Limits...
An siRNA Screen Identifies the U2 snRNP Spliceosome as a Host Restriction Factor for Recombinant Adeno-associated Viruses.

Schreiber CA, Sakuma T, Izumiya Y, Holditch SJ, Hickey RD, Bressin RK, Basu U, Koide K, Asokan A, Ikeda Y - PLoS Pathog. (2015)

Bottom Line: Genetic disruption of U2 snRNP and associated proteins, such as SF3B1 and U2AF1, also increased expression from AAV vector, suggesting the critical role of U2 snRNP spliceosome complex in this host-mediated restriction.In summary, we identify U2 snRNP and associated splicing factors, which are known to be affected during adenoviral infection, as novel host restriction factors that effectively limit AAV transgene expression.Concurrently, we postulate that pharmacological/genetic manipulation of components of the spliceosomal machinery might enable more effective gene transfer modalities with recombinant AAV vectors.

View Article: PubMed Central - PubMed

Affiliation: Department of Molecular Medicine, Mayo Clinic, Rochester, Minnesota, United States of America.

ABSTRACT
Adeno-associated viruses (AAV) have evolved to exploit the dynamic reorganization of host cell machinery during co-infection by adenoviruses and other helper viruses. In the absence of helper viruses, host factors such as the proteasome and DNA damage response machinery have been shown to effectively inhibit AAV transduction by restricting processes ranging from nuclear entry to second-strand DNA synthesis. To identify host factors that might affect other key steps in AAV infection, we screened an siRNA library that revealed several candidate genes including the PHD finger-like domain protein 5A (PHF5A), a U2 snRNP-associated protein. Disruption of PHF5A expression selectively enhanced transgene expression from AAV by increasing transcript levels and appears to influence a step after second-strand synthesis in a serotype and cell type-independent manner. Genetic disruption of U2 snRNP and associated proteins, such as SF3B1 and U2AF1, also increased expression from AAV vector, suggesting the critical role of U2 snRNP spliceosome complex in this host-mediated restriction. Notably, adenoviral co-infection and U2 snRNP inhibition appeared to target a common pathway in increasing expression from AAV vectors. Moreover, pharmacological inhibition of U2 snRNP by meayamycin B, a potent SF3B1 inhibitor, substantially enhanced AAV vector transduction of clinically relevant cell types. Further analysis suggested that U2 snRNP proteins suppress AAV vector transgene expression through direct recognition of intact AAV capsids. In summary, we identify U2 snRNP and associated splicing factors, which are known to be affected during adenoviral infection, as novel host restriction factors that effectively limit AAV transgene expression. Concurrently, we postulate that pharmacological/genetic manipulation of components of the spliceosomal machinery might enable more effective gene transfer modalities with recombinant AAV vectors.

No MeSH data available.


Related in: MedlinePlus

PHF5A and U2 snRNP component SF3B1 interact with AAV capsid.(A) HeLa or PHF5A-HA-expressing cell lysates were used to pull-down the HA-tagged PHF5A by anti-HA agarose beads. After 15 washes, the HA-tagged PHF5A was detected by anti-HA antibody. (B) Control or PHF5A-HA-expressing HeLa cells were transduced by AAV2 and AAV9 CMV-Luc vectors (MOI 4 x 105) and total AAV genome copies in the HA pulldown were determined by quantitative real-time PCR. (C) AAV2 CMV-Luc vector (3 x 1010 genome copies) was unheated or preheated for 30 min at 65°C. PHF5A-HA-over-expressing HeLa cell lysates were then incubated with vectors for 1 hour at 4°C, followed by pulldown of PHF5A-HA. AAV vector genome copies in the precipitates were determined by quantitative real-time PCR. (D) HeLa cells were infected with the AAV2 CMV-Luc vector (4 x 1010 genome copies/well) for 5 min, 4 or 12 hours. Confocal microscopy analysis was performed to detect the subcellular localizations of AAV vector particles (green) and PHF5A (red). Nuclei were counterstained by DAPI (blue). (E) HeLa cells were infected with AAV2 CMV-Luc vectors (4 x 1010 genome copies/well) or equivalent amounts of empty AAV2 vectors for 4 hours, and cells were analyzed for co-localization of AAV2 capsid and endogenous PHF5A signals. Prominent co-localized signals were indicated by white arrows. (F) HeLa cells were treated with the PHF5A siRNA for 24 hours, followed by transduction with the AAV2 vector as in E for 4 hours. AAV2 vector particles were detected by anti-AAV2 capsid A20 antibody, and the patterns of cytoplasmic and nuclear accumulations of AAV2 vector particles were compared between control and PHF5A-ablated cells. Representative Z-stack images of the middle sections (slices 3 and 4) from control and PHF5A knockdown cells are shown. (G) Schematic representation for the iodixanol cushion method to enrich cellular factors interacting with particulated AAV capsids. (H) HeLa cell lysates were incubated with AAV2 CMV-Luc vectors (5 x 1010 genome copies) for 1 hour at 4°C. After centrifugation over 25% iodixanol, three layers (the upper phase, lower phase, and pellet) were separately harvested for Western blotting. AAV capsid proteins VP1, 2 and 3, phospho-SF3B1, and endogenous PHF5A were detected by A20, anti-SF3B1, and anti-PHF5A antibodies, respectively. (I) Same as H for AAV capsid proteins, except that empty AAV2 VP3 only capsids were used for SF3B1 co-precipitation. (J) Control or AAV VP1-over-expressing 293T cell lysates were used to pull-down the AAV VP1 protein by A20 antibody. After 15 washes, the pellets were probed for SF3B1 enrichment by anti-SF3B1 antibody.
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4526370&req=5

ppat.1005082.g004: PHF5A and U2 snRNP component SF3B1 interact with AAV capsid.(A) HeLa or PHF5A-HA-expressing cell lysates were used to pull-down the HA-tagged PHF5A by anti-HA agarose beads. After 15 washes, the HA-tagged PHF5A was detected by anti-HA antibody. (B) Control or PHF5A-HA-expressing HeLa cells were transduced by AAV2 and AAV9 CMV-Luc vectors (MOI 4 x 105) and total AAV genome copies in the HA pulldown were determined by quantitative real-time PCR. (C) AAV2 CMV-Luc vector (3 x 1010 genome copies) was unheated or preheated for 30 min at 65°C. PHF5A-HA-over-expressing HeLa cell lysates were then incubated with vectors for 1 hour at 4°C, followed by pulldown of PHF5A-HA. AAV vector genome copies in the precipitates were determined by quantitative real-time PCR. (D) HeLa cells were infected with the AAV2 CMV-Luc vector (4 x 1010 genome copies/well) for 5 min, 4 or 12 hours. Confocal microscopy analysis was performed to detect the subcellular localizations of AAV vector particles (green) and PHF5A (red). Nuclei were counterstained by DAPI (blue). (E) HeLa cells were infected with AAV2 CMV-Luc vectors (4 x 1010 genome copies/well) or equivalent amounts of empty AAV2 vectors for 4 hours, and cells were analyzed for co-localization of AAV2 capsid and endogenous PHF5A signals. Prominent co-localized signals were indicated by white arrows. (F) HeLa cells were treated with the PHF5A siRNA for 24 hours, followed by transduction with the AAV2 vector as in E for 4 hours. AAV2 vector particles were detected by anti-AAV2 capsid A20 antibody, and the patterns of cytoplasmic and nuclear accumulations of AAV2 vector particles were compared between control and PHF5A-ablated cells. Representative Z-stack images of the middle sections (slices 3 and 4) from control and PHF5A knockdown cells are shown. (G) Schematic representation for the iodixanol cushion method to enrich cellular factors interacting with particulated AAV capsids. (H) HeLa cell lysates were incubated with AAV2 CMV-Luc vectors (5 x 1010 genome copies) for 1 hour at 4°C. After centrifugation over 25% iodixanol, three layers (the upper phase, lower phase, and pellet) were separately harvested for Western blotting. AAV capsid proteins VP1, 2 and 3, phospho-SF3B1, and endogenous PHF5A were detected by A20, anti-SF3B1, and anti-PHF5A antibodies, respectively. (I) Same as H for AAV capsid proteins, except that empty AAV2 VP3 only capsids were used for SF3B1 co-precipitation. (J) Control or AAV VP1-over-expressing 293T cell lysates were used to pull-down the AAV VP1 protein by A20 antibody. After 15 washes, the pellets were probed for SF3B1 enrichment by anti-SF3B1 antibody.

Mentions: Next, we assessed the interaction between PHF5A and AAV vector components. Upon pull-down of HA-tagged PHF5A from AAV vector-infected cells (Figs 4A and S5), total DNA in the pellets was assessed for AAV vector genomes. Quantitative real-time PCR detected significantly more AAV vector genome DNA in the HA pulldown from PHF5A-HA-over-expressing cells as opposed to control cells (Fig 4B). We also tested the influence of heat-mediated conformational changes of viral capsids, which lead to the exposure of the hidden VP1 N-terminal and viral genomic DNA [39,40], on the interaction with PHF5A. A three-fold increase in AAV genome copies was detected in the HeLa-PHF5A-HA pulldown when the cell lysates were incubated with pre-heated AAV2 CMV-Luc vector than non-heated vector (Fig 4C). Those data indicate interaction of PHF5A with AAV vector genome, directly or indirectly.


An siRNA Screen Identifies the U2 snRNP Spliceosome as a Host Restriction Factor for Recombinant Adeno-associated Viruses.

Schreiber CA, Sakuma T, Izumiya Y, Holditch SJ, Hickey RD, Bressin RK, Basu U, Koide K, Asokan A, Ikeda Y - PLoS Pathog. (2015)

PHF5A and U2 snRNP component SF3B1 interact with AAV capsid.(A) HeLa or PHF5A-HA-expressing cell lysates were used to pull-down the HA-tagged PHF5A by anti-HA agarose beads. After 15 washes, the HA-tagged PHF5A was detected by anti-HA antibody. (B) Control or PHF5A-HA-expressing HeLa cells were transduced by AAV2 and AAV9 CMV-Luc vectors (MOI 4 x 105) and total AAV genome copies in the HA pulldown were determined by quantitative real-time PCR. (C) AAV2 CMV-Luc vector (3 x 1010 genome copies) was unheated or preheated for 30 min at 65°C. PHF5A-HA-over-expressing HeLa cell lysates were then incubated with vectors for 1 hour at 4°C, followed by pulldown of PHF5A-HA. AAV vector genome copies in the precipitates were determined by quantitative real-time PCR. (D) HeLa cells were infected with the AAV2 CMV-Luc vector (4 x 1010 genome copies/well) for 5 min, 4 or 12 hours. Confocal microscopy analysis was performed to detect the subcellular localizations of AAV vector particles (green) and PHF5A (red). Nuclei were counterstained by DAPI (blue). (E) HeLa cells were infected with AAV2 CMV-Luc vectors (4 x 1010 genome copies/well) or equivalent amounts of empty AAV2 vectors for 4 hours, and cells were analyzed for co-localization of AAV2 capsid and endogenous PHF5A signals. Prominent co-localized signals were indicated by white arrows. (F) HeLa cells were treated with the PHF5A siRNA for 24 hours, followed by transduction with the AAV2 vector as in E for 4 hours. AAV2 vector particles were detected by anti-AAV2 capsid A20 antibody, and the patterns of cytoplasmic and nuclear accumulations of AAV2 vector particles were compared between control and PHF5A-ablated cells. Representative Z-stack images of the middle sections (slices 3 and 4) from control and PHF5A knockdown cells are shown. (G) Schematic representation for the iodixanol cushion method to enrich cellular factors interacting with particulated AAV capsids. (H) HeLa cell lysates were incubated with AAV2 CMV-Luc vectors (5 x 1010 genome copies) for 1 hour at 4°C. After centrifugation over 25% iodixanol, three layers (the upper phase, lower phase, and pellet) were separately harvested for Western blotting. AAV capsid proteins VP1, 2 and 3, phospho-SF3B1, and endogenous PHF5A were detected by A20, anti-SF3B1, and anti-PHF5A antibodies, respectively. (I) Same as H for AAV capsid proteins, except that empty AAV2 VP3 only capsids were used for SF3B1 co-precipitation. (J) Control or AAV VP1-over-expressing 293T cell lysates were used to pull-down the AAV VP1 protein by A20 antibody. After 15 washes, the pellets were probed for SF3B1 enrichment by anti-SF3B1 antibody.
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4526370&req=5

ppat.1005082.g004: PHF5A and U2 snRNP component SF3B1 interact with AAV capsid.(A) HeLa or PHF5A-HA-expressing cell lysates were used to pull-down the HA-tagged PHF5A by anti-HA agarose beads. After 15 washes, the HA-tagged PHF5A was detected by anti-HA antibody. (B) Control or PHF5A-HA-expressing HeLa cells were transduced by AAV2 and AAV9 CMV-Luc vectors (MOI 4 x 105) and total AAV genome copies in the HA pulldown were determined by quantitative real-time PCR. (C) AAV2 CMV-Luc vector (3 x 1010 genome copies) was unheated or preheated for 30 min at 65°C. PHF5A-HA-over-expressing HeLa cell lysates were then incubated with vectors for 1 hour at 4°C, followed by pulldown of PHF5A-HA. AAV vector genome copies in the precipitates were determined by quantitative real-time PCR. (D) HeLa cells were infected with the AAV2 CMV-Luc vector (4 x 1010 genome copies/well) for 5 min, 4 or 12 hours. Confocal microscopy analysis was performed to detect the subcellular localizations of AAV vector particles (green) and PHF5A (red). Nuclei were counterstained by DAPI (blue). (E) HeLa cells were infected with AAV2 CMV-Luc vectors (4 x 1010 genome copies/well) or equivalent amounts of empty AAV2 vectors for 4 hours, and cells were analyzed for co-localization of AAV2 capsid and endogenous PHF5A signals. Prominent co-localized signals were indicated by white arrows. (F) HeLa cells were treated with the PHF5A siRNA for 24 hours, followed by transduction with the AAV2 vector as in E for 4 hours. AAV2 vector particles were detected by anti-AAV2 capsid A20 antibody, and the patterns of cytoplasmic and nuclear accumulations of AAV2 vector particles were compared between control and PHF5A-ablated cells. Representative Z-stack images of the middle sections (slices 3 and 4) from control and PHF5A knockdown cells are shown. (G) Schematic representation for the iodixanol cushion method to enrich cellular factors interacting with particulated AAV capsids. (H) HeLa cell lysates were incubated with AAV2 CMV-Luc vectors (5 x 1010 genome copies) for 1 hour at 4°C. After centrifugation over 25% iodixanol, three layers (the upper phase, lower phase, and pellet) were separately harvested for Western blotting. AAV capsid proteins VP1, 2 and 3, phospho-SF3B1, and endogenous PHF5A were detected by A20, anti-SF3B1, and anti-PHF5A antibodies, respectively. (I) Same as H for AAV capsid proteins, except that empty AAV2 VP3 only capsids were used for SF3B1 co-precipitation. (J) Control or AAV VP1-over-expressing 293T cell lysates were used to pull-down the AAV VP1 protein by A20 antibody. After 15 washes, the pellets were probed for SF3B1 enrichment by anti-SF3B1 antibody.
Mentions: Next, we assessed the interaction between PHF5A and AAV vector components. Upon pull-down of HA-tagged PHF5A from AAV vector-infected cells (Figs 4A and S5), total DNA in the pellets was assessed for AAV vector genomes. Quantitative real-time PCR detected significantly more AAV vector genome DNA in the HA pulldown from PHF5A-HA-over-expressing cells as opposed to control cells (Fig 4B). We also tested the influence of heat-mediated conformational changes of viral capsids, which lead to the exposure of the hidden VP1 N-terminal and viral genomic DNA [39,40], on the interaction with PHF5A. A three-fold increase in AAV genome copies was detected in the HeLa-PHF5A-HA pulldown when the cell lysates were incubated with pre-heated AAV2 CMV-Luc vector than non-heated vector (Fig 4C). Those data indicate interaction of PHF5A with AAV vector genome, directly or indirectly.

Bottom Line: Genetic disruption of U2 snRNP and associated proteins, such as SF3B1 and U2AF1, also increased expression from AAV vector, suggesting the critical role of U2 snRNP spliceosome complex in this host-mediated restriction.In summary, we identify U2 snRNP and associated splicing factors, which are known to be affected during adenoviral infection, as novel host restriction factors that effectively limit AAV transgene expression.Concurrently, we postulate that pharmacological/genetic manipulation of components of the spliceosomal machinery might enable more effective gene transfer modalities with recombinant AAV vectors.

View Article: PubMed Central - PubMed

Affiliation: Department of Molecular Medicine, Mayo Clinic, Rochester, Minnesota, United States of America.

ABSTRACT
Adeno-associated viruses (AAV) have evolved to exploit the dynamic reorganization of host cell machinery during co-infection by adenoviruses and other helper viruses. In the absence of helper viruses, host factors such as the proteasome and DNA damage response machinery have been shown to effectively inhibit AAV transduction by restricting processes ranging from nuclear entry to second-strand DNA synthesis. To identify host factors that might affect other key steps in AAV infection, we screened an siRNA library that revealed several candidate genes including the PHD finger-like domain protein 5A (PHF5A), a U2 snRNP-associated protein. Disruption of PHF5A expression selectively enhanced transgene expression from AAV by increasing transcript levels and appears to influence a step after second-strand synthesis in a serotype and cell type-independent manner. Genetic disruption of U2 snRNP and associated proteins, such as SF3B1 and U2AF1, also increased expression from AAV vector, suggesting the critical role of U2 snRNP spliceosome complex in this host-mediated restriction. Notably, adenoviral co-infection and U2 snRNP inhibition appeared to target a common pathway in increasing expression from AAV vectors. Moreover, pharmacological inhibition of U2 snRNP by meayamycin B, a potent SF3B1 inhibitor, substantially enhanced AAV vector transduction of clinically relevant cell types. Further analysis suggested that U2 snRNP proteins suppress AAV vector transgene expression through direct recognition of intact AAV capsids. In summary, we identify U2 snRNP and associated splicing factors, which are known to be affected during adenoviral infection, as novel host restriction factors that effectively limit AAV transgene expression. Concurrently, we postulate that pharmacological/genetic manipulation of components of the spliceosomal machinery might enable more effective gene transfer modalities with recombinant AAV vectors.

No MeSH data available.


Related in: MedlinePlus