Repeated cell sorting ensures the homogeneity of ocular cell populations expressing a transgenic protein

Abstract

Transgenic proteins can be routinely expressed in various mammalian cell types via different transgenic systems, but the efficiency of transgene expression is constrained by the complex interplay among factors such as the temporal consistency of expression and compatibility with specific cell types, including ocular cells. Here, we report a more efficient way to express an enhanced green fluorescent protein (EGFP) in human corneal fibroblasts, corneal epithelial cells, and conjunctival epithelial cells through a lentiviral expression system. The relative transducing unit criterion for EGFP-expressing pseudovirions was first determined in HEK-293T cells. Homogeneous populations of EGFP-positive and EGFP-negative cells could be isolated by cell sorting. The half-maximal inhibitory concentration (IC50) value for puromycin was calculated according to viability curves for each cell type. The results revealed that cell types differed with respect to EGFP expression efficiency after transduction with the same amount of EGFP-encoding pseudovirions. Using a cell sorter, the homogeneity of EGFP-positive cells reached >95%. In the initial sorting stage, however, the efficiency of EGFP expression in the sorted cells was noticeably reduced after two rounds of sequential culture, but repeated sorting for up to four rounds yielded homogeneous EGFP-positive human corneal fibroblasts that could be maintained in continuous culture in vitro. The sorted EGFP-positive cells retained their proper morphology and cell type-specific protein expression patterns. Puromycin resistance was found to depend on cell type, indicating that the IC50 for puromycin must be determined for each cell type to ensure the isolation of homogeneous EGFP-positive cells. Taken together, repeated cell sorting is an efficient means of obtaining homogeneous populations of ocular cells expressing a transgenic protein during continuous culture without the potential confounding effects of antibiotics.

Fig 1. Differences in expression efficiency after transduction of various ocular cell types with EGFP-encoding pseudovirions.

Each cell type was seeded into a 10-cm culture dish: (A) 2.0 × 10⁵ HCFs, (B) 2.0 × 10⁶ HCnEs, (C) 2.0 × 10⁶ HCjEs, and (D) 2.0 × 10⁶ HEK-293T cells. At 24 h post-seeding, each cell type was treated with EGFP-encoding pseudovirions at 1.0 × 10⁶ R.T.U. or 5.0 × 10⁶ R.T.U. for 16 h. Subsequently, the culture media were refreshed for all transduced cells until >90% confluency was attained. Cell sorting was used to analyze the transduced cells with 1.0 × 10⁶ R.T.U. (left panels) or 5.0 × 10⁶ R.T.U. (middle panels). The right-most panels present data acquired after sorting of the cells transduced with 5.0 × 10⁶ R.T.U. The percentages represent the homogeneity of EGFP-positive cells in the gating regions. (E) The mean signal intensity was the average signal intensity of the gated EGFP-positive cells in the FL1 channel (S1 Fig). The analysis of transduction of each cell type with two different titers of EGFP pseudovirions was repeated three times with similar trends. The shown data represent the results of one of the experiments. Data are expressed as the mean ± S.D. from three independent experiments. Differences between two different titers of EGFP pseudovirions were analyzed with the Student’s t-test (***P < 0.001). FL1 is the green channel, and FL3 is the red channel. R.T.U., relative transducing unit.

Fig 2. Purity of the EGFP-positive cells after sequential passaging and sorting.

Each cell type was seeded into a 10-cm culture dish: (A) 2.0 × 10⁵ HCFs, (B) 2.0 × 10⁶ HCnEs, (C) 2.0 × 10⁶ HCjEs, and (D) 2.0 × 10⁶ HEK-293T cells. At 24 h post-seeding, each cell type was incubated EGFP-encoding pseudovirions at 5.0 × 10⁶ R.T.U. for 16 h. Subsequently, the culture media were refreshed for each of the transduced cells until >90% confluency was attained. The transduced cells were analyzed and sorted with a cell sorter at the indicated rounds. All rounds of cell sorting were repeated until the proportion of EGFP-positive cells was >95% in two successive passages. Box-whisker plots were drawn to show the proportion of sorted cells expressing EGFP. The n values on the box-whisker plots represent the number of different donor corneas from which HCFs were harvested (A). The n values on the box-whisker plots represent the number of experiments performed at the indicated sorting times for HCnEs (B), HCjEs (C), and HEK-293T cells (D). Horizontal lines represent the median, boxes denote the two inner quartiles, and whisker bars show the maximum and minimum values for expression efficiency. S2 Fig lists all the proportions of the EGFP-positive cells during cell sorting at the indicated rounds. Differences in the purity of EGFP-positive cells were analyzed separately by one-way analysis of variance followed by Tukey’s honestly significant difference post-hoc test (*P < 0.05; **P < 0.01; ***P < 0.001).

Fig 3. Viability of the sorted EGFP-expressing cells.

Each cell type was seeded into individual wells of a 96-well plate: (A) 2.0 × 10³ HCFs, (B) 2.0 × 10⁴ HCnEs, (C) 2.0 × 10⁴ HCjEs, and (D) 2.0 × 10⁴ HEK-293T cells. At 24 h post-seeding as the starting time (0 h), cell viability was analyzed at 0, 24, and 48 h by CCK-8 reagent. Data are expressed as the mean ± S.D. from three independent experiments. The upper bars represent the S.D. of the mean. Differences in the viability of EGFP-positive and EGFP-negative cells were analyzed with the Student’s t-test (*P < 0.05; **P < 0.01).

Fig 4. Morphology of the sorted EGFP-expressing cells.

(A-D) The morphology of the primary HCFs, HCnEs, HCjEs, and 293T cells without transduction by EGFP-expressing pseudovirions but after repeated cell sorting was captured under a bright field microscope at 100× magnification. (E-H) The morphology of the sorted EGFP-positive cells was captured using a light microscope at 100× magnification. (I-L) In the same fields, fluorescence images of the sorted EGFP-positive cells were captured using a fluorescence microscope. This experiment was performed three times with similar results. Scale bars, 100 μm. Other images for cell morphology are available at Figshare (DOI: 10.6084/m9.figshare.18134096).

Fig 5. Alteration of cell type-specific expression proteins in the sorted EGFP-expressing cells.

Non-transduced cells, sorted EGFP-negative cells, and sorted EGFP-positive cells of the four cell types were separately cultured in 6-cm dishes, and the cell lysates were prepared for immunoblotting analysis of cell type-specific protein profiles. This experiment was performed three times with similar results. Bcl-xL, B-cell lymphoma-extra large; PTEN, phosphatase and tensin homolog; FAK, focal adhesion kinase; CK13, cytokeratin-13; ZO-1, zonula occludens-1; α-SMA, alpha-smooth muscle actin; EGFP, enhanced green fluorescent protein.

Fig 6. Tolerance to puromycin of the sorted EGFP-expressing cells.

All EGFP-positive and EGFP-negative cells were subjected to repeated cell sorting to achieve >95% purity. Each cell type was seeded into individual wells of a 96-well plate: (A) 2.0 × 10³ HCFs, (B) 2.0 × 10⁴ HCnEs, (C) 2.0 × 10⁴ HCjEs, and (D) 2.0 × 10⁴ HEK-293T cells. At 24 h post-seeding, the medium was refreshed, and both EGFP-positive and EGFP-negative cells were incubated with a series of 3.16-fold dilutions of puromycin ranging from 0.01 to 316 μg/ml. After incubation for 48 h, cell viability was analyzed with the CCK-8 reagent. Data are expressed as the mean ± S.D. from three independent experiments. The bars correspond to the S.D. of the mean.