Science 6.03.2020

SCIENCE sciencemag.org

PHOTO: RAY ADKINSSS

The study of Chen et al. demonstrates

the challenge of identifying microproteins

by evolutionary conservation. Many short

ORFs that encode conserved peptides, in-

cluding some lncRNA ORFs identified by

Chen et al., have indeed been annotated

as such in some databases. However, most

microproteins found to be important for

cell growth do not show strong evolution-

ary conservation of their amino acid se-

quences, as canonical proteins typically do.

Perhaps not all the ORFs that regulate cell

growth in vitro would have similar impact

on organismal fitness ( 13 ). Nevertheless,

extrapolating from this study, many more

interesting peptides encoded by noncanon-

ical ORFs must exist and regulate far more

biological processes than currently known.

Their frequent usage of non-AUG start co-

dons may allow for translational control

that is distinct from canonical ORFs.

Noncanonical translation has been

linked to a variety of neurological diseases

caused by the expansion of short tandem

repeats in the genome ( 14 ). For example,

expanded GGC repeats within the 5 9 UTR

of human fragile X mental retardation 1

(FMR1) mRNA is translated into polygly-

cine by a uORF-like mechanism ( 15 ). It is

unclear whether most repeat-associated

non-AUG translation occurs this way.

However, the pervasive translation of cy-

tosolic RNAs and the widespread usage of

alternative initiation codons in producing

physiologically functional peptides pro-

vide a parsimonious explanation for the

pathological translation events in noncod-

ing regions. j

REFERENCES AND NOTES

1. I. Ulitsky, D. P. Bartel, Cell 154 , 26 (2013).


2. A. A. Bazzini et al., EMBO J. 33 , 981 (2014).


3. N. T. Ingolia et al., Cell Rep. 8 , 1365 (2014).


4. Z. Ji, R. Song, A. Regev, K. Struhl, eLife 4 , e08890 (2015).


5. D. M. Anderson et al., Cell 160 , 595 (2015).


6. T. Kondo et al., Nat. Cell Biol. 9 , 660 (2007).


7. A. Pauli et al., Science 343 , 1248636 (2014).


8. E. G. Magny et al., Science 341 , 1116 (2013).


9. S. R. Starck et al., Science 351 , aad3867 (2016).


10. J. Chen et al., Science 367 , 1140 (2020).


11. M. Guttman, P. Russell, N. T. Ingolia, J. S. Weissman, E. S.
    Lander, Cell 154 , 240 (2013).


12. T. G. Johnstone, A. A. Bazzini, A. J. Giraldez, EMBO J. 35 ,
    706 (2016).


13. W. F. Doolittle, T. D. Brunet, S. Linquist, T. R. Gregory,
    Genome Biol. Evol. 6 , 1234 (2014).


14. F. B. Gao, J. D. Richter, D. W. Cleveland, Cell 171 , 994
    (2017).


15. M. G. Kearse et al., Mol. Cell 62 , 314 (2016).

ACKNOWLEDGMENTS

We thank A. Giraldez, I. Ulitsky, N. DeRuiter, and other

members of the Guo lab for comments. J.U.G. is a NARSAD

Young Investigator and a Klingenstein-Simons Fellow in

Neuroscience. Our work is supported by the Yale Scholar in

Neuroscience Fund, the Ludwig Family Foundation, an NIH

New Innovator Award (DP2 GM132930), and the Muscular

Dystrophy Association (MDA602934).

10.1126/science.aba6117

LIQUID CRYSTALS

Rings rule three-dimensional

active matter

Dynamical defect networks are imaged in viral-particle

liquid crystals driven by biomotors

By Denis Bartolo

W

e are active matter, in that we move

and deform on our own without

resorting to external forces. Dur-

ing the past 10 years, fluids and

soft solids have been turned into

active matter animated by spon-

taneous flows and deformations, usually

by seeding a fluid with self-propelled units

that convert chemical or electric energy into

directed motion. Animated by spontaneous

flows and deformations driven from within,

active fluids are usually powered by dispers-

ing active particles that

self-assemble into coherent

dynamical structures. Ac-

tive materials—including

polymer gels, liquid emul-

sions, or colloidal suspen-

sions with mesmerizing

emergent dynamics ( 1 – 3 )—

have been mostly limited

to two-dimensional (2D)

materials. However, thin

films are unlikely to host

more than a small fraction

of the full complexity of ac-

tive matter. On page 1120 of

this issue, Duclos et al. ( 4 )

report the synthesis and

the observation of 3D ac-

tive liquid crystals (see the

image) and elucidate the

elementary excitations re-

sponsible for their complex inner dynamics.

Nematic liquid crystals are formed of

elongated molecules or colloidal particles

that fluctuate in their orientation, but on av-

erage, all locally point along the same direc-

tion, described as the director. In ordinary

passive nematics, shear deformations are

unhindered as in normal fluids, but bend-

ing deformations are suppressed by elastic

stresses in favor of homogeneously aligned

states (see the figure, top). In stark contrast,

when a nematic includes active components

that power internal stresses that distort the

fluid, the smallest bending deformations can

be exponentially amplified (see the figure,

top). This generic instability yields unsteady

bulk flows casually termed “active turbu-

lence” ( 5 ). Unlike turbulence in normal flu-

ids, the chaotic flows of active nematics are

devoid of self-similar structures (ones that

repeat identically at all scales). The flows

originate from the formation and disappear-

ance of vortices that have a well-defined size

set by the competition between active and

elastic stresses.

The spatial structure of 2D chaotic vor-

tices is easily inferred from a static picture

of the nematic director that defines its local

orientation. In 2D active

nematic films, each vortex

is centered on pointwise

singularities of the director

sketched in the figure, top.

Nematic defects are a sig-

nal of a finite local wind-

ing of the director field

and interact at a distance,

much like electric charges.

Topological charge being a

conserved quantity, no net

charge can develop in a

defect-free nematic, and in

two dimensions, no smooth

deformation can morph a

+½ into a –½ charge. This

topological constraint im-

plies that active stresses

can only nucleate pairs of

oppositely charged defects

in response to bending instabilities, and in

two dimensions (see the figure, top).

The mapping of the unsteady and hetero-

geneous flows of active nematics on the mo-

tion of interacting charged particles was the

key to understanding 2D active turbulence.

However, in 3D liquid crystals, the nature

of the fundamental excitations is drastically

different ( 6 ). The director field in 3D nemat-

ics can host spatially extended topological

excitations in the form of disclination lines

and loops. Although these singularities were

vastly explored in passive systems, they re-

mained out of reach of active matter until

the study of Duclos et al., who assembled

liquid crystals from micrometer-long fila-

mentous bacteriophage viruses instead of

Laboratoire de Physique, Ecole Normale Supérieure de Lyon,

Université de Lyon, France. Email: [email protected]

The active 3D nematic was imaged

with wide-field fluorescent microscopy.

6 MARCH 2020 • VOL 367 ISSUE 6482 1075

Published by AAAS