BIOMOD
Hokkaido University



Microscale Simulator
~ for human crowded motion ~

Abstract



 It has been known that human crowds often cause troubles thus controlling human collective motion has been one of the major goals to be achieved. A number of simulation theory and practices have been considered. One of them is to set pillars as “obstacles” in a human path which can make smooth people flow.
 Our project aim is to simulate human collective motion in a real dimension, not in virtual world. We reproduced human crowd situation in nano-scale by using gliding assay of motor-proteins and microtubules. By changing the arrangement of pillars, various types of microtubule flows could be seen.
 These results will be used for more practical simulation done by making microtubules’ movement similar to human one.

Intro


 In daily lives, we often face troubles caused by congestion and crowding. Crowding phenomena not only cost us much time, decrease the efficiency but also can become a critical problem in emergency situation. Japan is a country which has many earthquakes thus it is an important problem to find a solution to eliminate congestion and make a mass flows smoothly.
 Until today, simulations for reducing congestion have been done mainly by using software from standpoints of architecture and psychology. However, simulations in real dimension have not been done so many times because they would cost a lot of time and money.
 Cellular automaton is a representative software simulation model. It consists of many grid cells and simple rules and calculate collective behaviors to some extent. We got an idea from this simulation model.

 In this project, in order to observe characteristics of various types of collective behaviors and find a way to solve problems in congestion, we constructed a simulation model from a new standpoint which uses microtubules and proteins called molecular motors.
 Molecular motors are molecular machines which drive by converting chemical energy of ATP into kinetic energy. Molecular motors include myosins and kinesins which move above actins and microtubules respectively. In vivo these nano-scale molecular motors gather together and dynamics such as cell division and transportation occur.


background of our study


・Microtubule
 Microtubules are filamentous and cylindrical biopolymers of the heterodimer of tubulin. They are present in virtually all eukaryotic cells as a major component of an interconnected network of filaments known as the cytoskeleton. They perform various functions not only in cell motility, cell division, organization and orientation but also in transporting organelles as rail for motor proteins.


・Kinesin
 Kinesin is a motor protein found in eukaryotic cells. It moves along microtubule filaments fueled by the hydrolysis of ATP. It helps transport cellular cargoes and support several cellular functions including mitosis, meiosis and cytokinesis.


・Design for simulator
 Before making a microscale simulator, we had to find an idea or system to solve human crowded problem. Therefore, we focused one of a simulation model called “cellular automaton” and its simulation. Cellular automaton is a discrete model consisting with grid cells and some simple rules. Each cells changes their conditions by the time change and the interaction among surrounding cells.

 Floor Field model is a one of the cellular automaton model. This model is a two-dimensional model of cellular automaton and used for simulation such as congestion. It can be applied to simulations of evacuation from a building. Some studies discovered that obstacles in front of an exit can be effective for evacuation by using this model. This time, we adopted this result for smooth evacuation in our simulator.

Project


 Our goal is to reproduce crowded motion of human in micro-scale and simulate how people behave when they are crowded. Eventually, we want to apply the result to solving the congestion in the real world. In order to achieve this goal, we need a practical and reasonable simulator such as micro-scale environment and objects which can compare to a building and human beings. The detailed steps are introduced as follows;

 1. To reproduce human crowded motion by using gliding assay of motor-proteins and microtubules.

 2. To make micro-scale environment for the simulator by nanofabrication technique.

 3. To simulate a crowded situation and improve the precision of the simulator.





Design



 To achieve our goal, we combined gliding assay and nanofabrication for a simulator.

 1. Gliding Assay











 2. Microfabrication








 We used an experiment method called gliding assay to imitate human collective motion. In a gliding assay, motor-proteins such as kinesin are adsorbed to the glass plate and are immobilized. When microtubules are put on the surface of the plate, they are captured by the bound motors. Then, they are transported along the surface in a ‘gliding’ movement by adding ATPs to motors. Microtubules can interact each other when tagged some proteins or changing solvent.

 We created attractive interaction among microtubules by depletion force to resemble microtubules’ motion to human one. Furthermore, there was a big merit that we can observe phenomenon in the real world like human collective motion when we used gliding assay.






























 We used microfabrication technique to observe how pillars affect a microtubules’ motion. Microfabrication is a technique to make a tiny structure using chemical reactions or physical characters instead of processing by machine, which is difficult. Semiconductor manufacturing is a good example of microfabrication, and a method called lithography is utilized in this process. In our study, we made micro pillars by lithography and transcription technique.

 We designed any shape and arrangement of pillars by drawing software. Then a special laser processor read the design from the software and made a chrome mask. Ultraviolet rays are irradiated to a photo-resist-covered glass plate through the chrome mask, and a mold of photo-resist are developed. After that, PDMS(polydimethylsiloxane) is poured into the mold and cooled it down, and finally micro-pillars are made as you designed.




 Combining these two techniques, we succeeded to make a simple simulator.

experiments&results



Experiment 1

 Gliding assay (1st step for reproduction of collective motion)

Aim
 To investigate the effect of depletion force caused by methylcellulose on the behavior of microtubules

Materials
・Anti-GFP antibody (0.2 mg/mL)
・Motility buffer (80 mM PIPES, 1 mM EGTA, 2 mM MgCl2, 0.5 mg mL−1 casein, 1 mM DTT, 10 μM paclitaxel and ∼1% DMSO; pH 6.8)
・GFP-tagged kinesin solution (consisting of the first 560 amino acids of human kinesin1,80 mM PIPES, 1 mM EGTA, 2 mM MgCl2, 0.5 mg/mL casein, 1 mM DTT, 10 μM paclitaxel/DMSO, ∼1% DMSO; pH 6.8)
・Rhodamine-labelled tubulin solution (0.2 µM tubulin and 1 mM GTP.)
・ATP buffer (∼80 mM PIPES, 1 mM EGTA, 2 mM MgCl2, 0.5 mg/mL casein, 1 mM DTT, 10 μM paclitaxel/DMSO, ∼1% DMSO; pH 6.8)
・MC (0.1wt% or 0.3 wt% Methylcellulose)
・MAP4 (0.25 μM Microtubule-associated protein 4)

Methods
①Flow cells were assembled from two cover glasses, and a double-sided tape was used as the spacer.
②2 μL of anti-GFP antibody was applied to the flow cell. After incubation for 3 min, the flow cell was washed with 5 μL of motility buffer.
③Then, 3 μL of a GFP-tagged kinesin solution was introduced and incubated for 3 min to bind the kinesin to the antibody.
④The flow cell was washed with 5 μL of motility buffer.
⑤3 μL of microtubule solution of a prescribed concentration was introduced and incubated for 3 min, followed by washing with 5 μL of motility buffer 2 times.
⑥MC, 0.25 μM MAP4 and 10 μL of 10 mM ATP buffer was added to the flow cell.
⑦The samples were visualized with a fluorescence microscope using an oil-coupled Plan Apo 60 × 1.40 objective.

Result
 We observed a depletion force among microtubules by adding different concentrations of MC to a solution of microtubules. In lower density of microtubules, prepared from 200 nM tubulin, were suspended in 0.0, 0.1 and 0.3 wt% MC solution and observed under a fluorescence microscope. At the concentration of 0.1 and 0.3wt% of the MC solution, microtubules aggregated because of the depletion force. On the other hand, microtubules moved separately and showed little interaction each other in the absence of MC (0.0 wt%).




Fig.1 Fluorescence microscopy images of microtubules which were suspended in solution (left) without MC and (center and right) with MC of 0.1 and 0.3 wt%. Scale bar: 10 µm


Discussion
 This result indicates that the depletion force induced by MC causes the attractive interaction among the microtubules. We added 0.3% MC to the solution of microtubules in following experiments of gliding assay because we could observe aggregation the most in this condition.





Experiment 2

 Gliding assay (2nd step for reproduction of collective motion)

 Microtubules move separately and there is little interaction among them when there are few microtubules in a flow cell. We assume that the density of microtubules is a key element for the appearance of collective motion.

Aim
 To see the difference of collective motion whether the density of microtubules is high or low.

Materials
・Anti-GFP antibody
・Motility buffer
・GFP-tagged kinesin solution
・Rhodamine-labelled tubulin solution (Low concentration : 0.2 µM, High concentration : 5.0 µM in the presence of 1 mM GTP.)
・ATP buffer
・MC (0.3 wt% Methylcellulose)
・MAP4

Methods
①Flow cells were assembled from two cover glasses, and a double-sided tape was used as the spacer.
②2 μL of anti-GFP antibody was applied to the flow cell. After incubation for 3 min, the flow cell was washed with 5 μL of motility buffer.
③Then, 3 μL of a GFP-tagged kinesin solution was introduced and incubated for 3 min to bind the kinesin to the antibody.
④The flow cell was washed with 5 μL of motility buffer.
⑤3 μL of microtubule solution of a prescribed concentration was introduced and incubated for 3 min, followed by washing with 5 μL of motility buffer 2 times.
⑥0.3 wt% MC, 0.25 μM MAP4 and 10 μL of 10 mM ATP buffer was added to the flow cell.
⑦The samples were visualized with a fluorescence microscope using an oil-coupled Plan Apo 60 × 1.40 objective.

Result
 In order to find out how a density of microtubules affect their collective behavior, we prepared two types of tubulin solution which concentration were 0.2 and 5.0 µM. In high density, microtubules moved randomly but showed collective motion which made large streams at last. However, coordinated behavior was not observed in low concentration tubulin solution.




Fig.2 Fluorescence microscopy images of microtubules at low (left) and high (right) density. Scale bar: 50 µm



Discussion
 In low density, microtubules move more isotropically, that means they move separately.  And in high density, we can see nematic phase, which means they move as a group. Compared to the one with low density, now microtubules move more as a group making some roads. We successfully reproduce crowding situation in micro scale

*Experiment 1&2 were done on a flowcell (2×5×0.15mm3 W×L×H) which consists of double-sided tape and a cover glass.







Experiment 3

 Microfabrication (Flow cell design)

Aim
 Design flow cells by KLayout to estimate the optimum placement.

Methods
 We designed 4 groups of pillars which each group consisted with circle, two types of triangle, and quadrangle pillars. This time, we used a software named KLayout for design.
 A size of whole cell was 2000㎛×2000㎛, and a size of each pillar group was 7000×7000㎛, and distances between the groups were 2000㎛.
 The whole image of flow cell is as below;




 Each group consisted with 5 areas which distances among pillars were 100㎛~500㎛. The whole image of each group is as below (e.g. ❶circle);





 In each area, we arranged pillars, which diameters were 10~50㎛, in regular interval as following figures;








 We output the blueprint that designed by Klayout and ordered this flow cell from Nanotechnology Platform in University of Tokyo.


Results(microscopic observation)
 The flow cells were visualized with an optical microscope as following images.





Quadrangle pillars




Triangle pillars




Circle pillars





Experiment 4

 Hybrid experiment (Gliding assay on the flow cell we designed)

Aim
 To confirm whether pillars can control collective motion of microtubules

Materials
・PDMS (Dimethylpolysiloxane as flow cell)
・Anti-GFP antibody
・Motility buffer
・GFP-tagged kinesin solution
・Rhodamine-labelled tubulin solution (5.0 µM in the presence of 1 mM GTP)
・ATP buffer
・MC (0.3 wt% Methylcellulose)
・MAP4

Methods
①2 μL of anti-GFP antibody was applied to the flow cell. After incubation for 3 min, the flow cell was washed with 5 μL of motility buffer.
②Then, 3 μL of a GFP-tagged kinesin solution was introduced and incubated for 3 min to bind the kinesin to the antibody.
③The flow cell was washed with 5 μL of motility buffer.
④3 μL of microtubule solution was introduced and incubated for 3 min, followed by washing with 5 μL of motility buffer 2 times.
⑤0.3 wt% MC, 0.25 μM MAP4 and 10 μL of 10 mM ATP buffer was added to the flow cell.
⑥The samples were visualized with an fluorescence microscope using an oil-coupled Plan Apo 60 × 1.40 objective.

Result
 The results of gliding assay using a flowcell made of PDMS are following images. Collective motion was observed as in experiment 2. Around pillars microtubules showed a motion like avoiding obstacles. It was observed that specific distances of pillars made more clear streams.









Fig.3 Fluorescence microscopy images of microtubules in PDMS flowcell.



Discussion
 We succeeded in reproduction of the flow of a human crowd on a flowcell. It became clear that whether a stream close to human crowd appears depends on a distance between each pillar. From the result of this experiment, pillars at a distance of about 40㎛ are thought to be appropriate to reproduce a clear stream.


Future



 We believe that our work is just the beginning of an extra-ordinary project. Our simulator will become more practical and versatile one by achieving these goals.



  < Short-term Goal >

・Establishment an evacuation model
 In order to achieve this goal, we designed a further studies as follows:





  < Middle-term Goal >

・For more practical model
 In the real society, each person behaves and moves around with his/her own will. Therefore, some commands or rules are needed to microtubules and environment in order to reproduce human behavior in microscale. As we showed in our Project page, this study is formed by two factors, reproduction of people and environment. We are considering further arrangement for each as follows:

  ○ Modify microtubules using some proteins or DNA to give rules for their motion.
  ○ Make chemical stimuli and create a origin of panic as an environmental effector.


  < Long-term Goal >

・Application to a wide range of field
 We can design not only human evacuation model but a different kind of models for solution of traffic jam or effective transportation network. Previous research showed that a slime mold solve a maze and find out an optimized way. Our simulator may be used as a microscale biological model which can solve optimization problems like a slime mold.



TEAM MEMBER


member

Kento Nishii

School of Agriculture 2nd

Takahiro Yamamoto

School of Science 3rd

Miyuki Oka

School of Agriculture 2nd

Yuto Maeda

School of Agriculture 2nd

Ryo Shirakawa

School of Engineering 2nd

Tomoki Fukushima

School of Engineering 2nd

Haruki Otoi

School of Engineering 1st














advisor

Shinji Yamada

School of Science 3rd

Masaru Sakuma

School of Engineering 4th

Hiroki Sakai

School of Engineering 3rd







teacher

Akira Kakugo

Associate Professor







mentor

Arif Md. Rashedul Kabir

Post Doctor

Bulbul Mahmot

Doctor 1st