Introduction

Liquid-processing systems technology is involved in various fields of life sciences. Automated liquid handling systems are used day by day to develop protocols and for reliability in laboratory fields such as synthetic biology, microbiology, or genetics. LEGO Mindstorms, which is widely used in the creation of complex systems and offers a wide range of applications thanks to its openness to designing various codes for creating protocols and experimental setups, offers opportunities such as designing machine prototypes in the field of robotics1.

It has been reported in the literature that liquid-processing LEGO robotic systems can be used in biotechnology experiments2. In the future, these studies are thought to be useful for complex drug delivery experiments and many other applications in emerging fields such as regenerative medicine and tissue engineering, and biotechnological applications3. In addition to the high costs of fully automatic systems, they are also quite expensive to maintain4. It is beneficial to have systems that are affordable and easy to install. Furthermore, there is no reasonably priced robotic system that can cost-effectively apply molecular biology and genetic editing techniques to perform experiments in this specific field and meet the precision and programmability requirements for the sustainability of experiments in this field. These advantages enable the creation of a prototype-specific system that is much more cost-effective than commercially available systems, allowing the ability to add and remove parts to obtain a variety of shapes suitable for different experimental setups. With this study, machine learning, transgenic cell screening, analytical calculations, gene designs, genetic pathway selection, experiment design, cell cloning and genetic pathway analysis can be performed with a minimum margin of error by minimizing human intervention by developing remote and autonomous robotic systems.

LEGO Mindstorms robotic systems are a DIY (Do It Yourself) version that is automated, low-cost, easily programmable, and with a wide range of applications. As a result of the experiments, LEGO-based robotic devices were tested with various applications, and complex flow steps were carried out for protein purification, as well as for use as a feeding system for bacterial cultures, and uniform spherical microparticle production3. As another example, for adaptive radiotherapy (ART), it is possible to simulate the behaviors during therapy by making a humanoid ghost in which respiratory processes are simulated, by testing patients’ breaths and computerized tomography study analyses in clinical practice, and by re-establishing real patient conditions with the help of LEGO robot5. LEGO Mindstorms robots offer a cost-effective system that can provide the same experimental efficiency as autonomous robotic systems, as well as a programmable and remote-control system. It has been demonstrated that liquid-processing robotic systems like the CRISPR.BOT has the potential to be used in biotechnology experiments and laboratory routines2. We have built the prototype of an innovative technology that is cost-effective, practical, and free from human error in bioindustry applications. In our first experimental study, in which we showed all experimental stages of bacterial DNA transformation, production of transgenic human cells with lentiviruses, and CRISPR gene-engineered cell development with developed algorithms. The study suggests how the combination of autonomous technologies can accelerate genetic modification. By performing the first experiments with a programmed robotic system that can process liquids with remote control, transgenic bacterial or human cells that can express green fluorescent protein were created by performing genetic transfer in bacteria or human cells in an automated manner.

Results

Development of CRISPR.BOT liquid handling system

Construction of CRISPR.BOT autonomous and remotely controllable robotic system

The LEGO digital designer application was used to construct the CRISPR.BOT V1 robotic system with LEGO Mindstorms Education EV3 and the first entire installation benefitted from the study published by Gerber LC et al. in 20172. However, CRISPR.BOT V2 installation stages were completed from the beginning using the LEGO Digital Designer program as reported in the literature2. The first version of CRISPR.BOT (V1) was constructed by building a rail system using the LEGO Digital Designer program and the LEGO Mindstorms set (Fig. 1A, B). Next, the rail system was expanded to carry 10 cm plates (Fig. 1D and Supplementary Video 1). The rail was expanded in version 2 with the addition of components to allow the plates, Petri dishes, and cuvettes used in the planned experiments to be placed in a more robust and wider area compared to the system arrangement in version 1 (Fig. 1C). A wide area was created by increasing the rail system from one row to three rows (Fig. 1E). There is one color sensor in the CRISPR.BOT V1 robotic system. This color sensor in the LEGO Mindstorms EV3 set; can detect 8 different colors and measure the values of ambient light or reflected light. The primary purpose of the experiments with the color sensor was to ensure that the color sensor stops at the desired colors. By providing the color sensor integration with the syringe, the colored papers on the cuvettes stopped with the detection of the color sensor, and the syringe system provided the desired movement. Colored papers were stuck in the middle of the tubs for the color sensor to recognize (Fig. 1E). For the V1 robotic system, the syringe plunger was cut by the system in the robot and fixed to the LEGO piece with the help of glue and connected to the syringe (Fig. 1F). This work has enabled the study of micropipettes in CRISPR.BOT to experiment with the handling of microliter-sized liquids.

Fig. 1
figure 1

CRISPR.BOT V1 setup steps. (A) Installation of the rail system for CRISPR.BOT V1. (B) The assembled version of the V1 robotic system using the LEGO digital designer program and the spare parts in the LEGO Mindstorms set. (C) Expansion of the rail system. (D) By expanding the bottom area of the rail system, a wider sub-floor is provided for the materials to be used in the designed experiments. (E) There are integrated colored papers for the plate carrier rail system and color sensor. When the color sensor encounters paper of the color specified in the coding, it activates the stop command to stop in the current position. (F) The position indicating that the syringe is aligned with the well center as determined by the coding.

Full size image

Blocks have been prepared for use in experiments that will be conducted using the steps for creating blocks in the LEGO Mindstorms EV3 robotic coding program. The purpose of these blocks is to perform the commands in the blocks, respectively, when each block represents a movement and is added below the command that initiates the movement (Fig. 2A–C). Every movement of the robot is coded to process liquids in microliter volume with velocity and angular measurements (Fig. 2D, E). The degree reset block was aimed, it was aimed to reset the calibration of the robot before each experiment and to perform its movements more accurately. To ensure that the syringe draws the same amount each time in the experiments to be carried out, different experiments were made on the programming and the liquids were measured with the help of a pipette after each liquid draw. As can be seen from (Fig. 2D), because of the velocity-microliter measurements, it was observed that the velocity change did not change the measured microliter amount. As a result of the measurements, it was observed that the CRISPR.BOT robotic system can draw low microliter volumes of liquid, ranging from 0 to 40 µl, depending on the rotation angle of the motor controlling the syringe plunger, which operates between 1 and 10° (Fig. 2E). However, deviations in liquid withdrawal were observed, particularly at lower angles between 1 and 5°. After many trials and finding the right programming, experiments were carried out at different speeds and at different angles to understand the connection between velocity and measurement. Also, a chart was drawn up for each and it was determined that the different velocities do not affect the volume to any constant degree (Fig. 2D, E). By determining which angle to use in experimental programming from the graphics, its programming was arranged, and the first bacterial transformation experiment was carried out with CRISPR.BOT in the following experiment.

Fig. 2
figure 2

Blocks used in experiment programming and Power constant (= 300), speed-microliter table. (A) Liquid aspirate: It fulfills the liquid intake task of the pipette system. With the 40° movement of the large servo motor that provides the movement of the pipette system, the piston connected to the syringe goes up and aspirating liquid. (B) Liquid dispense: Performs the liquid dispensing function of the pipette system. It performs the reverse of the liquid aspiration process, allowing the liquid to be taken back. (C) Pipette movements: They are the blocks that perform the up and down movements of the pipette system. The power and speed are kept low so that the liquid enters slowly and precisely. (D) At a constant power of 300, microliter measurements were quantified at different speeds of Liquid Aspirate, Liquid Dispense, and Pipette Movements for 10, 20, 30, and 40°, and it was seen that there was not much difference in the withdrawn microliter even though the speed was increased. (E) Graph showing the relationship between pipetted volume and servo motor angle (1–10°) with a constant pipette clock speed (25) and power (300). Error bars represent the standard deviation of n > 6 replicates for each measurement. Statistical significance was assessed using an unpaired t-test; *p < 0.05 indicates significance, and non-significant results are labeled as NS.

Full size image

For the new version installed, the system mechanism has been completely changed and new pipette movements have been added to the robotic system. While the pipette mechanism can only move up and down in CRISPR.BOT V1, right and left movements can be provided in the installed CRISPR.BOT V2. With a rail system under the CRISPR.BOT V1 pipette, movement is provided and the well on the well plate to be processed is brought to the level of the pipette, while with CRISPR.BOT V2, the pipette can be taken to the desired location with its rail system. The kits in the LEGO Mindstorms series have the software and necessary hardware to create a programmable customizable robotic system5.The LEGO Mindstorms EV3 Home Edition software was downloaded from the LEGO website and certain functions were added to the robots (Fig. 2A–C). This study aimed to test the experimental approaches by providing the robot with the ability to process liquids with the help of LEGO digital designer and LEGO software2,6. The wheel at the top left in the image is connected to a servo motor and alone performs the task of pulling and releasing liquid. We call the syringe system and the mechanism linked to Fig. 3A, which can only pull and release fluid by fixing the syringe plunger, when placed inside Fig. 3B and connections are made, it has two more motion systems. The first of these provides up and down movements of the pipette tip so that it can enter the liquid. There is also a part that serves to hold the syringe system in a fixed position (Fig. 3C). On the other hand, the wheels at the four corners of the machine act as wheels and move on the gears and can move on the robotic system by being placed on the system on the frame. To provide the piston movement connected to the syringe, the piston must be connected to a LEGO piece and combined with the wheels and provide up-and-down movement (Fig. 3D, E). In the CRISPR.BOT V2 robotic system, there are 3 different motors integrated into the pipette module. Two of them are large servo motors and one is a medium servo motor. One of the large servo motors provides the movement of the robot on the frame. Another large servo motor provides the movement of the pipette system, which provides the liquid exchange process (Fig. 3F). With the operation of the medium servo motor, the movement of the impellers allows the pipette, which is integrated into a rail mechanism, to enter the liquid, providing up and down movements (Fig. 3G). The movement of the wheels is provided by the toothed rail plates and the upward and downward movement of the straw is provided (Fig. 3H). The pipette system, which performs pipette movements and liquid exchange operations, is placed on the frame. The movements are carried out by the gear parts on the frame and the wheels at the bottom of the pipette system. The frame is fixed and does not have any mobility (Fig. 3I, J). It is the mechanism that provides the movement of the pipette system to the right and left and back and forth movements on the frame (Fig. 3K, L). For the new version that was installed, the system mechanism was completely changed, and new pipette movements were added to the robotic system. While the pipette mechanism provided only up and down movements in the previous version, right and left movements can be provided in the new version that has been installed. In the previous version, while the well on the well plate to be processed was brought to the level of the pipette by a rail system under the pipette, with this version, the pipette can be taken to the desired location with its rail system (Fig. 3M). In the speed measurement optimization phase of CRISPR.BOT V1 version, many trials were made on power and speed in its programming, and since it was seen that the use of high power was more effective, the optimization studies of CRISPR.BOT V2 was continued over this and the power and speed were kept constant at high degrees. For the 6-well plate, 12-well plate, and 96-well plate planned to be used in the experiments, separate algorithms were required. First, the rail movement was divided into two. The first stage is the rail movement of the mechanism that takes the straw on the frame to the desired position, that is, the back-and-forth movements. In this, separate algorithms were performed for each cell culture plate.

Fig. 3
figure 3

CRISPR.BOT V2 setup steps. (A) It is the mechanism that provides the movement of the syringe and is placed in the middle part of (B) The movement of the wheels is provided by toothed rail plates and the up-and-down movement of the straw is provided. (C) It is a part that keeps the syringe connected to the system. (D) Alternative parts are used to provide the syringe movement connected to the plunger. By placing the syringe plunger inside the red part, the movement of the part connected to the piston is provided by the movement of the impellers connected to the servo motor. With this system, the movement of the piston up and down is provided, and the fluid exchange process is carried out. (E) The prepared mechanism is placed in the pipette system. (F) Integration of the CRISPR.BOT pipetting system into the servo motor. The movement of the impeller connected to the large servo motor on the rail plate provides the movement of the pipette system, which performs the liquid exchange process. (G) Side view of the pipette system. The up-down movements of the pipette system are provided by the medium servo motor moving the wheels. (H) The gear piece enables the wheels to move on it. (I) CRISPR.BOT is the bottom part of the robotic system, called the frame. Image of the frame setup made from the LEGO digital designer program. (J) It is the bottom part of the robotic system, called the frame. The pipette system, which performs pipette movements and liquid exchange operations, is placed on the frame. Movements take place thanks to the toothed parts on the frame and the wheels at the bottom of the pipette system. The frame is fixed and does not have any mobility. (K) The system provides the liquid transition movements of the pipette between the other wells. The mechanism that provides the movement of the pipette system on the frame, was designed with the LEGO digital designer program. (L) It is the mechanism that enables the pipette system to move right, left, forward, and backward on the frame. The pipette provides fluid passage movements between other wells. (M) Pipette system that provides the right and left movements of the pipette.

Full size image

Development of algorithms per plates used in molecular biology experiments

CRISPR.BOT V2 autonomous robotic system can perform molecular biology and gene engineering experiments in a more automated, fast, and practical way. For the robotic system to perform its movements on the x and y axis, codes were created over the LEGO Mindstorms programming application. Servo motors, which provide the movement of the robotic system used in CRISPR.BOT V2, are required to perform their movements sequentially. In programming, commands are selected from columns designed to run a single engine. A block is created by adding the selected movement columns one after the other, and the prepared block movement is added under the start command of the programming, and the robotic system performs the movements in the block sequentially. In CRISPR.BOT V2, specific command blocks have been prepared for each plate, taking into account the programming of the 6-well plate (Fig. 4A–C), 12-well plate (Fig. 4D–F), and 96-well plate (Fig. 4G–I) used in current molecular biology experiments, and to develop protocols in an optimized way, taking into account the properties and fixed positions of the plates (Supplementary Video 5). During the block design phase, changes were made to the degrees according to the specific properties of the plates over an optimized block diagram in such a way as to reduce the error risk of the movements of the motors per a different kind of cell culture plate (Fig. 4). For this reason, the first commands in the diagram for the operation of the servo motor only provided the movement ensures that the other motors remain motionless. Other commands in the block diagram provided power to the servo motor that performs the desired movement in the robotic system (Fig. 4 and Supplementary Video 5). By starting the servo motors in the robotic system at optimum power, the block prepared for a single servo motor movement realizes the desired movement by providing the specified degree of movement in the blocks. These algorithm blocks were used in future molecular biology experiments.

Fig. 4
figure 4

Experiment blocks specially prepared for 6-well plate, 12-well plate, and 96-well plate. The first stop commands in the block stop the motors in the robotic system, except for the pipette module. With the operation of the motor that will move the pipette module, power, and speed are added to ensure the movement of the pipette in optimum condition. The movement of the pipette module between distances is determined in degrees. (A) Block for the 6-well plate that allows the pipette to move between two wells on the x-axis. (B) The block that the pipette system performs on the y-axis, allowing the 6-well plate to move from the beginning to the end. (C) For the 6-well plate, the block that provides the movement of the pipette on the y-axis at the distance between the two wells. (D) The block that provides the movement of the pipette on the x-axis at the distance between the two wells for the 12-well plate. (E) The block that the pipette system performs on the y-axis, allowing the 12-well plate to move from the beginning to the end. (F) The block that provides the movement of the pipette system on the y-axis at the distance between the two wells for the 12-well plate. (G) The block for the 96-well plate allows the pipette system to move between two wells on the x-axis. (H) The block that the pipette performs on the y-axis, allows the 96-well plate to move from the beginning to the end. (I) The block that provides the movement of the pipette system on the y-axis at the distance between the two wells for the 96-well plate.

Full size image

Molecular biology applications using the CRISPR.BOT system

Genetic transformation in bacteria

Based on the velocity-microliter measurement charts prepared for the targeted bacterial transformation experiment, the degrees required to experiment without error by drawing the required amount of liquid were determined. To prepare for the experimental programming, the bacterial transformation test stages to be carried out were schematized and planned (Fig. 5, Supplementary Video 2, and the flow algorithm of the operation related to this experiment is indicated in Supplementary Fig. 1). White colonies formed as a result of DNA transformation with an autonomous remotely controlled robotic system were determined in red and blue circles (Fig. 5E).

Fig. 5
figure 5

The schematization of bacterial transformation experiment steps and bacterial colonies after transformation using CRISPR.BOT. (A) Take 10–20 µl of DNA at an angle of 5° and add 50 µl of E.coli in the transwell at 4 °C. Then wait 30 min. (B) Take 60 µl of DNA + E.coli at an angle of 15° and wait for 20–30 s by immersing the pipette in 42 °C hot water. (C) Remove the pipette, which is immersed in 42 °C hot water, and leave E.coli + DNA in an 800 µl medium. Then shaking is done 3 times at intervals of 5 min for 15 min. (D) Take 205 µl 3 times at an angle of 40°, leave it in different parts of the Petri dish, and allow it to spread by shaking for 10 s. (E) Transgenic white bacterial colonies formed as a result of the transformation of plasmid DNA encoding GFP and Ampicillin (Amp) resistance gene into E.coli bacteria. µl, microliter; GFP, Green Fluorescent Protein; E.coli, Escherichia coli.

Full size image

Genetic transfer in human cells

With the 2nd version of the robotic system, the first experiments were carried out to prove that human cells can be genetically modified with viruses. The experiment was carried out; It was aimed to prove that GFP lentivirus can genetically modify human cells with an automated system (Fig. 6A–E). In the experiment planning, appropriate coding was performed for the 4 wells of the 12-well plate, side by side, with the first well as a control and increasing amounts of GFP lentivirus transfer to the other three wells (Fig. 6F–H). Cells were seeded in 4 wells of the 12-well plate to be used first, with 2,5 × 104 Human Jurkat cell lines (Clone E6-1, TIB-152, ATCC) in each well, and each well was completed with RPMI medium with a total volume of 500 µl. Plates were placed in the frame of the robotic system by adding 1 ml of GFP encoding recombinant lentivirus to one well of the other 12-well plate (Supplementary Video 3 and the flow algorithm of the operation related to this experiment is indicated in Supplementary Fig. 2). The robot performed GFP lentivirus extraction at a 70° angle (60–80 µl) and added the virus to the 1st well in which Jurkat cells (Clone E6-1, TIB-152, ATCC) were seeded in the first 12-well plate (Fig. 6I, J). Then, the CRISPR.BOT returned and added GFP lentivirus with a 100 ° angle (90–120 µl) and added the virus to the second well in which Jurkat cell (Clone E6-1, TIB-152, ATCC) was cultivated. Finally, the robot added the virus to the 3rd well by shooting with a 150° angle (160 µl) from the same virus. As a control, RPMI medium was added to the 4th well at an angle of 100 degrees. The plate was placed in a cell culture incubator at 37 C containing 5% CO2. For the control of cell number and viability, cells were analyzed with the trypan blue and a Cell Counting device (Biorad TC20) (Table 1). This shows that CRISPR.BOT does not have cytotoxicity on the cells. With this experiment, it was found that there was no GFP expression in the control group, but an increased GFP expression at 70-100-150-degree angles, respectively.

Fig. 6
figure 6

Genetic transfer in human cells, experiment programming and analysis of GFP protein expression. (A) Performing Cell Sowing, Jurkat Cells (Clone E6-1, TIB-152, ATCC) were cultivated in 4 wells at 2.5 × 104 cells/well. It was cultured in RPMI medium with a total volume of 500 µl. (B) Virus sample preparation, 1 ml of GFP (green fluorescent protein) Lentivirus was placed in one well of the 12-well plate. (C) Integration of Plates into the Autonomous Robot System, 12-well plates were placed in the frame of the Robot. (D) Addition of GFP Lentivirus to Cells, 70, 100, and 150° extractions were made from the well containing GFP Lentivirus (GFP LV) in a 12-well plate and added into wells 1, 2, and 3. (E) Incubation, after the robot performed the experiment, the 12-Well plate was placed in the incubator. (F) 70, 100, and 150° blocks were used for genetic transfer in human cells experiment programming. (G) 100°, and (H) 150° blocks that activate the pipette system are prepared in the experiment programming and perform the liquid withdrawal process. (I) Analysis of GFP protein expression in CRISPR.BOT modified transgenic Jurkat cell (Clone E6-1, TIB-152, ATCC) line using flow cytometry and (J) Fluorescence microscopy at 4X.

Full size image
Table 1 Cell count and viability rates of Jurkat cells (Clone E6-1, TIB-152, ATCC) after 72 h of transduction.
Full size table

GFP expression levels were determined 72 h after experimenting with Flow Cytometry analysis and fluorescent microscope imaging (Fig. 6I, J). Cells exhibited normal growth under all experimental conditions. While control wells were also pipetted using CRISPR.BOT, there was no observable difference in cell viability between wells pipetted with CRISPR.BOT and those pipetted manually. The absence of cytotoxicity was also supported by the absence of GFP expression in the control group and a noticeable increase in GFP expression in the experimental group at angles of 70, 100 and 150 degrees, respectively. As a result of the experiment, it was observed that there was no virus expression in the control group, and an increased virus expression was observed in the injection process made by the robot with the pipette system at 70–100–150° angles (Fig. 6I, J). Thus, it has been proven that the genetics of cells can be changed with a cost-effective autonomous system with viruses transferred in different amounts. The experiment aimed to automatically produce transgenic cells and work remotely with autonomous systems without touching the viruses, which are dangerous to study. The results obtained from the experiments carried out, it is the prototype of a system in which viruses can be produced and used without human touch, and vaccine and drug studies can be carried out in studies to be carried out with viruses such as SARS-CoV2 which is infectious.

CRISPR gene modification

In this experiment, it was aimed to transfer three different CRISPR guide RNAs of the robot to the cells. The main purpose of the evidential experiment for the future is to experiment by understanding which guide RNA should be added to which cell with the CRISPR system to perform the required genetic modification in the cell. Cells were seeded in 4 wells of the first 12 well plate with 5 × 104 Human Jurkat cells (Clone E6-1, TIB-152, ATCC) in each well, and the total volume was completed with RPMI medium to be 500 µl in each well. Lentiviruses encoding guide RNA 1 (gRNA1), guide RNA 2 (gRNA2), and guide RNA 3 (gRNA3) were added to the wells of the other 12 well plates, respectively, by CRISPR system, diluted 1:10 in RPMI medium. The plate was lifted into a cell culture incubator containing 370C 5% CO2 (Fig. 7A–E) (For watching: Supplementary Video 4 and the flow algorithm of the operation related to this experiment is indicated in Supplementary Fig. 3). With this experiment, it was observed that there was no virus expression in the control group, and gRNA1, gRNA2, and gRNA3 were added at 100 degrees angles, respectively, providing 35–40% GFP and guide RNA expression rate. Expression levels were determined by Flow Cytometry analysis on the 3rd and 7th days after the experiment was carried out (Fig. 7F). The results show that the CRISPR.BOT has the capability of editing cells using CRISPR encoding lentivirus. Optimization studies were carried out for the pipette system in CRISPR.BOT V1 to effectively draw low microliter liquids that have also been applied to the CRISPR.BOT V2 robotic system. With these studies, the sensitivity of the pipette system of CRISPR.BOT V2 and its ability to work with low microliters have been seen. Before the “Single-cell sub-cloning procedures of genetically modified cells” experiment, it was ensured that CRISPR.BOT V2 could perform a Single-cell Sub-Cloning process with a 96-well plate depending on the pipette precision and transfer appropriate microliters (Fig. 8A).

Fig. 7
figure 7

CRISPR gene modification and flow cytometry analysis. (A) Seeding of Jurkat Cells (Clone E6-1, TIB-152, ATCC) into 12 well plate at 5 × 104 cells/well. (B) virus sample preparation. gRNA 1,2,3 Lentivirus was placed in 12 well plate. (C) Integration of plates into the autonomous robot system, 12-well plates were placed in the frame of the Robot. (D) Addition of gRNA lentivirus to cells. The wells containing gRNA Lentivirus 1,2,3 in the 12-well plate were drawn at 100° and added to wells 1, 2, and 3 separately. (E) Incubation. After the robot performed the experiment, the plate was placed in the incubator. (F) flow cytometry analysis of GFP protein expression in transgenic human cells encoding CRISPR.BOT modified CRISPR guide RNAs. gRNA guide RNA, GFP green fluorescent protein.

Full size image
Fig. 8
figure 8

Single-cell sub-cloning procedures of genetically modified cells, GFP expression and cell viability in subcloned cells. (A) Precise microliter liquid intake performance trial studies with pipette system for CRISPR.BOT V2. (B) Seeding of Cells. CRISPR + gRNA 1-2-3 Jurkat Cells (Clone E6-1, TIB-152, ATCC) were cultivated at 5 × 104 cells/well in a 6-well. (C) Cell dilution. 10 µl of 3 different viruses in the wells of the 6-well plate were taken separately. 1 ml/500 cells were taken from the cell + RPMI mixture with a total volume of 1 ml and added onto 9 ml RMPI medium and placed in 6-well. (D) Integration of plates into the autonomous robot system. 6-well and 96-well plates were placed on the Robot’s frame. (E) Sub-cloning of cells into 96-well plate. CRISPR + gRNA 1-2-3 Jurkat Cells (Clone E6-1, TIB-152, ATCC) in the 6-well plate were pulled at 100° and added to a whole 96-well plate. (F) Incubation of the 96-Well plate. (G) CRISPR.BOT-based subcloning of the transgenic cells. (H) Shows the positive and negative rates of GFP expression in cells sorted in the experiment performed with CRISPR.BOT. The positive rate of GFP expression (more than 90%) is defined as “GFPhi”, while the negative rate (less than 10%) is called “GFPlo”. At the end of the experiment, the majority of the samples were sorted, resulting in a statistically significant result. Furthermore, some samples showed moderate GFP expression (between 10% and 90%), which is referred to as “GFPint”. (I) Shows the percentage of viability of subcloned cells and indicates that the statistical significance of these results was confirmed by unpaired t-test, *p < 0.05. Not significant, NS. µl, microliter; GFP green fluorescent protein, gRNA guide RNA. This indicates that the results are reliable and that the experiment was successful.

Full size image

Single-cell sub-cloning procedures of genetically modified cells

The main goal of single-cell sub-cloning is to ensure that each of the genetically modified cells has a consistent genotype and phenotype. This minimizes variation in experiments and allows for more stable and reproducible results. One of the most frequently used cell lines in our studies is Jurkat cells (Clone E6-1, TIB-152, ATCC) a CD4 + T cell lymphoma line. Gene editing was performed on these cells using CRISPR guide RNAs. In this process, it was aimed to isolate cells with different levels (high, medium, low) of CRISPR guide RNA expression and clone them separately. To achieve this goal, single-cell sub-cloning method was preferred. This is normally performed by plating cells into individual wells using fluorescent signals through a FACS device. However, since FACS devices are costly and complex to use, it is planned to make the manual single-cell sub-cloning process more automated and precise in robotic systems. The aim is to increase the accuracy and efficiency of the system by ensuring that a single cell is placed in each well. Furthermore, single cell dropping allows researchers to better explore the details of cells at the molecular level. Combined with the use of robotic autonomous systems, interactions between cells are minimized, allowing researchers to comprehensively and reliably study the characteristics of the targeted single cell population7,8. The robot is intended to release cells one by one without the need for single cell sorting techniques such as FACS. The main goal of this future evidence-based experiment is to enable researchers to better understand the properties of a single cell population without the influence of other cells using robotic autonomous systems. Jurkat cell (Clone E6-1, TIB-152, ATCC) encoding CRISPR guide RNAs were inoculated into a 6-well plate with 5 × 104 cells in each well. Prepared cell samples were placed on a 6-well plate and placed in the frame of the Robot. Then, the 96-well U Plate was placed on the frame of the Robot. Jurkat cells (Clone E6-1, TIB-152, ATCC) encoding CRISPR guide RNAs in a 6-well plate at 100° were seeded into a whole 96-well plate. After cultivation, the plates were transferred to an incubator at 370C and 5% CO2 (Fig. 8B–F and Supplementary Video 5, and the flow algorithm of the operation related to this experiment is indicated in Supplementary Fig. 4). After 24 h, plate scans were performed with an inverted microscope and cell photos were taken. After 7 days, plate scans were performed again with an inverted microscope and cell photographs were taken (Fig. 8G). These observations show that single or up to three-cell could be cultured per well as single-cell subcloning. Evaluation of the CRISPR.BOT-based subclones of transgenic cells, as seen in Fig. 8H, I, showed that the GFP expression was between 90 and 100% in the positive cell population and between 0 and 10% GFP expression in the negative cell population, the difference was clear, and a statistically significant value was found (Fig. 8H, I). These results showed that CRISPR.BOT accomplished single cell subcloning with high efficiency and low GFP expression.

Discussion

Liquid-processing robotic systems are also needed in life sciences laboratories to increase practicality and efficiency9,10. Today, systems that can handle automatic liquids and provide sample dispensing functions have become used in most life science laboratories11. However, human intervention does not disappear, it only reduces the risk of error. Robotic systems, on the other hand, can operate independently without the need for intervention at the start of the experiment12. It reduces the volume of the samples used in the experiments and increases the efficiency of the experiments both financially and in terms of workmanship7. Studies of integrating automation systems with artificial intelligence (AI) provide the ability to optimize targeted therapeutic nanoparticles for unique cell types and patients13. Apart from the high costs of fully automatic systems, their maintenance is also quite expensive. There is a great cost loss because it usually requires system-specific protocols and laboratory equipment4. Compared to the robotic systems available on the market, LEGO Mindstorms EV3 can be purchased at an average cost of around 1000 dollars, focusing on robotic coding. This system offers precision, easy installation and the opportunity to replace parts at affordable prices when necessary. The main purpose of the future-proof experiments carried out in CRISPR.BOT study is machine learning, transgenic cell screening, analytical calculations, gene designs, genetic pathway selection, experimental design, cell cloning, and genetic pathway analysis through the development of remote and autonomous robotic systems in the field of molecular biology and genetics (Fig. 9). It can be performed without human touch, with a very low margin of error. Various versions of CRISPR systems, such as CRISPR-Cas9, CRISPR-Cas12 and CRISPR prime editing, are being studied in our laboratory. The main goal of the CRISPR.BOT study, to test the operability of different CRISPR systems on an automated robotic platform as a ‘proof of concept’. In the “CRISPR Gene Modification” experiment, it was shown that it is possible to transfer guide RNA and CRISPR-Cas system into cells. Although not all experiments conducted at this stage are directly CRISPR-oriented, our long-term goal is to develop the CRISPR.BOT robotic system for CRISPR-based gene editing.

Fig. 9
figure 9

CRISPR.BOT autonomous systems. Experimental approaches that can be performed with CRISPR.BOT autonomous systems.

Full size image

The utilization of fully automated commercial robotic systems necessitates adherence to system-specific protocols and the use of specialized laboratory equipment. In contrast, CRISPR.BOT, facilitated by the user-friendly LEGO Mindstorms EV3 kit, offers a distinctive advantage by providing an easily installable solution for daily laboratory routines. Notably, it delivers precise and sensitive results at microscale liquid addition levels. A crucial aspect of CRISPR.BOT’s functionality is its ability to ensure experiment replicability and sensitivity consistency. Rigorous programming adjustments were implemented to guarantee that the pipette consistently withdrew the intended volume. Post-liquid draw, manual measurements using a pipette confirmed the accuracy of each draw. Repeated experiments demonstrated the robot’s capability to consistently deliver the programmed volumetric value with precision, showcasing its reliability in the 5 –200 µl liquid range. However, a noted increase in error rates occurred in liquid withdrawals below 5 µl.

In the transition to CRISPR.BOT V2, significant enhancements were introduced to the system mechanism. The pipette movements evolved beyond the up-and-down motion of CRISPR.BOT V1, now enabling movements along the x and y axes on diverse plates (Fig. 4). Unlike the previous version, CRISPR.BOT V2, with its rail system, allows the pipette to reach desired plates directly, providing a 2-dimensional movement capability. This development facilitates simultaneous work on multiple solutions, marking a significant advancement. The evolution of the robotic system is complemented by algorithmic advancements, empowering the creation of open code tailored to specific experiments. Specialized codes were devised for different plates (6-well, 12-well, and 96-well) used in molecular biology experiments, considering their distinct characteristics and fixed positions. Crucially, experiments conducted with CRISPR.BOT robotic systems within a closed system demonstrated the absence of contamination, affirming its sensitivity comparable to other commercial robotic systems in cell culture studies. In the optimization phase of CRISPR.BOT V1, numerous trials focusing on power and speed in programming revealed the efficacy of higher power usage. This informed the optimization studies of CRISPR.BOT V2, maintaining constant high degrees of power and speed. Remarkably, prior to the development of the LEGO Mindstorms EV3-based liquid handling robotic system, algorithms for molecular biology experiments were non-existent. The introduction of CRISPR.BOT algorithms encompass bacterial transformation experiment programming (Supplementary Fig. 1), automatic genetic modification of human cells (Supplementary Fig. 2), CRISPR gene modification experiment programming (Supplementary Fig. 3), and cloning procedures of genetically modified cells (Supplementary Fig. 4). This innovative approach encompasses not only cloning procedures, but also antibody staining and flow cytometry analyses, including mixing of cells with antibodies and incubation systems. The robot we developed is equipped with special kits that allow colorimetric analyses to be performed with an ELISA reader. In addition, the preparation of pre- and post-PCR solutions for RT-PCR experiments is also performed with the robotic system; the feasibility of these developments is envisaged. The CRISPR.BOT system not only demonstrates consistent accuracy and sensitivity in liquid processing but also evolves with enhanced functionalities in its V2 version. The integration of open-code algorithms adds a layer of versatility to its applications, showcasing its potential as a transformative tool in molecular biology experiments and laboratory automation.

We conducted a comparative analysis with a conventional autonomous liquid handling system, exemplified by the Thermo Fisher TriPlus RSH Automatic Sampling Device, a widely recognized tool for precise microliter-level experimental studies. Renowned for processing numerous samples rapidly and reliably, Thermo Fisher TriPlus RSH autosamplers significantly boost laboratory productivity through flexible automation and robust software. Similar to the Thermo Fisher system, the CRISPR.BOT robotic system ensures reliable production and operates 24/7. While the CRISPR.BOT can perform liquid processing on up to four side-by-side plates, the TriPlus RSH autosamplers demonstrate flexibility with various sample plates, efficiently handling low microliters. Through optimization studies, the CRISPR.BOT robotic system has demonstrated precision in processing liquids ranging from 5 µl to 200 µl, though improvements are needed for liquids below 5 µl. In contrast, the Thermo Fisher TriPlus RSH autosamplers can handle samples ranging from 20 mL to as low as 1 µL, offering direct injection capability and a loading capacity of 6912 samples using a maximum of 18 well plates. The CRISPR.BOT autonomously processes a maximum of 384 samples using four well plates. Process reliability in the TriPlus RSH system is ensured through temperature control monitoring for missing bottles and recording injector positioning, position, and temperature by confirming the correct drawer position. Similarly, the CRISPR.BOT robotic system integrates various sensors for position, distance analysis, or color detection on plates, enabling command movements based on colors. Currently, syringe tip exchange is not possible in the CRISPR.BOT system; this is due to the specific syringe tip design. However, this limitation is offset by advantages such as the cost-effectiveness of the system and the accessibility of the design. The main objective of the project is to make laboratory automation more accessible and affordable. The current design allows users to face less complexity in setting up and using the system. However, future developments aim to address the syringe tip exchange feature. These improvements will increase the flexibility of the system, providing more options for different types of experiments and applications.

Looking ahead, these studies are envisioned to be valuable for diverse applications, including complex drug delivery experiments and advancements in fields like regenerative medicine and tissue engineering, in addition to biotechnological applications. The CRISPR.BOT study underscores our ability to diminish human intervention and establish foundational infrastructure for molecular biology and genetic engineering executed by robots in a closed system. In our work, human cells, DNA and viruses are integrated into a robotic system with minimal human intervention. Human intervention takes place only in the initial phase, i.e. during the insertion of viruses and cells into the system. With the start of the experimental protocols, all operations are carried out completely automatically by the robotic system. After the cell manipulation is completed using robotic arms, transferring the cells to the incubator and centrifuge has been identified as one of the main objectives of our project. Importantly, this prototype system allows for virus production and experimentation without direct human contact. This is especially critical for research on pathogens such as SARS-CoV-2. Furthermore, the advances of the study include recombinant virus synthesis, characterization, testing and predictive experiments in both living and non-living environments. The integration of artificial intelligence enhances the CRISPR.BOT system’s autonomy, offering substantial potential for self-determining and executing various molecular biology experiments. This synergy of artificial intelligence and robotics points towards a future where molecular biology and genetic methods operate 24/7 in a more automated, rapid, and practical manner. In terms of cost-effectiveness, LEGO Mindstorms-based robotic systems, including CRISPR.BOT, present a notable advantage. The cost comparison reveals a range of $10,000 to $30,000 for liquid handling robots such as Beckman Coulter Biomek 4000 and Opentrons NGS workstation, whereas the LEGO Mindstorms EV3 kit utilized in the CRISPR.BOT installation costs $1,000. Beyond cost, the unique feature of editable algorithms without additional charges positions CRISPR.BOT as a cost-effective solution, bridging a gap not addressed by many commercial robotic systems that lack the capability to create open code specific to experiments.

In addition to biotechnological applications in vaccine studies, transgenic cell generation such as recombinant viruses or dangerous viruses, genetic therapies, cancer studies, laboratory fields such as synthetic biology, microbiology, or genetics, and in developing fields such as regenerative medicine and tissue engineering, complex drug delivery experiments and it is thought to be useful for many other applications (Fig. 10). The production of an end-to-end, high-efficiency laboratory automation system with ready-made workflow algorithms with CRISPR.BOT Robotic systems, of which we produced the prototypes, will enable us to establish a development company on molecular autonomous systems in the next stage. There are companies in the world (e.g. Beckman Coulter, Opentrons, Thermo Fisher, and Molecular Devices) that manufacture devices based on existing automation and experimental customizations. It is anticipated that robotic systems that meet the application requirements of laboratories will increasingly continue in the near future. CRISPR.BOT Robotic systems will be a prominent development company in the future with the advantages of cost-effective, user-friendly, and remote control by researchers compared to their current competitors.

Fig. 10
figure 10

CRISPR.BOT closed the laboratory automation system. Development of CRISPR.BOT closed laboratory autonomous systems and their integration with laboratory analyzer (RT-PCR, ELISA Plate Reader, Flow Cytometry, Incubator, and Thermal Cycler). RT-PCR real time-PCR.

Full size image

Materials & methods

For molecular biology experiments with CRISPR.BOT, we first prepared plasmid DNA, and GFP or CRISPR-Cas9/gRNA encoding lentivirus with the following methods.

Transformation and isolation

Competent E. coli DH5α strain (NEB® 5-alpha Competent E. coli (High Efficiency) NEB C2987H) were cultured in a liquid LB medium at 37oC. The bacterial strain was stored at -80oC as stock in LB medium with added glycerol. Bacteria were transformed with the plasmid DNAs to be used in the CRISPR.BOT studies using the heat shock method according to the manufacturer’s instructions. Bacteria were inoculated on LB agar plates with ampicillin and incubated at 37oC overnight. Plasmid isolation was performed from selected bacterial colonies with the CompactPrep Plasmid Maxi Kit (#12863, QIAGEN). DNA concentration was measured in ng/µl with Microplate ELISA Reader (FLUOstar Omega) and it was evaluated whether the purity value was between 1.8 < A260/A280 < 2.0.

Lentivirus production

The previously designed 3 gRNA sequences were synthesized with the pU6 promoter by GenScript and cloned into the lentiviral pHIV-EGFP (Addgene, #21373) plasmid14 For lentivirus production, CRISPR-Prime Editing system encoding lentiviral pLenti-PE2-BSD plasmid DNAs (Addgene, # 161514) encoding CRISPR-PE containing the blasticidin resistance gene, control pHIV-EGFP non-coding pegRNA1/2/3, and coding GFP and pHIV EGFP coding pU6-pegRNA1/2/3 and GFP were transfected with pSPAX2 and pVSV-G envelope and packaging plasmid DNAs. After all plasmids (2:1 pSPAX2: 1 pVSV-G) were treated with P-PEI (Polyethylenimine) transfection reagent15, lentivirus production was performed using the host cells, HEK293T14. Packed recombinant lentiviruses were obtained by collecting HEK293T cell supernatants 72 h after transfection. Produced lentiviruses were concentrated 100x with Lenti-X Concentrator (#631232, Takara) to increase virus concentration. For the titration of lentiviruses, titration steps using Jurkat (Clone E6-1, TIB-152, ATCC) (RPMI medium containing 10% fetal bovine serum, 200 U/ml penicillin/streptomycin antibiotic, 2 mM L-glutamine, 1X MEM vitamin solution, and 1X NEAA) cells in a previously published study were performed16. After the cells were transferred to 96-well plates, Jurkat cells (Clone E6-1, TIB-152, ATCC) containing pLenti-PE2-BSD (Addgene, #161514) were treated with 10 µg/ml blasticidin for selection of blasticidin and incubated for 72 h. Afterward, cell viability was tested by staining with a 7-AAD antibody in flow cytometry. Finally, cells encoding CRISPR-PE with > 99% blasticidin resistance were selected and cultured. Titration of pegRNA-GFP and control GFP lentiviruses was calculated by measuring the GFP expression level in flow cytometry. Subsequently, using CRISPR.BOT, three different pegRNA-GFP or control GFP lentivirus were individually transduced into CRISPR-PE (+) Jurkat cells (Clone E6-1, TIB-152, ATCC).

Cell survival analysis

Survival analysis was performed to control the cell viability of Jurkat cells (Clone E6-1, TIB-152, ATCC). Trypan blue (Biological Industries, #03-102-1B) was applied to identify and count surviving cells. Cell counting and viability analysis were performed with the BIO-RAD TC20 Automated Cell Counter.

Statistical approach

Data recorded by flow cytometry were analyzed using CytExpert (Beckman Coulter). Statistical analyses were performed using GraphPad Prism 8.0.1 software (GraphPad Software, Inc., San Diego, CA, USA) with a two-tailed unpaired t-test using independent mean values. Error bars represent the standard error of means (SEM). For all experiments, significance was defined as *p < 0.05 and NS non-significant.