Nature - USA (2020-08-20)

412 | Nature | Vol 584 | 20 August 2020

Article

Data Fig. 3d, Supplementary Table 1). Modulation of LYS-7 aggregation
by CLEC-1 was shown by overexpression (Extended Data Fig. 3e). Thus,
the ECRs identified in the present screen are likely to be constitutively
active against a broad range of proteins aggregating in the extracel-
lular space.
To assess whether overexpression of ECRs improves extracellular
proteostasis, we randomly chose to overexpress in the body-wall
muscle C36C5.5 and F56B6.6—two uncharacterized proteins with
cysteine-rich sequences—and the C-type lectin CLEC-1. In addition,
we overexpressed the lysozyme-like protein LYS-3. The ECRs tagged
with mVenus were secreted and taken up by coelomocytes (Fig. 2a).
C36C5.5 and F56B6.6 were mainly diffusely localized in the extracel-
lular space, whereas CLEC-1 and LYS-3 had a more punctate pattern

(Fig. 2a). Notably, overexpression of all four candidates significantly

prevented LBP-2 aggregation, whereas secreted GFP alone had no effect

(Fig. 2b, Extended Data Fig. 4a). To understand how ECRs modulate

extracellular protein aggregation, we assessed whether they interact

with LBP-2. Indeed, we observed co-localization of all four ECRs with

LBP-2 aggregates (Fig. 2c), indicating a specific interaction as secreted

GFP did not accumulate in LBP-2 aggregates (Extended Data Fig. 4b).

Examination of the interaction between C36C5.5 and LBP-2 showed

that LBP-2 efficiently co-purifies with C36C5.5 (Fig. 2d, Extended Data

Fig. 4c). C36C5.5 maintained excessive LBP-2 diffusely distributed in

the pseudocoelom of animals that lack coelomocytes (Fig. 2e, f). Thus,

the previously uncharacterized C36C5.5 acts as a holdase chaperone

by directly binding to and stabilizing aggregation-prone proteins and,

b

a F56B6.6::mVenus

C36C5.5::mVenus

CLEC-1::mVenus

LBP-2::tagRFP aggregation

50

35

kDa

Flow- Elution

through

Anti-His

PF

Rg

at-i

tn

A

LBP-2::tagRFP

C36C5.5::mVenus

::histag

c

d

CLEC-1 LBP- 2

C36C5.5 LBP- 2

LYS- 3 LBP- 2

50

35

F56B6.6 LBP-2

++

+

++

+

LYS-3::mVenus

ssGFP

OE

ssGFP

OE

C36C5.5

OE

C36C5.5

OE

+ + – – Coelomocytes

eLBP-2::tagRFP aggregation

Co-purication f LBP-2::tagRFP + ssGFP OE

LBP-2::tagRFP + C36C5.5 OE

LBP-2::tagRFP + ssGFP OE w/o coelomocytes

LBP-2::tagRFP + C36C5.5 OE w/o coelomocytes

LBP-2::tagRFP

C36C5.5::mVenus::histag

>10 aggregates

1–10 aggregates

Animals withaggregates (%)

n = 9086 7572 85 83

P = 0.37

P < 0.0001

P < 0.0001

P < 0.0001

5441 8681

P < 0.0001

Ctrl

F56B6.6

OE Ctrl

C36C5.5

OE Ctrl

LY

Ctrl S-3OE

ssGFP

OE Ctrl

CLEC-1

OE

>10 aggregates

1–10 aggregates

Animals withaggregates (%

)

0

20

40

60

80

100

n=9 5748275

0

20

40

60

80

100 P<0.0001P<0.0001

Fig. 2 | Overexpression of extracellular regulators prevents LBP-2
aggregation. a, Localization of mVenus-labelled ECRs, F56B6.6 (left n = 6, right
n = 12 worms), C36C5.5 (left n = 6, right n = 13), CLEC-1 (left n = 7, right n = 6) and
LYS-3 (left n = 12, right n = 11) in young animals. Arrowheads denote
coelomocytes in whole animal (left). Maximum projection of head region
(right). b, Quantification of LBP-2::tagRFP aggregation in animals
overexpressing ECRs at day 6 (n = 2 independent experiments). ‘Ctrl’ indicates
non-overexpressing siblings of ECR transgenics. c, Colocalization of LBP-
2::tagRFP (magenta) and mVenus-tagged (green) F56B6.6 (n = 11 worms),
C36C5.5 (n = 7), CLEC-1 (n = 6) or LYS-3 (n = 10) at day 3, 25 °C, to accelerate LBP-2
aggregation (single plane). d, Pull-down of C36C5.5::mVenus::histag

co-purifies with LBP-2::tagRFP (n = 2 independent experiments). e, ECR

C36C5.5 efficiently prevents LBP-2 aggregation at day 6 without coelomocytes

(n = 2 independent experiments). f, ECR C36C5.5 maintains excessive LBP-2

diffuse in absence of coelomocytes (day 8; LBP-2 + secreted (ss) GFP

overexpression (OE) n = 17 worms, LBP-2 + ssGFP OE without (w/o)

coelomocytes n = 22, LBP-2 + C36C5.5 OE n = 24, LBP-2 + C36C5.5 OE w/o

coelomocytes n = 17). Maximum projection of head region. Laser intensity 10%.

Scale bars, 15 μm (a), 5 μm (c) or 20 μm (f). P values determined by two-sided

Fisher’s exact test (b) and chi-square test (e). For blot source image, see

Supplementary Fig. 1.