1
5
Ti le: MEG3 ac i a es nec op osis in human neu on xenog a s modeling
Alzheime ’s disease
Au ho s: S i am Balusu1,2*, Ka ien Ho é1,2, Nicola Th upp1,2, An Snellinx1,2, Lu ga de
Se neels1,2, Io dana Ch ysidou1,2, Amaia M. A anz3,4, Anne ieke Sie ksma1,2, Joel Sim én5,6,
Thomas K. Ka ika i5,6,7, Hen ik Ze e be g5,6,8,9,10, Wei-Ting Chen1,2, Die ma Rudol Thal11,12,
E genia Sal a13, Ma k Fie s1,2,9, Ba De S oope 1,2,9,14*
A ilia ions:
10 1VIB Cen e o B ain & Disease Resea ch; 3000 Leu en, Belgium.
2KU Leu en, Leu en B ain Ins i u e ; 3000 Leu en, Belgium.
3Achuca o Basque Cen e o Neu oscience; 48940 Leioa, Spain.
4Ike basque Basque Founda ion o Science; 48009 Bilbao, Spain.
5Clinical Neu ochemis y Labo a o y, Sahlg enska Uni e si y Hospi al; 431 80 Möndal, 15
Sweden.
6Depa men o Psychia y and Neu ochemis y, Sahlg enska Academy a he Uni e si y o
Go henbu g; 431 80 Möndal, Sweden.
7Depa men o Psychia y, Uni e si y o Pi sbu gh; Pi sbu gh, PA, USA.
8Depa men o Neu odegene a i e Disease, UCL Ins i u e o Neu ology; London, WC1N 20
3BG, UK.
9UK Demen ia Resea ch Ins i u e a UCL; London, WC1E 6BT, UK.
10Hong Kong Cen e o Neu odegene a i e Diseases; Hong Kong, China.
11Labo a o y o Neu opa hology, Depa men o Imaging and Pa hology, Leu en B ain
Ins i u e (LBI), KU Leu en (Uni e si y o Leu en); 3000 Leu en, Belgium. 25
12Depa men o Pa hology, Uni e si y Hospi al Leu en; 3000 Leu en, Belgium.
13Labo a o y o Neu ogenesis and Neu odegene a ion, Ne he lands Ins i u e o Neu oscience;
1105BA Ams e dam, he Ne he lands.
14Lead con ac
30 * Co esponding au ho s: [email p o ec ed] o [email p o ec ed]
35
2
Abs ac : Neu onal cell loss is a de ining ea u e o Alzheime 's disease (AD), bu i emains
unclea how neu ons die and how his ela es o o he de ining cha ac e is ics o he disease. He e
we demons a e ha human neu ons xenog a ed in mouse b ain exposed o amyloid pa hology
de elop sa kosyl-insoluble au ilamen s, posi i e Gallyas sil e s aining, elease phospho yla ed
5
au (p- au181 and p- au231) in o he blood, and display conside able neu onal cell loss, p o iding
a model o he induc ion o ull Tau pa hology by simple exposu e o amyloid pa hology in AD.
The al e a ions a e speci ic o human neu ons and con as wi h he mild e ec s exhibi ed in he
ci cum en ing mouse neu ons o in ansplan ed mouse neu ons. A co e ansc ip ional p og am
in he human neu ons is cha ac e ized by s ong up egula ion o MEG3, a neu on-speci ic long
10
noncoding RNA. MEG3 is also up egula ed in neu ons om AD pa ien s in si u. MEG3 exp ession
alone is su icien o induce nec op osis in human neu ons in i o. Inhibi ing nec op osis using
o ally adminis e ed small molecule ecep o -in e ac ing p o ein (RIP) kinase -1 and -3 inhibi o s
o RIPK3 knockou escues neu onal cell loss in his no el AD model. Thus, xenog a ed human
neu ons in con as o mouse neu ons a e uniquely sensi i e o amyloid pa hology, ecapi ula e
15
neu opa hological ea u es o AD, and ul ima ely die by nec op osis.
One-Sen ence Summa y: MEG3 ac i a es neu onal nec op osis and inhibi ion o he nec op osis
pa hway escues neu onal loss in Alzheime ’s disease.
20
3
Main Tex : The majo ques ion o how he de ining hallma ks o Alzheime ’s disease (AD)
amyloid-β (Aβ) plaques, neu onal au angles, g anulo acuola neu odegene a ion and neu onal
cell loss ela e o each o he has ne e been sol ed, la gely owing o he lack o good models ha
encompass all neu opa hological aspec s o he disease. Aβ pa hology s ongly in luences au
pa hology kine ics in se e al mouse models, bu only i au pa hology is a i icially induced by
5
ex a on o empo al demen ia (FTD) causing mu a ions o by injec ing au-seeds isola ed om
AD pa ien b ains(1–3). The undamen al unanswe ed ques ions emain whe he amyloid-β can
induce Tau pa hology and how neu ons die in AD.
The many a emp s o gene a e a comple e model o AD ha e ailed likely because o (unknown)
10
human-speci ic ea u es ha escape modeling in oden s. Two models ha e made g ea s ides
o wa d in using human cells o model ea u es o AD, one econs i u ing a 3D cell cul u e based
on di e en ypes o human cells(4) and one using xenog a ed human neu ons in mouse b ain
(5). We gene a ed a much-imp o ed mouse model o xeno ansplan a ion o human neu ons using
a Rag2-/- (Rag2 m1.1Cgn) gene ic backg ound and a single AppNL-G-F (App m3.1Tcs/App m3.1Tcs) knock-
15
in gene o d i e Aβ pa hology, ins ead o he combined APP-PS1 (Tg(Thy1-APP*Swe, Thy1-
PSEN1*L166P)21Jck ) ansgene o e exp ession used be o e(6). Human s em cell-de i ed
neu onal p ogeni o cells (NPCs) ansplan ed in he con ol Rag2-/- mice in eg a e well and
de elop o e ime dend i ic spines ( ig. S1A and B). Two mon hs pos - ansplan a ion (2M PT),
he xenog a ed neu ons display cha ac e is ic ma u e neu onal (NEUN and MAP2) and co ical
20
ma ke s (CTIP2, SATB2, TBR1, and CUX2). Fu he mo e, unlike oden neu ons, human neu ons
display equal 3R and 4R au splice o ms a 6-mon hs pos - ansplan a ion (6M PT) ( ig. S1C and
ig. S5H). Compa ed o he p e iously used NOD-SCID animals(6), Rag2-/- animals show a
conside ably inc eased li e span (>18 mon hs), allowing he s udy o heal hy human neu ons
du ing b ain aging.
25
We xenog a ed GFP-labelled H9-de i ed human co ical neu onal p ecu so cells (100,000
NPCs/mouse) in o Rag2-/-/AppNL-G-F ( u he e e ed o as amyloid mice) o Rag2-/-/Appmm/mm
( u he e e ed o as con ol mice). β-shee s aining wi h dye X34 e ealed obus plaque
pa hology in he amyloid mice (Fig. 1A & ig. S1D-E). RNA sequencing con i med ha he
30
ansc ip ional p o ile o he human ansplan s a 2M in con ol and amyloid mice a e e y simila
(Fig. 2A). Thus, human neu ons in eg a e and di e en ia e simila ly in con ol and amyloid mice,
in line wi h he ac ha amyloid plaque pa hology appea s only om 3M onwa ds in his model.
Numbe o cells a e also no signi ican ly di e en in amyloid and non-amyloid xenog a ed mice
a 2M ( ig. S6F).
35
Full blown amyloid plaque pa hology is seen a 18M pos ansplan a ion. A his la e poin in ime,
human neu ons in he con ol b ain appea o e all heal hy, display neu onal p ojec ions and
dend i ic spines in e mingled wi h hos mic oglial (IBA1) and as oglia (GFAP) cells ( ig. S1D).
In con as , g a ed neu ons in he amyloid mice display se e e dys ophic neu i es nea Aβ
40
plaques, associa ed wi h mic ogliosis (6-7 IBA1-posi i e mic oglia pe Aβ plaque) and as ogliosis
4
(2-3 GFAP-posi i e as ocy es pe plaque) (Fig. 1A, and ig. S2A-B). The glial cell ec ui men o
plaques is e y simila o non-g a ed con ol animals ( ig. S3A and B). Immunohis ochemis y
wi h phospho yla ed au (P- au) an ibodies, AT8 (P- au Se 202, Th 205), PHF1 (P- au Se 396 and
Se 404), and MC1 (pa hological con o ma ional au epi ope) demons a e conside able deposi ion
o neu i ic plaque au (NP- au)(1) in he human neu ons o he g a ed amyloid mice (Fig. 1C and
5
D) (Su ace s ained a ound Aβ plaques (20 µm ing) was ˜2% (AT8), ˜5% (PHF1), and ˜3% (MC1),
Fig. 1E-G). In e es ingly, bo h NP- au- and AT8-posi i e neu ons appea ed as ea ly as 6-mon hs
pos - ansplan a ion ( ig. S1E), indica ing ha amyloid deposi ion d i es au phospho yla ion ea ly
on in his model. Signi ican au pa hology is no obse ed in mouse neu ons in he same animal,
no in human g a ed con ol, no in non-g a ed o mouse neu ons g a ed amyloid mice (Fig. 1C-
10
D, ig. S1E, ig. S3C, H and ig. S4A-C).
Pa hological, β-shee X-34-posi i e au- eac i i y is seen in he xenog a ed human neu ons ( ig.
S2C). Fu he mo e, Gallyas sil e s aining and au-immunogold labeling o PHF ib il-like
s uc u es ex ac ed wi h sa kosyl om g a ed neu ons in he amyloid mice (Fig. 1H and I)
15
con i med he p og ession o P- au in o pa hological s a es (> 20 au- ib ils/EM g id (con ol n=4,
amyloid n=4) ( ig. S6F). Such pa hology is la gely absen in g a ed con ol o non-g a ed amyloid
mice (Fig. 1H, ig. S6F & ig. S3I-J). Finally, and clinically ele an , we ound ha p- au181 and
p- au231(7, 8) is signi ican ly inc eased in he plasma o he g a ed amyloid mice bu no in
con ol g a ed o non-g a ed mice (Fig. 1J-K), e lec ing an inc eased sec e ion o soluble P- au
20
species om neu ons in o he bloods eam in esponse o amyloid pa hology, mimicking wha is
obse ed in humans wi h AD.
I is possible o quan i y accu a ely up o 0.1 ng o human DNA in a mix u e wi h 100 ng mouse
genomic DNA using qPCR(9) ( ig. S6D). Using his assay, we es ima e ha up o 50% ewe
25
human neu ons we e p esen in he amyloid animals compa ed o con ol animals (Fig. 1L and ig.
S6G) a 6M and 18M pos ansplan a ion. The e is no signi ican inc ease in o al numbe o
human cells in con ol mice be ween 6M and 18M pos ansplan a ion ( ig. S6E). The numbe o
mouse neu ons was also no signi ican ly a ec ed in he non-g a ed con ol and amyloid mice
( ig. S3E-G). Thus, human neu ons, bu no mouse neu ons, de elop obus neu opa hological A
30
AD-like au pa hology changes including p- au and au ib ils upon exposu e o amyloid plaques
in i o and degene a e much like hei coun e pa s in he human b ain by an unknown dea h
p ocess.
P e ious wo k has shown (opposing) e ec s o Rag2-/- immunode iciency on amyloid plaque
35
pa hology and cellula esponses(10, 11). Compa ison o soluble Aβ (Aβ40 & Aβ42) and
guanidine-ex ac able Aβ (Aβ40 & Aβ42) e ealed no signi ican di e ences be ween Rag2-/-
/AppNL-G-F and AppNL-G-F mice (no shown). Bulk RNA sequencing indica ed ha main aspec s o
he inna e mic oglia and as oglial cellula ansc ip omic esponses o amyloid plaques a e
main ained in he Rag2-/-/AppNL-G-F immune-de icien mouse compa ed o a p e iously
40
cha ac e ized non-immune supp essed amyloid mouse (APP/PS1)(12, 13) (Fig. 1B) (R2=0.81 o
5
di e en ially exp essed genes). Thus, while we acknowledge ha ou model does no ha e an
adap i e immune sys em, key ea u es o he neu opa hology o AD such as amyloid accumula ion,
glial esponses, au agg ega ion, au phospho yla ion, and neu onal cell loss, a e ep oduced in
ansplan ed human neu ons by simple exposu e o amyloid pa hology.
5
T ansc ip ional changes in xenog a ed neu ons e eal induc ion o nec op osis
We isola ed he ansplan ed neu on-posi i e b ain egions guided by he g een luo escen p o ein
(GFP) ma ke om mice a 2-, 6-, and 18-mon hs pos - ansplan a ion and ex ac ed o al RNA o
sequencing (mean human eads ˜13.3 million) ( ig. S5A-B). Mapping o mouse and human
e e ence da abases gene a ed species-speci ic da ase s. As expec ed, up egula ion o he mouse
10
genes Axl, B2m, Cs 7, C ss, I gax, T em2, Ty obp, Lyz2, Ccl3, and Ccl4 con i ms mic oglial
ac i a ion when exposed o amyloid in he Rag2-/-/AppNL-G-F mice (Fig. 1B, ig. S7A-C and
Supplemen a y Table 2)(13). Human samples clus e ed acco ding o geno ype and age ( ig. S5C).
Di e en ial exp ession (DE) analysis o he human g a s a wo mon hs showed only a ew
signi ican ly down egula ed DE genes, o unknown ele ance (Fig. 2A, Supplemen a y Table 1).
15
In con as , DE o g a s in amyloid and con ol condi ions a 6M and 18M e ealed 916 up and
73 down- egula ed genes, and 533 up and 34 down- egula ed genes, espec i ely (Fig. 2B-C, and
Supplemen a y Table 1). Gene al e a ions o in e es include CD74, a ma ke o au angle-
con aining cells(14); HBB, a co ical py amidal neu onal ma ke (15–17) and A2M, which is
associa ed wi h neu i ic plaques in AD human b ains(18) ( ig. S5E). In addi ion o he neu onal
20
signa u e, in line wi h ou p e ious indings(19), we also obse e as oglial and oligodend ocy ic
genes, including MOBP, MBP, OPALIN, S100A6, APOE, among o he s, sugges ing ha he
g a ed NPC cells no only gene a e human neu ons bu also glia ( ig. S5F). The o e all exp ession
p o iles o he g a s exposed o amyloid plaques a 6M and 18M show s ong co ela ions (R2 =
0.88 o genes di e en ially exp essed a one o bo h ime poin s), sugges ing ha he pa hological
25
cell s a es o neu ons a 6M and a 18M a e simila ( ig. S5G).
We in es iga ed whe he he gene signa u es o he human neu ons exposed o amyloid in
xenog a ed mice a e o e lapping wi h ansc ip ional changes obse ed in human AD b ains
(Supplemen a y Table 3). We ound a ema kable en ichmen using gene se en ichmen analysis
30
(GSEA) be ween p e iously published AD da ase s, including he ROSMAP coho (20) and ou
da a in ansplan ed neu ons a 6M and 18M PT (P.adj <0.05) bu no a 2M ( ig. S8A-B).
Fu he mo e, GSEA also e ealed a s iking en ichmen in up egula ed genes om ansplan ed
neu ons and neu ons di ec ly ep og ammed om ib oblas s o AD pa ien s (M6: P.adj=1.64e-05,
M18: P.adj=2.19e-04). These da a con i m ha ou ansplan ed neu ons cap u e AD- ele an
35
ansc ip ional signa u es(21).
Func ional gene on ology (GO) en ichmen analysis o signi ican ly up egula ed genes (P.adj
<0.05) a 6M using DAVID and REVIGO co e ed seman ic space a ound posi i e egula ion o
ansc ip ion, p o ein phospho yla ion, posi i e egula ion o MAPK cascade, in lamma o y
40
esponses, including TNF and in e e on signaling, cell p oli e a ion, aging, issue egene a ion,
6
and myelina ion. (Fig. 2D and Supplemen a y Table 1). Some epo s ha e sugges ed ha neu ons
in AD display signa u es o hypo-ma u i y, dedi e en ia ion, and cell cycle e-en y(21–24). Using
p e iously published da ase s (Supplemen a y Table 1) and Gene se en ichmen analysis (GSEA),
we ind ha cellula signa u es o hypo-ma u i y, dedi e en ia ion, P53, HIF signaling, NF-kB,
Myc, TGFβ-signaling, egula ion o cell cycle (posi i e and nega i e), and cell-cycle e-en y a e
5
indeed en iched in he 6M and 18M g a s bu no a 2M (Fig. 2E). Con e sely, ma u e neu onal
ma ke pa hways, such as synap ic plas ici y, synap ic ansmission, long- e m po en ia ion, and
axon guidance we e no signi ican ly en iched (Fig. 2E).
How neu ons die in AD emains a highly con o e sial discussion(25, 26). As shown in Fig. 1L
and ig. S6G, abou ˜50% o xenog a ed human neu ons a e los in he amyloid mice. We did no
10
obse e any signi ican al e a ions in he exp ession o genes associa ed wi h cell dea h
mechanisms such as apop osis o e op osis, bu ound signi ican up egula ion o MLKL, he gene
encoding he execu o p o ein o nec op osis (Fig. 2B-C, ig. S2E & ig. S6A-C). This is no
obse ed in he ansc ip ome o he hos issue ( ig. S7A-C). We used pRIPK1 (Se 166), pRIPK3
(Se 227), and pMLKL (Se 358) speci ic an ibodies o s ain b ain issue om 18M g a ed and
15
non-g a ed mice. This esul ed in in ense punc ua e s aining wi h a esicula pa e n in he soma
o he neu ons in g a ed amyloid mice. These esicula s uc u es co-s ained wi h casein kinase 1
del a (CK1), a ma ke o g anulo acuola degene a ion (Fig. 2L and ig. S6I). Non-g a ed
con ol and amyloid animals o mouse neu ons de i ed om NPC and simila ly ansplan ed in o
mice did no display hese pa hologies ( ig. S3D and ig. S4D).
20
We alida ed his obse a ion by qPCR on RNA ex ac ed om he empo al gy us o AD and
age-ma ched con ol b ain samples, which e ealed signi ican up egula ion o MLKL and RIPK3,
an ups eam kinase in he nec op osis pa hway in he human b ain (Fig. 2F-H). Based on ini ial
obse a ions made in he xenog a model desc ibed he e, we ha e al eady ex ensi ely con i med
he p esence o nec op osis ma ke s in g anulo acuola degene a ion in neu ons o AD pa ien s in
25
a p e ious publica ion(27). Thus, ou model has p edic i e alue o he neu opa hology o AD.
MEG3 modula es he neu onal nec op osis pa hway
We iden i ied 36 up and 21 down- egula ed long noncoding RNAs (lncRNA) in he g a s o 6-
mon h-old animals (Supplemen a y Table 1). Some o hese non-coding RNAs (e.g., NEAT1) ha e
30
been implica ed in AD(28), bu o e all i is unclea whe he and how hey con ibu e o
pa hogenesis. In ou da a, he long non-coding RNA MEG3 (Ma e nally Exp essed 3) was he mos
s ongly (˜10 old) up egula ed gene in human neu ons exposed o amyloid pa hology (Fig. 2B-C
& ig. S2D), bu no in he hos mouse neu ons ( ig. S7D-F). MEG3 has been linked o cell dea h
pa hways(29) ia p53(30), is in ol ed in he TGFβ pa hway(31), and has been associa ed wi h
35
Hun ing on’s disease(32). In a ce eb al ischemia- epe usion inju y model, ep ession o MEG3
has bene icial e ec s(33).
7
While MEG3 has no been p e iously implica ed in AD, we ound up egula ion o MEG3 in AD
pa ien b ains in a single nucleus ansc ip omic da abase(34). We con i med a 2-3- old
up egula ion o MEG3 in RNA ex ac ed om he empo al gy us o human AD b ains
(Supplemen a y Table 5) using qPCR (Fig. 2I). MEG3 in si u hyb idiza ion combined wi h cell-
speci ic immunohis ochemical ma ke s NeuN (neu ons), IBA1 (mic oglia), and GFAP
5
(as ocy es) demons a ed i s exclusi e exp ession in he nucleus o neu ons ( ig. S9A-B) and i s
s ong en ichmen in AD b ains (2-3 punc a/nucleus in con ol and 8-10 punc a/nucleus in AD
b ain (P= <0.0001), (Fig. 2J-K). In AD, MEG3-exp essing neu onal nuclei appea ed blo ched wi h
educed DAPI in ensi y, and MEG3-posi i e neu ons also displayed high le els o he nec op osis
ma ke pMLKL ( ig. S9C).
10
MEG3 has 15 known ansc ip ional a ian s in humans(35). We pe o med Sange sequencing on
b ain cDNA and ound ha MEG3 ansc ip a ian -1 (accession numbe NR_002766.2) is mos
abundan ly exp essed in he human adul b ain. A len i i us was used o exp ess MEG3 V1 in H9-
de i ed ma u e co ical neu ons (Fig. 3A-B). Exp ession o MEG3 in he neu ons esul ed in a
s ong educ ion o cell iabili y a 9 days a e ansduc ion compa ed wi h con ol- ansduced
15
human neu ons (Fig. 3C). The MEG3 induced cell loss can be escued by he nec op osis inhibi o s
pona inib, dab a enib o nec osul onamide (NSA) o by CRISPR media ed dele ion o RIPK1,
RIPK3, and MLKL in i o (Fig. 3I-J and ig. S10H-J). Immunohis ochemis y analysis e ealed
he p esence o ac i a ed nec op o ic ma ke s, pRIPK1 (S166), pRIPK3 (S227), and pMLKL
(S358) (Fig. 3D-F). Nec op osis-posi i e neu ons displayed a educed amoun o cy oskele al
20
ilamen neu o ilamen -H (NF-H). Wes e n blo analysis con i med inc eased pRIPK1 (S166)
le els (Fig. 3G), showing ha he neu ons we e dying om nec op osis.
Nex , o unde s and o wha ex en he MEG3 exp ession con ibu es o ansc ip ional changes
iden i ied in he xenog a ed neu ons, we pe o med RNA sequencing o MEG3 ansduced
neu ons 7-days pos - ansduc ion. To compa e ansc ip omic signa u es o he MEG3 ansduced
25
neu ons wi h hose iden i ied in xenog a ed neu ons, we pe o med gene se en ichmen analysis
(gene se en ichmen analysis). We anked he genes acco ding o hei log old change di e en ial
exp ession be ween amyloid-exposed and non-exposed xenog a ed neu ons (a M6, see Fig. 2C),
and compa ed hose o he op genes up egula ed in esponse o MEG3 ansduc ion in neu ons in
i o. We ound a signi ican en ichmen (NES=1.5, P.adj=1.2e-05) (Fig. 3H and Supplemen a y
30
Table 4). This sugges s ha some ansc ip ional changes in he xenog a ed neu ons migh be
seconda y o he up egula ion o MEG3. DE analysis o MEG3 ansduced neu ons e ealed no
changes in nec op osis genes ( ig. S10A-C) bu DAVID Gene On ology (GO) analysis o leading
edge genes om GSEA on up egula ed genes om 6M ansplan ed human neu ons and in i o
MEG3 exp essing neu ons e ealed signa u es o NF-kB and TNF signaling which ha e been
35
ela ed o nec op osis induc ion(36). In addi ion, signa u es o lipid anspo , in e e on-
signaling, posi i e egula ion o pep idyl- y osine phospho yla ion, and apolipop o ein-L signaling
a e obse ed ( ig. S10D and Supplemen a y Table 4).
8
Inhibi ion o nec op osis by small molecules o abla ion o RIPK3 blocks human neu onal
loss in he xenog a ed mice
We i s used a p e iously cha ac e ized shRNA an isense cons uc (PMID: 31656536 and PMID:
32201430) agains MEG3 o con i m he di ec ole o MEG3 exp ession in he ansplan ed
neu ons when exposed o amyloid plaques. Down egula ion o MEG3 signi ican ly imp o ed he
5
neu onal su i al and was associa ed wi h down egula ed exp ession o nec osome p o eins (Fig.
3K-L and ig. S10E-G). I emains howe e unknown whe he he neu ons ha degene a e du ing
AD o in ou neu onal ansplan s a e all dying by nec op osis o whe he nec op osis ma ke s a e
limi ed o he neu ons ha a e emaining a he s age o pa hological analysis. We hus ea ed
xenog a ed mice (3 g oups o n=5) om 2M un il 6M pos - ansplan a ion wi h he o ally
10
a ailable nec op osis kinase inhibi o s, pona inib (30 mg/kg), and dab a enib (50 mg/kg).
Pona inib inhibi s RIPK1(37) and RIPK3(38) and is an FDA-app o ed d ug o he ea men o
acu e lymphoid leukemia (ALL) and ch onic myeloid leukemia (CML). Dab a enib is a mo e
speci ic inhibi o o RIPK3 (39). Immunos aining a six mon hs e ealed a s iking educ ion in
he le els o pRIPK1, pRIPK3, and pMLKL in bo h ea men g oups compa ed o un ea ed mice
15
(Fig. 4A-D), wi hou al e ing he glial esponse o amyloid ( ig. S9D). qPCR analysis o neu onal
cell numbe s e ealed a signi ican inc ease in neu ons in he ea ed g oups compa ed o con ol
g oups (Fig. 4E). In u he expe imen s, we ha e ansplan ed RIPK1 o RIPK3 knockou human
s em cells gene a ed using CRISPR/Cas9 simila o he ones used in he in i o expe imen s abo e.
In con as o he in i o expe imen s, none o he neu ons ansduced wi h guide RNA agains
20
RIPK1 su i ed, indica ing ha knock ou o RIPK1 is le hal in i o. P e ious wo k has indeed
shown ha cons i u i e dele ion o RIPK1 in mice leads o emb yonic le hali y (PMID: 27819682).
In con as , dele ion o he RIPK3 gene wi h gRNA signi ican ly imp o ed neu onal su i al (Fig.
4 F-H). Immuno luo escence s aining con i med he signi ican dec ease o nec op osome signals
in RIPK3 de icien human neu ons.
25
9
Conclusion
Ou esul s demons a e ha amyloid pa hology is su icien o induce ull hallma k
neu opa hological ea u es o AD in non-gene ically manipula ed human neu ons, including
dys ophic neu i es, au ib ils, g anulo acuola degene a ion, cell loss, and inc eased plasma p-
au181 and -231 in he g a ed mice. One o he majo unsol ed ques ions in AD esea ch is how
5
neu ons die. Some epo s ha e p oposed apop o ic mechanisms, bu i has ne e been
demons a ed con incingly ha his d i es neu onal cell loss in he disease(40). While p e ious
publica ions sugges ed ac i a ed nec op osis in he AD b ain(27, 41), we demons a e he e ha a
nec op osis media ed neu onal loss in he xenog a model can be escued by ea men wi h
clinically ele an nec op osis inhibi o s o by abla ing he RIPK3 gene in he ansplan ed
10
neu ons(39). Thus, we sugges ha neu onal dea h in AD is la gely d i en by nec op osis, linking
he loss o neu ons o in lamma o y p ocesses ha a e ups eam o his well-s udied dea h
pa hway(13, 42, 43). The e o e, he apies ha p e en neu onal cell loss, in combina ion wi h mo e
mains eam Aβ and au a ge ed in e en ions, migh be use ul addi ions o he cu en e o s o
de elop disease-modi ying s a egies o AD(44–46). Ou da a sugges ha nec op osis is
15
downs eam o he accumula ion o pa hological au and is induced by he up egula ion o he non-
coding RNA MEG3 possibly ia TNF-in lamma o y pa hway signaling. Long non-coding RNAs
(lncRNAs) a e impo an egula o s o gene exp ession and in luence a a ie y o biological
p ocesses, including b ain aging and neu odegene a i e disease(47). The la ge numbe o non-
coding RNAs ha a e di e en ially exp essed in ou no el AD model wa an u he
20
in es iga ion.
I is deeply in iguing ha human ansplan ed neu ons display an AD pheno ype, while mouse
neu ons in e spe sed wi hin he g a o NPCs de i ed om mouse neu ons ansplan ed in a
simila way as hei human coun e pa s, do no display such signs. This s ongly sugges s ha ye
unknown human-speci ic ea u es de ine he sensi i i y o he neu ons o amyloid pa hology. We
25
no ice ha ansc ip ional analysis o he ansplan ed human neu ons indica es signa u es o
down egula ion o ma u e neu onal p ope ies and up egula ion o imma u e signaling pa hways.
This ag ees wi h obse a ions in a ecen s udy ha analyzed he ansc ip ional p o iles in neu ons
di ec ly de i ed om ib oblas s o AD pa ien s (iNs)(21). The neu ons in ha s udy did no show
angles, amyloid, o nec op osis ma ke s ypical o AD, and i was no clea how hose obse a ions
30
ela e o he classical hallma ks o AD. While he con ibu ion o hese imma u e signa u es o he
disease p ocess in ou model needs u he in es iga ion, ou da a indica e ha nec op osis is an
impo an con ibu o o cell dea h in Alzheime ’s Disease. Nec op osis is an ac i e a ea o d ug
de elopmen in cance ( e ) and ALS ( e ), and we he e o e sugges ha i is wo hwhile o explo e
his pa hway u he o po en ial d ug de elopmen in AD as well.
35
Finally, i is o c ucial impo ance o be e unde s and wha p ecisely makes ha ansplan ed
human neu ons become diseased when exposed o amyloid pa hology, while mouse neu ons
emain p ese ed. Unde s anding he esilience o mouse neu ons o amyloid pa hology will no
Figu e 2:
A B C
F G H
Con ol AD
DAPI/MEG3
J KL
Con ol AD
0
10
20
30
40
MEG3 punc a/nucleus
✱✱✱✱
P= <0.0001
Con olAmyloid
GFP/pRIPK1 GFP/pRIPK3 GFP/pMLKL
Amyloid s Con ol 2M (human) Amyloid s Con ol 6M (human) Amyloid s Con ol 18M (human)
D E
I
Ma u e neu onal
gene se s
Neu onal dedi e en ia ion gene se s
M2 M6 M18
Aging
Tissue egene a ion
Myelina ion
Posi i e egula ion o gene exp ession
Posi i e egula ion o ansc ip ion
Pep idyl- y osine phospho yla ion
Posi i e egula ion o MAPK cascade
Nega i e egula ion o cell popula ion
Cell p oli e a ion and g ow h
Cellula esponse o in e e on-β
Cellula esponse o umo nec osis ac o
Cellula esponse o g ow h ac o s
Immune esponse
An igen p ocessing and p esen a ion
Response o hypoxia
Response o in e e on-γ
Con ol AD
0
1
2
3
4
Rela i e RIPK1 exp ession
ns
P= 0.0947
Con ol AD
-1
0
1
2
3
4
Rela i e RIPK3 exp ession
✱✱
P= 0.0076
Con ol AD
-1
0
1
2
3
4
Rela i e MLKL exp ession
✱✱
P= 0.0043
Con ol AD
0
1
2
3
4
Rela i e MEG3 exp ession
✱
P= 0.0043
Figu e 2: T ansc ip ional changes in xenog a ed neu ons.
Bland-Al man MA plo showing di e en ial exp ession o genes om RNA sequencing o he human
g a s isola ed om con ol and amyloid mice a (A) 2mon hs (con ol n=5, amyloid n=7), (B) 6
mon hs (con ol n=5, amyloid n=5), and (C) 18mon hs (con ol n=4, amyloid n=3) pos -
ansplan a ion.Red, signi ican ly up egula ed genes.Blue, signi ican ly down egula ed genes (FDR <
0.05). FC= old change, CPM=coun s pe million.Compa isons o logFC a mon h 6and mon h 18 a e
shown in Ex ended da a igu e 5g.
(D) Gene On ology (GO) analysis showing e ms associa ed wi h genes up egula ed in amyloid mice
a 6M. Poin size ep esen s he old en ichmen o up egula ed genes in he e m, and colo
ep esen s he –log10FDR.Only e ms wi h FDR < 0.1 a e shown.The xand yaxes ep esen he
'seman ic space' - ha is, how simila wo e ms a e o each o he (calcula ed by Re igo,PMID:
21789182).
(E) Hea map o no malised en ichmen sco es (NES) in neu onal dedi e en ia ion gene se s anked
along he di e en ially exp essed genes in he xenog a s (amyloid s con ol) shown in (Fig. 2a-c).
Posi i e en ichmen s a e shown in ed, nega i e en ichmen s a e shown in blue.Signi ican FDR
alues (P.adj < 0.01)a e shown as numbe s.
(F) Analysis o RIPK1gene exp ession using qRT-PCR on AD (n=11)and con ol (n=10)pos -mo em
human b ain samples.
(G) Analysis o RIPK3gene exp ession using qRT-PCR on AD (n=11)and con ol (n=10)pos -mo em
human b ain samples.
(H) Analysis o MLKL gene exp ession using qRT-PCR on AD (n=11)and con ol (n=10)pos -mo em
human b ain samples.
(I) Analysis o MEG3gene exp ession using qRT-PCR on AD (n=11)and con ol (n=10)pos -mo em
human b ain samples.
(J) Rep esen a i e con ocal images showing MEG3RNAScope in empo al gy us o AD (n=3) and
con ol (n=2) pos -mo em human b ain samples.Nucleus (DAPI, blue), MEG3( ed).Scaleba 10
µM.
(K) Quan i a i e ep esen a ion o he numbe o MEG3punc a pe nucleus in con ol (>100
nuclei/sample, n=2), and AD (>100 nuclei/sample, n=3) pos -mo em human b ain samples.
(L) Rep esen a i e con ocal images showing he exp ession o ac i a ed nec op osis pa hway
ma ke s in 18-mon hs old human neu ons in con ol (n=4) and amyloid (n=4) mice.G a ed human
neu ons (GFP, g een), pRIPK1, pRIPK3o pMLKL ( ed).Whi e a ows indica e pRIPK1, pRIPK3, o
pMLKL posi i e cells.Scaleba 30 µm.
Values a e p esen ed as mean ±SEM.S uden ’s - es in (F-I, and K) o measu e he s a is ical
signi icance.
Figu e 3:
D E
A B C
G
F
pRIPK1 Me ge
MEG3 LV Con ol LV
pRIPK3 Me ge
MEG3 LV Con ol LV
MEG3 LV Con ol LV
pMLKL Me ge
H
Con ol LV MEG3 LV
-15
-10
-5
0
5
ΔΔCT alues
✱✱
P=0,0043
Day 7 Day 9 Day 11
0
50
100
150
% Cell su i al
ns ✱✱ ✱✱✱
Con ol LV
(n=8)
MEG3 LV
(n=8)
P=0,0016 P=0,0001
I
J
DAPI pRIPK1GFP HUNU Me ge
MEG3 shRNA Con ol LV
K L
105
106
107
Numbe o human cells
To al numbae o cells MEG3 shRNA_Ti in
✱✱
0.0094
Con ol MEG3
shRNA
0
50
100
150
200
Cell su i al (%)
Con ol
MEG3
✱✱ ns ns ns
0.0099
DMSO Pona. Dab . NSA
Figu e 3: Long noncoding RNA MEG3induces nec op osis in human neu ons.
(A) Schema ic ep esen a ion o he MEG3exp ession s a egy using len i i al ec o s (LV) in H9-
de i ed human neu ons.
(B) Analysis o he MEG3exp ession in H9-de i ed neu ons a DIV75,se en days pos - ansduc ion
wi h ei he con ol LV (n=5) o MEG3LV (n=6).
(C) Analysis o neu onal cell su i al using CellTi e -Glo eagen a e ansducing wi h ei he con ol
LV (n=8) o MEG3LV (n=8) on day-7, day-9, and day-11 pos - ansduc ion.
(D) Rep esen a i e con ocal image showing ac i a ed nec op o ic ma ke (pRIPK1, Se 166)in H9-
de i ed neu ons a DIV75 ansduced wi h ei he con ol LV (n=3) o he MEG3LV (n=3). Nucleus
(DAPI, blue), pRIPK1(g ey, indica ed wi h ed a ows), ansduced neu ons (GFP, g een), neu ons
(NF-H, ed).Scaleba 30 µm.
(E) Rep esen a i e con ocal image showing ac i a ed nec op o ic ma ke (pRIPK3, Se 227)in H9-
de i ed neu ons a DIV75 ansduced wi h ei he con ol LV (n=3) o he MEG3LV (n=3). Nucleus
(DAPI, blue), pRIPK3(g ey, indica ed wi h ed a ows), ansduced neu ons (GFP, g een), neu ons
(NF-H, ed).Scaleba 30 µm.
(F) Rep esen a i e con ocal image showing ac i a ed nec op o ic ma ke (pMLKL,Se 358)in H9-
de i ed neu ons a DIV75 ansduced wi h ei he con ol LV (n=3) o he MEG3LV (n=3). Nucleus
(DAPI, blue), pMLKL (g ey, indica ed wi h ed a ows), ansduced neu ons (GFP, g een), neu ons
(NF-H, ed).Scaleba 30 µm.
(G) Immunoblo analysis o pRIPK1(Se 166)le els.P o eins we e ex ac ed se en days pos -
ansduc ion om H9-de i ed neu ons a DIV75 ansduced wi h ei he con ol LV (n=4) o MEG3LV
(n=8). GAPDH is used as he loading con ol.Two independen di e en ia ions (Di ).
(H) The same Bland-Al man MA plo showing di e en ial exp ession o bulk RNA sequencing o
human g a s om 6mon hs as in Fig. 2b. Genes highligh ed in ed a e he leading-edge genes
iden i ied in agene-se en ichmen analysis (GSEA), aking he op 400 up egula ed genes om he
bulk sequencing o he p ima y neu ons exp essing MEG3(Ex ended Da a Fig. 10) and plo ing hem
agains he amyloid s con ol old changes o he human g a s (FDR<0.05). FC= old change,
CPM=coun s pe million.
(I) Schema ic ep esen a ion o he nec op osis inhibi ion in i o using H9-de i ed human neu ons
and pona inib (0.5 µM), dab a enib (0.9 µM) and nec osul onamide (0.5 µM).
(J) Analysis o neu onal cell su i al using CellTi e -Glo eagen .Expe imen was pe o med in
iplica es wi h 6 echnical eplica es o each eplica e.
(K) Immunohis ochemis y analysis o pRIPK1le els in samples ob ained om he 6-mon h-old
ansplan ed animals ansplan ed wi h con ol LV (n=5) o he MEG3shRNA (n=5) ansduced
human NPCs.Nuclei we e s ained wi h DAPI (blue), GFP was ep esen ed in g een, pRIPK1was
ep esen ed in ed, and human nuclei we e depic ed in g ey.Whi e a ows indica e pRIPK1posi i e
cells.Scaleba 30 µm.
(L) The human cell numbe was es ima ed om 6-mon h-old xenog a ed mice om con ol (n=5)
and MEG3shRNA (n=12)using q-PCR.Values a e p esen ed as mean ±SEM.S uden ’s - es in (B
and L), One-way ANOVA wi h Tukey’s pos hoc es o mul iple compa isons was used in (C and J) o
measu e he s a is ical signi icance.
Figu e 4:
X34/GFP/pMLKL/HUNU X34/GFP/pRIPK1/HUNU
Con ol Pona inib Dab a inib
X34/GFP/pRIPK3/HUNU
Amyloid mice
A B C
D E
Con ol
Pona inib
Dab a enib
0
20
40
60
% pRIPK1 posi i e cells
✱✱✱✱
✱✱✱
✱
Con ol
Pona inib
Dab a enib
0
10
20
30
40
% pRIPK3 posi i e cells
✱
✱✱✱
ns
Con ol
Pona inib
Dab a enib
0
10
20
30
40
% pMLKL posi i e cells
✱✱✱
✱✱✱
ns
Con ol
Pona inib
Dab a enib
0
1×107
2×107
3×107
To a l n um be o h um an ce ll s
ns
✱✱✱✱
✱✱
F G
0
1×107
2×107
3×107
Numbe o human cells
✱✱
✱✱
Cas9 - + +
RIPK1 sgRNA + + -
RIPK3 sgRNA + + -
0.0042
0.0016
H
-Cas9
DAPI/GFP/pRIPK3/HUNU
+ Cas9
DAPI/GFP/pRIPK1/HUNU
Figu e 4: Inhibi ion o nec op osis p e en s human neu onal cell loss in i o.
(A) Rep esen a i e con ocal images showing he exp ession o he ac i a ed nec op o ic ma ke s
pRIPK1, pRIPK3, and pMLKL in amyloid animals (n=5), amyloid animals ea ed wi h ei he pona inib
(n=5) o dab a enib (n=6). Amyloid (X34,blue), human neu ons (GFP, g een), pRIPK1, pRIPK3, and
pMLKL ( ed), human nuclei (HUNU, g ey).Whi e a ows indica e nec op osis ma ke posi i e neu ons.
Scale ba 30 µm.
(B) Quan i a i e ep esen a ion o he pe cen numbe o human cells showing immuno eac i i y o
pRIPK1in con ol (n=5), pona inib (n=5), and dab a enib (n=6) ea ed animals.
(C) Quan i a i e ep esen a ion o he pe cen numbe o human cells showing immuno eac i i y o
pRIPK3in con ol (n=5), pona inib (n=5), and dab a enib (n=6) ea ed animals.
(D) Quan i a i e ep esen a ion o he pe cen numbe o human cells showing immuno eac i i y o
pMLKL in con ol (n=5), pona inib (n=5), and dab a enib (n=6) ea ed animals.
(E) Quan i a i e es ima ion o he numbe o human cells p esen in con ol (n=5), pona inib (n=5), and
dab a enib (n=6) ea ed g a ed mice a 6mon hs pos - ansplan a ion.
(F) Schema ic ep esen a ion o he RIPK1and RIPK3KO gene a ion and ansplan a ion.NPCs de i ed
om he H9-inducible Cas9line a e ansduced wi h he sgRNAs co esponding o ei he RIPK1o
RIPK3. 100,000 cells we e g a ed in o each mouse a P1-P2 s age o each condi ion.
(G) Quan i a i e ep esen a ion o he numbe o neu ons su i ing om 6mon hs old mice
ansplan ed wi h RIPK1sgRNA wi hou doxycycline (n=7) o RIPK1sgRNA wi h doxycycline (n=7) o
RIPK3sgRNA wi h doxycycline (n=10)using qPCR.
(H) Rep esen a i e immuno luo escence analysis o phospho yla ed nec op osis ma ke s pRIPK1(n=3)
pRIPK3(n=3) om 6mon hs old mice ansplan ed wi h RIPK1sgRNA wi hou Cas9o RIPK3sgRNA
wi h Cas9.Whi e a ows indica es ei he pRIPK1o pRIPK3posi i e cells.Scaleba 30µM.Values a e
p esen ed as mean ±SEM.One-way ANOVA wi h Tukey’s pos hoc es o mul iple compa isons was
used in B-E and G o measu e he s a is ical signi icance.
A B
C
D
GFP (human g a )
GFP (human g a )
Supplemen a y Figu e 1:
X34 GFP IBA1 GFAP Me ge
Con olAmyloid Con ol
3R 4R Me geGFP
Amyloid
Amyloid
Magni ied
X34 GFP AT8 TOPRO3 Me ge
E
Supplemen a y Figu e 1: In eg a ion o he ansplan ed neu ons in he Rag2-/- animals.
(A) Co onal sec ion o 18-mon hs old g a ed mouse b ain showing he g a loca ion (g een,
indica ed wi h whi e a ows). Scale ba 1000 µm.
(B) Highe magni ica ion image o neu onal p ocess displaying dend i es and ma u e spines ( igh
boxes).Whi e a ows indica es dend i ic spines.Scale ba 10 µm.
(C) Rep esen a i e con ocal images o g a ed neu ons 6mon hs pos - ansplan a ion s ained wi h
3R (magen a) and 4R ( ed) TAU iso o ms (examples o human neu ons indica ed wi h whi e a ows).
Scale ba 30 µm.
(D) Rep esen a i e con ocal images showing glial esponses a 18 mon hs pos - ansplan a ion.
Amyloid (X34 in blue), human neu ons in g een (GFP), GFAP (g ey), IBA1( ed).Scale ba 30 µm.
(E) Rep esen a i e con ocal images 6-mon hs old g a ed mice showing dys ophic neu i es and
neu i ic plaque au (NP- au) pa hology (AT8) a ound he X34 plaques 6mon hs pos - ansplan a ion.
The op panel shows NP- au, indica ed wi h whi e a ows.The bo om panel shows amagni ied
egion o one plaque showing dys ophic neu i es and NP- au.Amyloid (X34 in blue), human
neu ons in g een (GFP), P- au (AT8, ed), TOPRO3(nucleus, magen a).Scale ba 30 µm.
Supplemen a y Figu e 2:
A B
C
D
E
3D segmen a ion
X34/AT8/HUNU X34/PHF1/HUNU X34/MC1/HUNU
0
10
20
30
Numbe 'o 'mic oglia'in'
20'µm'X34' ing
N.D
Con ol Amyloid
0
2
4
6
8
10
Numbe 'o 'as oglia'in'
20'µm'X34' ing
N.D
Con ol Amyloid
Con ol Amyloid
0.0
0.5
1.0
1.5
No malised log
MEG3 exp ession
ns
Con ol Amyloid
0.0
0.5
1.0
1.5
No malised log
MLKL exp ession
ns
Con ol Amyloid
0
2
4
6
8
No malised log
MEG3 exp ession
P adj.=5.38e-136
Con ol Amyloid
0
1
2
3
4
No malised log
MLKL exp ession
P adj. =7.88e-13
Con ol Amyloid
0
2
4
6
8
No malised log
MEG3 exp ession
P adj.=4.16e-181
Con ol Amyloid
0
1
2
3
4
No malised log
MLKL exp ession
P adj.=2.22e-05
2 mon hs 6 mon hs 18 mon hs
2 mon hs 6 mon hs 18 mon hs
Supplemen a y Figu e 2: Cha ac e iza ion o he 18 mon hs old g a ed animals.
(A) Quan i ica ion o he numbe o hos mic oglia a ound he Aβ plaque, wi hin 20 mic ons
diame e (n=4, >100 plaques/mice).
(B) Quan i ica ion o he numbe o hos as ocy es a ound he Aβ plaque, wi hin 20 mic ons
diame e (n=4, >100 plaques/mice).
(C) Neu ons immuno eac i e o AT8, PHF1o MC1a e segmen ed in Ima is so wa e and ende ed
o e eal in acellula X34 s aining (β-shee ib illa y s uc u es), an indica i e o he in acellula au
ib ils.Ex acellula X34 (blue) s aining ep esen s Aβ plaques whe e as in acellula X34 (blue)
ep esen s au β-shee s uc u es.Whi e a ows indica e X34 s aining in neu onal somas.Amyloid
(X34,blue), P- au ( ed), human nucleus (HUNU, g ey).Scale ba 5µm.
(D) Human MEG3exp ession om 2-mon hs (con ol n=5, amyloid n=7),6-mon hs (con ol n=5,
amyloid n=5), and 18-mon hs (con ol n=4, amyloid n=3) pos - ansplan a ion.
(E) Human MLKL exp ession om 2-mon hs (con ol n=5, amyloid n=7),6-mon hs (con ol n=5,
amyloid n=5), and 18-mon hs (con ol n=4, amyloid n=3) pos - ansplan a ion.
A
Supplemen a y Figu e 6:
E
B
C
D
012345
20
22
24
26
28
30
huTi in egion qPCR s anda d cu e
log numbe o human/mouse cells
C alue
human+mouse mix gDNA inpu
human alone gDNA inpu
Amyloid s Con ol (6M)
Amyloid s Con ol (6M)
Amyloid
Amyloid magni ied
GFP pMLKL CK1𝛅Me ge
I
Con ol Amyloid
0
10
20
30
Numbe o ib ils/EM g id
Con ol Amyloid
0
1×107
2×107
3×107
4×107
To al n um be o hu ma n ce ll s
✱
P=0.0153
F G
Con ol
(6M)
Con ol
(18M)
0
1×107
2×107
3×107
4×107
To al nu mb e o hu ma n ce ll s
ns
Con ol Amyloid
0
50
100
150
200
250
2 mon hs TITIN nomalized (%)
% Human cells
ns
H
Supplemen a y Figu e 6: Neu onal cell dea h analysis in ansplan ed human neu ons.
(A) Ingenui y pa hway analysis (IPA) o canonical nec op osis pa hway o e laid wi h di e en ially
exp essed gene om he 6-mon hs old neu onal xenog a s.Red indica es up egula ion, and g een
indica es down egula ion.
(B) Bland-Al man MA plo showing exp ession o apop osis pa hway genes in he 6mon h old neu onal
xenog a s.
(C) Bland-Al man MA plo showing exp ession o e op osis pa hway genes in he 6mon h old neu onal
xenog a s.
(D) Es ablishmen o he s anda d cu e o qPCR using TITIN wi h human genomic DNA ( ed line) and
human plus mouse genomic DNA (blue line).
(E) Quan i a i e ep esen a ion o he human neu ons in con ol mice (6Mand 18M)(n=6) using qPCR.
(F) Quan i ica ion o he numbe o immunogold posi i e ib ils iden i ied in he g a ed con ol (n=4)
and amyloid mice (n=4).
(G) Quan i a i e ep esen a ion o he human neu ons a 18Mpos - ansplan a ion in con ol (n=5) and
amyloid (n=6) mice using qPCR.
(H) Quan i a i e ep esen a ion o he human neu ons a 2M pos - ansplan a ion in con ol (n=6) and
amyloid (n=6) mice using qPCR.
(I)Rep esen a i e con ocal images showing co-localiza ion o pMLKL wi h g anulo acuola degene a ion
(GVD) ma ke casein kinase 1del a (CK1𝛅)in 18Mold amyloid mice.Human g a (GFP in g een), pMLKL
( ed), CK1𝛅(g ey).Scale ba 30 µM.
Amyloid s Con ol 2M (mouse) Amyloid s Con ol 6M (mouse) Amyloid s Con ol 18M (mouse)
A B C
Supplemen a y Figu e 7:
Supplemen a y Figu e 7: Gene exp ession p o iles o mouse hos issue om g a ed
mice a 2, 6 and 18 mon hs.
(A) Bland-Al man MA plo showing he di e en ial gene exp ession o mouse
ansc ip ome a 2-mon hs (con ol n=5, amyloid n=7), (B), 6-mon hs (con ol n=5,
amyloid n=5), and (C), 18-mon hs (con ol n=4, amyloid n=3) old mice.Red,
signi ican ly up egula ed genes.Blue, signi ican ly down egula ed genes (FDR < 0.05).
Selec ed mic oglial ac i a ion genes a e labelled.FC= old change, CPM=coun s pe
million.Analysis o mouse Meg3exp ession 2- (D), 6- (E) and 18 (F) mon hs pos
ansplan a ion.Bulk RNA sequencing eads om he RNA sequencing is ex ac ed om
he 2-, 6-, and 18-mon hs old animals and p esen ed as no malized coun s (log scale)
om ei he con ol o amyloid animals.ns, no signi ican .
Con ol Amyloid
0
5
10
15
No malised log
Meg3 exp ession
ns
Con ol Amyloid
0
5
10
15
No malised log
Meg3 exp ession
ns
Con ol Amyloid
0
5
10
15
No malised log
Meg3 exp ession
ns
D E F
Supplemen a y Figu e 8:
A B
Supplemen a y Figu e 8: Xenog a ed neu ons ecapi ula e signa u es iden i ied in human pos -
mo em issue and o he xenog a ed neu ons.
(A) Hea map o no malized en ichmen sco es (NES) in indica ed gene se s anked along he human
g a old changes (up egula ed genes, amyloid s con ol) shown in igu e 2a-c. Posi i e
en ichmen s a e shown in ed, nega i e en ichmen s a e shown in blue.Signi ican FDR alues (p.adj
< 0.01)a e shown as ex in he boxes. M2, up egula ed genes in human neu ons 2mon hs pos -
ansplan a ion (PT). M6, up egula ed genes in human neu ons 6M PT. M18,up egula ed genes in
human neu ons 18MPT.
(B) Hea map o no malised en ichmen sco es (NES) in indica ed gene se s anked along he human
g a old changes (down egula ed genes, amyloid s con ol) shown in igu e 2a-c. Posi i e
en ichmen s a e shown in ed, nega i e en ichmen s a e shown in blue.Signi ican FDR alues (p.adj
< 0.05)a e shown as ex . M2, down egula ed genes in human neu ons 2mon hs pos -
ansplan a ion (PT). M6, down egula ed genes in human neu ons 6M PT. M18,down egula ed
genes in human neu ons 18MPT.
Da ase s used in he abo e compa ison:
Espuny-Camacho e al., 2017 xN AD young, DE up egula ed genes o m human neu ons 4mon hs
pos - ansplan a ion (PT).Espuny-Camacho e al., 2017 xN AD old, 6M-8M PT.(PMID:28238547).
Ma ens e al., 2021 iN AD, DE genes om neu ons di ec ly ep og ammed om he ib oblas -
de i ed om AD pa ien s compa ed o neu ons ep og ammed om con ol ib oblas s.
(PMID:33910058).
Annese (2018)HC AD, DE genes om La e-Onse AD (LOAD) compa ed o con ols om he
hippocampus.(PMID:29523845)
Blalock (2011)LC NFT, DE genes om lase cap u ed P- au posi i e neu ons.(PMID:21756998)
Blalock (2004)MMSE, DE genes om MiniMen al S a us Examina ion (MMSE>20) om AD b ains.
(PMID:14769913).
Mos a a i 2018 DLPFC, DE genes om do sola e al p e on al co ex om AD (PMID:29802388).
DAPI GFAP MEG3 NEUN Me ge
As
Nu
Nu
As
Nu
Nu
As
Nu
Nu
As
Nu
Nu
As
Nu
Nu
DAPI IBA1 MEG3 NEUN Me ge
Mg
Nu
Nu
Nu Mg
Nu
Nu
Nu Mg
Nu
Nu
Nu Mg
Nu
Nu
Nu Mg
Nu
Nu
Nu
Supplemen a y Figu e 9:
A
B
C
DAPI/pMLKL/MEG3/NeuN
Con ol AD
Vehicle Pona inib Dab a inib
X34/IBA1/GFAP
D
Supplemen a y Figu e 9: MEG3is selec i ely exp essed in neu ons.
(A) Rep esen a i e con ocal images om FFPE ixed human hippocampal b ain samples s ained wi h
MEG3(RNA scope, ed) and combined wi h immunos aining o GFAP (g een), NeuN (g ey), and
DAPI (blue).Nu=Neu on, As=As ocy es.Scale ba 30 µm.
(B) FFPE ixed human hippocampal b ain samples we e used o MEG3RNA scope ( ed) and
combined wi h immunos aining o IBA1(g een), NeuN (g ey), and DAPI (blue).Nu=Neu on,
Mg=mic oglia.Scale ba 30 µm.
(C) Rep esen a i e con ocal images showing he colocaliza ion o pMLKL (g een) in he MEG3
posi i e neu ons in AD (n=3) and con ol (n=3) hippocampal b ain samples.Scale ba 30 µm.
(D) Analysis o he glial cell a e om he mice adminis e ed wi h con ol die (n=5) o die mixed
wi h Pona inib (n=5) o Daba a enib (n=6). Mice ea ed wi h he indica ed d ug om 2mon hs o 6
mon hs added o he egula mouse main enance die .B ain samples we e isola ed a e pe usion
wi h PBS ollowed by 4% PFA pos ixa ion o e nigh .Vib a ome sec ions o 40 mic ons a e made
and s ained wi h a X34 (blue, amyloid plaques), IBA1( ed, mic oglia), and GFAP (g ay,as ocy es).
Uppe panel is ala ge ield o iew om he co ex and he lowe panel is amagni ied ield.
Scaleba 30 um.
A
C D
Supplemen a y Figu e 10:
Leading edge (human g a ) MEG3 up egula ed genes
B
Con ol
MEG3 shRNA
0
1
2
3
MEG3 exp ession in i o
Fold MEG3 exp ession
✱✱✱
0.0002
Con ol LV
GFP MEG3 (RNASCOPE) Me ge
MEG3 shRNA
E GF
0
50
100
150
200
%cell su ial
RIPK1 sgRNA + + + +
Cas9 - + - +
Con ol LV + + - +
MEG3 LV - - + +
✱
ns
✱
0
100
200
300
%cell su ial
RIPK3 sgRNA + + + +
Cas9 - + - +
Con ol LV + + - +
MEG3 LV - - + +
ns
ns
✱✱
0
50
100
150
%cell su ial
MLKL sgRNA + + + +
Cas9 - + - +
Con ol LV + + - +
MEG3 LV - - + +
✱✱✱
ns
✱✱✱
0
50
100
150
200
%cell su ial
RIPK1 sgRNA + + + +
Cas9 - + - +
Con ol LV + + - +
MEG3 LV - - + +
✱
ns
✱
0
100
200
300
%cell su ial
RIPK3 sgRNA + + + +
Cas9 - + - +
Con ol LV + + - +
MEG3 LV - - + +
ns
ns
✱✱
0
50
100
150
%cell su ial
MLKL sgRNA + + + +
Cas9 - + - +
Con ol LV + + - +
MEG3 LV - - + +
✱✱✱
ns
✱✱✱
0
50
100
150
200
%cell su ial
RIPK1 sgRNA + + + +
Cas9 - + - +
Con ol LV + + - +
MEG3 LV - - + +
✱
ns
✱
0
100
200
300
%cell su ial
RIPK3 sgRNA + + + +
Cas9 - + - +
Con ol LV + + - +
MEG3 LV - - + +
ns
ns
✱✱
0
50
100
150
%cell su ial
MLKL sgRNA + + + +
Cas9 - + - +
Con ol LV + + - +
MEG3 LV - - + +
✱✱✱
ns
✱✱✱
H JI
Supplemen a y Figu e 10: RNA sequencing o he MEG3exp essing neu ons.
(A) Ba plo s o he o al numbe o eads pe sample ob ained om RNA sequencing o MEG3oexp ession (OE,
con ol n=6, o e exp ession n=6).
(B) Mul i-dimensional scaling (MDS) o he RNA sequencing samples o MEG3exp ession (OE).Neu ons de i ed
om wo independen di e en ia ions (Di ), con ol LV (n=6), MEG3LV (n=6).
(C) Bland-Al man MA plo showing di e en ial exp ession o bulk RNA sequencing o MEG3exp ession (con ol
LC n=6, MEG3LV n=6). Signi ican ly up egula ed genes a e shown in ed and down egula ed genes in blue (FDR <
0.05). Full di e en ial exp ession esul s a e shown in Supplemen a y Table 6.
(D) DAVID gene on ology analysis o leading-edge up egula ed genes om he GSEA analysis (indica ed wi h ed
colo in ig.3h). Top 10 GO e ms we e displayed. - e=nega i e, + e=posi i e, eg.= egula ion ,TF= ansc ip ion
ac o , TNF= umo nec osis ac o , APP=an igen p ocess and p esen a ion.
(E) Schema ic ep esen a ion o he MEG3knockdown app oach using MEG3shRNA.Len i i al ec o s con aining
he MEG3shRNA sequence we e cloned in o he len i i al backbone and packaged. H9-de i ed NPCs we e
ansduced wi h ei he he con ol GFP len i i us o he MEG3shRNA, ollowed by pu omycin selec ion o selec
posi i ely ansduced cells.T ansduced cells we e hen ansplan ed in o amyloid mice.
(F) qRT-PCR analysis was conduc ed o assess human MEG3exp ession using RNA ex ac ed om xenog a ed
mouse b ains a six mon hs o age (con ol n=5; MEG3shRNA n=13).
(G) MEG3knockdown alida ion was pe o med using RNA Scope analysis on samples ob ained om
ansplan ed animals a six mon hs o age.The whi e a ow indica es he MEG3RNA Scope punc a ( ed).
Analysis o nec op osis inhibi ion in i o using he RIPK1, RIPK3, and MLKL knock-ou cells.Neu op ogeni o cells
(NPCs), de i ed om he H9 s em cell line ansduced wi h RIPK1(H),RIPK3(I),and MLKL (J),we e pla ed in la -
bo om black 96-well pla es.Cells we e ea ed wi h doxycycline o induce he Cas9in he indica ed condi ions.
On app oxima ely day 60 in i o (~DIV60), he cells we e ansduced wi h ei he acon ol len i i us o MEG3
len i i us and cells we e main ained in he NMM un il he CellTi e -Glo® assay.Ten days pos ansduc ion o
he MEG3CellTi e -Glo® luminescen cell iabili y assay was pe o med o measu e cell iabili y.The assay was
conduc ed in iplica es wi h six echnical eplica es o each eplica e.The da a om he con ol wells we e
no malized, and he esul s a e p esen ed as he pe cen age o cell su i al be ween he con ol and ea ed
condi ions.Bon e oni mul iple compa ison es us used o measu ing he s a is ical signi icance.
