Nuclear pore complex remodeling by p75NTR ...

B r i e f c o m m u n i c at i o n s

© 2015 Nature America, Inc. All rights reserved.

Nuclear pore complex remodeling by p75NTR cleavage controls TGF-β signaling and astrocyte functions Christian Schachtrup1,2, Jae Kyu Ryu1,14, Könül Mammadzada2,3,14, Abdullah S Khan1, Peter M Carlton4,13, Alex Perez5,6, Frank Christian7, Natacha Le Moan1,13, Eirini Vagena1,13, Bernat Baeza-Raja1, Victoria Rafalski1, Justin P Chan1, Roland Nitschke8,9, Miles D Houslay10, Mark H Ellisman5,6, Tony Wyss-Coray11, Jorge J Palop1,12 & Katerina Akassoglou1,12 Astrocytes modulate neuronal activity and inhibit regeneration. We show that cleaved p75 neurotrophin receptor (p75NTR) is a component of the nuclear pore complex (NPC) required for glial scar formation and reduced gamma oscillations in mice via regulation of transforming growth factor (TGF)-b signaling. Cleaved p75NTR interacts with nucleoporins to promote Smad2 nucleocytoplasmic shuttling. Thus, NPC remodeling by regulated intramembrane cleavage of p75NTR controls astrocyte–neuronal communication in response to profibrotic factors. TGF-β is a major growth factor promoting the transition of quiescent to reactive astrocytes, and it is implicated in neuronal functions and neurodegeneration1. However, the mechanisms that control astrocyte responses to TGF-β and their contribution to astrocyte-neuronal communication remain elusive. After brain injury, p75NTR is upregulated in astrocytes2,3. To investigate the role of p75NTR in astrocytes, we crossed NGFR−/− (named p75NTR−/−) mice with glial fibrillary acidic protein (GFAP)-TGF-β mice, which spontaneously develop astrocytosis, hydrocephaly and neuronal dysfunction4. Remarkably, loss of p75NTR prevented astrocyte activation and rescued the hydrocephaly in GFAP-TGF-β mice (Fig. 1a,b and Supplementary Fig. 1a,b). p75NTR was expressed in astrocytes and its deletion did not affect TGF-β levels in GFAP-TGF-β mice (Supplementary Fig. 1c,d). Astrocytes contribute to gamma oscillations5, which control learning, memory and attention6. TGF-β-induced astrocyte activation decreased gamma oscillations and altered locomotor activity; these were rescued upon genetic depletion of p75NTR (Fig. 1c,d). In accordance, after brain

trauma, p75NTR−/− mice had reduced astrocyte activation and deposition of neurocan, a component of the glial scar that inhibits neural regeneration7,8 (Supplementary Fig. 2). TGF-β treatment of wild-type (WT) primary astrocytes stimulated protein secretion and gene expression of neurocan, whereas treatment of p75NTR−/− primary astrocytes failed to do so (Supplementary Fig. 3a). Indeed, treatment of cortical neurons with conditioned medium from TGF-β-treated p75NTR−/− astrocytes did not inhibit neurite outgrowth, unlike WT conditioned medium (Supplementary Fig. 3b). Neurocan expression in TGF-β-treated WT cells was unaffected by neutralization of neurotrophins or inhibition of tropomyosin-related receptor kinase (Trk) signaling (Supplementary Fig. 3c). Loss of p75NTR also reduced TGF-β-induced expression of GAT1 (Slc6a1) and S100b (Supplementary Fig. 3d), which regulate astrocyte–neuronal crosstalk9. These data reveal an unanticipated role for p75NTR in regulating astrocyte activation and neuronal activity in response to TGF-β. TGF-β signals by nuclear accumulation of phosphorylated Smad (P-Smad) transcriptional regulators10. p75NTR−/− astrocytes had reduced nuclear, but not cytosolic, TGF-β-dependent accumulation of P-Smad2 compared to WT (Fig. 2a and Supplementary Fig. 4). p75NTR undergoes regulated intramembrane cleavage by γ-secretase resulting in the liberation of the p75 intracellular domain (p75ICD)11. Nuclear P-Smad2 and neurocan expression were reduced in WT astrocytes treated with γ-secretase inhibitors (Fig. 2b and Supplementary Fig. 5a). In p75NTR−/− astrocytes, transient transfection of p75-FasTM, a p75NTR mutant resistant to γ-secretase cleavage12, increased P-Smad2 nuclear accumulation by 2.8-fold upon TGF-β treatment, compared to 4.7- and 7.7-fold increases induced by full-length p75NTR (p75FL) or p75ICD, respectively (Supplementary Fig. 5b). Nuclear accumulation of Stat-1 was not affected (Supplementary Fig. 5c). These results suggest that γ-secretase cleavage of p75NTR and its proteolytic product p75ICD regulate P-Smad2 nuclear accumulation in astrocytes. Smad nuclear translocation depends on direct binding to FG-repeat-containing nucleoporins (FG-Nups) that triggers the opening of the nuclear pore13. Since stabilization of the natively unfolded FG domains by binding of nuclear cofactors is necessary for opening of the pore13, we examined whether p75NTR regulates nuclear import by interacting with FG-Nup complexes. In astrocytes, p75NTR showed perinuclear localization and colocalized with FG-Nups, while in neurons p75NTR localized primarily to the cytoplasm and plasma membrane, as expected (Supplementary Fig. 6a,b). The distinct ring-like perinuclear localization was specific for the p75ICD, and absorption of the p75NTR antibody with recombinant p75ICD abolished the staining (Supplementary Fig. 6c–e).

1Gladstone

Institute of Neurological Disease, University of California, San Francisco, California, USA. 2Institute of Anatomy and Cell Biology, Department of Molecular Embryology, University of Freiburg, Freiburg, Germany. 3Faculty of Biology, University of Freiburg, Freiburg, Germany. 4Department of Biochemistry and Biophysics, University of California, San Francisco, San Francisco, California, USA. 5Department of Neurosciences, University of California, San Diego, La Jolla, California, USA. 6National Center for Microscopy and Imaging Research, University of California, San Diego, La Jolla, California, USA. 7Institute of Cardiovascular and Medical Sciences, College of Medical, Veterinary and Life Sciences, University of Glasgow, Glasgow, UK. 8Life Imaging Center, Center for Biological Systems Analysis, University of Freiburg, Freiburg, Germany. 9BIOSS Centre for Biological Signalling Studies, University of Freiburg, Freiburg, Germany. 10Institute of Pharmaceutical Science, King’s College London, London, UK. 11Department of Neurology and Neurological Sciences, Stanford University School of Medicine, Palo Alto, California, USA. 12Department of Neurology, University of California, San Francisco, San Francisco, California, USA. 13Present addresses: Institute for Integrated Cell-Material Sciences, Kyoto University, Kyoto, Japan (P.M.C.), Omniox, San Francisco, California, USA (N.L.M.), and Diabetes Center, University of California, San Francisco, San Francisco, California, USA (E.V.). 14These authors contributed equally to this work. Correspondence should be addressed to C.S. ([email protected]) or K.A. ([email protected]). Received 25 September 2014; accepted 4 June 2015; published online 29 June 2015; doi:10.1038/nn.4054

nature neuroscience advance online publication

b r i e f c o m m u n i c at i o n s *

c


Gamma power (µV)

40 30 20

GFAP-TGF-β: p75NTR–/–

n.s.

n.s.

*** *

120 100 80

n.s. n.s.

**

n.s. *** n.s.

******

6–25

26–100

60 40 20

0 –5

0

Minutes

41

0

–4 31

–3

–2

0

10

21

*

GFAP-TGF-β

**

50

10

*

WT

140

11

15 12 9 6 3

*

d

60

W p7 T G 5 NT R FA –/ – PG T FA GF -β p7 P5 NT TG R F –/ -β – :

Astrocyte activation (% area)

GFAP / DAPI

n.s.

WT GFAP-TGF-β GFAP-TGF-β: p75NTR–/–

1–

GFAP-TGF-β: p75NTR–/–

*

*

Movements per min

GFAP-TGF-β

n.s. 10 8 6 4 2

W p7 T G 5 NTR FA –/ – PG FA T G p7 P- F-β 5 NT TG R F –/ -β – :

Hydrocephaly

a

Ventricle size (% area)

b

1–5

>100

Movements per minute

Figure 1 p75NTR deficiency rescues TGF-β-induced hydrocephalus, astrocyte activation and neuronal dysfunction. (a) Hematoxylin and eosin stain and GFAP immunostaining (red) of representative brain sections of 4-week-old WT, p75NTR−/−, GFAP-TGF-β and GFAP-TGF-β: p75NTR−/− mice. Nuclei are stained with 4′,6′-diamidino-2-phenylindole (DAPI) (blue). Representative images from n = 4 mice. (b) Quantification of ventricle size (top) and GFAP intensity (bottom) (n = 4 mice per group). Mean ± s.e.m. Ventricle size: *P = 6.083 × 10−6, *P = 7.792 × 10−6, *P = 5.849 × 10−5 (from left to right) and n.s., P = 0.467; GFAP intensity: *P = 5.136 × 10−6, *P = 1.337 × 10−6, *P = 5.138 × 10−5 (from left to right) and n.s., P = 0.275; one-way ANOVA and Fisher’s least significant difference test. Scale bars, 750 µm (top) and 15 µm (bottom). (c) Locomotor activity in the open field during the 50 min of EEG recordings (n = 7 WT mice, 8 GFAP-TGF-β mice and 6 GFAP-TGF-β: p75NTR−/− mice per group). **P = 0.0025 by repeated-measures ANOVA and Bonferroni post hoc multiple comparisons test. (d) Gamma oscillatory power (30–80 Hz) during different locomotor activity intervals in an open field from c. GFAP-TGF-β mice, but not GFAP-TGF-β: p75NTR−/−, had impaired activity-dependent inductions of gamma oscillatory activity (n = 7 WT, 8 GFAP-TGF-β and 6 GFAP-TGF-β: p75NTR−/− mice). n.s., P = 0.4566, 0.3024, 0.1249, 0.0670, 0.9999 and 0.9999, left to right; *P < 0.05, **P < 0.01, ***P < 0.001: 1–5 movements/min, P = 0.0066; 6–25, P = 6.3841 × 10−5; 26–100, P = 4.0256 × 10−12 and P = 1.2077 × 10−12 (left to right); >100, P = 0.0002 and P = 0.0334; one-way ANOVA and Bonferroni post hoc multiple comparisons test.

Super-resolution imaging of p75NTR and FG-Nups in astrocytes by threedimensional structured illumination microscopy (3D-SIM) identified p75NTR at the nuclear outer membrane adjacent to FG-Nups and inside the nucleus (Fig. 2c, Supplementary Fig. 7 and Supplementary Movie 1). In astrocytes, p75NTR interacted with FG-Nups located throughout the NPC14, including the cytoplasmic filament Nup358, the inner center Nup62, and the nuclear basket Nup153 (Fig. 2d and Supplementary Fig. 8a,b). In situ proximity ligation assay, which allows sensitive singlemolecule detection of protein interactions, showed that TGF-β increased the interaction of both Smad2 and p75ICD with Nup153 and Nup358 (Supplementary Fig. 8c). Scanning peptide array analysis and deletion mutagenesis indicated that the Nup153 FG region and the death domain of p75ICD are required for the interaction (Supplementary Fig. 9a–c). Transfection of p75-∆83 in p75NTR−/− astrocytes increased P-Smad2 nuclear accumulation upon TGF-β treatment by only 2.2-fold, compared to 4.7- and 7.7-fold increases induced by p75FL or p75ICD, respectively (Supplementary Fig. 5b), suggesting that the p75 death domain, which is required for interaction with the FG domain, also regulates the nuclear transport of P-Smad2. These results suggest that, in astrocytes, p75NTR binds to the FG-repeat-containing Nups to facilitate Smad2 translocation into the nucleus. We further investigated how TGF-β regulates NPC remodeling in astrocytes. TGF-β induced γ-secretase cleavage of p75NTR (Fig. 2e). In WT astrocytes, TGF-β induced interaction of Nup153 with p75ICD specifically in the nuclear fraction, while p75FL was detected only in the cytosolic fraction (Supplementary Fig. 10). TGF-β induced redistribution of the p75ICD within the NPC, as assessed by increased colocalization of p75ICD with Nup358 that was prevented by inhibition of γ-secretase (Fig. 2f). In accordance, TGF-β treatment of p75FL-transfected WT astrocytes or p75FL-transfected p75NTR−/− fibroblasts induced γ-secretasedependent p75ICD formation (Supplementary Fig. 11a,b). TGF-β induced rapid phosphorylation of presenilin-1 (Supplementary Fig. 11c),

suggesting that TGF-β activates a key component of the γ-secretase complex localized to the nuclear envelope15. To detect spatiotemporal dynamics of p75NTR intramembrane cleavage, we generated a Cherry-p75NTR-EGFP reporter, which can differentiate between uncleaved p75FL (yellow) and cleaved p75NTR (green) (Fig. 3a). Untreated astrocytes, but not fibroblasts, transfected with Cherry-p75NTREGFP showed primarily green signal (Fig. 3b and Supplementary Fig. 12). Inhibition of α-secretase, but not γ-secretase, increased the yellow signal, suggesting that in unstimulated astrocytes the p75NTR reporter was cleaved by α-secretase (Fig. 3b and Supplementary Fig. 12a). Time-lapse imaging in astrocytes transfected with the Cherry-p75NTR-EGFP reporter showed that TGF-β induced the release and translocation of cleaved p75NTR reporter into the nucleus (Fig. 3b,c and Supplementary Movie 2). Similarly, TGF-β increased the release of endogenous p75ICD into the nucleus by ~2.75-fold (P = 0.016) (Fig. 3d). In accordance with the biochemical data for endogenous p75NTR (Fig. 2e), inhibition of γ-secretase abolished the TGF-β-induced accumulation of cleaved p75NTR in the nucleus, suggesting that TGF-β induced the release and nuclear accumulation of the p75ICD (Fig. 3b,c and Supplementary Movie 2). Inhibition of p75NTR cleavage by long (24 h) treatment with α- or γ-secretase inhibitors prevented the nuclear localization of p75ICD (Supplementary Fig. 13). To examine the redistribution of the p75ICD at the NPC upon TGF-β treatment at high resolution, we performed simultaneous imaging by 3D-SIM of endogenous p75NTR at ~800 individual nuclear pores. Upon TGF-β stimulation, the p75ICD traveled from the outer nuclear membrane to the inner center of the NPC (Fig. 3e). Quantification of the p75ICD, nucleoporin and DNA signals (Supplementary Fig. 14) showed that under basal conditions the p75ICD correlated with the DNA signal16 (Fig. 3e,f), suggesting that at baseline the p75ICD is excluded from the inner center of the NPC. In contrast, upon TGF-β stimulation the p75ICD correlated with the FG-Nup signal (Fig. 3e,f), suggesting redistribution of the p75ICD into the inner center of the NPC. As expected, the DNA advance online publication nature neuroscience

a

Cytosolic fraction

Primary astrocytes

Nuclear fraction

p75NTR–/–

WT

Cytosolic fraction

p75NTR–/–

WT

Compound E TGF-β

P-Smad2

60

P-Smad2

kDa 60

Smad2

60

Smad2

60

α-tubulin

50

GAPDH

37

H3

15

H3

15

+

–

+

–

+

1.0 27.4 0.8 34.2 1.0 37.2 1.3 14.2 1.0

d

0.5

1.2

0.7

1.0

0.5

1.0

TGF-β

–

IP: p75NTR –

+

e

IP: p75NTR IB: Nup358

IP: p75NTR IB: p75NTR

+

+

–

–

+

+

–

+

–

+

–

+

Nuclear fraction kDa

TGF-β Compound E

– –

+ –

+ +

f

2.5

75

Untreated

p75ICD

20

H3

15

75

Primary astrocytes

kDa

153 1.0

Astrocyte nuclear center

Primary astrocytes

358

IP: p75NTR IB: Nup153 Fold

+

– +

P-Smad2: fold 1.0 16.3 0.7 18.2 1.0 26.8 1.1 11.6

IP: IgG –

– –

0.4

Primary astrocytes Whole cell lysates Input

Astrocyte outer nuclear membrane

TGF-β

TGF-β + Compound E

p75NTR

–

Nup358

Smad2: fold

+

Merge

P-Smad2: fold

–

c

Nuclear fraction

kDa

TGF-β


b

Primary astrocytes

p75NTR / FG-Nup / DAPI

b r i e f c o m m u n i c at i o n s

Figure 2 p75NTR is a component of the NPC regulating TGF-β-induced P-Smad2 nuclear translocation. (a) P-Smad2 in the cytosolic and nuclear fractions of WT and p75NTR−/− astrocytes. Representative immunoblots from three independent experiments. Underlined numbers indicate the fold changes of nuclear P-Smad2 in TGF-β-treated WT and p75NTR−/− astrocytes. (b) P-Smad2 in cytosol and nuclear fractions of TGF-β-treated astrocytes in the presence of the γ-secretase inhibitor compound E. Representative immunoblots from three independent experiments. (c) 3D-SIM of WT astrocytes stained for p75NTR (green) and FG-Nup (red) shows abundant p75NTR staining at the nuclear surface (top) and sparse p75NTR staining in the nuclear center (bottom). Enlargements of regions indicated by rectangles show p75NTR adjacent to FG-Nup proteins at the nuclear surface (top inset) and the nuclear center (bottom inset). Scale bars: 250 nm, top and bottom panels; 100 nm, insets. Nuclei are stained with DAPI (blue). Representative images from three independent experiments. (d) Coimmunoprecipitation (IP) of individual FG-Nup proteins with endogenous p75NTR in whole cell lysates. Representative immunoblots (IB) from two independent experiments. (e) Nuclear fraction of WT astrocytes treated with TGF-β for 1 h, or after 4 h pretreatment with compound E. Blots were developed with anti-p75NTR and histone H3 antibodies. Representative immunoblots from three independent experiments. (f) Three-dimensional stimulated emission depletion microscopy (3D-STED) of WT astrocytes stained for p75NTR (green) and FG-Nup358 (red). TGF-β-induced relocalization of p75NTR signal within the NPC is blocked by compound E. Scale bar, 100 nm. Nuclei are stained with 1,5-bis{[2-(dimethylamino)ethyl]amino}-4,8-dihydroxyanthracene9,10-dione (DRAQ5) (blue). Representative images from two independent experiments. Full-length blots are shown in Supplementary Figure 15.

signal was unaltered under basal conditions or upon TGF-β stimulation (Fig. 3e,f). Finally, three-dimensional electron tomography17 showed increased nuclear pore size in p75NTR−/− astrocytes (Fig. 3g), consistent with the function of nucleoporins as regulators of the structure and size of the nuclear pore18. Together these results suggest that in astrocytes a part of the intracellular pool of p75NTR might be readily available for TGF-β-induced γ-secretase cleavage to facilitate the transport of P-Smad2 into the nuclear pore. We found that γ-secretase-dependent cleavage of p75NTR was required for TGF-β signaling in astrocytes by regulating dynamic changes in the composition of the NPC to facilitate the nuclear translocation of P-Smad2. It is therefore possible that the p75ICD functions as a nuclear factor that stabilizes FG domains of Nups to allow P-Smad2 nuclear import. Intramembrane proteolysis is mostly associated with the release of DNA-binding intracellular domains into the nucleus, as for example for Notch. We propose a different mechanism whereby intramembrane proteolysis generates nucleoporin-binding intracellular domains to promote nucleocytoplasmic shuttling of transcription factors. In accordance with biophysical models proposed to explain NPC selectivity and transport14, it is possible that TGF-β-induced regulated intramembrane cleavage of p75NTR generates proteolytic fragments that bind FG-Nups along the NPC to generate affinity gradients to modulate P-Smad2 transport. The cleaved p75NTR is detected in the NPC of astrocytes, but not neurons, which nature neuroscience advance online publication

is in line with the differences in NPC composition between cells19. Thus, cell-specific regulation of intramembrane proteolysis might contribute to the cell-specific composition of the NPC to create differences in growth factor signal transduction pathways between neurons and astrocytes. Further, we found that p75NTR-mediated TGF-β signaling altered activity-dependent gamma oscillations, suggesting an unanticipated function for p75NTR in the regulation of neural information processing and cognition. Identification of γ-secretase-mediated cleavage of p75NTR as a molecular link between TGF-β signaling, astrocyte activation and neuronal functions could provide therapeutic targets for resolving the gliotic scar and promoting neuronal activity. Methods Methods and any associated references are available in the online version of the paper. Note: Any Supplementary Information and Source Data files are available in the online version of the paper. Acknowledgments We thank J.W. Sedat for 3D-SIM microscopy, B.D. Sachs, D. Davalos, R.Y.H. Lim and M.J. Moore for discussions, R. Margolis (New York University) for anti-neurocan antibody, M.V. Chao (New York University) for p75NTR antibodies and constructs, W. Fouquet and U. Schwarz at Leica Microsystems and the University of Freiburg Life Imaging Center (LIC) for microscopy support, and B. Cabriga, A. Naumann

b r i e f c o m m u n i c at i o n s a

c

d 1.5

Untreated

DAPI

Untreated

p75ICD/FG-Nup

TGF-β

1.0

5

10

15 min

g

0.5 5

10

15 min

0

x

WT p75NTR–/–

y Pore size (nm)

1

DAPI p75 NPC

0.2

0

TAPI + TGF-β

1.0

0.4

.0 5 –0 0 .0 5 –0 .1 5 –0 .2 5

1

0.2

0.6

.1 5

1.5

DAPI p75 NPC

–0

0.5

0.4

TGF-β

Compound E + TGF-β

1.0

0.8

0.6

.2 5

15 min

1.0

0.8

–0

10

1.0

TGF-β

1.5

5

Untreated

1

f

–0

0.5

.0 5 –0 0 .0 5 –0 .1 5 –0 .2 5

15 min

–0

10

TGF-β

.1 5

5

–0

1

1.0

.2 5

1.5

Untreated

0.5

–0

TAPI-2 + TGF-β

Compound E + TGF-β

TGF-β

Untreated

Cherry-p75NTR-EGFP

p75FL (yellow) p75ICD (green) p75ICD/CTF (green) Compound E

GFP intensity GFP intensity (nucleus) (nucleus)

p75

GFP intensity GFP intensity (nucleus) (nucleus)

EGFP NTR

b

p75NTR

FG-Nup

γ-secretase

Cherry


e

Cherry-p75NTR-EGFP reporter α-secretase

100

*

** **

n.s.

75 Figure 3 TGF-β-induced p75NTR intramembrane cleavage regulates the NPC structure and function. (a) Cherry-p75NTR-EGFP fusion protein. p75NTR N terminal, 50 Cherry (red); p75NTR C terminal, EGFP (green); uncleaved p75FL (yellow); 25 cleaved p75ICD or p75 C-terminal fragment (CTF) (green). EGFP signal reduction TGF-β – + – + after γ-secretase inhibition indicates reduced p75ICD formation. (b) Real-time imaging of TGF-β-treated astrocytes transfected with Cherry-p75NTR-EGFP in the presence of the γ-secretase inhibitor compound E or the α-secretase inhibitor TAPI-2. Representative merged Cherry/EGFP images from three independent experiments are shown. Scale bar, 5 µm. (c) Quantification of nuclear EGFP signal; mean of 20 nuclei obtained from three independent experiments. (d) TGF-β-treated WT astrocytes show increased nuclear p75ICD (right) compared to control (left). Scale bar, 200 nm. Representative images from five independent experiments. (e) Averaged signal for p75NTR (white) reveals relocalization of p75NTR into the NPC in TGF-β-treated astrocytes. p75NTR signal is absent (black) in the NPC in untreated astrocytes. Scale bar, 100 nm. (f) Quantification of p75NTR (green), FG-Nup (red) and DAPI (blue) signals in ~800 NPCs obtained from three independent experiments. (g) Nuclear pore diameter measurements by three-dimensional electron tomography in ~16 nuclear pores per condition. Arrowheads, individual nuclear pores; yellow line, nuclear pore diameter measurement. mean ± s.e.m. *P < 0.01, ** P < 0.0001; n.s., not significant; by one-way ANOVA and Bonferroni post hoc multiple comparisons; WT untreated versus WT + TGF-β, P = 0.0003; WT untreated versus p75NTR−/− untreated, P = 3.79 × 10−7; WT + TGF-β versus p75NTR−/− + TGF-β, P = 8.77 × 10−10; p75NTR−/− untreated versus p75NTR−/− + TGF-β, P = 0.054.

and M. Ast-Dumbach for technical support. Supported in part by US National Center for Research Resources 5P41RR004050-24 and US National Institute of General Medical Science 8P41GM103412-24 to M.E. and the BIOSS – Centre for Biological Signaling Studies EXC 294 for the Life Imaging Center to R.N. Supported by US National Multiple Sclerosis Society postdoctoral fellowships to J.K.R. and N.L.M., an American Heart Association fellowship to V.R., a German Academic Exchange Service fellowship to K.M., US National Institute on Aging AG047313 to J.J.P., the European Commission FP7 PIRG08-GA-2010-276989 and the German Research Foundation SCHA 1442/3-2 to C.S., and US National Institute for Neurologic Diseases and Stroke R01NS051470, R01NS052189, R01NS066361 and R21NS082976 to K.A. AUTHOR CONTRIBUTIONS C.S. performed the majority of the experiments. J.K.R. performed histology and surgeries, K.M. performed STED microscopy and biochemical experiments, A.S.K. performed EEG recordings and behavioral measurements, P.M.C. performed 3D-SIM microscopy, A.P. performed electron tomography, F.C. performed peptide arrays, N.L.M. contributed to live cell imaging and histology, and E.V. maintained mouse colonies. B.B.-R., V.R. and J.P.C. contributed to animal colonies and histology. R.N. contributed to image analysis. M.D.H., M.H.E., T.W.-C. and J.J.P. contributed to the experimental design, data analysis and interpretation. C.S. and K.A. designed the study, analyzed data, coordinated the experimental work and wrote the manuscript with contributions from all authors. COMPETING FINANCIAL INTERESTS The authors declare no competing financial interests.

Reprints and permissions information is available online at http://www.nature.com/ reprints/index.html.

1. Krieglstein, K., Zheng, F., Unsicker, K. & Alzheimer, C. Trends Neurosci. 34, 421–429 (2011). 2. Hanbury, R. et al. J. Comp. Neurol. 444, 291–305 (2002). 3. Oderfeld-Nowak, B. et al. Neuroscience 120, 31–40 (2003). 4. Wyss-Coray, T. et al. Am. J. Pathol. 147, 53–67 (1995). 5. Lee, H.S. et al. Proc. Natl. Acad. Sci. USA 111, E3343–E3352 (2014). 6. Verret, L. et al. Cell 149, 708–721 (2012). 7. Silver, J. & Miller, J.H. Nat. Rev. Neurosci. 5, 146–156 (2004). 8. Schachtrup, C. et al. J. Neurosci. 30, 5843–5854 (2010). 9. Araque, A. et al. Neuron 81, 728–739 (2014). 10. Schmierer, B. & Hill, C.S. Nat. Rev. Mol. Cell Biol. 8, 970–982 (2007). 11. Parkhurst, C.N., Zampieri, N. & Chao, M.V. J. Biol. Chem. 285, 5361–5368 (2010). 12. Le Moan, N., Houslay, D.M., Christian, F., Houslay, M.D. & Akassoglou, K. Mol. Cell 44, 476–490 (2011). 13. Frey, S., Richter, R.P. & Gorlich, D. Science 314, 815–817 (2006). 14. Terry, L.J. & Wente, S.R. Eukaryot. Cell 8, 1814–1827 (2009). 15. Annaert, W.G. et al. J. Cell Biol. 147, 277–294 (1999). 16. Schermelleh, L. et al. Science 320, 1332–1336 (2008). 17. Kim, C.E., Perez, A., Perkins, G., Ellisman, M.H. & Dauer, W.T. Proc. Natl. Acad. Sci. USA 107, 9861–9866 (2010). 18. Allen, T.D., Cronshaw, J.M., Bagley, S., Kiseleva, E. & Goldberg, M.W. J. Cell Sci. 113, 1651–1659 (2000). 19. Tran, E.J. & Wente, S.R. Cell 125, 1041–1053 (2006).

advance online publication nature neuroscience

ONLINE METHODS

Animals. C57BL/6J mice (Jackson Laboratory), C57BL/6J-inbred p75NTR−/− mice20 and the transgenic mouse line GFAP-TGF-β (ref. 4) were used. Transgenic GFAP-TGF-β mice and p75NTR−/− mice were bred to obtain GFAP-TGF-β: p75NTR−/− mice. All mice were in C57Bl/6 background crossed for more than ten generations and littermates were used in all experiments. Mice were housed under a 12-h light/dark cycle. Up to five animals per cage were housed. They were fed standard chow and had access to food and water ad libitum. All animal procedures were performed under the guidelines set by the University of California, San Francisco, Institutional Animal Care and Use Committee and are in accord with those set by the US National Institutes of Health.


Cortical stab wound injury. Cortical stab wound injury was performed as described previously in male mice8. For the expression analysis of p75NTR in astrocytes and for the analysis of the role of p75NTR in astrocyte activation after stab wound injury, the following stereotaxic coordinates were used: sagittal: anteroposterior (AP), −1.0 mm; mediolateral (ML), −1.0 mm; dorsoventral (DV), −1.7 mm, respectively from bregma according to Paxinos and Franklin. Three-dimensional structured illumination microscopy (3D-SIM). Astrocytes were grown to 60–80% confluency on 22 × 22 mm cover glasses (Corning) coated with poly-d-lysine, cultured in DMEM supplemented with 10% heat-inactivated FBS and penicillin/streptomycin. Image stacks of astrocytes were taken using 3D-SIM microscope16. Briefly, super-resolution imaging was performed using a 3D-SIM microscope21 equipped with a 100×, 1.4 NA, oil-immersion objective (Olympus). For 3D-SIM recording, image stacks with a z distance of 0.125 µm were acquired. Exposure times were between 100 and 200 ms, yielding typically 3,000–10,000 counts in a raw image of 16-bit dynamic range. Raw images containing 3D-SIM illumination pattern were processed to reconstruct high-resolution information as described16 with Volocity 5 software (Perkin Elmer). Reconstructed images were registered to compensate for slightly differing camera positions and orientations by calibration with images of 100-nm fluorescent beads. For deconvolution of confocal data sets, a maximumlikelihood estimation algorithm was applied using theoretic point spread functions and a maximum number of ten iterations. To determine numbers and volumes of RF, a threshold-based segmentation was applied using the object separation option of Volocity. Despite the slightly bigger RF sizes within dense clusters possible owing to incomplete separation, the average volume of RF was found to correspond well to the twofold increase in resolution that 3D-SIM provides over confocal laser scanning microscopy. Stimulated emission depletion (STED) microscopy. Astrocytes were grown to 60–80% confluency on 24 × 24 mm cover glasses (Assistent), cultured in DMEM supplemented with 10% heat-inactivated FBS and penicillin/streptomycin. Astrocytes were treated with 2 ng/ml of TGF-β (R&D Systems) or were left without treatment. They were pretreated with 1 µM compound E (Calbiochem), an inhibitor of γ-secretase activity, 3 h before TGF-β treatment or were left without pretreatment. Images were obtained using a commercial gated STED microscope (TCS SP8 STED-WS or TCS SP8 STED 3x) with a HCX PL APO 100×/1.4 oil objective (Leica Microsystems, Germany). Detection of the STAR440 and STAR488 fluorophores (Abberior, Germany) was performed in sequential acquisition mode using HyD detectors in the gating mode and wavelengths ranges of 470–510 nm and 530–560 nm, respectively. The image stacks were Nyquistsampled with a pixel size of 30 nm, a z-step size of 210 nm and a scan speed of 400 Hz using six-line averages. The pinhole was set to one Airy unit. All image stacks were deconvolved using the Huygens STED deconvolution module (Scientific Volume Imaging, Netherlands). In the further analysis to obtain orthogonal sections, IMARIS 7.6 (Bitplane, Switzerland) was used. EEG recordings and behavioral locomotor activity. Mice were implanted for video EEG monitoring after anesthesia with intraperitoneal ketamine (75 mg/kg) and medetomidine (1 mg/kg). Both male and female mice were used and experiments were performed during the light cycle. Teflon-coated silver wire electrodes (0.125 mm diameter) soldered to a multichannel electrical connector were implanted into the subdural space over the left frontal cortex (1 mm lateral and anterior to the bregma) and the left and right parietal cortex (2 mm

doi:10.1038/nn.4054

lateral and posterior to the bregma). The left frontal cortex electrode was used as a reference. All EEG recordings were carried out at least 10 d after surgery on freely moving mice in an open field chamber (40 × 40 × 30 cm). Exploratory locomotor activity was measured with an automated Flex-Field/Open Field Photobeam Activity System (San Diego Instruments). Mice were placed in one of four identical open field chambers for 50 min. The apparatus was cleaned with 70% ethanol between trials. Total movements included fine and ambulatory horizontal movements as well as vertical movements. EEG activity was recorded with Harmonie software (version 5.0b, Stellate, Canada) for 50 min. For spectral analysis of EEG recordings, segments of 50 min of unfiltered EEG recordings (sampling rate 200 s−1) from freely moving animals were imported and analyzed using LabChart 7 Pro software (AD Instruments). The 30–90 Hz band was used for the gamma power calculation with a 58–62 Hz notch filter applied to remove electrical noise. Spectral power was obtained by subjecting the recordings to a fast Fourier transform (FFT) using a Hann cosine-bell window with 50% overlap between windows. FFT was performed with a 512-point FFT size to obtain a resolution of 0.39 Hz. The gamma frequency band (gamma activity) represents the average of the spectral values in the 30–90 Hz range. Construction of Cherry-p75NTR-EGFP. For generation of Cherry-p75NTR-EGFP, a rat p75NTR-EGFP plasmid22 was used. A forward (5′) primer containing a BglII restriction site immediately 5′ to the initiating Cherry sequence and a reverse primer designed to delete the Cherry stop codon with an EcoR1 restriction site immediately 3′ was used to amplify the Cherry sequence and cloned in-frame into the p75NTR-EGFP plasmid, generating Cherry-p75NTR-EGFP. Live cell imaging. Rat astrocytes were transfected by electroporation using Amaxa Astrocyte Nucleofector kit (Lonza) following manufacturer’s instructions with the Cherry-p75NTR-EGFP fusion construct. Cells were plated in 500 µl of DMEM with 10% heat-inactivated FBS and penicillin/streptomycin at a density of 50,000 cells per well in eight-well Nunc plates coated with poly-d-lysine. 24 h after transfection, rat astrocytes were serum-deprived for 3 h before imaging. Astrocytes were treated with 2 ng/ml of TGF-β (R&D Systems), having been pretreated with 1 µM compound E (Calbiochem), an inhibitor of γ-secretase activity, or with 20 µM TAPI-2 (Calbiochem), an inhibitor of α-secretase, after transfection and 3 h before TGF-β treatment. TGF-β was added and the Cherryp75NTR-EGFP intracellular dynamic was imaged in living cells under 5% CO2 and 95% air through a 63× oil-immersion lens, NA 1.4, on a TCS SP5 confocal system (Leica) with a heated stage (37 °C). EGFP and Cherry were excited with the 488-nm line of an argon laser and the 543-nm line of a helium-neon laser, respectively. The FITC-TRITC filter combination was used to detect the EGFP and Cherry signal. Fluorescence emissions were collected between 500 and 560 nm for the EGFP and, to prevent any bleedthrough, we selected a more efficient emission filter for TRITC corresponding to 600–640 nm. Images were acquired in the green and red channels at 11-s intervals over 15 min. The images were processed using LASAF software (Leica). The digitized images were analyzed using ImageJ software (National Institutes of Health). The number of pixels per image with an intensity above a predetermined threshold level within the astrocyte nucleus was quantified by measurement of the EGFP signal of the Cherry-p75NTR-EGFP fusion protein representing the cleaved p75ICD. All quantitative analyses were performed in a blinded manner. Astrocyte conditioned medium and neurite outgrowth assay. Astrocyte conditioned medium (ACM) and neurite outgrowth assays were performed as described8 using WT or p75NTR−/− primary astrocytes treated with 20 ng/ml TGF-β for 2 d. For inhibitor studies, astrocytes were pretreated with 10 µM TGF-β receptor type I inhibitor (SB431542, Calbiochem, catalog no. 616461) 1 h before TGF-β treatment. Cortical neurons were cultured with 80% ACM, allowed to extend processes for 24 h and stained with β-tubulin (Sigma). Neurite outgrowth was determined as the proportion of total cells bearing neurites longer than the diameter of the cell body, an indication of successful initiation of neurite outgrowth23. The number of neurite-bearing cells was counted and percentage of the total number of cells was calculated. The number of neurite-bearing cells was measured from 400 to 500 neurons per condition. Ten representative images per well were taken. All experiments were repeated four times and were performed in triplicate.

nature neuroscience


p75NTR cleavage. For TGF-β induced cleavage of p75NTR, primary mouse astrocytes were pretreated with 1 µM compound E (Calbiochem) 3 h before TGF-β treatment. Primary mouse astrocytes were treated with 20 ng/ml TGF-β (R&D Systems) for 1 h. Rat astrocytes were electroporated using Amaxa Astrocyte Nucleofector kit (Lonza) with the p75NTR full-length construct. Cells were plated in DMEM with 10% heat-inactivated FBS and penicillin/streptomycin at a density of 500,000 cells per 75-cm2 flask coated with poly-d-lysine. Primary p75NTR−/− MEFs were electroporated and cultured in the same conditions, but were plated at a density of 300,000 cells per well of a six-well plate. Fibroblasts were pretreated with 1 µM compound E (Calbiochem) or with 20 µM TAPI-2 (Calbiochem), 14 h and 1 h before TGF-β treatment. 10 µM MG132 (Calbiochem) was added to astrocytes and fibroblasts 1 h before TGF-β treatment. 24 h after electroporation, cells were treated with 20 ng/ml TGF-β (R&D Systems) for 12 h for astrocytes and for 3 h for embryonic fibroblasts, and cell lysates were harvested and processed for western blotting. The primary mouse astrocyte nuclear fraction was prepared using the Active Motif kit (Active Motif 54001). Cell lysates were probed with rabbit anti-p75ICD (1:300, 9992, kind gift of M.V. Chao, New York University)11, rabbit anti-p75NTR (1:1,000, 07-476, Millipore), rabbit anti-H3 (1:1,000, 9715, Cell Signaling) and rabbit anti-GAPDH (1:1,000, 2118, Cell Signaling). Peptide array mapping. Peptide array mapping was performed as described24. Peptide libraries were synthesized by automated SPOT synthesis25. Synthetic overlapping peptides (25 amino acids in length) of Nup153 were spotted on Whatman 50 cellulose membranes according to standard protocols by using Fmoc chemistry with an AutoSpot Robot ASS 222 (Intavis Bioanalytical Instruments AG). Membranes were overlaid with 10 µg/ml recombinant GST-p75NTR ICD. Bound recombinant GST-p75NTR ICD was detected using goat anti-GST (1:2,000, 27-4557-01, GE Healthcare) followed by secondary rabbit anti–horseradish peroxidase antibody (1:2,500, RPN 4031, GE Healthcare). Electron tomography. For electron microscopy, primary astrocytes were grown on poly-d-lysine-coated 75 cm2 tissue culture flasks. Astrocytes were treated with 2 ng/ml of TGF-β1 (R&D Systems) for 1 h. Astrocytes were fixed in 2.5% glutaraldehyde, 2% paraformaldehyde and 0.1 M sodium cacodylate pH 7.4 and pelleted, then postfixed in 1% osmium tetroxide, 0.8% potassium ferrocyanide and 3 mM calcium chloride in the same buffer, en bloc stained in 2% uranyl acetate, dehydrated in ethanol, and infiltrated and embedded in Durcupan. Samples were sectioned on a Reichert Ultracut S ultramicrotome at a thickness of 300 nm and sections picked up on 100-mesh clamshell grids. These sections were poststained for 15 min in a 1% uranyl acetate solution followed by 5 min in a Sato lead solution. To ensure stability in the beam, a thin coat of carbon was applied to each side of the sections. Colloidal gold particles with a diameter of 15 nm were deposited on opposite sides of the section to serve as fiducial cues. For each tilt series, a single series of images was collected with a JEOL (Tokyo, Japan) 4000EX intermediate-voltage electron microscope operated at 400 kV. The specimens were irradiated before collecting data to limit anisotropic specimen thinning during tilt series collection. Using a CCD camera, tilt series were recorded at 20,000× magnification in angular increments of 2° from −60° to +60° about an axis perpendicular to the optical axis of the microscope. Precise angular increments were achieved using a computer-controlled goniometer. The pixel dimensions of the CCD camera were 3448 × 3448 and the pixel resolution was 0.865 nm. The IMOD software package26 was used for rough alignment of the tilt series and the TxBR software package was used for fine alignment and reconstruction27. Nuclear pore diameter measurements were made using NIH ImageJ. Measurements were taken in the x-y plane of multiple slices about the presumed center of the pore and the maximum value was taken to be the true center. Proximity ligation assay (PLA) analysis. PLA analysis was performed following the manufacturer’s instructions (Olink Bioscience). Briefly, astrocytes were plated in 300 µl of DMEM with 10% heat-inactivated FBS and penicillin/streptomycin at a density of 30,000 cells per well in eight-well Nunc plates coated with poly-d-lysine and cultured for 2 d at 37 °C in 5% CO2. Cells were fixed with 4% PFA at 4 °C for 30 min, permeabilized using 0.1% Triton/PBS for 10 min at 4 °C, and blocked by using the blocking solution from the Olink PLA kit for 30 min at 37 °C. Cells were incubated with primary antibody diluted in 1% BSA/PBS overnight at 4 °C. The following primary antibodies were used: rabbit anti-p75ICD (1:300, 9992, kind gift of M.V. Chao, New York University)11, anti–mouse nuclear

nature neuroscience

pore complex (mouse monoclonal (Mab414) to nuclear pore complex proteins) (1:2,000, ab24609, Abcam), anti–mouse Nup358 (1:100, sc-74518, Santa Cruz Biotechnology), anti–rat Nup153 (1:100, sc-101544, Santa Cruz Biotechnology), anti-rabbit P-Smad2 (1:500, 3101, Cell Signaling). Cells were incubated with the appropriate PLA probes (secondary anti-mouse, anti-rat and anti-rabbit antibodies, respectively) for 60 min at 37 °C. For ligation and circularization of the DNA oligonucleotides, cells were incubated with ligase-solution for 30 min at 37 °C. For rolling circle amplification, cells were incubated with amplification solution containing a fluorophore with the excitation wavelength of 594 nm for 100 min at 37 °C. Cells were mounted with a coverslip using a minimal volume of Duolink In situ Mounting Medium with DAPI and analyzed by confocal microscopy. Images were acquired on a Leica TCS SP8 laser confocal microscope with a 63× oil immersion objective, NA 1.4 and LAS AF image analysis software. For quantification, the PLA signal per nucleus was counted for 70 nuclei per each condition. All quantitative analyses were performed in a blinded manner. Immunohistochemistry. P28 mice were transcardially perfused with ice-cold saline under avertin anesthesia and brain samples were removed, embedded in OCT (Tissue-Tek) and frozen on dry ice. Immunohistochemistry on sagittal brain cryostat sections was performed as described8. The primary antibodies used were rat anti–glial fibrillary acidic protein (GFAP) (1:1,000; 13-0300, Invitrogen), rabbit anti-neurocan (1:500)28 and rabbit anti-p75NTR (1:300; AB1554, Millipore) and secondary antibodies were conjugated to Alexa Fluor 488 or 594 (1:200; Jackson ImmunoResearch Laboratories). Quantitative image analysis for the immunostained mouse sagittal sections was performed on three separate tissue sections through the body of the lateral ventricle. For measurement of ventricle sizes in the brain, cresyl violet–stained sagittal sections an equivalent distance from the midline were chosen on the basis of common morphological landmarks. Ventricle size was calculated from the ventricle area divided by the total brain area in a blinded manner using ImageJ. For quantification of GFAP intensity at the lateral ventricle, five nonoverlapping rectangular boxes (100 × 100 µm) were placed along the subventricular zone of the lateral ventricle. The digitized images were analyzed using ImageJ software (US National Institutes of Health). The number of pixels per image with an intensity above a predetermined threshold level was quantified by measurement of the immunoreactive areas for GFAP. The measurement of total immunoreactivity is represented as percent area density defined as the number of pixels (positively stained areas) divided by the total number of pixels (sum of positively and negatively stained area) in the imaged field. All quantitative analyses were performed in a blinded manner. Cell culture. HEK293T (ATCC), NIH3T3 (ATCC) and isolated MEF cells were maintained in DMEM supplemented with 10% heat-inactivated FBS and penicillin/streptomycin. MEFs and primary cortical astrocytes were isolated from p75NTR−/− mice as we described8,29. Astrocytes were used for experiments after they reached confluency. Immunocytochemistry. Primary astrocytes were plated for 2 d at a density of 20,000 cells per well in eight-well Nunc plates coated with poly-d-lysine in 300 µl of DMEM with 10% heat-inactivated FBS and penicillin/streptomycin. Astrocytes were treated with 2 ng/ml of TGF-β1 (R&D Systems) for 1 h. Immunocytochemistry was performed as described8 using goat anti-p75ECD (1:100, sc-6189, Santa Cruz Biotechnologies), rabbit anti-p75ICD (1:300, 9992, kind gift of M.V. Chao, New York University)11, mouse anti–nuclear pore complex protein (1:2,000, ab24609, Abcam), rabbit anti–phospho-presenilin-1 (1:100, ab78914, Abcam), rabbit anti–phospho-Smad2 (1:500, ab3849, Millipore) in PBS with 1% BSA overnight. RNA isolation and quantitative PCR. RNA was isolated from primary astrocyte cultures and quantitative real-time PCR was performed as described8. The following primers were used:

Neurocan (Ncan) Fwd: 5′-TGC AAC CAC GGC TAA GCT C-3′ Neurocan Rev: 5′-GGG GAT AAG CAG GCA ATG AC-3′ Fibronectin (Fn1) Fwd: 5′-GCA GTG ACC ACC ATT CCT G-3′ Fibronectin Rev: 5′-GGT AGC CAG TGA GCT GAA CAC-3′ GAT1 (Slc6a1) Fwd: 5′-GAAAGCTGTCTGATTCTGAGGTG-3′ GAT1 Rev: 5′-AGCAAACGATGATGGAGTCCC-3′

doi:10.1038/nn.4054


S100b Fwd: 5′-TGGTTGCCCTCATTGATGTCT-3′ S100b Rev 5′-CCCATCCCCATCTTCGTCC-3′ Gapdh Fwd: 5′-CAA GGC CGA GAA TGG GAA G-3′ Gapdh Rev: 5′-GGC CTC ACC CCA TTT GAT GT-3′

Immunoblots. Immunoblots to detect neurocan in the supernatant of WT or p75NTR−/− astrocytes were performed as described8. Astrocytes were treated with 20 ng/ml TGF-β (R&D Systems) for 2 d and pretreated with 10 µM SB431542 (Sigma), an inhibitor of TGF-β receptor type I, 1 h before TGF-β treatment as indicated. For rescue experiments, p75NTR−/− astrocytes were infected with lentivirus expressing p75NTR constructs, medium was changed 2 d after infection and cells were treated on the third day after infection. Astrocyte cytoplasmic and nuclear fractions were prepared using the Active Motif kit (Active Motif 54001). For detection of P-Smad2, WT and p75NTR−/− astrocytes were treated with 2 ng/ml TGF-β (R&D Systems) for 1 h or, in inhibitor studies, WT astrocytes were pretreated with 1 µM compound E (Calbiochem), an inhibitor of γ-secretase activity, 1 h before TGF-β treatment. For detection of P-presenilin-1, WT primary astrocytes were treated with 2 ng/ml TGF-β. The following primary antibodies were used: rabbit anti-histone H3 (1:1,000, 9715, Cell Signaling), rabbit anti-P-Smad2 (1:1,000, 3101, Cell Signaling), mouse anti-Smad2 (1:1,000, 3103, Cell Signaling), rabbit anti-GAPDH (1:1,000, 2118, Cell Signaling), rabbit anti-neurocan (1:500)28, rabbit anti–P-presenilin-1 (1:1,000, ab78914, Abcam), rabbit anti-presenilin-1 (1:1,000, ab71181, Abcam), mouse anti-α-tubulin (1:1,000, T6199, Sigma). Coimmunoprecipitation. Coimmunoprecipitation was performed as described24. For mapping experiments, immunoprecipitation was performed with anti-HA antibody (Cell Signaling) or anti-Nup153 antibody (Santa Cruz Biotechnologies). For endogenous coimmunoprecipitation, cell lysates were incubated with rabbit anti-mouse p75NTR antibody11 (1:100; 9992, kind gift of M.V. Chao, New York University) and rat anti-mouse Nup153 antibody (5 µg, Santa Cruz Biotechnology) bound to A-agarose beads for 4 h at 4 °C. Cell lysates were

doi:10.1038/nn.4054

probed with the following antibodies: mouse anti–nuclear pore complex protein antibody (1:1,000, ab24609, Abcam), rat anti-Nup153 antibody (1:500, sc-101544, Santa Cruz Biotechnology), mouse anti-Nup358 antibody (1:500, sc-74518, Santa Cruz Biotechnology), rabbit anti-HA (1:1,000, 3724, Cell Signaling), rabbit antip75NTR (1:1,000, AB1554, Millipore). Statistics. Statistical significance was calculated using GraphPad Prism (GraphPad Software) by unpaired or paired two-sided Student’s t-test to analyze significance between two experimental groups or by one-way ANOVA, Fisher’s least significant difference test or Bonferroni’s post-test, or two-way ANOVA for multiple comparisons. Data are presented as mean ± s.e.m. Data distribution was assumed to be normal, but this was not formally tested. No statistical methods were used to predetermine sample sizes, but our sample sizes are similar to those reported in previous publications6,8,23. No randomization was used to assign experimental groups or to collect data, but mice and cells were assigned to specific experimental groups without bias. All animals were present at the end of study. No data points were excluded. All histopathological analyses were performed in a blinded manner. A Supplementary Methods Checklist is available.

20. Lee, K.F. et al. Cell 69, 737–749 (1992). 21. Gustafsson, M.G. et al. Biophys. J. 94, 4957–4970 (2008). 22. Kreitzer, G., Marmorstein, A., Okamoto, P., Vallee, R. & Rodriguez-Boulan, E. Nat. Cell Biol. 2, 125–127 (2000). 23. Schachtrup, C. et al. Proc. Natl. Acad. Sci. USA 104, 11814–11819 (2007). 24. Sachs, B.D. et al. J. Cell Biol. 177, 1119–1132 (2007). 25. Frank, R. J. Immunol. Methods 267, 13–26 (2002). 26. Kremer, J.R., Mastronarde, D.N. & McIntosh, J.R. J. Struct. Biol. 116, 71–76 (1996). 27. Lawrence, A., Bouwer, J.C., Perkins, G. & Ellisman, M.H. J. Struct. Biol. 154, 144–167 (2006). 28. Milev, P., Maurel, P., Haring, M., Margolis, R.K. & Margolis, R.U. J. Biol. Chem. 271, 15716–15723 (1996). 29. Baeza-Raja, B. et al. Proc. Natl. Acad. Sci. USA 109, 5838–5843 (2012).

nature neuroscience

a

b

Fibronectin relative expression

p75NTR-/-

WT

GFAP / DAPI

3

GFAP-TGF-β

GFAP-TGFβ:p75NTR-/-

*

*

2

1

WT

d

GFAP-TGF-β

2.0 1.0

n.d.

GFAP

3.0

Merge

TGF-β relative expression

p75NTR

c

*

Supplementary Figure 1 NTR

Loss of p75

rescues TGF-β-induced astrocyte activation.

(a) GFAP immunoreactivity in the brains of WT, p75NTR-/-, GFAP-TGF-β and GFAP-TGF-β:p75NTR–/– mice. Scale bar, 90 μm. Representative images from n = 4 mice are shown. (b) Decreased fibronectin gene expression in whole brain extracts of 4-week-old GFAP-TGF-β:p75NTR–/– compared to GFAP-TGF-β mice (n = 3 mice per genotype) by RT-PCR analysis performed in duplicate. Values are mean ± SEM [*P ˂ 0.001 by one-way ANOVA and Bonferroni post-hoc multiple comparisons; WT vs GFAP-TGF-β, P = 0.0001; p75NTR-/- vs GFAP-TGF-β, P = 3.888 x 10-5; GFAP-TGF-β vs GFAP-TGF-β:p75NTR–/–, P = 0.0007]. (c) TGF-β gene expression in brain cortex of WT, GFAP-TGF-β and GFAP-TGF-β:p75NTR–/– mice. Results are from three independent experiments. Bar graphs represent means ± SEM [not significant (P = 0.284) by unpaired Student’s t test]. n.d., not detectable. (d) p75NTR expression in GFAP-positive astrocytes in GFAP-TGF-β transgenic mice. Immunolabeling for p75NTR (green) and GFAP (red) revealed p75NTR colocalizing with astrocytes in brain sections of 4 week old GFAP-TGF-β transgenic mice. Representative images from n = 4 mice per genotype are NTR positive astrocytes in brain sections of WT mice were detected. Scale bar: 100 µm. shown. No p75

Nature Neuroscience: doi:10.1038/nn.4054

b

c

Stab Wound Injury Day 3

Stab Wound Injury Day 3

GFAP

Uninjured

WT

d

p75NTR-/GFAP-IR

Stab Wound Injury

p75NTR / GFAP

a

15 12 9 6 3

GFAP *

Uninjured WT p75NTR -/-

Neurocan-IR

Neurocan

SWI

4 3 2 1

Neurocan *

Uninjured WT p75NTR -/-

SWI

Supplementary Figure 2 Loss of p75NTR reduces astrocyte activation and neurocan expression after cortical stab wound injury. NTR (green) and GFAP (red) of representative (a) Cortical stab wound injury (SWI), a model for brain trauma. (b) Immunostaining for p75 NTR deficiency decreases astrocyte activation and neurocan brain sections of two WT mice 3 days after SWI. Scale bar, 750 μm. (c) p75 expression after cortical stab wound injury. GFAP (red) and neurocan (green) immunostaining of representative brain sections of eight adult (10 - 15 weeks old) WT and p75NTR-/- mice 3 days after stab wound injury. (d) Quantification of GFAP and neurocan immunoreactivity (IR). Values are mean ± SEM. n = 8 per group, GFAP-IR (top), *P = 0.026; Neurocan-IR (bottom), *P = 0.028, by unpaired Student’s t test. Scale bar, 60 μm.


WT p75NTR-/-

TGF-

-

+ - + -

+

neurocan

kDa 250

GAPDH

37

Neurocan relative expression

a

Fold 1.0 1.3 0.7 0.7 0.5 0.3

b

WT CM

3 2 1

hr

TGF- -

3 6 12 -

3

6 12

ns Neurite Outgrowth (%)

β-tubulin / DAPI

p75NTR -/- CM

WT astrocyte CM

**

*

100

p75NTR -/astrocyte CM

75 50 25

TGF-

c

* *

-

+

-

+

ns

*

3

ns

ns

ns

2

Fc-TrkB

Fc-p75NTR

Fc 1 µg/ml

Fc 20 µg/ml

goat anti-NGF

K252a

goat IgG

TGF-β

DMSO

1

untreated

Relative expression

Neurocan expression

+ TGF-β 3

WT p75NTR-/-

GAT1

*

2 1

TGF- -

3

6 12

-

hr 3

6 12

relative expression

relative expression

d 3

WT p75NTR-/-

S100b *

2 1

TGF- -

3

6 12

-

hr 3

6 12


Loss of p75

rescues TGF-β-induced astrocyte activation in vitro.

NTR–/– astrocytes. Representative (a) Immunoblot for neurocan secretion and gene expression analysis in TGF-β-treated WT and p75 immunoblot is shown from three independent experiments. Gene expression results are from two independent experiments. Bar graphs represent means ± SEM. WT 6 h TGF-β treatment vs p75NTR–/– 6 h TGF-β treatment, *P = 0.042; WT 12 h TGF-β treatment vs p75NTR–/– 12 h TGF-β treatment, *P = 0.003 by two-way ANOVA. (b) Cultures of cortical neurons stained with β-tubulin (red) treated with CM from NTR–/– astrocytes. Nuclei are stained with DAPI (blue). Scale bar, 20 μm. A minimum of 400-500 neurons per TGF-β-treated WT or p75 condition were analyzed from at least three different experiments. Values are mean ± SEM. *P 0.001 and ns, not significant, by oneway ANOVA and Bonferroni post-hoc multiple comparisons; WT untreated vs WT + TGF-β, P = 4.698 x 10-7; WT + TGF-β vs p75NTR-/untreated, P = 1.346 x 10-7; WT + TGF-β vs p75NTR-/- + TGF-β, P = 4.698 x 10-7; WT untreated vs p75NTR-/- untreated, P = 0.568; WT NTR-/+ TGF-β, P = 1.000. (c) Trk inhibition or neurotrophin blocking have no effect on TGF-β induced neurocan untreated vs p75 expression in primary astrocytes. Primary astrocytes were pre-treated 1 h prior to TGF-β treatment with the following: Trk inhibitor K252a (10 nM) or DMSO vehicle control; NGF-blocking antibody goat anti-NGF (2 µg/mL) or goat IgG control; neurotrophin scavenger NTR or Fc fragment control (20 µg/mL); or BDNF scavenger Fc-TrkB or Fc fragment control (1 µg/mL). No effect of inhibitors on Fc-p75 the expression of neurocan mRNA in TGF-β treated primary astrocytes when compared to vehicle treated cells as determined by quantitative PCR and normalized to GAPDH. Results are from three independent experiments. Bar graphs represent means ± SEM. *P = 0.0004 and ns, not significant, by unpaired Student’s t test. (d) GAT1 and S100b gene expression analysis in TGF-β-treated WT and p75NTR–/– astrocytes. Results are from three independent experiments. Bar graphs represent means ± SEM. GAT1 (left), *P = 0.01; S100b (right), *P = 0.004 by two-way ANOVA and Bonferroni post-hoc multiple comparisons. Full-length blots are shown in Supplemental Figure 15.


WT astrocytes

Untreated

+TGFβ

+TGFβ

P-Smad2 / DAPI

Untreated

p75-/- astrocytes

WT

P-Smad2-IR (nucleus)

p75NTR-/-

**

ns

***

150 100 50

TGF-

-

+

-

+


p75

regulates TGF-β-induced P-Smad2 nuclear translocation in vitro.

P-Smad2 immunostaining (green) in WT and p75NTR–/– primary astrocytes. Nuclei are stained with DAPI (blue). Scale bar: 35 µm. Representative images are shown from three independent experiments. Quantification of nuclear P-Smad2 signal. Values are derived from the mean of 60 nuclei obtained from three independent experiments quantified in a blinded manner. Values are mean ± SEM. WT -5 NTR–/– + TGF-β, **P = 0.0002; p75NTR–/– untreated vs p75NTR–/– + untreated vs WT + TGF-β, ***P = 7.887 x 10 ; WT + TGF-β vs p75 TGF-β, P = 0.718 (ns, not significant), by two-way ANOVA and Bonferroni post-hoc multiple comparisons.


Neurocan expression

a

P 0.05

Relative expression

2.0

P 0.03

1.5 1.0 0.5 0

untreated DMSO compound E + TGF-β

nuclear fraction

b

kDa

P-Smad2

60

Smad2

60

H3

15

Fold

c

1.0

4.7

2.8

2.2

7.7

cytosol fraction nuclear fraction WT IFN - γ

-

p75NTR -/-

+

-

+

p75NTR -/-

WT

-

+

-

+

kDa

P-Stat1

84

Stat1

84

GAPDH

37

H3

15


p75

-mediated TGF-β signaling depends on γ-secretase release of p75ICD and does not regulate Stat nuclear localization.

(a) Inhibition of γ-secretase by compound E decreases neurocan gene expression in TGF-β-treated astrocytes. Results are from three independent experiments. Bar graphs represent means ± SEM. Untreated cells vs TGF-β treated cells, P = 0.011; TGF-β treated cells vs TGF-β + compound E treated cells, P = 0.046 by unpaired Student's t test. (b) Immunoblot for P-Smad2 of the nuclear fraction of p75NTR-/- astrocytes transfected with GFP, p75FL, p75ICD, p75∆83, and p75-FasTM treated with TGF-β for 1 h. Representative immunoblot is shown from two independent experiments. (c) Immunoblot for P-Stat1 of the cytosolic and nuclear fractions of WT and p75NTR–/– astrocytes. No difference in P-Stat1 accumulation in the nuclear fraction between WT and p75NTR–/– astrocytes. Full-length blots are shown in Supplemental Figure 15.


a

b

p75NTR / DAPI

p75NTR / FG-Nup / DAPI

Astrocyte

Neuron

d

c

p75ICD / p75ECD / FG-Nup

p75ICD / p75ECD / DAPI

p75ICD

p75ECD

e p75NTR

p75NTR / DAPI

p75ICD / p75ECD / FG-Nup

1st Ab absorption

1st Ab omission


p75

is a component of the NPC in astrocytes.

(a) The p75NTR co-localizes with phenylalanin-glycin- (FG-) repeat containing nucleoporins (FG-Nups) in astrocytes. Co-localization of p75NTR (green) and FG-Nups (red) in primary astrocytes. Nuclei are stained with DAPI (blue). Scale bar: 5 µm. Representative images are shown from two independent experiments. (b) p75NTR immunostaining (green) in primary astrocytes and cortical neurons. Nuclei are stained with DAPI (blue). Scale bar: 5 µm. Representative images are shown from two independent experiments. (c) Immunolabeling for p75ECD (red) and p75ICD (green, 9992 antibody) revealed nuclear staining for p75ICD in primary astrocytes. Nuclei are stained with DAPI (blue). Scale bar: 13.5 µm. Representative images are shown from two independent experiments. (d) Confocal imaging of p75ICD (green), p75ECD (red) and FG-Nup proteins (blue) in primary astrocytes revealed a perinuclear localisation of p75ICD in astrocytes. Scale bar: 2.5 µm. Representative images are shown from two independent experiments. (e) Specificity of the antibody rabbit anti-p75NTR clone 9992 recognizing the p75NTR intracellular domain (ICD). Immunocytochemistry of primary mouse astrocytes for p75NTR (green, left panel), p75NTR staining after primary antibody absorption (middle panel), and p75NTR staining after 1st antibody omission (right panel) revealed antibody specificity for p75NTR staining. Nuclei are stained with DAPI (blue). Scale bar, 10 µm. Representative images are shown from two independent experiments.


Nuclear center

p75ICD / FG-Nup / DAPI

Outer nuclear membrane

Supplementary Figure 7 p75NTR is localized together with FG-Nups to the astrocytic nuclear membrane. Superresolution microscopy (3D-SIM) of WT astrocytes stained for p75NTR (green) and FG-Nup proteins (red) show abundant p75NTR staining at the outer nuclear membrane (left) and sparse p75NTR staining in the nuclear center (right). Scale bar: 1 µm. Nuclei are stained with DAPI (blue). Representative images are shown from three independent experiments.


a

i n n e r c e n t e r

FG-Nups

Nup358 cytoplasm

Nup62

outer membrane nuclear envelope inner membrane

Nup153

b

nucleus

Primary Astrocytes - Whole Cell Lysates Lysate IP:p75NTR IgG TGF-

-

+

-

+

-

+

kDa

Nup153

150

Nup62

60

p75NTR

75

c

+ TGF-β

Nup358 / p75NTR

****

*

1000

Dots per Nucleus

Nup153 / p75NTR

Untreated

FG-Nup / P-Smad2

800 600 400 200 TGF-β

-

+

-

+

-

+

Supplementary Figure 8 The p75ICD interacts with FG-Nups in astrocytes. (a) Schematic diagram of the localization of three FG-Nups at the NPC. The three FG-Nups included in this study are found in discrete substructural locations of the NPC: Nup358 is localized at the cytoplasmic filaments, Nup62 is distributed centrally at the NPC, and NTR with Nup153 and Nup62 in whole cell lysates. Primary Nup153 is localized at the nuclear basket. (b) Endogenous co-IP of p75 NTR antibody and western blots were developed with anti-nucleoporin protein astrocyte lysate was immunoprecipitated with anti-p75 recognizing all FG-containing Nups and anti-p75NTR protein antibody, revealing a complex of Nup153 and Nup62 with p75NTR. (c) NTR NTR interaction Proximity Ligation Assay (PLA) to detect single molecule interactions between FG-Nups and p75 . TGF-β induced p75 with individual FG-Nups in primary astrocytes. TGF-β-induced interaction of FG-Nup with P-Smad2 is shown as positive control. PLA (red); DAPI (blue). Representative images are shown from six independent experiments. Scale bar, 5 µm. Values are mean ± SEM. -6 Nup358 / p75ICD, *P = 0.0497; Nup153 / p75ICD, ****P = 6.406 x 10 , and ns, not significant, by unpaired Student's t test.


a F

Control

01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30

p75-ICD Control

G

p75-ICD

H

p75-ICD

I

p75-ICD

J

p75-ICD

Control

Control

Control

01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30

* *

* *

** * * Nup153

aa 1

400 650 N-terminal region

b

Nup153 FL Δ595

+

IP:Nup153 IB:HA IP:Nup153 IB:Nup153

IB:HA

+ -

+ + -

1475

FG region

c FG region

N-terminal region Zn finger HA-FL p75NTR Nup153 Δ595 Nup153 GFP

880

Zn finger region

+ + -

kDa 75 150 130 75

FL Δ83 Δ151

HA

HA-FL p75NTR HA-Δ83 p75NTR HA-Δ151 p75NTR Nucleoporin153 GFP IP:HA IB:Nup153

p75NTR TM

Extracellular

+

+ + -

DD

Intracellular

+ + -

IP:HA IB:HA

IB:Nup153

+ + -

kDa 150 75 50

150

Supplementary Figure 9 The p75ICD interacts with the natively unfolded phenylalanine-glycine-rich FG domain of FG-Nups. (a) Nup153 peptide library screened with recombinant GST-p75ICD revealed distinct interaction domains of Nup153 (asterisks) that NTR compared to GST control. The peptide array shown (spots F1 – J20) contains aa 751 – 1475 of Nup interact with the ICD of p75 153. Spots G3 (aa 911 – 935), G8 (aa 936 – 960), G16 (aa 976 – 1000), H9 (aa 1091 – 1115), H13 (aa 1111 – 1135), I21 (aa 1301 – 1325) all contain FG repeats. (b) Schematic diagram of the Nup153 FG domain deletion (Δ595). Nup153∆595 mutant, which lacks the NTR C-terminal FG region (aa 881-1475) abolished the interatcion with p75 . Lysates were immunoprecipitated with anti-Nup153 and probed with anti-Nup153 and anti-HA. Representative immunoblot is shown from two independent experiments. (c) Schematic diagram of HA-tagged p75NTR intracellular domain deletions. Representative immunoblot is shown from two independent experiments. TM, transmembrane domain; DD, death domain. Lysates were immunoprecipitated with anti-HA and probed with anti-Nup153 and anti-HA. Full-length blots are shown in Supplemental Figure 15.


kDa

WT

p75NTR-/-

b p75FL

75

Cytosolic fraction

kDa

Nuclear fraction

WT p75NTR-/- WT p75NTR-/-

75

WT whole cell lysate

Nuclear fraction

WT whole cell lysate

a

p75FL

50 50

37

37 25 25 20

20

c

Cytosolic fraction Nuclear fraction

TGF-β - + - + kDa 75 50 37 25 20 15 10

- + - +

p75ICD

Supplementary Figure 10 p75FL is not detected in the nuclear fraction, and TGF-β induces the interaction of Nup153 with p75ICD. NTR in the nuclear fraction of WT and p75NTR-/- primary astrocytes or whole cell lysate of WT astrocytes using an (a) Western blot for p75 antibody against the p75ICD (9992). p75FL band is detected only in the WT astrocyte lysate and is indicated with arrow. Bands detected in both WT and p75NTR-/- extracts are non-specific bands. (b) Western Blot for p75NTR in the cytosolic and nuclear fractions of WT and p75NTR-/- primary astrocytes or whole cell lysate of WT astrocytes. p75FL band is detected only in the WT astrocyte lysate and NTR-/extracts are non-specific bands. (c) in the cytosolic fraction and is indicated with arrow. Bands detected in both WT and p75 NTR Endogenous co-IP of p75 with Nup153. Cytosolic (2.5 mg) and nuclear (1.5 mg) fractions were immunoprecipitated with anti-Nup153 or IgG control antibody and western blots were developed with an antibody against the p75ICD (9992). Full-length blots are shown.


a

b

TGF- compound E TAPI-2

-

+ - + - -

+ + -

+

+ +

TGF-

p75FL

75

p75CTF p75ICD

25 20

GAPDH

37

control

P-presenilin / nucleoporin

c

-

+

kDa

kDa p75FL

75

p75CTF p75ICD

25 20

GAPDH

37

TGF-β

TGF-β

-

10

30

60

min

P-presenilin

53 kDa

presenilin

53 kDa

Supplementary Figure 11 TGF-β induces -secretase-dependent cleavage of the p75ICD and phosphorylation of presenilin NTR-/embryonic mouse fibroblasts were transfected with p75FL construct and treated with TGF-β for 3 h. Representative (a) p75 immunoblot is shown from three independent experiments. (b) Astrocytes were transfected with p75FL construct and treated with TGFNTR and anti-GAPDH antibodies. p75FL: p75NTR full length (p75FL), p75NTR Cβ for 12 h. Western blots were developed with anti-p75 NTR intracellular domain (p75ICD). Representative immunoblot is shown from three independent terminal fragment (p75CTF), p75 experiments. (c) Immunocytochemistry for P-presenilin (green) and FG-Nups (red) and immunoblot for P-presenilin in TGF-β-treated astrocytes. Representative images and immunoblot are shown from two independent experiments. Scale bar: 8 µm. Full-length blots are shown in Supplemental Figure 15.


a

b

Astrocytes Cherry

p75NTR

p75FL (yellow) p75ICD/CTF (green)

EGFP

Merge

Fibroblasts Cherry

Cherry

p75NTR

EGFP

EGFP

p75FL (yellow) p75ICD/CTF (green)

Merge

TAPI-2

compound E untreated

Cherry

EGFP

Supplementary Figure 12 TGF-β induces -secretase-dependent cleavage of p75NTR. NTR (a) Rat primary astrocytes transfected with cherry-p75 -EGFP construct revealed high abundance of the C-terminus (green) and low NTR abundance of the N-terminus (red) of the p75 full length fusion protein in untreated cells. Treatment of astrocytes with TAPI-2, which inhibits α-secretase cleavage of p75NTR, increased the abundance of p75NTR full length fusion protein (yellow). Scale bar, 3 µm. Representative images are shown from three independent experiments. (b) Cherry-p75NTR-EGFP expression in NIH3T3 cells. NIH3T3 cells transfected with cherry-p75NTR-EGFP construct revealed expression of the fusion protein (yellow) associated with similar abundance of the C-terminus (green) and N-terminus (red) signals, indicative of full length p75NTR. Nuclei are stained with DAPI (blue). Scale bar, 10 µm. Representative images are shown from two independent experiments.


compound E

TAPI-2

p75NTR / DAPI

control


Regulated intramembrane cleavage determines the nuclear localization of p75

in astrocytes.

Nuclear localization of p75ICD (green) in astrocytes is reduced after 24 h incubation with the α-secretase inhibitor TAPI-2 or the γsecretase inhibitor compound E. Nuclei are stained with DAPI (blue). Representative images are shown from two independent experiments. Scale bar, 8 µm.


DAPI

p75ICD

TGF-

untreated

FG-Nup

1 5 50 250 Averaged nuclear pore number Supplementary Figure 14 Quantification method for the p75NTR signal at individual nuclear pores. 3D-SIM images of primary mouse astrocytes stained for p75NTR and FG-Nup proteins revealed redistribution of p75ICD from the outer nuclear membrane to the inner center of the NPC only upon TGF- treatment. Analysis was done on 10 μm2 areas of nuclei containing NPC areas, which are defined by positive signals for FG-Nup staining exactly in the center. Representative images of the average of 1, 5, 50 or 250 individual nuclear pores are displayed as small boxes, final analysis as shown in figure 3d was done using the average signal of 800 individual nuclear pores. Positive signals are displayed by white color, absent staining by black. DNA signal (DAPI), which is known not to colocalise with the NPC is absent from the inner center of the NPC (black center). p75NTR signal is absent from the inner center of the NPC of untreated primary astrocytes (black center), however redistributes into the NPC upon TGF-β treatment (white signal in the center).


WT TGF-

-

p75NTR –/– +

-

WT

-

+

+

IB Images corresponding to Fig. 2b

-

+

kDa compound E -

-

+

+

-

-

+ +

TGF- -

+

-

+

-

+

-

150

+

75

P-Smad2

IB Images corresponding to Fig. 2d Input IP:p75NTR + TGF-β -

cytosolic fraction nuclear fraction

p75NTR –/–

50

kDa 150

-

+

kDa

150

50 75 75 15

75 50

75

Smad2

50

150

50

15 150

75

25

75

50

150

50 GAPDH 15 150

150

H3

15

-

+

-

+

25

IB Images corresponding to Suppl. Fig. 9b HA-FL p75NTR - + + + - + Nup153 + - Δ595 Nup153 + - kDa GFP

IB Images corresponding to Suppl. Fig. 5c

kDa IFN - γ -

+

-

+

nuclear fraction WT p75NTR -/-

+

-

+

150

kDa IP:Nup153 IB:HA

250 P-Stat1

neurocan

IB Images corresponding to Suppl. Fig. 9c HA-FL p75NTR HA-Δ83 p75NTR HA-Δ151 p75NTR Nucleoporin153 GFP

75

+

+ + -

+ + -

+ + kDa

50 75 50

25 IP:HA IB:Nup153

25 15

IP:Nup153 IB:Nup153

50 37

GAPDH

25 20 15

H3

15

cytosol fraction WT p75NTR -/+

75 50

50

50

-

50

75

75

TGF-

75

150

IP: p75NTR IB: p75NTR

15

IB Images corresponding to Suppl. Fig. 3a

25 20 15

p75ICD

75

15 150

50

50

150

IP: p75NTR IB: Nup 153

150

H3

kDa

358

15

α-tubulin

Nuclear fraction + + + -

TGF-β compound E

75

P-Smad2

Smad2

IB Images corresponding to Fig. 2e

IP:IgG

IP: p75NTR IB: Nup 358

IB Images corresponding to Fig. 2a cytosolic fraction nuclear fraction

150

150 75 50

25 Stat1

15

50 75

25

50

150

IB Images corresponding to Suppl. Fig. 5b

25 15

nuclear fraction

150 IB:HA

75 50

IP:HA IB:HA

kDa 25

75 50

75 75

P-Smad2

50 25

GAPDH

15

IB Images corresponding to Suppl. Fig. 11c TGF-β - 10 30 60 min kDa

25 15

75 P-presenilin

Smad2

75 50

25

75 50 25

15

H3

15

25 15

25


150 75

25 75 50

presenilin H3

IB:Nup153

50

25 75

25

Supplementary Figure 15 Full blots of the western blots shown in the figures This Figure contains the full blots of the western blots shown in the Figures and Supplementary Figures of the paper
