Content uploaded by Mousumi Akter
Author content
All content in this area was uploaded by Mousumi Akter on Aug 23, 2023
Content may be subject to copyright.
https://doi.org/10.20965/jrm.2023.p1047
Review:
Biomolecular Motor-Based Swarm Robot:
An Innovation in Molecular Delivery
Mousumi Akter∗and Akira Kakugo∗∗
∗Institute of Molecular Biology, University of Oregon
1229 University of Oregon, 1318 Franklin Blvd, Eugene 97403, USA
E-mail: mousumia@uoregon.edu
∗∗Department of Physics, Graduate School of Science, Kyoto University
Kitashirakawa Oiwake-cho, Sakyo-ku, Kyoto 606-8502, Japan
E-mail: kakugo.akira.8n@kyoto-u.ac.jp
[Received January 20, 2023; accepted March 29, 2023]
Biomolecular motor-based micro-sized robots have re-
cently created an innovation in the field of science and
technology as molecular transporters. Groups of these
tiny robots can work substantially better than individ-
ual ones in terms of the transported distance and num-
ber or size of cargo. Site-specific molecular delivery,
the main feature of these robots, has helped to improve
the workability of robots in a more controllable man-
ner.
Keywords: biomolecular motor, swarm robot, molecular
delivery, microtubule, DNA
1. Introduction
Group formation allows the execution of tasks more ef-
ficiently than a single entity by providing robustness and
flexibility to the group as a swarm [1]. Socialized ani-
mals such as bees, ants, or fish often adopt this strategy
to accomplish complex tasks, e.g., building nests, carry-
ing and gathering large loads to a destination, and task
allocation. In recent years a system composed of multiple
robots named “swarm robots” has attracted the attention
of researchers working in the field of science and technol-
ogy [2, 3].
Swarm systems made of mechanical robots have al-
ready been observed to work for medical treatment or
disaster management [4]. However, such swarm-based
task achievement is challenging in the microscopic world,
which could be beneficial to molecular sensing and com-
putations as well as molecular robotic systems. Re-
cently developed technology overcomes all the challenges
to create tiny robots with higher scalability and con-
trollability. Self-propelled systems, driven by chemi-
cal [5], enzymatic [6], or photochemical reactions [7],
have been promising in the construction of microscale
swarm robots over colloidal systems because of their ca-
pability of autonomous actuation. Among the enzymatic
self-propelled systems, adenosine triphosphate (ATP)-
fueled biomolecular motor systems, for example, micro-
tubulekinesin/dynein, have attracted considerable atten-
tion in swarm robotics because of their translational mo-
tion and engineering properties [8,9].
In 2018, for the first time, we created biomolecu-
lar motor-based swarm robots by utilizing the molecular
recognition ability of DNA in controlling local interac-
tions [10, 11]. It took another five years to get the molec-
ular robots to work cooperatively [12]; in this review, we
will focus on the different aspects of our molecular swarm
robots.
2. Biomolecular Motor Based Swarm Robot in
Molecular Delivery
2.1. Biomolecular Motor System-Based Swarm
Robot
Biomolecular motor systems are the smallest natural
machines, which are known as the active workhorses
of cells [13]. Consisting of cytoskeletal filaments and
biomolecular motors, these systems can convert chemical
energy, usually in the form of the high-energy phosphate
bond of ATP, into directed motion. The biomolecular mo-
tor system was successfully reconstructed using biotech-
nology and an in vitro motility assay system was estab-
lished. Cytoskeletal filaments (microtubules and actins)
with an “inverted” geometry (relative to the intracellular
arrangement) are propelled by surface-immobilized mo-
tors (kinesin/dynein and myosin) over distances ranging
from micrometers to centimeters using ATP as the chemi-
cal fuel. Owing to these unique features, biological motor
systems have attracted attention as actuators in the fields
of chemistry, physics, biology, robotics, etc. [14]. Actua-
tors work as power generators for machines or robots. The
small size and engineering properties of the microtubule
(MT)-kinesin system make it suitable for application as a
molecular actuator for a wide range of applications.
Generally, three basic components are essential for
constructing molecular robots: the actuator, processor,
and sensor [15]. Sato et al. constructed an amoeba-like
shape changing molecular robot composed of a body, an
Journal of Robotics and Mechatronics Vol.35 No.4, 2023 1047
© Fuji Technology Press Ltd. Creative Commons CC BY-ND: This is an Open Access article distributed under the terms of
the Creative Commons Attribution-NoDerivatives 4.0 InternationalLicense (https://creativecommons.org/licenses/by-nd/4.0/)
Akter, M. and Kakugo, A.
Fig. 1. Schematic illustration of micro-sized molecular
swarm robot constructed from kinesin-powered DNA modi-
fied MTs.
Fig. 2. Light-controlled reversible swarming of molecular
robots. (a) Reversible hydrogen bonding of photoresponsive
DNAs by light-induced cis-trans isomerization of azoben-
zene. (b) Reversible control of the swarming of robots un-
der UV and VIS light for three repeated cycles. Scale bar:
20
μ
m. (c) High resolution AFM image of the orientation
of MTs in a circular swarm robot. The figures are used with
permission from [10, 18].
actuator, and an actuator controlling processor. The body
was made of a lipid bilayer based vesicle, whereas the ac-
tuator was composed of MT, kinesin, and protein and the
processor was a DNA molecule [16]. Our swarm robot
also consists of the three important components. We con-
structed our molecular robots using MT-kinesin as an ac-
tuator, the photoresponsive molecule as a sensor to sense
the environment to further process the information, and
DNA as a processor to process that information for con-
trolling the robots (Fig. 1). Inserting azobenzene, a pho-
toresponsive molecule, into DNA permits the swarm for-
mation of robots under visible light (
λ
=480 nm) and dis-
sociation under UV light (
λ
=365 nm) (Fig. 2(a)) [17].
Swarming of the molecular robots was initiated by DNA
duplex formation between the robots, which was induced
by the trans isomerization of the azobenzene units in the
single-stranded DNA under visible (VIS) light irradiation.
The dissociation of the swarms was triggered by trans-to-
cis isomerization after UV light irradiation by dissociat-
ing the double-stranded DNA into single-stranded DNA
(Fig. 2(b)). The size of the swarms in terms of the
length and width grows with the VIS irradiation time.
The morphology of the swarm (circular ring-shaped and
bundle shaped) could also be controlled by changing the
flexibility and length of the MTs. MTs in a swarm
Fig. 3. (a) Schematic illustration of the loading and trans-
port of cargo (spheres) by the swarm of molecular robots
under VIS light irradiation and unloading under UV light ir-
radiation. (b) Time-lapse fluorescence microscopy images
of the loading and transportation of cargo by the swarm
robot. Scale bar: 20
μ
m. Thefigureisusedwithpermis-
sion from [12].
system could be oriented by piling up in multiple lay-
ers instead of being organized in a single layer, which
was recently investigated by high speed atomic force mi-
croscopy (Fig. 2(c)) [18].
2.2. Molecular Swarm Robots as Cargo
Transporters
After the successful optimization of our swarm robots,
we focused on their application. However, several reports
demonstrated cargo transportation by a single MT in with
a cargo diameter ranging between 20 nm to 1
μ
m [19,20].
Moreover, loading and transporting cargo for a distance
exceeding 50
μ
m was a great challenge. We were in-
terested in implementing our molecular swarm robots as
cargo transporters to overcome all the existing challenges
in the field. Fig. 3 demonstrates the design and concept of
cooperative cargo transportation by the molecular swarm
robot, in which VIS light irradiation triggers cargo load-
ing and transportation, and UV light irradiation performs
cargo unloading. Two types of molecular robots were
prepared by modifying two fluorescent dyes (ATTO550
and ATTO488)-labeled MTs with two photoresponsive
single-stranded DNAs: pDNA1 and pDNA2. These two
robots were introduced into a flow cell and propelled by
surface-adhered motors, recombinant kinesins in the pres-
ence of ATP. The model cargo was prepared by modifying
a1.1
μ
m diameter of polystyrene beads with pDNA2 and
the fluorescent dye ATTO647. The gliding swarms loaded
the cargo followed by collision owing to the duplex for-
mation between pDNA1 on pDNA1-MT and pDNA2 on
the cargo surface and continued gliding. The swarm-
loaded cargo traveled a total distance of ∼1100
μ
m with-
out falling off. The swarm robots unload the cargo under
UV light irradiation. The azobenzene units in the photore-
sponsive DNA isomerized from the trans to cis conforma-
1048 Journal of Robotics and Mechatronics Vol.35 No.4, 2023
Biomolecular Motor-Based Swarm Robot
Fig. 4. (a) Trajectories of the transported cargo by the
swarm and single robot. Bottom shows the longer distance
traveled by the swarm compared to the single robots. Scale
bar: 20
μ
m. (b) Higher cargo transport efficiency of swarm
robot than single under UV and VIS light. (c) Change in
cargo attachment rate to swarm robot with different cargo
size. The figure is used with permission from [12].
tion, which caused duplex dissociation due to the lower
melting temperature of the cis form of the photorespon-
sive DNA. The dissociation of the swarm robots into in-
dividuals and cargo unloading from the transporters were
occurred simultaneously. The unloaded cargo remained
stationary and was no longer transported under UV light
irradiation.
2.3. More Robots for Higher Transport Efficiency
The groups of molecular robots demonstrated better
cargo transport efficiency compared to the single molec-
ular robot. To demonstrate the advantages of the swarm
over the single robot, cargo transport by gliding pDNA1-
MTs was attempted under similar conditions to those
applied to the swarms. The gliding single transporters
loaded the cargo by collision and transport under VIS
light irradiation; however, the travel distance was much
shorter than that observed with the swarm robots owing
to the cargo dropping off (Fig. 4(a)). The performance
of swarms depends on the number of MTs in the swarms.
The cargo transportation efficiency of the swarm robots
was approximately four times higher than that of the sin-
gle robots (Fig. 4(b)). The better performance of the MT
swarms in loading a large number of cargo and transport-
ing it for a longer distance could be attributed to the high
attachment rate with cargo by the larger surface area and
larger number of binding sites on the swarms compared
to the single robots (Fig. 4(c)). The advantage of the
swarms over individual transporters was that they could
carry large cargo by varying the size of the cargo diam-
eters in the range of 3.4–30.0
μ
m. Under the same con-
ditions, the swarms could load and transport cargo with
diameters up to 20.0
μ
m. However, the single robots
could not load and transport cargo with a diameter larger
than 3.4
μ
m. The single robots encountered cargo of all
sizes but could hardly load them. Notably, a diameter of
Fig. 5. Light controlled cargo delivery. (a) Schematic rep-
resentation showing cargo unloading by the swarm robots at
a designated place under UV light irradiation. (b) Fluores-
cence microscopy images of the UV-irradiated place (white
area) and VIS light-irradiated place (black area). The right
images show the time-lapse fluorescence microscopy images
of cargo unloading and concentration after dissociation of
swarm robots in the UV irradiated area. Scale bar: 50
μ
m.
(c) Dependence of time to unload cargo with different swarm
size. The figure was used after modification with the permis-
sion of [12].
20.0
μ
m is so large that it is approximately several hun-
dred times larger than that of a single MT. Transporting
this large size cargo is considered beneficial because it
can be harnessed only through cooperative tasks accom-
plished by groups of transporters.
2.4. Light-Controlled Molecular Delivery
The light-controlled molecular delivery of cargo to a
designated destination by a swarm robot was challenging
and it was performed under external and dynamic user
control. UV light was applied at a designated place as
the cargo destination, along with VIS light, as shown in
Fig. 5(a). The transported cargo was unloaded at the des-
ignated destination due to the dissociation of the DNA du-
plex between the swarm robot and cargo. The time-lapse
fluorescence microscopy images in Fig. 5(b) illustrate that
the entry of the swarm transporter carrying cargo from the
VIS light region to the UV light region induced dissoci-
ation to form single robots and caused cargo unloading.
The dissociated single robots then came back to the VIS
light area, parting the cargo in the UV light region. The
rate of cargo concentration in the UV region was much
larger than that in the other region without UV light irra-
diation, which was because of the cargo unloading from
the swarm robot. For light-controlled cargo delivery by
the swarm robot, the reversibility of cargo loading and
transport was also observed under VIS light irradiation
and cargo unloading under UV light irradiation. The time
required to unload cargo or the distance traveled to un-
load cargo in the unloading zone depends on the swarm
size (Fig. 5(c)). The accuracy of site-specific transporta-
tion was also estimated by quantifying the mean distance
required to unload a cargo and was approximately 30
μ
m
depending on the specific transporter velocity. To enable
Journal of Robotics and Mechatronics Vol.35 No.4, 2023 1049
Akter, M. and Kakugo, A.
more efficient cargo transport to a predetermined place,
the motion and delivery of the cargo should be controlled
along predesigned tracks to well-designed cargo drop-off
stations [21]. However, our work lacks this predesigned
infrastructure that can offer more flexible and efficient
transportation.
3. Conclusions
We constructed a group of molecular robots and ap-
plied their advantages over individuals to transport large
cargo in a cooperative manner combining DNA nanotech-
nology, material science, and bioengineering. Molecu-
lar swarm robots have the advantage of being very small
in size, which enables the creation of a large number
of robots and managing them to perform tasks coher-
ently. Allowing small individuals to interact with each
other and produce a more complex, collective task, has
been motivated by the collective behavior of living organ-
isms. We also presented the potential to construct such
groups in the future to further intensify these coopera-
tive functions of the robots. Recently, several computa-
tion studies have demonstrated the potential of molecu-
lar swarm robots for practical applications by their ability
to participate in logical operations and molecular compu-
tations [22]. Prospects of the molecular swarm robots in
utilizing the emergent functions through group formation
are also emphasized alongside future perspectives. How-
ever, our findings on the cooperative task achievement
of a group of robots overcame all the recent challenges;
it will encourage researchers working in molecular ma-
chinery to create autonomic molecular robots. Simulta-
neously, molecular machines working in a group can be
applied in molecular delivery. Intensive drug delivery to a
specific location or the collection of micro-contaminants
from environments is the expected application for our
swarm robots. We also assumed that our swarm of molec-
ular robots would be useful for a micro-devicethat can de-
tect pathogens or genetic information or as micro-reactors
by assembling nano-parts more efficiently.
Our next step is to implement artificial intelligence
in the molecular swarm system to integrate autonomous
functions in robotics [23, 24]. Another novel molecular
computing approach based on reservoir computing could
also functionalize our molecular robots to execute more
complex task sequencing. Reservoir computing uses a
fixed, pre-determined dynamic system referred to as a
reservoir to process input signals and has been applied to
various fields, including molecular computing, to execute
more complex task sequencing [25]. However, some is-
sues still need to be addressed in molecular robotics such
as the low energy efficiency, short lifetime, thermal denat-
uration etc. Researchers from the field of biology, math-
ematics, physics, and biocomputational engineering can
contribute to making the molecular robot system smarter,
sustainable, and more active.
Acknowledgments
The authors would like to acknowledge the JPNP20006 project,
commissioned by the Future AI and Robot Technology Research
and Development Project from the New Energy and Industrial
Technology Development Organization (NEDO), JSPS Grant-in-
Aid for Scientific Research on Innovative Areas “Molecular En-
gine” JP18H05423 and Grant-in-Aid for Scientific Research (A)
JP21H04434 for the funding support.
References:
[1] S. Bahar, “Flocking, Swarming, and Communicating,” in “The Es-
sential Tension,” The Frontiers Collection, Springer, Dordrecht,
pp. 127-152, 2018. https://doi.org/10.1007/978- 94-024-1054- 9 8
[2] M. Rubenstein, A. Cornejo, and R. Nagpal, “Programmable self-
assembly in a thousand-robot swarm,” Science, Vol.345, pp. 795-
799, 2014. https://doi.org/10.1126/science.1254295
[3] R. Vicerra, E. Dadios, A. Bandala, and L. Lim, “Swarm Robot
System for Underwater Communication Network,” J. Adv. Com-
put. Intell. Intell. Inform., Vol.18, No.5, pp. 769-775, 2014. https:
//doi.org/10.20965/jaciii.2014.p0769
[4] M. Schranz, M. Umlauft, M. Sende, and W. Elmenreich, “Swarm
Robotic Behaviors and Current Applications,” Front. Robot. AI,
Vol.7, Article No.36, 2020. https://doi.org/10.3389/frobt.2020.
00036
[5] J. Katuri, W. E. Uspal, M. N. Popescu, and S. S´anchez, “Infer-
ring non-equilibrium interactions from tracer response near con-
fined active Janus particles,” Sci. Adv., Vol.7, No.18, 2021. https:
//doi.org/10.1126/sciadv.abd0719
[6] P. GranˇciˇcandF.ˇ
Stˇep ´anek, “Swarming behavior of gradient-
responsive colloids with chemical signaling,” J. of Physical Chem-
istry B., Vol.117, No.26, pp. 8031-8038, 2013. https://doi.org/10.
1021/jp400234n
[7] B. Dai, J. Wang, Z. Xiong, X. Zhan, W. Dai, C.-C. Li, S.-P. Feng,
and J. Tang, “Programmable artificial phototactic microswimmer,”
Nat. Nanotechnol., Vol.11, pp. 1087-1092, 2016. https://doi.org/10.
1038/nnano.2016.187
[8] T. Vicsek, “Swarming microtubules,” Nature, Vol.483, pp. 411-412,
2012. https://doi.org/10.1038/483411a
[9] M. Akter, J. J. Keya, A. M. R. Kabir, H. Asanuma, K. Murayama,
K. Sada, and A. Kakugo, “Photo-regulated trajectories of glid-
ing microtubules conjugated with DNA,” Chem. Comm., Vol.57,
pp. 7953-7956, 2020. https://doi.org/10.1039/D0CC03124K
[10] J. J. Keya, R. Suzuki, A. M. R. Kabir, D. Inoue, H. Asanuma, K.
Sada, H. Hess, A. Kuzuya, and A. Kakugo, “DNA-assisted swarm
control in a biomolecular motor system,” Nat. Commun., Vol.9, Ar-
ticle No.453, 2018. https://doi.org/10.1038/s41467- 017-02778-5
[11] J. J. Keya, A. M. R. Kabir, D. Inoue, K. Sada, H. Hess, A.
Kuzuya, and A. Kakugo, “Control of swarming of molecular
robots,” Sci. Rep., Vol.8, Article No.11756, 2018. https://doi.org/
10.1038/s41598-018- 30187-1
[12] M. Akter, J. J. Keya, K. Kayano, A. M. R. Kabir, D. Inoue, H.
Hess, K. Sada, A. Kuzuya, H. Asanuma, and A. Kakugo, “Cooper-
ative cargo transportation by a swarm of molecular machines,” Sci.
Robot., Vol.7, Article No.eabm0677, 2022. https://doi.org/10.1126/
scirobotics.abm0677
[13] H. Hess and G. D. Bachand, “Biomolecular motors,” Materials To-
day, Vol.8, No.12, pp. 22-29, 2005. https://doi.org/10.1016/S1369-
7021(05)71286-4
[14] H. Hess, “Engineering applications of biomolecular motors,” Annu.
Rev. Biomed. Eng., Vol.13, pp. 429-450, 2011. https://doi.org/10.
1146/annurev-bioeng-071910-124644
[15] S. Hamada, “Molecular Robotics,” M. H. Ang, O. Khatib, and B.
Siciliano (Eds.), “Encyclopedia of Robotics,” Springer, Berlin, Hei-
delberg, 2021. https://doi.org/10.1007/978-3- 642- 41610- 1 189- 1
[16] Y. Sato, Y. Hiratsuka, I. Kawamata, S. Murata, and S. I. M. Nomura,
“Micrometer-sized molecular robot changes its shape in response
to signal molecules,” Sci. Robot., Vol.2, No.4, Artcle No.ieaal3735,
2017. https://doi.org/10.1126/scirobotics.aal3735
[17] H. Asanuma, X. Liang, H. Nishioka, D. Matsunaga, M. Liu, and M.
Komiyama, “Synthesis of azobenzene-tethered DNA for reversible
photo-regulation of DNA functions: hybridization and transcrip-
tion,” Nat. Protoc., Vol.2, No.1, pp. 203-212, 2007. https://doi.org/
10.1038/nprot.2006.465
[18] M. R. Rashid, C. Ganser, M. Akter, S. R. Nasrin, A. M. R. Kabir,
K. Sada, T. Uchihashi, and A. Kakugo, “3D structure of ring-
shaped microtubule swarms revealed by high-speed atomic force
microscopy,” Chem. Lett., Vol.52, No.2, pp. 100-104, 2023. https:
//doi.org/10.1246/cl.220491
1050 Journal of Robotics and Mechatronics Vol.35 No.4, 2023
Biomolecular Motor-Based Swarm Robot
[19] S. M. Fr ¨uh, D. Steuerwald, U. Simon, and V. Vogel, “Covalent
cargo loading to molecular shuttles via copper-free “click chem-
istry”,” Biomacromolecules, Vol.13, No.2, pp. 3908-3911, 2012.
https://doi.org/10.1021/bm301437c
[20] B. Nitzsche, F. Ruhnow, and S. Diez, “Quantum-dot-assisted char-
acterization of microtubule rotations during cargo transport,” Nat.
Nanotechnol., Vol.3, pp. 552-556, 2008. https://doi.org/10.1038/
nnano.2008.216
[21] C. Brunner, C. Wahnes, and V. Vogel, “Cargo pick-up from engi-
neered loading stations by kinesin driven molecular shuttles,” Lab
on a Chip, Vol.7, No.10, pp. 1263-1271, 2007. https://doi.org/10.
1039/B707301A
[22] Y. Yang and M. A. Bevan, “Cargo capture and transport by colloidal
swarms,” Sci Adv., Vol.6, No.4, Article No.eaay7679, 2020. https:
//doi.org/10.1126/sciadv.aay7679
[23] T. Yasuda and K. Ohkura, “Sharing Experience for Behavior Gen-
eration of Real Swarm Robot Systems Using Deep Reinforcement
Learning,” J. Robot. Mechatron., Vol.31, No.4, pp. 520-525, 2019.
https://doi.org/10.20965/jrm.2019.p0520
[24] Z. Zou, Y. Liu, Y.-N. Young, O. S. Pak, and A. C. H. Tsang, “Gait
switching and targeted navigation of microswimmers via deep rein-
forcement learning,” Commun. Phys., Vol.5, Article No.158, 2022.
https://doi.org/10.1038/s42005-022-00935- x
[25] T. Lymburn, S. D. Algar, M. Small, and T. J¨ungling, “Reservoir
computing with swarms,” Chaos: An Interdisciplinary J. of Nonlin-
ear Science, Vol.31, No.3, Article No.033121, 2021. https://doi.org/
10.1063/5.0039745
Name:
Mousumi Akter
ORCID:
0000-0002-1555-2444
Affiliation:
Postdoctoral Research Scholar, Institute of
Molecular Biology, University of Oregon
Address:
1229 University of Oregon, 1318 Franklin Blvd, Eugene 97403, USA
Brief Biographical History:
2017 Received M.S. degree in Chemistry from University of Dhaka
2017- Ph.D. Student, Hokkaido University
2020- Postdoctoral Researcher, Faculty of Science, Hokkaido University
2022- Postdoctoral Research Scholar, Institute of Molecular Biology,
University of Oregon
Main Works:
•“Cooperative cargo transportation by a swarm of molecular machines,”
Science Robotics, Vol.7, Issue 65, Article No.eabm0677, 2022.
•“Photo-regulated trajectories of gliding microtubules conjugated with
DNA,” Chemical Communications, Vol.56, Issue 57, pp. 7953-7956, 2020.
Membership in Academic Societies:
•Biophysical Society (BPS)
•The Biophysical Society of Japan (BSJ)
•The Society of Polymer Science, Japan (SPSJ)
Name:
Akira Kakugo
ORCID:
0000-0002-1591-867X
Affiliation:
Division of Physics and Astronomy, Graduate
School of Science, Kyoto University
Address:
Kitashirakawa Oiwake-cho, Sakyo-ku, Kyoto 606-8502, Japan
Brief Biographical History:
2003 Received Doctor of Science from Hokkaido University
2003- Assistant Professor, Hokkaido University
2011- Associate Professor, Graduate School of Science, Hokkaido
University
2022- Professor, Department of Physics and Astronomy, Graduate School
of Science, Kyoto University
Main Works:
•Active matter
•Young Researcher Award for Science and Technology Minister of
Education, Culture, Sports, Science and Technology (2012)
•Academic Award for Polymer Science (2016)
•HFSP (Human Frontier Science Program) Research Grants Award (2021)
Membership in Academic Societies:
•The Biophysical Society of Japan (BSJ)
•The Society of Polymer Science, Japan (SPSJ)
Journal of Robotics and Mechatronics Vol.35 No.4, 2023 1051
