Limits...
Physiologically generated presenilin 1 lacking exon 8 fails to rescue brain PS1-/- phenotype and forms complexes with wildtype PS1 and nicastrin.

Brautigam H, Moreno CL, Steele JW, Bogush A, Dickstein DL, Kwok JB, Schofield PR, Thinakaran G, Mathews PM, Hof PR, Gandy S, Ehrlich ME - Sci Rep (2015)

Bottom Line: Also, PS1(∆exon8) did not rescue Aβ generation in PS1/2 double knockout cells indicating its identity as a severe loss-of-function splice form.Further co-IP demonstrates that PS1(∆exon8) interacts with nicastrin, participating in the γ-secretase complex formation.These data support that catalytically inactive PS1(∆exon8) is generated physiologically and participates in protein-protein interactions.

View Article: PubMed Central - PubMed

Affiliation: Fishberg Department of Neuroscience and Friedman Brain Institute, Icahn School of Medicine at Mount Sinai, New York, NY 10029.

ABSTRACT
The presenilin 1 (PSEN1) L271V mutation causes early-onset familial Alzheimer's disease by disrupting the alternative splicing of the PSEN1 gene, producing some transcripts harboring the L271V point mutation and other transcripts lacking exon 8 (PS1(∆exon8)). We previously reported that PS1 L271V increased amyloid beta (Aβ) 42/40 ratios, while PS1(∆exon8) reduced Aβ42/40 ratios, indicating that the former and not the exon 8 deletion transcript is amyloidogenic. Also, PS1(∆exon8) did not rescue Aβ generation in PS1/2 double knockout cells indicating its identity as a severe loss-of-function splice form. PS1(∆exon8) is generated physiologically raising the possibility that we had identified the first physiological inactive PS1 isoform. We studied PS1(∆exon8) in vivo by crossing PS1(∆exon8) transgenics with either PS1- or Dutch APP(E693Q) mice. As a control, we crossed APP(E693Q) with mice expressing a deletion in an adjacent exon (PS1(∆exon9)). PS1(∆exon8) did not rescue embryonic lethality or Notch-deficient phenotypes of PS1- mice displaying severe loss of function in vivo. We also demonstrate that this splice form can interact with wildtype PS1 using cultured cells and co-immunoprecipitation (co-IP)/bimolecular fluorescence complementation. Further co-IP demonstrates that PS1(∆exon8) interacts with nicastrin, participating in the γ-secretase complex formation. These data support that catalytically inactive PS1(∆exon8) is generated physiologically and participates in protein-protein interactions.

No MeSH data available.


Related in: MedlinePlus

PS1 and PS1∆exon8 co-immunoprecipitate in vitro.Co-IP immunoblots of different epitope-tagged PS1 and PS1∆exon8 proteins. Constructs were cotransfected in six different combinations: (A) (1) empty vectors containing either HA or FLAG only, (2) HA-tagged PS1 and FLAG-tagged PS1, and (3) HA-tagged PS1∆exon8 and FLAG-tagged PS1∆exon8. (B) (4) No transfection, (5) HA-tagged PS1 and FLAG-tagged PS1∆exon8, and (6) FLAG-tagged PS1 and HA-tagged PS1∆exon8. Transfected constructs were co-immunoprecipitated with anti-HA antibody (1.5 μg per 300 μg cell lysate) and blotted with the anti-FLAG antibody (top immunoblot) and the anti-human PS1 antibody NT.1 (directly below FLAG immunoblot). The co-immunoprecipitated band is indicated by arrows and corresponds to full-length PS1 (~50 kDa). Note there is no band present in the eluate lane for Empty Vectors and No Transfection. (C,D) For pooled lysates, constructs were transfected in separate wells and then pooled after lysis. The pooled lysed cells were co-immunoprecipitated with anti-HA antibody and blotted for anti-FLAG as described above. No PS1 band was detected in the eluate (indicated by arrows). (E,F) The experiment in (A,B) was repeated but transfected constructs were co-immunoprecipitated with either IgG rabbit as a negative control or the anti-HA antibody. No PS1 band was detected in the IgG control but was detected in the HA lane, except in the Empty Vectors and No Transfection lanes (indicated by arrows, asterisk marks non-specific bands corresponding to the IgG heavy chain). I = input, E = eluate, IgG = rabbit IgG, and HA = HA antibody.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4660297&req=5

f6: PS1 and PS1∆exon8 co-immunoprecipitate in vitro.Co-IP immunoblots of different epitope-tagged PS1 and PS1∆exon8 proteins. Constructs were cotransfected in six different combinations: (A) (1) empty vectors containing either HA or FLAG only, (2) HA-tagged PS1 and FLAG-tagged PS1, and (3) HA-tagged PS1∆exon8 and FLAG-tagged PS1∆exon8. (B) (4) No transfection, (5) HA-tagged PS1 and FLAG-tagged PS1∆exon8, and (6) FLAG-tagged PS1 and HA-tagged PS1∆exon8. Transfected constructs were co-immunoprecipitated with anti-HA antibody (1.5 μg per 300 μg cell lysate) and blotted with the anti-FLAG antibody (top immunoblot) and the anti-human PS1 antibody NT.1 (directly below FLAG immunoblot). The co-immunoprecipitated band is indicated by arrows and corresponds to full-length PS1 (~50 kDa). Note there is no band present in the eluate lane for Empty Vectors and No Transfection. (C,D) For pooled lysates, constructs were transfected in separate wells and then pooled after lysis. The pooled lysed cells were co-immunoprecipitated with anti-HA antibody and blotted for anti-FLAG as described above. No PS1 band was detected in the eluate (indicated by arrows). (E,F) The experiment in (A,B) was repeated but transfected constructs were co-immunoprecipitated with either IgG rabbit as a negative control or the anti-HA antibody. No PS1 band was detected in the IgG control but was detected in the HA lane, except in the Empty Vectors and No Transfection lanes (indicated by arrows, asterisk marks non-specific bands corresponding to the IgG heavy chain). I = input, E = eluate, IgG = rabbit IgG, and HA = HA antibody.

Mentions: We first established that the HA- and FLAG-tags on PS1 did not alter PS1 subcellular localization. We observed that all the tagged proteins localized to the endoplasmic reticulum (ER) and plasma membrane, where PS1 is normally found in the cell (Fig. 5). To assay for interactions via co-IP, plasmids were co-transfected in six different combinations: (1) HA-tagged PS1 and FLAG-tagged PS1; (2) HA-tagged PS1∆exon8 and FLAG-tagged PS1∆exon8; (3) HA-tagged PS1 and FLAG-tagged PS1∆exon8; (4) FLAG-tagged PS1 and HA-tagged PS1∆exon8; (5) empty HA-tagged vector and empty FLAG-tagged vector (neither containing PS1 nor PS1∆exon8, as a negative control); and (6) GFP only (as a positive transfection control), in order to determine whether an HA-tagged PS1 or PS1∆exon8 protein could co-immunoprecipitate with a FLAG-tagged PS1 or PS1∆exon8 protein. Proteins from total cellular extracts were immunoprecipitated with anti-HA antibody (1.5 μg per 300 μg lysate) and then immunoblotted with anti-FLAG antibody, anti-human PS1 NT.1 antibody, and anti-HA antibody. We show that the HA-tagged PS1 and FLAG-tagged PS1 protein and HA-tagged PS1∆exon8 and FLAG-tagged PS1∆exon8 protein complexes co-immunoprecipitated, providing evidence for physical interaction and the possible formation of homodimers (Fig. 6A). We also found that the HA-tagged PS1 protein co-immunoprecipitated with FLAG-tagged PS1∆exon8 protein and that the FLAG-tagged PS1 protein co-immunoprecipitated with HA-tagged PS1∆exon8 protein, potentially forming heterodimers (Fig. 6B). To confirm that the interaction between the tagged PS1 and PS1∆exon8 proteins was occurring in live cells and was not an artifact of cell lysis, the plasmids were individually transfected in separate tissue culture wells and extracts were pooled after lysis. The pooled, lysed cells were co-immunoprecipitated with 1.5 μg anti-HA antibody per 300 μg lysate and blotted for anti-FLAG as described above. A PS1 band was not detected in the eluate (Fig. 6C,D). A final control experiment was performed to ensure that the interaction between the differently tagged PS1 proteins was specific to the antibody and not to the IgG control. The co-IP experiment was repeated but transfected proteins were co-immunoprecipitated with either rabbit IgG as a negative control or the HA antibody. Again, a PS1 band was not detected in the IgG control (Fig. 6E,F).


Physiologically generated presenilin 1 lacking exon 8 fails to rescue brain PS1-/- phenotype and forms complexes with wildtype PS1 and nicastrin.

Brautigam H, Moreno CL, Steele JW, Bogush A, Dickstein DL, Kwok JB, Schofield PR, Thinakaran G, Mathews PM, Hof PR, Gandy S, Ehrlich ME - Sci Rep (2015)

PS1 and PS1∆exon8 co-immunoprecipitate in vitro.Co-IP immunoblots of different epitope-tagged PS1 and PS1∆exon8 proteins. Constructs were cotransfected in six different combinations: (A) (1) empty vectors containing either HA or FLAG only, (2) HA-tagged PS1 and FLAG-tagged PS1, and (3) HA-tagged PS1∆exon8 and FLAG-tagged PS1∆exon8. (B) (4) No transfection, (5) HA-tagged PS1 and FLAG-tagged PS1∆exon8, and (6) FLAG-tagged PS1 and HA-tagged PS1∆exon8. Transfected constructs were co-immunoprecipitated with anti-HA antibody (1.5 μg per 300 μg cell lysate) and blotted with the anti-FLAG antibody (top immunoblot) and the anti-human PS1 antibody NT.1 (directly below FLAG immunoblot). The co-immunoprecipitated band is indicated by arrows and corresponds to full-length PS1 (~50 kDa). Note there is no band present in the eluate lane for Empty Vectors and No Transfection. (C,D) For pooled lysates, constructs were transfected in separate wells and then pooled after lysis. The pooled lysed cells were co-immunoprecipitated with anti-HA antibody and blotted for anti-FLAG as described above. No PS1 band was detected in the eluate (indicated by arrows). (E,F) The experiment in (A,B) was repeated but transfected constructs were co-immunoprecipitated with either IgG rabbit as a negative control or the anti-HA antibody. No PS1 band was detected in the IgG control but was detected in the HA lane, except in the Empty Vectors and No Transfection lanes (indicated by arrows, asterisk marks non-specific bands corresponding to the IgG heavy chain). I = input, E = eluate, IgG = rabbit IgG, and HA = HA antibody.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4660297&req=5

f6: PS1 and PS1∆exon8 co-immunoprecipitate in vitro.Co-IP immunoblots of different epitope-tagged PS1 and PS1∆exon8 proteins. Constructs were cotransfected in six different combinations: (A) (1) empty vectors containing either HA or FLAG only, (2) HA-tagged PS1 and FLAG-tagged PS1, and (3) HA-tagged PS1∆exon8 and FLAG-tagged PS1∆exon8. (B) (4) No transfection, (5) HA-tagged PS1 and FLAG-tagged PS1∆exon8, and (6) FLAG-tagged PS1 and HA-tagged PS1∆exon8. Transfected constructs were co-immunoprecipitated with anti-HA antibody (1.5 μg per 300 μg cell lysate) and blotted with the anti-FLAG antibody (top immunoblot) and the anti-human PS1 antibody NT.1 (directly below FLAG immunoblot). The co-immunoprecipitated band is indicated by arrows and corresponds to full-length PS1 (~50 kDa). Note there is no band present in the eluate lane for Empty Vectors and No Transfection. (C,D) For pooled lysates, constructs were transfected in separate wells and then pooled after lysis. The pooled lysed cells were co-immunoprecipitated with anti-HA antibody and blotted for anti-FLAG as described above. No PS1 band was detected in the eluate (indicated by arrows). (E,F) The experiment in (A,B) was repeated but transfected constructs were co-immunoprecipitated with either IgG rabbit as a negative control or the anti-HA antibody. No PS1 band was detected in the IgG control but was detected in the HA lane, except in the Empty Vectors and No Transfection lanes (indicated by arrows, asterisk marks non-specific bands corresponding to the IgG heavy chain). I = input, E = eluate, IgG = rabbit IgG, and HA = HA antibody.
Mentions: We first established that the HA- and FLAG-tags on PS1 did not alter PS1 subcellular localization. We observed that all the tagged proteins localized to the endoplasmic reticulum (ER) and plasma membrane, where PS1 is normally found in the cell (Fig. 5). To assay for interactions via co-IP, plasmids were co-transfected in six different combinations: (1) HA-tagged PS1 and FLAG-tagged PS1; (2) HA-tagged PS1∆exon8 and FLAG-tagged PS1∆exon8; (3) HA-tagged PS1 and FLAG-tagged PS1∆exon8; (4) FLAG-tagged PS1 and HA-tagged PS1∆exon8; (5) empty HA-tagged vector and empty FLAG-tagged vector (neither containing PS1 nor PS1∆exon8, as a negative control); and (6) GFP only (as a positive transfection control), in order to determine whether an HA-tagged PS1 or PS1∆exon8 protein could co-immunoprecipitate with a FLAG-tagged PS1 or PS1∆exon8 protein. Proteins from total cellular extracts were immunoprecipitated with anti-HA antibody (1.5 μg per 300 μg lysate) and then immunoblotted with anti-FLAG antibody, anti-human PS1 NT.1 antibody, and anti-HA antibody. We show that the HA-tagged PS1 and FLAG-tagged PS1 protein and HA-tagged PS1∆exon8 and FLAG-tagged PS1∆exon8 protein complexes co-immunoprecipitated, providing evidence for physical interaction and the possible formation of homodimers (Fig. 6A). We also found that the HA-tagged PS1 protein co-immunoprecipitated with FLAG-tagged PS1∆exon8 protein and that the FLAG-tagged PS1 protein co-immunoprecipitated with HA-tagged PS1∆exon8 protein, potentially forming heterodimers (Fig. 6B). To confirm that the interaction between the tagged PS1 and PS1∆exon8 proteins was occurring in live cells and was not an artifact of cell lysis, the plasmids were individually transfected in separate tissue culture wells and extracts were pooled after lysis. The pooled, lysed cells were co-immunoprecipitated with 1.5 μg anti-HA antibody per 300 μg lysate and blotted for anti-FLAG as described above. A PS1 band was not detected in the eluate (Fig. 6C,D). A final control experiment was performed to ensure that the interaction between the differently tagged PS1 proteins was specific to the antibody and not to the IgG control. The co-IP experiment was repeated but transfected proteins were co-immunoprecipitated with either rabbit IgG as a negative control or the HA antibody. Again, a PS1 band was not detected in the IgG control (Fig. 6E,F).

Bottom Line: Also, PS1(∆exon8) did not rescue Aβ generation in PS1/2 double knockout cells indicating its identity as a severe loss-of-function splice form.Further co-IP demonstrates that PS1(∆exon8) interacts with nicastrin, participating in the γ-secretase complex formation.These data support that catalytically inactive PS1(∆exon8) is generated physiologically and participates in protein-protein interactions.

View Article: PubMed Central - PubMed

Affiliation: Fishberg Department of Neuroscience and Friedman Brain Institute, Icahn School of Medicine at Mount Sinai, New York, NY 10029.

ABSTRACT
The presenilin 1 (PSEN1) L271V mutation causes early-onset familial Alzheimer's disease by disrupting the alternative splicing of the PSEN1 gene, producing some transcripts harboring the L271V point mutation and other transcripts lacking exon 8 (PS1(∆exon8)). We previously reported that PS1 L271V increased amyloid beta (Aβ) 42/40 ratios, while PS1(∆exon8) reduced Aβ42/40 ratios, indicating that the former and not the exon 8 deletion transcript is amyloidogenic. Also, PS1(∆exon8) did not rescue Aβ generation in PS1/2 double knockout cells indicating its identity as a severe loss-of-function splice form. PS1(∆exon8) is generated physiologically raising the possibility that we had identified the first physiological inactive PS1 isoform. We studied PS1(∆exon8) in vivo by crossing PS1(∆exon8) transgenics with either PS1- or Dutch APP(E693Q) mice. As a control, we crossed APP(E693Q) with mice expressing a deletion in an adjacent exon (PS1(∆exon9)). PS1(∆exon8) did not rescue embryonic lethality or Notch-deficient phenotypes of PS1- mice displaying severe loss of function in vivo. We also demonstrate that this splice form can interact with wildtype PS1 using cultured cells and co-immunoprecipitation (co-IP)/bimolecular fluorescence complementation. Further co-IP demonstrates that PS1(∆exon8) interacts with nicastrin, participating in the γ-secretase complex formation. These data support that catalytically inactive PS1(∆exon8) is generated physiologically and participates in protein-protein interactions.

No MeSH data available.


Related in: MedlinePlus