsinogen, and this cannot be overcome by increasing mRNA levels. To investigate whether protein folding might be rate limiting, we measured markers of endoplasmic reticulum (ER) stress in HEK 293T cells transfected with cDNA and minigene PRSS1 constructs. We found that relative to cells transfected with vector only, BiP (HSPA5) mRNA levels were increased by 2.0 ± 0.7-fold and 2.7 ± 1.3-fold, respectively, and CHOP (DDIT3) mRNA levels were increased by 1.2 ± 0.2-fold and 1.5 ± 0.4-fold, respectively (mean ± SD, n = 6). The higher ER stress elicited by the minigene construct is consistent with the western blot data showing elevated trypsinogen levels in cell lysates (Fig 2C).

Expression of human cationic trypsinogen in HEK 293T cells transfected with PRSS1 cDNA and minigene constructs.A, Secreted trypsinogen protein (arrow) in the conditioned medium 24 hours and 48 hours after transfection was assessed by SDS-PAGE and Coomassie Blue staining. Representative gels from 3 independent transfections are shown. B, Trypsinogen levels in the conditioned medium were determined by trypsin activity measurement after activation with enteropeptidase. Individual values from 4 transfections with duplicates (n = 8) are shown, with the mean and SD indicated. C, Western blot analysis of trypsinogen (arrow) levels in cell lysates 48 hours post-transfection. Alpha-tubulin was measured as loading control. Representative blots are shown. D, PRSS1 mRNA levels were measured 48 hours after transfection by reverse-transcription quantitative PCR and expressed as fold change relative to the average value of the cDNA construct. Individual values from 3 transfections with duplicates (n = 6) are shown, with the mean and SD indicated. E, Splicing of PRSS1 mRNA expressed from cDNA and minigene constructs was analyzed 48 hours after transfection by reverse-transcription PCR and agarose gel electrophoresis. The arrow indicates the correctly spliced PRSS1 band. The smaller faint band in the minigene samples indicated by the asterisk corresponds to an aberrant splice product in which nucleotide c.40 was spliced to c.114 resulting in the deletion of the mini-intron and an extra 74 nucleotides from exon 2.

Mouse cationic trypsinogen (Prss3b) minigene constructs. We generated two minigene constructs expressing mouse cationic trypsinogen by placing a 200-nucleotide mini-intron from human SPINK1 [24] in the Prss3b cDNA either between exons 1 and 2 (minigene v1) or between exons 2 and 3 (minigene v2) (Fig 1, Supporting Information). HEK 293T cells were transiently transfected with the Prss3b cDNA and minigene constructs. Conditioned media were collected 24 hours and 48 hours after transfection and analyzed by SDS-PAGE and Coomassie Blue staining (Fig 3A, Supporting Information). Trypsinogen content of the conditioned media was also measured by trypsin activity assays after enteropeptidase-mediated activation (Fig 3B). In contrast to human cationic trypsinogen, secretion of mouse cationic trypsinogen was increased by the minigene constructs both at the 24-hour (2.6-fold) and 48-hour (2.9-fold) timepoints, relative to the cDNA construct. The location of the intron had no impact on the stimulatory effect, as minigenes v1 and v2 yielded identical results. RT quantitative PCR analysis 48 hours after transfection demonstrated 4-fold (v1) and 5-fold (v2) increases in Prss3b mRNA levels in cells transfected with the minigene constructs relative to those with the Prss3b cDNA (Fig 3C). RT-PCR followed by agarose gel electrophoresis and DNA sequencing confirmed that the minigene constructs were spliced correctly with no mis-spliced byproducts (Fig 3D, Supporting Information). The results indicate that the minigene approach is suitable for increasing heterologous expression of mouse cationic trypsinogen, both at the mRNA and secreted protein levels.

Expression of mouse (T7) cationic trypsinogen in HEK 293T cells transfected with Prss3b cDNA and minigene constructs.A, Secreted trypsinogen protein (arrow) in the conditioned medium 24 hours and 48 hours after transfection was assessed by SDS-PAGE and Coomassie Blue staining. Representative gels from 3 independent transfections are shown. The faint bands in the 12-14 kDa region are due to autolysis of trypsinogen at Arg123. B, Trypsinogen levels in the conditioned medium were determined by trypsin activity measurement after activation with enteropeptidase. Individual values from 4 transfections with duplicates (n = 8) are shown, with the mean and SD indicated. C, Prss3b mRNA levels were measured 48 hours after transfection by reverse-transcription quantitative PCR and expressed as fold change relative to the average value of the cDNA construct. Individual values from 3 transfections with duplicates (n = 6) are shown, with the mean and SD indicated. D, Splicing of Prss3b mRNA expressed from cDNA and minigene constructs was analyzed 48 hours post-transfection by reverse-transcription PCR and agarose gel electrophoresis. The arrow indicates the correctly spliced Prss3b band.

Discussion

In this study, we tested the impact of a short mini-intron on the heterologous expression of human and mouse cationic trypsinogen in transfected HEK 293T cells. We previously found that a 100-nucleotide long intronic segment derived from intron 1 of human SPINK1 markedly increased the expression of human SPINK1 mRNA and SPINK1 protein in HEK 293T cells [23]. Similarly, 100–200-nucleotide long mini-introns from human SPINK1 robustly enhanced expression of mouse Spink1 mRNA and SPINK1 protein not only in HEK 293T cells but also in the AR42J rat acinar cell line and primary mouse acini [24]. Here, we tested this approach in the context of human and mouse cationic trypsinogens. We found that minigene constructs increased mRNA expression of both trypsinogen isoforms, however, the elevated mRNA levels resulted in higher protein secretion only for mouse cationic trypsinogen. Although we used different size introns for the PRSS1 (100 nt) and Prss3b (200 nt) minigenes, this cannot explain the difference seen in trypsinogen secretion. In our previously published experiments, the 100 nt and 200 nt introns stimulated expression of mouse Spink1 mRNA and mouse SPINK1 protein comparably, while complete or partial loss of effectiveness was seen with shorter (50 nt) and longer (400 nt) introns, respectively [24]. In case of human cationic trypsinogen, the most likely explanation for the inability of the minigene construct to increase protein secretion is ineffective translation and/or folding, which becomes rate limiting. In contrast, translation and/or folding of mouse cationic trypsinogen is less constrained, and elevated mRNA levels translate to higher protein secretion. We note, however, the discrepancy in the extent of the minigene-induced increase of Prssb3 mRNA (minigene v1 4-fold, minigene v2 5-fold) versus secreted trypsinogen protein (2.9-fold), suggesting a partial block at the translation and/or folding level.

The positive effect of introns on gene expression has been documented extensively [26,27]. The removal of introns via splicing is required for increased gene expression, which may be mediated by improved transcription, more efficient mRNA export, and slower mRNA decay [28,29]. Furthermore, exon-junction complexes stimulate expression by suppressing aberrant splice sites, preventing re-splicing of the mRNA, and increasing translation efficiency [30–33]. Finally, intron-dependent gene-looping can promote interactions between the splice sites and the promoter or terminator regions, thereby enhancing gene expression [34]. While our results confirm that the use of minigene constructs can generally increase heterologous expression in transfected cells, the effect size is target specific and requires experimental testing. Higher mRNA levels may not always result in elevated recombinant protein yield, which limits the utility of minigenes for enhanced expression of proteins that are difficult to translate and/or fold.

Supporting information

S1 FileThe file contains the DNA sequence of cDNA and minigene constructs and the uncropped gel and blot pictures.(PDF)