Proteomic and functional analyses of protein-DNA complexes during gene transfer

Abstract

One of the barriers to successful nonviral gene delivery is the crowded cytoplasm, which plasmids need to actively traverse for gene expression. Relatively little is known about how this process occurs, but our lab and others have shown that the microtubule network and motors are required for plasmid movement to the nucleus. To further investigate how plasmids exploit normal physiological processes to transfect cells, we have taken a proteomics approach to identify the proteins that comprise the plasmid-trafficking complex. We have developed a live cell DNA-protein pull-down assay to isolate complexes at certain time points post-transfection (15 minutes to 4 hours) for analysis by mass spectrometry (MS). Plasmids containing promoter sequences bound hundreds of unique proteins as early as 15 minutes post-electroporation, while a plasmid lacking any eukaryotic sequences failed to bind many of the proteins. Specific proteins included microtubule-based motor proteins (e.g., kinesin and dynein), proteins involved in protein nuclear import (e.g., importin 1, 2, 4, and 7, Crm1, RAN, and several RAN-binding proteins), a number of heterogeneous nuclear ribonucleoprotein (hnRNP)- and mRNA-binding proteins, and transcription factors. The significance of several of the proteins involved in protein nuclear localization and plasmid trafficking was determined by monitoring movement of microinjected fluorescently labeled plasmids via live cell particle tracking in cells following protein knockdown by small-interfering RNA (siRNA) or through the use of specific inhibitors. While importin β1 was required for plasmid trafficking and subsequent nuclear import, importin α1 played no role in microtubule trafficking but was required for optimal plasmid nuclear import. Surprisingly, the nuclear export protein Crm1 also was found to complex with the transfected plasmids and was necessary for plasmid trafficking along microtubules and nuclear import. Our results show that various proteins involved in nuclear import and export influence intracellular trafficking of plasmids and subsequent nuclear accumulation.

Figure 1

A comparison of total numbers of proteins identified in the plasmid complexes. The total number of unique identified proteins for each condition was determined, using the criterion that it must be present in atleast two of three replicates. From these lists, the constructs and timepoints were compared using Venn diagrams to represent the number of proteins unique to that group (outer segments), in common (overlapping segments), or total (addition of each segment). (a) There were more unique identified proteins at 30 minutes in complexes with pCMV-DTS compared with pBR322. (b) Identified protein numbers at 30 minutes in complexes with pCMV-DTS, pBR322, or the no DNA control. (c) Total numbers of proteins in complex with pCMV-DTS change over the time course of transfection. (d) Total numbers of proteins in complex with pBR322 at 30 and 240 minutes. (e) Total numbers of proteins in complex with pCMV-DTS versus pBR322 at 240 minutes. DTS, Simian Virus 40 DNA nuclear-targeting sequence; pCMV, cytomegalovirus immediate early promoter.

Figure 2

Proteins are more abundant in pCMV-DTS complexes. Biotinylated plasmids were electroporated into adherent A549 cells, cross-linked at 30 minutes post-transfection, and cells were lysed. Plasmid–protein complexes were isolated via streptavidin-coated beads, cross-links were reversed by boiling beads with Laemmli sample buffer, and the resulting lysates were run in western blots using antibodies against various proteins of interest. “Lysate” represents crude lysates with no bead precipitation. DTS, Simian Virus 40 DNA nuclear-targeting sequence; HSP, heat shock protein; pCMV, cytomegalovirus immediate early promoter.

Figure 3

Importin β1 is required for plasmid trafficking and nuclear import. (a) Western blot images show reduced levels of certain nuclear import proteins in lysates from cells transfected with siRNA against either importin (imp) 7, α, or β1, compared to a scramble (scr) control siRNA. Average protein knockdown was 73% for imp 7, 70% for imp α, and 75% for imp β1. (b) Quantum dot-labeled plasmids were cytoplasmically microinjected into adherent A549 cells after transfection with various siRNAs (24 hours for imp α, 48 hours for imp β1, and 72 hours for imp 7). Cells were imaged at 1-second intervals over 5–10 minutes, and the net (mean) velocity of individual microinjected plasmids was determined using particle tracking software (PolyParticleTracker, MATLAB). The frequency distribution histograms are plotted as the number of plasmids moving at certain velocities for each condition. The pBR322 plasmid was tracked in untreated cells as a control. (c) Individual plasmid net velocities were averaged for each of the conditions from atleast three separate experiments. Error bars represent mean ± SEM, n = 70–80. **P < 0.001 compared with scramble siRNA. (d) A549 cells were plated on coverslips and then cytoplasmically microinjected with CY3-PNA–labeled pBR322 or pCMV-DTS plasmids. Cells were incubated and fixed with 4% paraformaldehyde 4 hours later. Representative images show cells scored for nuclear import or no import. Bar, 10 µm. (e) Injected cells were imaged and scored for nuclear import. Experiments were performed in triplicate, and atleast 50–100 cells were scored per condition. *P < 0.01, **P < 0.001 compared with scramble siRNA. Error bars represent mean ± SD. DAPI, 4′,6-diamidino-2-phenylindole; DTS, Simian Virus 40 DNA nuclear-targeting sequence; KD, knockdown; pCMV, cytomegalovirus immediate early promoter; PNA, peptide nucleic acid; siRNA, small-interfering RNA.

Figure 4

Exportin 1 activity is required for plasmid trafficking and nuclear import. (a) A549 cells were plated on coverslips and incubated with or without 50 nmol/l leptomycin B (LMB) for 1 hour, fixed with 4% paraformaldehyde, and immunofluorescence was carried out using antibodies against the p65 subunit of NFκB. After treatment with LMB, p65 is retained in the nucleus (bottom panel). (b) Adherent A549 cells were treated with or without LMB as in a, and individual plasmids were tracked for 5–10 minutes and velocities analyzed as in Figure 3b. (c) Individual plasmid net velocities were averaged for each of the conditions. Error bars represent mean ± SEM, n = 70–80. **P < 0.001 for no LMB versus LMB treatment (− or +LMB). (d) A549 cells were plated on coverslips and incubated with or without LMB as in a. Cells were then cytoplasmically microinjected with either GST-fluorescein–labeled M9 (Fl-M9), NLS (Fl-NLS), or reverse NLS control proteins (Fl-SLN) and fixed 2 hours later with 4% paraformaldehyde. Representative images show cytoplasmic (top panel) and nuclear import (bottom panels) of proteins. Injected cells were imaged and scored for nuclear import as in Figure 3d,e to show that LMB treatment does not block general nuclear import. Experiments were performed in duplicate and atleast 50 cells were scored per condition. (e) Cells were treated with and without LMB as in a, microinjected with CY3-labeled plasmids, fixed, and scored for nuclear import as in Figure 3d,e. Error bars represent mean ± SD. **P < 0.001 compared to pCMV-DTS with no LMB. DAPI, 4′,6-diamidino-2-phenylindole; DTS, Simian Virus 40 DNA nuclear-targeting sequence; NFκB, nuclear factor κB; NLS, nuclear localization signal; pCMV, cytomegalovirus immediate early promoter.