Science - USA (2021-12-17)

INSIGHTS | PERSPECTIVES

PHOTO: CEES DEKKER LAB TU DELFT/SCIXEL

science.org SCIENCE

the development of a simi-
lar approach for amino acids
(AAs), whereby nanopore mea-
surements of single AAs were
identified, even distinguishing
leucine and lysine, which have
the same molecular weight ( 5 ).
However, the technique relied
on the incorporation of target
AAs into predesigned charged
peptide chains.
In addition to the enhanced
chemical complexity of pro-
teins compared with nucleic
acids, they also come in folded
configurations that usually pre-
vent readout in a linear fash-
ion. Here, nature may provide
a helping hand in the form of
molecular machines, such as
unfoldases that could be com-
bined with nanopores to facili-
tate peptide linearization. A few
strategies implemented such
guided protein translocation
through a nanopore, including
using negatively charged DNA
( 1 , 6 ), carrier peptides ( 5 ), unfol-
dases ( 7 , 8 ), or denaturants ( 9 ).
These approaches improve pro-
tein identification by nanopores
and may aid understanding of
isoform diversity on a polypep-
tide level, but longer peptide
reads remain desirable.
The first attempts to use
nanopore sequencing for pep-
tide discrimination were re-
cently reported, but the ap-
proach is limited to reading
very short peptides ( 6 ). By con-
trast, Brinkerhoff et al. present a method to
increase the accuracy by using a helicase
that is present in solution and is allowed to
bind repeatedly to the DNA molecules. This
ingenious design ensures that the same
polypeptide can be read, pulled back into
the nanopore by an electric field, and then
reread with the help of a new helicase. The
ability to read out the same molecule in a
nanopore with such high resolution is a ma-
jor breakthrough in the field. Crucially, the
rereading of the same peptide allows the
authors to reduce nanopore read errors to
one in a million with only 30 rereads.
The approach of Brinkerhoff et al. is not
only elegant but also compatible with com-
mercially available nanopore platforms.
However, the path toward proteomics will
remain challenging. There are ~40 million
proteins in a simple eukaryotic cell ( 10 ). At
an average length of 400 AA, the complete
proteome yields 16 billion AAs. For a low er-
ror rate, 20 rereads for a 26-AA polypeptide

requires 2 min. A single nanopore system

would need 20.5 million hours to analyze

the protein in an entire cell. Even current

commercial systems with 144,000 pores

would only reduce the read time of the cel-

lular proteome to ~1 week. As in genomics,

the question is likely to become whether the

full proteome is needed or whether a more

targeted approach will suffice.

A combination approach with other pro-

tein analysis methods seems inevitable. One

method involves single-molecule fluores-

cence on immobilized proteins, which can

identify hundreds of proteins by labeling

specific AAs ( 11 ). Additionally, protein iden-

tification has been achieved through pro-

tein fragmentation and subsequent nano-

pore readout ( 12 ). However, it is challenging

to adopt these approaches for quantitative

proteomics.

PTMs are another obstacle for nanopore-

based protein sequencing. There are hun-

dreds of PTMs that increase the potential

complexity of nanopore signals

( 13 ). Nevertheless, identifica-

tion of PTMs should be attain-

able with nanopore sensing and

has the potential to be combined

with PTM-specific labeling ( 14 ).

However, the molecular mass

of some modifications, such as

glycosylation, can substantially

surpass that of AAs and stop

peptide translocation through

nanopores. An enzymatic gly-

can-removal step could be nec-

essary before nanopore analysis

of native proteins, but with the

inevitable loss of information

on PTMs. Nanopore sensing

for PTMs seems possible with

biological nanopores where the

diameter is adapted to the tar-

get. Simultaneous sensing with

a range of different nanopores

could lead to a comprehensive

understanding of proteome and

protein isoform diversity.

For the goal of single-mole-

cule proteomics, more devel-

opments are required. Spatial

proteomics with single-cell

resolution will require the

positioning of target cells on

nanopore arrays. The design of

new nanopores with improved

chemical recognition to aid AA

and PTM discrimination seems

to move closer to reality with

the advent of protein-folding

predictions. The abundant

ionic current data will pose

new challenges for data analy-

sis. There are still many barri-

ers to overcome; however, the future looks

bright for true single-molecule readout-

based proteomics. j

REFERENCES AND NOTES

1. H. Brinkerhoff, A. S. W. Kang, J. Liu, A. Aksimentiev, C.
   Dekker, Science 374 , 1509 (2021).


2. Y. Liu, A. Beyer, R. Aebersold, Cell 165 , 535 (2016).


3. J. A. Alfaro et al., Nat. Methods 18 , 604 (2021).


4. J. Clarke et al., Nat. Nanotechnol. 4 , 265 (2009).


5. H. Ouldali et al., Nat. Biotechnol. 38 , 176 (2020).


6. S. Yan et al., Nano Lett. 21 , 6703 (2021).


7. S. Zhang et al., Nat. Chem. 13 , 1192 (2021).


8. J. Nivala, D. B. Marks, M. Akeson, Nat. Biotechnol. 31 , 247
   (2013).


9. P. Tripathi et al., Proc. Natl. Acad. Sci. U.S.A. 118 ,
   e2016262118 (2021).


10. B. Ho, A. Baryshnikova, G. W. Brown, Cell Syst. 6 , 192
    (2018).


11. C. V. de Lannoy, M. Filius, R. van Wee, C. Joo, D. de Ridder,
    iScience 24 , 103239 (2021).


12. F. L. R. Lucas, R. C. A. Versloot, L. Yakovlieva, M. T. C.
    Walvoort, G. Maglia, Nat. Commun. 12 , 5795 (2021).


13. G. A. Khoury, R. C. Baliban, C. A. Floudas, Sci. Rep. 1 , 90
    (2011).


14. L. Restrepo-Pérez, C. H. Wong, G. Maglia, C. Dekker, C.
    Joo, Nano Lett. 19 , 7957 (2019).

10.1126/science.a bn0001

A nanopore sequencer composed of a nanopore (turquoise) and a helicase (red) can

reread DNA-protein conjugates (yellow-purple) with single–amino acid resolution.

1444 17 DECEMBER 2021 • VOL 374 ISSUE 6574