Science - USA (2022-01-28)

Interneuron progenitors in TSC tumors
To characterize the cellular composition of
tumors in TSC organoids, we performed single-
cell transcriptomic analysis on 220-day-old
organoids grown in H-medium. At this age,
organoids consist almost exclusively of tumor
tissues (fig. S5A). To compare intertumoral
heterogeneity, three organoids were dissected
into three tumor regions each and barcoded
separately ( 25 ). Unsupervised clustering in
UMAP projection identified four main clusters:
interneurons (cluster 1), interneuron progeni-
tors (cluster 2), dividing interneuron progenitors

(cluster 3), and excitatory neurons (cluster 4)

(Fig. 2A). Interneurons were characterized

by the expression of canonical regulators of

interneuron development such asDLX2,DLX5,

andDLX6-AS1(Fig. 2C and fig. S5B). Inter-

neuron progenitors expressedDLX2;EGFR;

and progenitor markers such asHES1,SLC1A3,

orVIM(Fig. 2C and fig. S5B). Dividing pro-

genitors of the interneuron lineage were char-

acterized by additionally expressing markers

such asMKI67andTOP2A(Fig. 2C and fig.

S5B). Only a very small number of excitatory

lineage cells (cluster 4) were detected (3%) (Fig.

2B), marked by the expression ofNEUROD2

andNEUROD6(fig. S5B). Thus, tumors in

TSC2+/−H-organoids consist mainly of pro-

genitors and interneurons of ventral origin

(Fig. 2C).

To investigate intertumoral heterogeneity,

we compared the barcoded tumor regions. The

cell type composition was highly consistent

among the ventral lineage clusters, with all

barcodes being evenly distributed (Fig. 2B and

fig. S5, C and D). Whereas in the traditional

view of tumorigenesis, in which TSC inactiva-

tion of the second allele is thought to be a

Eichmülleret al.,Science 375 , eabf5546 (2022) 28 January 2022 2 of 10

Reprogramming

Patient 2 TSC2+/-

Patient 2 Ctrl

TSC2+/-

mosaic

TSC2+/-

TSC Patient 1

TSC Patient 2

Patient 1 TSC2+/-

Patient 1 Ctrl

CRISPR/Cas9

Repair

Pat. 1

TSC2

+/-

110 days

Pat. 2

TSC2

+/-

105 days

H pS6 Ki67 Map2

Pat.1

TSC2

+/-

130d

Pat.2

TSC2

+/-

130d

L

Map2

TSC Patient 2y

hiPSCmTeSR NI High N.

BDNF, GDNF, cAMP

Day 40

Transition

Low Nutrients

High Nutrients

05 10

Tumor-like

Tuber-like

GFAP

TSC Patient 35GW

Vimentin

Vimentin, Ki67

Pat.1

TSC2

+/-

230d

pS6, GFAP

Pat.2

TSC2

+/-

230d

L

Fetal TSC 35GW

H&E

**** ****

0

200

400

600

1000

1400

Pa

t.1 Ctrl

Pat.1

TSC2

+/-^

Pat.2

TSC2

+/-^

Pat.2 Ctrl1 Pat

.2 Ctrl2

App. Soma Size Map2 (μm

2 )

dysmorphic

Neurons Giant Cells

0

500

1000

10001500

2500

App. Cell Size pS6 (μm

2 )

Pat.1 Ctrl

Pat

.1^

TSC2

+/-^

Pat.

2 TSC2

+/-^

Pa

t.2 Ctrl1

**** ****

C

DEF

G H

IJ

A B

Tumor Tumor

Pat.1 Ctrl

Pat.1

TSC2

+/-^

Pat.2

TSC2

+/-^

Pat.3

TSC2

+/-^

Pat.2

Ctrl1

Pat.2

Ctrl

2

0

5

10

15

# of tumors per section

Pat.1

TSC2

+/-^

Pat.2

TSC2

+/-^

Pat.

3 TSC2

+/-^

0

20

40

60

% of Tumor area

Fig. 1.TSC2+/Ð-derived organoids recapitulate histopathology of TSC.
(A) Control (Ctrl) andTSC2+/−cell lines derived from two patients (supplementary
materials, materials and methods). (B) High- and low-nutrient organoid protocols used
to model distinct TSC phenotypes. (C) Hematoxylin and eosin (H&E) staining of
35GW fetal brain depicts histopathology of a fetal SEN. (D) pS6 and Ki67 staining on
110- and 105-day-oldTSC2+/−-derived organoids in high-nutrient medium identifies
SEN-like structures. (Bottom) Higher magnification of inset. (E) (Top) Map2 staining
on 130-day-old organoids in L-medium shows dysmorphic neurons, with
morphology comparable with those in (bottom) a resected tuber of a 2-year-old
patient. Nuclear counterstain was performed with 4′,6-diamidino-2-phenylindole
(DAPI) or hematoxylin. (F) pS6 and GFAP identifies GCs in 230-day-old organoids
comparable with GCs in patient tubers. GCs in organoids express Vimentin, as
shown in patients. GCs can be distinguished from tumors by their lower expression
of Ki67. Nuclear counterstain was performed with DAPI or hematoxylin. (G) Tumors
identified as pS6- and Ki67-positive areas are found inTSC2+/−-derived organoids
of all three patients. Control organoids of patient 1 and two clones of repaired

patient 2 showed no tumors. Color of dots indicate independent batches of

experiments (a summary of replicates is available in fig. S3E). (H) Percentage

of tumor area of the total organoid area reveals similar tumor burden for organoids

derived from threeTSC2+/−patients. Color of dots indicate independent batches

of experiments (a summary of replicates is available in fig. S3F). (I) Area of the

soma in Map2+neurons shows that dysmorphic neurons inTSC2+/−-derived

organoids are roughly fourfold larger then Map2+neurons in control organoids

[patient 1 control versusTSC2+/−P< 0.0001; patient 2 control 1 versusTSC2+/−

and patient 2 control 2 versusTSC2+/−bothP< 0.0001; ordinary one-way analysis

of variance (ANOVA)] a summary of replicates is available in fig. S4D). (J) Cell

area of pS6-positive cells shows enlarged pS6 cells inTSC2+/−-derived organoids.

(patient 1 control versusTSC2+/−P< 0.0001, patient 2 control versusTSC2+/−

P< 0.0001, patient 1 control versus patient 2 controlP> 0.9999, patient 1TSC2+/−

versus patient 2TSC2+/−P> 0.9999; Kruskal-Wallis test with Dunn’s multiple

comparisons test) (a summary of replicates is available in fig. S4E) Scale bars,

(C) and (D) 500mm; (E) 20mm; and (D), inset, and (F), inset, 50mm.

RESEARCH | RESEARCH ARTICLE