Team:SUSTech Shenzhen/Hardware

Load Tools

Overview

To support the manipulating model of light-controllable mammalian cell secretion, the quality and quantity of experiment data are vital for it limits how accurate and reliable then control could be. And this quality and quantity can be improved by developing related hardware setups. The Automatic Illumination, Culturing, and Collection System (AICCS) is so designed and built up, to apply high quality of light and maintain cell situation as well as avoiding disturbance caused by operators and operations.

Automatic Illumination, Culturing, and Collection System (AICCS)

What's the huge trouble a team studying mammalian cells with influence with light would face? For the SUSTech-Shenzhen-2019, the problem comes from our extremely time-consuming and repeating illumination experiment. Maybe with a light bulb and a dark room, you can give light on cells in a Petri dish.

But when demanding for illumination of 60~150 hours under dozens of different light conditions, things are quite different.

How do you ensure that the light is uniform on all cells and stable for a long time, or even change in need every second?
How do you maintain the cells’ healthy situation so the secretion data is reliable?
How do you collect cell secretions to analyze from different time points without much disturbance on cell and who can keep focus when it lasted 60 hours?

That’s why we spend our time designing and building up the Automatic Illumination, Culturing, and Collection System (AICCS) which fulfilled all our requirements above. The AICCS consists of three compartments, including devices we designed and used and programs to automate them. They are

Illumination Compartment that applies high quality of changeable light;
Culturing Compartment where cells are cultured with continuously refreshed medium and being illuminated;
Collecting Compartment that stores the effluent medium which has cell secretions.

The three compartments can work individually for different purposes or composes quite as the AICCS. Here’s how it works:

The three compartments are connected by tiny PE tubes, where mediums flow. The fresh medium flows into the incubator and reaches the culturing plate which is under the illumination compartment. Then old medium with cell secretion is pushed out of the culturing plate and transported to the collection compartments where they are frozen on the cooling surface.

Illumination Compartments

To apply uniform, stable, and changeable light signals to cells, we designed and built up the illumination compartment. The compartment is designed to give light on cells seeded in 24-well plates. Light sources are controlled individually and the wells are illuminated respectively. This compartment can be coupled with a culturing compartment, or work with a commercialized μ-plate 24-well plate.

Control Light Intensity individually

We give out 465nm blue light with commercialized LED bulbs (cost less than 10 cents per piece). The LED bulbs are soldered on a customized printed circuit board (PCB). The light intensity of bulbs can be controlled individually by applying different currents to different bulbs through the 24 circuits of the PCB.

The light intensities of LED bulbs are controlled by a constant current source, so a programmable constant current source is used to control the illumination. The self-written python program commanding the current source will read CSV files whose columns represent current supported to each channel, and each row gives signal within a time interval of 1~120s.

Calibration

In the incubator, humidity and temperature accelerate units’ aging. If so, the input signal should be changed frequently since the resistance of units may change the next second. To minimize the influence of aging, we use constant current sources. But the PCB and bulbs still require replacement every 3 months. Luckily, one set of light sources including 24 LED bulbs and one PCB only takes 3 dollars. However, when light sources are changed, their electrical characteristics will change.
How we know the exact intensity that current is corresponding to when bulbs and PCBs are changed. The answer is to calibrate the input data and find out the new relation between input current and output intensity. We use a camera lens and a self-written image analyzing program in MATLAB, and a commercial silicon photometer. We photographed the output light as in below figure three times, with different exposure times, under a series of the current input. Images are converted into grayscale value data with the scale of the pixel, then three photos’ data under the same input are averaged.

Each well’s grayscale average value at different input is fitted into a relative curve. Then we use a commercial photometer to search and record the intensity of the brightest light spot in each well at a certain current. We compared the sum of the brightest pixels that add up to the space of the photometer’s photosensitive element and the intensity recorded, normalize the relative curve.

Wells Illuminated Respectively

To avoid leakage of light from other wells, a surface blackened aluminum Light hole block with 24 dark tubes is used within the bulbs and the cells. We also used a commercialized μ-plate 24-well block to distribute light between wells. We also added a water cool-down plate to cool down the LED bulbs which warm up during illumination and maintain cell culture medium temperature which is heated by the blue light.

In the figure above, the blue light plate is the self-made LED PCB mentioned above. It is assembled with other components through a self-designed aluminum frame. The thermostat plate cools down the LED plate and cell culture medium with the same water flow provided by a water pump inside the incubator. By lowing down the incubator’s culturing temperature by 0.5~1 degrees Celsius, the thermostat plate can maintain the culture medium at 37. The LED plate which provides parallel light beams and the light hole block can project 24 round spots on the 24-well plate, and the μ-plate 24-well block will prevent light influences from other wells.

We also designed a special intrigue to make the cell illumination operation easier.

Previously, we have to disassemble the whole illumination system frame to insert and detach the 24-well block, so we added this little intrigue to make it easier. The intrigue has two levels. With springs, the meable thermal conduct supporter can be pressed up or kept down as shown in the above figures. When it’s down, inserting the 24-well block is easy. Then change the level and the supporter is pressed back to keep the block at the location where 24 light beams can project suitably at wells bottom where cell lies.

Here's an example of cell illumination experiment using the illumination compartment and 24-well block:

The experiment used illumination compartment to study the relationship between illumination duration and the light-on system’s reaction.

Culturing compartment

While going through the cell illumination experiments that last for 60~150 hours and requires tens of medium obtaining in between, we noticed the difficulties of maintaining cell situation and avoiding operators’ disturbance due to personnel change and attention decrease. Considering the model’s demands for continuous and large quantities of data from repeating experiments, we decided to build up an auto cell culturing device, with common materials found in Biology laboratories. This device allows us to conduct hundreds of hours of experiments, with prepared cells and culture medium.

Medium Applying

The fresh medium and the refreshing force are applied by a commercialized syringe pump. It provides continuous but slow flow (e.g., 3μL/min).

Automatic Culturing Plate (ACP)

To ensure the refreshment of the medium in cell culture, we made an airtight culturing plate (ACP) with Polydimethylsiloxane(PDMS) and slide. PDMS is a commonly used material in making a microfluid chip, and the slide is used to hold samples for microscope observation. We made a PDMS shaping mold by attaching a cut acrylic plate and slide with UV glue. (A & B). Hollows are produced on the PDMS surface by mold which forms airtight lumens with the slide. The PDMS and the slide are bounded by plasma-activated bounding. (D)

When the medium comes to the bounded plate, the old medium will be pushed out of the lumen and new medium comes in. To seed cells on the slide surface, the lumen is treated with gelatin, which is derived from animal body parts.

To change the culture plate from 24-well block to our automatic culturing plate (ACP), a 3D-printed part is paired with the meable thermo conduct plate.

Here's an example of cell culturing using the automatic culturing plate (ACP):

It’s an image of our Hela cells cultured in the automatic culturing plate (ACP) 24 hours after seeded. Comparing with cells seeded on the commercial 24-well plate, cell attachment rate is close.

Collection Compartment

Our Illumination Compartment can automatically illuminate and culture cells for a long time now, but one flaw remains. When the culturing compartment has mediums refreshed continuously, the effluents still requires collection, for it contains cell secretions we need to measure. But the reporter proteins will not stand tens of hours at room temperature. We need a refrigerator here. But as mentioned above, the compartments are connected through PE tubes. If we put a tube directly into the refrigerators, the end of the tube would be blocked by frozen fluids. We also need a method to distinguish cell secretions from the same culturing well but different time intervals. So we introduced the semiconductor refrigerator to storage effluents, and three-axis stepper motor to store effluents respectively.

Semiconductor Refrigerator

We use a semiconductor refrigerator to preserve the protein solution. The main part of the refrigerator is a semiconductor chilling plate, which works typically at 60W. The plate is made from metal. The surface was covered by glass, which is bonded by conductive silicone grease with thermal conductivity at 8.5 W/(m·K). A water-fan cooling device is inside the refrigerator to cool down the hot surface of the chilling plate. The cool surface can be maintained at temperature -20℃ ~ -10℃, which can satisfy for our project’s need.

Three-axis Stepper Motor

Base on the requirement of moving material accurately, we bought a 3D move platform. It comes with a very simple control system and some modules (Fig.A) for driving step-motor. The old control system is a black box, it's hard to use and lacking customizability. After it is broken, we decided to build our own control system.
We replace the driving module to DRV8825 (Fig.D), it's small physical size but still can drive up to 2A current with 32 subdivision steps. We used Arduino Mega 2560 for the control center and connect parts with a hand made circuit board (Fig.B). We also add a small OLED screen and a rotary encoder for more friendly control (Fig.B). We use CSV file to store the pre-defined control command then read and decode the file from a removable Micro SD card(Fig.C). The pre-defined commands include move command, delay command and nestable loop control command. It's total cost less than 15 US dollar which is only one-tenth of the original control and driving system.
The reason we not used those famous open-source control system such as Marlin is they are built for G-code, it's to complex to define some simple movement and also lack of some customizability.

Here's an example of effluents frozen on the cooling plate of collection compartment:

Summary

Light from the illumination compartment is uniform, stable, and controllable. It irradiates the airtight culturing plate (ACP) where cells are seeded. Mediums in ACP are refreshed continuously, push’s out the cell secretions to the collection compartment. Three-axis Stepper Motor moves the end of the PE tube so that effluents are frozen at different locations on the semiconductor refrigerator’s frozen surface. All the three compartments make our Automatic Illumination, Culturing, and Collection System (AICCP).