Science - USA (2022-04-22)

25. S. Zhanget al.,Prog. Mater. Sci. 68 ,1– 66
    (2015).


26. Z. Chenet al., Giant tuning of ferroelectricity in single crystals
    by thickness engineering.Sci. Adv. 6 , eabc7156 (2020).


27. M. Müller, E. Soergel, M. C. Wengler, K. Buse,Appl. Phys. B 78 ,
    367 – 370 (2004).


28. F. G. Wakim,J. Chem. Phys. 49 , 3738–3739 (1968).


29. A. Jovanović, M. Wöhlecke, S. Kapphan, A. Maillard,
    G. Godefroy,J. Phys. Chem. Solids 50 , 623– 627
    (1989).


30. L. Kovács, M. Wohlecke, A. Jovanović, K. Polgár, S. Kapphan,
    J. Phys. Chem. Solids 52 , 797–803 (1991).


31. C. Heet al.,J. Appl. Phys. 110 , 083513 (2011).


32. F. Wuet al.,J. Mater. Sci. 47 , 2818–2822 (2012).


33. E. Sunet al.,J. Appl. Phys. 107 , 113532 (2010).


34. Materials and methods are available as supplementary
    materials.


35. S. Ducharme, J. Feinberg, R. Neurgaonkar,IEEE J. Quantum
    Electron. 23 , 2116–2121 (1987).


36. H. Takenaka, I. Grinberg, S. Liu, A. M. Rappe,Nature 546 ,
    391 – 395 (2017).


37. F. Liet al.,Science 364 , 264–268 (2019).


38. M. Veithen, X. Gonze, Ph. Ghosez,Phys. Rev. B Condens.
    Matter Mater. Phys. 71 , 125107 (2005).


39. A. K. Hamze, A. A. Demkov,Phys. Rev. Mater. 2 , 115202
    (2018).


40. A. K. Hamze, M. Reynaud, J. Geler-Kremer, A. A. Demkov,
    Npj Comput. Mater. 6 , 130 (2020).


41. M. Veithen, Ph. Ghosez,Phys. Rev. 71 , 132101
    (2005).


42. W. Zhaoet al.,Ceram. Inter. 42 , 11909– 11914
    (2022)


43. W. Koechner,Solid-State Laser Engineering, vol. 1 (Springer,
    1988).

ACKNOWLEDGMENTS
Funding:F.L., Z.X., and X.Y.W. acknowledge the support of the
National Natural Science Foundation of China (grants 51922083,
51831010, and 51761145024), the development programme of Shaanxi
province (grants 2019ZDLGY04-09), the National Key R&D Program
of China (grant 2021YFE0115000), and the 111 Project (B14040).
H.T. acknowledge the support of the National Natural Science
Foundation of China (grants 12074092 and 12004085) and the
fellowship of China National Postdoctoral Program for Innovative
Talents (grant BX20200111). B.X. acknowledges financial support from
the National Natural Science Foundation of China (grant 12074277)
and the Natural Science Foundation of Jiangsu Province
(BK20201404). D.W. acknowledges the support of the Australian
Research Council (FT180100541). B.W. and L.Q.C. acknowledge the
support of the US National Science Foundation under grant DMR-
1744213 and Materials Research Science and Engineering Center
(MRSEC) grant DMR-1420620.Author contributions:The work was
conceived and designed by S.Z., H.T., and F.L.; X.L. prepared the
samples, with assistance from C.Q. and L.Q.; X.L. and P.T. performed
the optical experiments, with assistance from W.Z.; P.T. and X.J.
fabricated the device and characterized the performances, with
assistance from C.W. and X.L.; K.S. and H.G. grew the crystals; F.L. and
Z.X. supervised crystal growth; F.L., X.W., and H.T. supervised the
optical experiments; H.T. supervised device fabrication and
characterization; X.M. and B.X. performed the DFT calculation;
Y.L. and B.W. performed the phase field simulations with the
supervisionofF.L.andL.Q.C.;F.L.andX.L.draftedthe
manuscript; S.Z., D.W., S.L., and L.Q.C. revised the manuscript,
and all authors discussed the results.Competing interests:The
authors declare that they have no competing interests.Data
and materials availability:All relevant data are available in the
main text or the supplementary materials.

SUPPLEMENTARY MATERIALS
science.org/doi/10.1126/science.abn7711
Materials and Methods
Supplementary Text
Figs. S1 to S12
Tables S1 to S7
References ( 44 Ð 63 )
Movies S1 and S2

19 December 2021; accepted 23 March 2022
10.1126/science.abn7711

CELL BIOLOGY

ESCRT-mediated membrane repair protects

tumor-derived cells against T cell attack

Alex T. Ritter^1 *, Gleb Shtengel^2 , C. Shan Xu^2 , Aubrey Weigel^2 , David P. Hoffman^2 †, Melanie Freeman^2 †,

Nirmala Iyer^2 , Nensi Alivodej^2 ‡, David Ackerman^2 , Ilia Voskoboinik^3 , Joseph Trapani^3 ,

Harald F. Hess^2 , Ira Mellman^1 *

Cytotoxic T lymphocytes (CTLs) and natural killer cells kill virus-infected and tumor cells through

the polarized release of perforin and granzymes. Perforin is a pore-forming toxin that creates a

lesion in the plasma membrane of the target cell through which granzymes enter the cytosol and

initiate apoptosis. Endosomal sorting complexes required for transport (ESCRT) proteins are involved

in the repair of small membrane wounds. We found that ESCRT proteins were precisely recruited in

target cells to sites of CTL engagement immediately after perforin release. Inhibition of ESCRT

machinery in cancer-derived cells enhanced their susceptibility to CTL-mediated killing. Thus, repair

of perforin pores by ESCRT machinery limits granzyme entry into the cytosol, potentially enabling

target cells to resist cytolytic attack.

C

ytotoxic lymphocytes, including cytotoxic

T lymphocytes (CTLs) and natural killer

(NK) cells, are responsible for identify-

ing and destroying virus-infected or tu-

morigenic cells. To kill their targets,

CTLs and NK cells secrete a pore-forming

toxin called perforin through which apoptosis-

inducing serine proteases (granzymes) are

delivered directly into the cytosol. Successful

killing of target cells often requires multiple

hits from single or multiple T cells ( 1 ). This has

led to the idea that cytotoxicity is additive, often

requiring multiple rounds of sublethal lytic

granule secretion events before a sufficient

threshold of cytosolic granzyme activity is

reached to initiate apoptosis in the target ( 2 ).

Loss of plasma membrane integrity induced

by cytolytic proteins or mechanical damage

leads to a membrane repair response. Dam-

age results in an influx of extracellular Ca2+,

which has been proposed to lead to the re-

moval of the membrane lesion by endocytosis,

resealing of the lesions by lysosomal secretion,

or budding into extracellular vesicles ( 3 ). Per-

forin pore formation was initially reported to

enhance endocytosis of perforin ( 4 ), but sub-

sequent work has challenged this claim ( 5 ).

Endosomal sorting complexes required for

transport (ESCRT) proteins can repair small

wounds and pores in the plasma membrane

caused by bacterial pore-forming toxins, mech-

anical wounding, and laser ablation ( 6 , 7 ).

ESCRT proteins are transiently recruited

to sites of membrane damage in a Ca2+-

dependent fashion, where they assemble bud-

ding structures that shed to eliminate the

wound and restore plasma membrane integ-

rity. ESCRT-dependent membrane repair has

been implicated in the resealing of endoge-

nous pore-mediated plasma membrane dam-

age during necroptosis ( 8 ) and pyroptosis ( 9 ).

Localization of target-derived ESCRT proteins

to the cytolytic synapse

To investigate whether ESCRT-mediated mem-

brane repair might be involved in the removal

of perforin pores during T cell killing, we first

determined whether ESCRT proteins in cancer-

derived cells were recruited to sites of CTL

engagement after perforin secretion. We used

CTLs from OT-I mice that express a high-

affinity T cell receptor (TCR) that recognizes

the ovalbumin peptide SIINFEKL (OVA257-264)

bound to the major histocompatibility com-

plex (MHC) allele H-2Kb ( 10 ). We performed

live-cell microscopy of OT-I CTLs engaging

SIINFEKL-pulsed target cells that express en-

hanced green fluorescent protein (EGFP)–tagged

versions of Tsg101 or Chmp4b, two ESCRT

proteins implicated in membrane repair ( 6 ).

To correlate recruitment of ESCRT proteins

with perforin exposure in time, we monitored

CTL-target interaction in media with a high

concentration of propidium iodide (PI), a cell-

impermeable fluorogenic dye that can rapidly

diffuse through perforin pores to bind and il-

luminate nucleic acids in the cytosol and nu-

cleus of the target ( 5 ). EGFP-tagged ESCRT

proteins were consistently recruited to the site

of CTL engagement within 30 to 60 s after PI

influx (Fig. 1, A and B). EGFP-Tsg101 and EGFP-

Chmp4b in target cells accumulated at the

cytolytic synapse after PI influx in 25 of 27

(92.6%) and 31 of 33 (93.9%) of conjugates

monitored, respectively, compared with a cy-

tosolic EGFP control, which was not recruited

SCIENCEscience.org 22 APRIL 2022¥VOL 376 ISSUE 6591 377

(^1) Genentech, Inc., South San Francisco, CA 94080, USA. (^2) Janelia
Research Campus, Howard Hughes Medical Institute, Ashburn,
VA 20147, USA.^3 Rosie Lew Cancer Immunology Program, Peter
MacCallum Cancer Centre, Melbourne VIC, Australia.
*Corresponding author. Email: [email protected] (A.T.R.);
[email protected] (I.M.)
†Present address: 10X Genomics, Inc., Pleasanton, CA 94566, USA.
‡Present address: Neurovascular Interface, Goethe University
Frankfurt and Max Planck Institute for Brain Research, 60438
Frankfurt am Main, Germany.
RESEARCH | RESEARCH ARTICLES