Assembly of a Custom-made Device to Study Spreading Patterns of Pseudomonas putida Biofilms

引用 收藏 提问与回复 分享您的反馈 Cited by



Journal of Industrial Microbiology & Biotechnology
Oct 2018



Biofilms are bacterial communities in the shape of exopolysaccharide matrix-encased aggregates attached onto interphases able to resist environmental aggressions. The development of bacteria in the shape of biofilms deeply affects the performance of many industrial processes which work with fluidic systems, where bacteria may settle and prosper. As a consequence industrial equipment experiments low performance issues and substantial maintenance costs.

The study of how bacteria of industrial interest such as Pseudomonas putida spread in these fluidic systems is highly dependent on the chosen experimental system to retrieve such data, thus using scaled prototypes becomes an essential step towards the design of a more efficient system to handle biofilms, either to control them or to prevent them. This protocol describes how to assemble, operate and maintain a device to grow and monitor the biofilm spreading pattern of this bacterium (as a function of the fluid hydrodynamics) in a custom-made chamber larger than those typically used in laboratory environments, and how to analyze the information gathered from it in a straightforward fashion. Description of the protocol was thought to be used as a working template not only for the presented case study but for any other potential experiment in different contexts and diverse scales following similar design principles.

Keywords: Biofilm (生物膜), Custom-made device (定制设备), Fluidic system (流体系统), Spreading pattern (扩散模式), Image analysis (成像分析), Pseudomonas putida (恶臭假单胞菌)


The characterization of bacterial biofilm spreading pattern dynamics in fluidic systems is a key issue when trying to gain control of the proliferation of these living organisms. Especially relevant is the case of biofilm presence in industrial environments, where an uncontrolled growth may lead to large economic loses. Typical studies on this area focus their efforts on creating laboratory-scale setups to investigate the effect of different variables to test. These parameters are controlled to monitor subsequent biofilm expansion for every tested experimental condition. However extrapolating these results to an industrial-like fluidic environment is not so straightforward, as equipment, materials and operation conditions may deeply affect biofilm behavior. Furthermore the designs performed for laboratory-scale devices are in many cases incompatible with typical industrial equipment and procedures. As a consequence, the number of studies in the literature using industrial-like devices is very limited.

In this protocol we describe a procedure for assembling, testing and operating an industrial-like device to study the biofilm spreading pattern of platform strain Pseudomonas putida mt-2. The design and operation procedures of the device were specifically chosen to have into account constrains and features typically found when using industrial materials and equipment, delivering a viable solution for given research purposes.

Materials and Reagents

  1. Laboratory consumables
    1. 0.4 μm pore size nitrile filter (Millipore, catalog number: SLHA033SS)
    2. 0.45 μm PES filter unit (VWR, catalog number: 514-0335)
    3. Microscope slide (VWR, catalog number: 16004-422)
    4. Microscope cover slip (VWR, catalog number: 48393-026)
    5. Aluminum foil
    6. Marker pen

  2. Tubing
    1. 1 x 25 m 5.5 mm inner diameter flexible polyurethane tubing (Legris, catalog number: 1025U0805)
    2. 1 x Silicone tubing, 25 m, 2 mm (id) x 4 mm (od), (VWR, catalog number: 228-0704)
    3. 1 x Silicone tubing, 4 m, 4 mm (id) x 6 mm (od), (VWR, catalog number: 228-0709)

  3. Valves and clamps
    1. 8 x F-F ¼’ BSPP ball valve (RS, RS Pro, catalog number: 733-5202)
    2. 1 x ISMATEC 2 stop coded tubes, id: 1.65 mm (blue-blue, catalog number: SC0018), Mat: Tygon R-3607 (Cole Parmer, catalog number: EW-96460-38)
    3. 2 x hosecock clamp (VWR, catalog number: 21716-102)

  4. Connectors and fittings
    1. 1 x 8 mm–8 mm–8 mm Tee Legris (RS, catalog number: 445-4495)
    2. 14 x Legris Pneumatic Straight Threaded-to-Tube Adapter, R 1/4" Male, Push In 8 mm (RS, catalog number: 182-4784)
    3. 3 x SMC Pneumatic Straight Threaded-to-Tube Adapter, Rc 1/8" Female, Push In 8 mm (RS, catalog number: 771-5002)
    4. 2 x SMC Pneumatic Elbow Threaded-to-Tube Adapter, Push In 8 mm 1/4" thread (RS, catalog number: 771-6030)
    5. 2 x Legris Pneumatic Elbow Threaded Adapter, R 1/4" Male x R 1/4" Male (RS, catalog number: 367-5792)
    6. 2 x Legris Brass 1/4" in. BSPT Male Plug Threaded Fitting (RS, catalog number: 231-5045)
    7. 1 x Legris Pneumatic Tee Threaded-to-Tube Adapter, 1/4". x 8 mm x 8 mm, 30 bar (RS, catalog number: 225-0617)
    8. 1 x Barbed fittings, Straight Connector, Natural PP, 1/8" ID; 10/pk (Cole-Parmer, catalog number: EW-51518-05)
    9. Barbed Tee Connector, Polypropylene, 1/8"; 25/pk (Cole-Parmer, catalog number: SI-50623-66)
    10. 1 x ADCF Male Luer to 1/8" L Barb Adapter, Polypropylene, 25/pk (Cole-Parmer, catalog number: SI-30800-24)
    11. 1 x ADCF Female Luer to 1/8" L Barb Adapter, Polypropylene, 25/pk (Cole-Parmer, catalog number: SI-30800-08)

  5. Accessories
    1. Electrical Tape, 19 mm x 20m (RS, catalog number: 511-4306)
    2. Socket extension lead (RS, catalog number: 122-1105)
    3. 1 x Clear Polycarbonate PC Sheet, 1.25 m x 610 mm x 6 mm (RS, catalog number: 681-665)
    4. 1 x Nitrile rubber O-Ring Cord, 2 mm diameter (RS, catalog number: 138-1600)
    5. 1 x White PTFE Tape 12 mm x 12 m x 0.075 mm (RS, catalog number: 512-238)
    6. 1 x Pan Head Bright Zinc Plated Steel Machine Screw bag, M3, 20 mm (RS, catalog number: 560-625)
    7. 1 x Stainless Steel, Hex Nut bag, M3 (RS, catalog number: 189-563)
    8. 1 x Stainless Steel Plain Washer, 0.5 mm Thickness, M3 (RS, catalog number: 189-620)
    9. 1 x Cyanoacrylate glue (RS, catalog number: 533-478)

  6. Chemicals
    1. Gentamycin (Sigma-Aldrich, catalog number: PHR1077-1G)
    2. Absolute Ethanol (Sigma-Aldrich, catalog number: 1009832500)
    3. NaClO (VWR, catalog number: 470302-586)
    4. Glycerol 85% (Sigma-Aldrich, catalog number: 1040942500)
    5. Na2HPO4·2H2O (Sigma-Aldrich, catalog number: 71643-250G)
    6. KH2PO4 (Sigma-Aldrich, catalog number: P5655-500G)
    7. NaCl (Sigma-Aldrich, catalog number: S3014-1KG)
    8. NH4Cl (Sigma-Aldrich, catalog number: A9434-500G)
    9. MgSO4 (Sigma-Aldrich, catalog number: M2643-500G)
    10. 10x M9 salts stock (see Recipes)
    11. 20% Glycerol (see Recipes)
    12. 1 M MgSO4 (see Recipes)
    13. M9 minimal medium supplemented with 0.2% (w/v) glycerol (see Recipes)


  1. 2 x 2 L Pyrex® narrow-mouth Erlenmeyer flask (Sigma-Aldrich, catalog number: CLS49802L)
  2. 3 x 1 L Pyrex® narrow-mouth Erlenmeyer flask (Sigma-Aldrich, catalog number: CLS49801L)
  3. 1x DuranTM 4-Port Assembled HPLC Screw Caps GL-45 (Fisher Scientific, catalog number: 10740834)
  4. 1 x 1 L Laboratory bottles with screw cap, DURAN® GL-45 (VWR, catalog number: 215-1517)
  5. 1 x 5 L BRAND® PP beaker with spout, low form (Sigma-Aldrich, catalog number: BR87826-4EA)
  6. 1 x 16 mm length magnetic stirring bar (VWR, catalog number: 442-0366)
  7. 1 x 250 x 250 mm magnetic stirrer (VWR, catalog number: 444-0572)
  8. Orbital shaker
  9. 1 x Flowmeter (Aalborg P-150 mm, 1/4" FNTP fitting, catalog number: P11A3-BB0)
  10. 2 x 12VDC Pumps (EKWD, EK-DCP 2.2, catalog number: 3831109862506)
  11. 3 x 220V AC/12V DC adaptor (RS Online, catalog number: 737-8149)
  12. 1 x 19-37.8 L aquarium air pump (Hagen, A842 Aquaclear Air pump 10, catalog number: 15561108447)
  13. 1 x Ismatec IPC-N ISM936D IPC-N (ISM936D) Low-Speed Digital Peristaltic Pump with Click-'N-Go Cartridges; 8-Channel, 230V (Cole-Parmer, catalog number: EW-78001-16)
  14. 1 x Adjustable Spanner (RS, catalog number: 469-7018) 
  15. 1 x EMCO Concept Mill 105 milling machine
  16. 1x Fluorescence microscope with the following features:
    1. Light source and light filter to allow visualization of GFP and Texas Red fluorochromes
    2. Monochrome digital camera with EMCCD chip (minimum requirement)
    3. Motorized tray for automatic XY displacement and automatic tile-scan acquisition
    4. Automatic Z autofocus


  1. Scientific programming language (i.e., MATLAB) or general purpose programming language (i.e., Python, C+) able to work with scientific computing packages supporting basic statistic functions (i.e., mean, standard deviation), array/structure manipulation (array sorting, length) and element selection operators (unique elements)
  2. Image treatment library for selected program language (i.e., scikit-image for Python, Image processing toolbox for MATLAB)
  3. Image treatment software (ImageJ or equivalent) containing a proper library to open microscope files with native formats (i.e., *.lif), or microscope-related software able to perform regular format file storage


The complete experiment could be divided into the following 6 sections (A to F).

  1. Flow cell/Biofilm chamber design and manufacturing
    The design and manufacturing of the biofilm chamber involved the use of a milling machine to generate the geometry specified in Figure 1. The biofilm chamber geometry was designed with the aim of promoting a laminar flow through a rectangular square duct of variable Reynolds number, being this last parameter adjustable by setting the flow rate at the entrance. The central part of the chamber was divided in smaller ducts having the proper hydraulic diameter to create the hydrodynamic conditions to study within the channels (a laminar flow with Re~100-1,000). The ducts additionally served as a way of taking technical replica of different sections of the chamber placed at different positions. Two lateral chambers were also added at the input and output sides to hydrodynamically stabilize the fluid pattern before entering the channels.
      However, alternative methods to create such chambers are also available. Please read Note 1 for a more detailed explanation.

    Figure 1. Custom-made biofilm chamber blueprint showing (A) side, (B) front and (C) top views. A polycarbonate sheet was carved using a milling machine following this blueprint. The current state-of-the-art in manufacturing technology allows the use of other techniques such as 3D printing to create this design or any other conceptualization in a different type of materials (i.e., PP, PC, etc.) aligned to the purpose of the experiment. Dimensions expressed in mm.

  2. Assembly and setting up of the fluidic circuit
    The fluidic circuit was designed to work at steady state regime and room temperature while maintaining a constant flow rate passing through the biofilm chamber. Losses of humidity due to constant bubbling can be minimized by using a humidity chamber to increase the relative humidity of incoming air flow. Additionally, a redundancy criteria was applied when designing the fluidic circuit for culture aeration (by using mechanical stirring and bubbling) and liquid pumping (use of two pumps in series) to avoid an eventual insufficiency in oxygen demand and increased head loss due to growing biomass within the circuit surfaces (i.e., tubing, equipment, etc.).
    1. Organize the equipment according with the distribution depicted in Figure 2 (up) taking into account the available space in the room where the device will be placed. Figure 2 (down) shows an oblique and top view of an assembled device. By doing this, you will realize how much tube is needed to connect the different elements and how to place the different elements to make easy all the assembly, operation, inspection and maintenance tasks.

      Figure 2. Schematic layout (up) and real views (down) of the experimental device. The system consists of a culture vessel containing a large volume of planktonic bacteria, a fresh input of nutrient and air to feed and aerate the culture, a waste line to work in steady state and a recirculating line to transport culture to the biofilm chamber. This recirculating line was designed to allow the flow of large amounts of liquid culture through the biofilm chamber without consuming unaffordable amount of nutrient medium.

    2. Screw tube connectors with valves, flowmeter and flow cell as indicated in Figure 3 using PTFE tape in male threads. Tight firmly the threads without breaking the tape. Then, cut pieces of tubing with the desired length and connect them as illustrated.

      Figure 3. Assembly scheme. The device was assembled following this scheme using the different components described in the protocol. Equipment and containers were labeled following an alphanumeric coding (i.e., BR-1) to help providing detailed instructions at element level of how to manipulate the device during the start-up and operation.

    3. Assemble the custom O-Ring to use for sealing the biofilm chamber lid: cut 400 mm of the 2 mm nitrile rubber O-Ring cord and join both ends using a drop of cyanoacrylate glue. Wait at least 15 min to let the glue being completely dry.
    4. Cover the flow cell/biofilm chamber with the lid and seal it using the most appropriate system (see Note 3 for more details).
    5. Check that all pumps and air compressor are connected according with designed flow direction in the circuit, and label the biofilm chamber inflow and outflow parts (i.e., a mark on the lid with a pen, a sticker, etc.) to have a reference and know the direction of the flow inside of it (because of the geometric symmetry of the chamber).
    6. Prepare electric connections of centrifugal pumps. Manually connect the centrifugal pumps and the AC/DC connectors. To do this, check that AC/DC converters are not connected to the electric grid and cut the extreme of the wire to let it free. Separate positive and negative wires and peel their extremes to allow the connection between AC/DC converter and the pump. Link together the wires of the same polarity coming from the centrifugal pump and the converter, twisting them. Cover the peeled part with electrically isolated tape. Alternatively, if the user has a crimp tool (24 AWG), a butt wire splice connector can be used to connect electrical wiring.
    7. Connect all electric equipment to the socket extension lead. All electric wiring must be safely placed out of water contact, thus check that wires are in an elevated position. Use any method available (tape to hang them, elevated supports to separate them from the surface where water) to avoid potential contact with any liquid in case of leaking or accidental spill.
    8. It is recommended to perform a leaking test to detect bad/poor/failed connections:
      1. Start by closing all valves of the assembled circuit.
      2. Fill a 2 L Erlenmeyer flask with distilled water: this would work as culture vessel (labeled as BR-1 in Figure 3) for the whole experiment. Open valve V-1 and prime the centrifugal pumps (P-1, P-2) by manually filling with distilled water the tube connecting them with the culture vessel (BR-1) to purge as much air as possible. This last step will not be necessary if peristaltic pumps are used to pump the water.
      3. Open the flowmeter FM-1 at maximum capacity (if present) and all the valves in the recirculation line (V-1, V-3, V-4, V-6). Switch on the pumps P-1, P-2, checking that the liquid fills the circuit completely.
      4. Let the liquid flow for a minimum period of time of 30 min and check periodically that there are no leaks by inspecting the presence of small drops of water in thread connectors or water spills below any element of the circuit. If a leak is detected, stop the flow and try first to tight the connection, dry the zone and restart the flow to see if water still leaks. If the leaking persists, switch off the current, purge the water through the auxiliary line 2 (see Figure 2) and dismount the affected part. Disconnect screws and replace PTFE tape if broken. Add additional layers of PTFE tape to increase isolation when required.
      5. Once the leaking test is ended, close valves V-1, V-3, V-4 and V-6, and manually empty culture vessel.

  3. Initial bacterial culture and stocks
    1. One day before the start of the experiment (Day 0), inoculate a 50 ml Erlenmeyer flask with 10 ml of M9 minimal medium supplemented with 0.2% (w/v) glycerol as carbon source and 0.5 μg/ml of gentamycin. Add directly from a -80 °C frozen glycerol stock of P. putida mt-2.
    2. Incubate overnight at 30 °C in an orbital shaker set at 170 rpm.
    3. On the day of the experiment start (Day 1), measure the OD600 of the culture and dilute it in 1 L of fresh M9 + 0.2% (w/v) glycerol to reach an OD600 of 0.0025.
    4. Prepare a sterile stock of 4 L of M9 + 0.2% (w/v) glycerol supplemented with gentamycin (0.5 μg/ml) prior to start the device.

  4. Starting-up
    1. The fluidic circuit must be sterilized prior to beginning the experiment. In order to do that:
      1. Connect ISMATEC tubing corresponding to nutrient/waste peristaltic pumping lines PP-1 with 2 mm silicone tubing parts using barbed–barbed straight connectors. Connect also the 2 mm tube line through which the air will be conducted to the main flask with a Male Luer to 1/8" L Barb Adapter. Use aluminum foil to cover the extremes of both tube lines separately, and autoclave them.
      2. Fill 500 ml of 1 L Erlenmeyer flask FL-3, and 700 ml of BR-1 vessel with ethanol 70% (v/v). Introduce the magnetic stirring bar in BR-1 and connect FL-3 to auxiliary line 1. Check that valves V2, V-3, V-4 and V-6 in the recirculating loop are opened and V-1 and V-5 are closed, leaving also the flowmeter open at maximum capacity and switch on the centrifugal pumps. Transfer the stored ethanol in FL-3 to BR-1 until most of the flask is empty and then switch off pumps P-1, P-2. This will fill the auxiliary line 1 with ethanol. Close V-2 and open V-1. Then switch on again pumps P-1 and P-2, enabling the accomplishment of the sterilization of the system along 60 min.
      3. Stop centrifugal pumps P-1, P-2, open valve V-5, close V-6 and switch on P-1 and P-2 to let the ethanol leave the system. The removal of the ethanol has to be performed by letting the pumps to empty the culture vessel BR-1 as much as possible. Avoid its total empty while pumps are still working, as this would introduce air in recirculating line and cause priming problems in the pumps. 
      4. Disconnect P-1 and P-2 once BR-1 is totally empty (and no air has still entered in the circuit, as said above), and replace the 1 L Erlenmeyer flask FL-3 by another one filled with 1 L of sterile water. Connect again P-1, P-2 to wash with approximately 100 ml of water the remaining ethanol within recirculating line. Close V-5 and open V-6 to let the fresh water enter in the main tank and wash the ethanol during 1 h.
      5. Repeat the previously described operations for ethanol–water replacement, but using 1.2 L of fresh M9 medium supplemented with 0.2% (w/v) glycerol to replace water in the system, finally leaving 1 L of fresh medium inside BR-1. 
    2. Microscope settings will be adjusted at this point. Choose a 12-bit or 16-bit resolution for images to obtain a larger sensitivity in grayscale values. Select the channel to detect the fluorescent signal generated by the studied strain (here we recommend GFP because of its stability and performance). Take 10 μl of saturated overnight culture and spread it onto a microscope slide with the tip of the pipette. Place a cover slip onto the sample and examine it with the microscope. Adjust the gain of the microscope and laser intensity to obtain a non-saturated image. Keep these settings, as they will be constant through the experiment 
    3. Switch off pumps P-1, P-2 and PP-1, close all valves, disassemble biofilm chamber BC-1 and place it under the microscope following the steps detailed in the image acquisition section. Gather images in the biofilm chamber (if any signal was detected, it would be caused by material and/or liquid medium self-fluorescence). These images will be used to set the background noise threshold.
    4. Remove aluminum foil and connect nutrient stock FL-1 and waste tank FL-2 with peristaltic pump PP-1, and their respective outputs with BR-1. Avoid any contact of the tubing with any surface when placing them inside the flasks to prevent contamination. Place the ISMATEC stop coded tubes in the pump cassettes and check that they are properly secured.
    5. Connect a 4 mm silicone tubing piece with the output exit of the air pump. Add a tee and connect in the 90° branch a silicone tube with a clamp CL-1, which will serve to regulate air inflow. Then connect the syringe filter by its Luer side with humidity chamber HC-1. Finally, connect the output tube of HC-1 with BR-1 by introducing the free side of the tube in the flask. See Note 2 for additional comments.
    6. Switch on the magnetic stirrer at 150 rpm, and connect air pump. With clamp CL-1 regulate the flow of air entering in BR-1 to avoid an excessive foam formation.
    7. Connect peristaltic pump at desired target flow rate.

  5. Operation of the system
    1. Check daily that nutrient stock is not empty. Refill it manually as necessary within a sterile environment. 
    2. The spread of biofilms through the inner surface of the recirculating line (both tubing and equipment) increases the hydrodynamic resistance to overcome, and may partially generate bottlenecks by partial clogging phenomenon. This can generate drops of flow rate in the recirculating line despite the fact flowmeter is set at a desired flow rate at the beginning of the experiment. Check daily that the flow rate value given by the flowmeter FM-1 is the chosen one for the experiment and adjust it as necessary. 
    3. Look for any leak that could appear during the operation due to loose connections. Carefully inspect unexpected spills or liquid drops close to joints. If any, stop the pumps and tighten the joints using the adjustable spanner. Wet the leak with ethanol 70% (v/v), dry it and reconnect the system to check that the issue is fixed.
    4. Once the experiment is finished, replace the culture with a mixture of sterile water and sodium hypochlorite (NaClO at 10% [w/v]) using auxiliary lines 1 and 2 as described previously. Let the system recirculate the bleach for 1 h to sterilize the recirculating loop. Purge the circuit and refill it with distilled water during 1 h to wash the bleach of the system.
    5. Disassembly the circuit, dispose of tubing and biofilm chamber, and store spare parts and pumps.
    6. Repeat all previous steps for every biological replica and every tested experimental condition.

  6. Image acquisition process
    1. At the desired time, stop centrifugal pumps P-1, P-2 and peristaltic pumps PP-1, close valves V-1, V-6 and open valves V-2, V-5. Under sterile conditions, prepare 300 ml of fresh M9 medium + 0.2% (w/v) glycerol in a 500 ml Erlenmeyer flask and connect it in FL-3. Reduce the set flow rate of FM-1 by 50% of the current value and switch on the centrifugal pumps P-1, P-2. This will wash the flow cell chamber with clear nutrient medium, reducing the unspecific bright produced by planktonic bacteria when gathering images. Switch off P-1, P-2 when 400 ml of the liquid has been introduced, avoiding the entrance of air in the circuit, and close valves V-2, V-5.
    2. Close the valves V-3 and V-4 and disconnect them from the circuit using the adjustable spanner. Fill a beaker with ethanol 70% (v/v) and immerse both elbow threaded-to-tube adapters to avoid outer contamination. If the biofilm chamber is placed on the microscope at the beginning of the experiment, skip this step.
    3. Place the biofilm chamber under the microscope and use the bright field to locate the bottom of the material by looking its irregularities. If the chamber material is smooth and transparent, it can be difficult focusing the bottom of the sample (there is no visual reference to guide the user about where to find the position of the bottom): make a small spot with a marker pen at the outer face of the chamber to create a visual reference to follow with the microscope when inspecting the sample.
    4. Images can be taken by picking random snapshots or by performing composed images from merging images spatially distributed through a regular grid of n x m images (i.e., see Figure 4). Choose the region of the biofilm chamber to gather it and spot it with a marker if an evolution of the region is necessary. Otherwise, choose a random point in the chamber.
    5. Once the images have been acquired, reconnect the biofilm chamber to its original position, checking that the orientation of the chamber matches with flow direction, reset the initial flow rate in FL-1 and reconnect the device to continue the experiment (when applies).

      Figure 4. Tile-scan image showing a section of a custom-made chamber. The shape of the device and the whole biofilm pattern colonizing it can be observed only after merging a set of pictures forming a regular grid covering the whole width of the channel.

Data analysis

Once the experiment is finished, a set of images  showing the studied biofilm morphology will be available for analysis, where k, q subscripts and t superscript indicates the k-th image of the experimental set q taken at time t. Variable names in bold style refer to vector type variable, and regular style refers to scalar quantities. Each image will be treated as an n x n matrix in which operations of subtraction, counting and averaging will be performed for every element of these matrices. Every pixel contained in reports a GFP fluorescence associated intensity value captured by the microscope in a certain spatial position. Larger values of intensity are thus directly correlated with larger amounts of detected GFP in every pixel.

  The procedure first treats every image to remove noise and extract relevant variables containing information about the morphology of the biofilm (see Figure 5), which will be further exploited to gather quantitative statistic parameters for analysis. Here is described how to proceed:

  1. Using microscope software or any other image software (i.e., ImageJ), save the images as *.tiff files.
  2. Convert *.tiff files to grayscale images using the chosen image analysis software.
  3. For every set of experimental conditions q, open the t = 0 image () and compute the average (“AVG” operator) of intensity detected in the image, which will be the background signal of the biofilm chamber and nutrient medium. Round the result to the nearest integer (“round” operator) to obtain an integer to avoid decimal numbers when discounting its effect from intensity profiles.

  4. For each available image , do the following loop:
    1. Remove the autofluorescent contribution of the sample by subtracting the auto-fluorescent value  (computed from t = 0 sample) to all pixels of the image.

    2. Assign a value of 0 to all the pixels with negative value after the subtraction.

    3. Compute the basal part  of the image, which is defined as a matrix of the same size that the processed image whose values are all equal to the average of the intensity value registered in all the pixel images (). Conceptually it would be the fraction of the intensity of constant value detected in the image related to the existence of a homogeneous biofilm layer at the bottom of the chamber and the accumulation of GFP within biofilm cells. is computed by first averaging the intensity value of all the pixels in the image and multiplying the obtained scalar value by an n x n matrix filled with ones. is thus a matrix with a homogeneous signal value equal to AVG().

    4. Filter the noise value produced by GFP accumulation within cells and the presence of planktonic cells within biofilm pores. Here, different criteria can be applied depending on the user’s preference. In our case, noise is computed as:

      where, STD() operator computes the standard deviation of the intensity values present in .

    5. The difference between basal layer signal and noise contribution will be the estimation of the real biofilm signal covering the image ().

    6. Drop to zero all negative values of .

    7. Now, compute the fluctuating part;  of the signal by discounting the basal signal from the image , and again drop to zero all negative values.

Figure 5. Filtering and morphological data extraction of grayscale images. Box diagram showing the calculation scheme to extract relevant morphological parameters ( and  variables) for further statistical analysis.

  1. Copy  into a new variable called , in which a binary mask will be created by dropping to zero all values smaller than a chosen threshold th, and setting a value of one otherwise. Choose a threshold value equal to 1.

  2. Label the connected pixels in the binary image using a labeling algorithm included in image processing software, and calculate the area in pixels of every connected region detected in label using a regionprops-like algorithm. Note that the result of this operator is a list with different measurements of detected bodies, and then it is required to select the ‘area’ property. See Note 4 for additional information.

  3. Compute the parameters of interest just by performing selection and counting of labeled bodies. Here four metrics are evaluated: Average intensity of biofilm layer (p1), average cluster size (p2), average percent area occupied by clusters (p3) and number of clusters (p4).

    where phys_ratio is the factor employed to convert a square pixel unit to μm2, which is given once an objective and a zoom is chosen when gathering images with the microscope, the factor n2 is the maximum possible area in pixels to be covered in an image (the total number of pixels forming the image), and the length() operator provides the number of elements of a given list.
  4. For any picture of interest, compute the cluster size distribution in pixels () of labeled objects base of their occupied area. To do this, get the unique values of areas appearing in the area distribution in pixels () and micron2 () and then iterate through a loop to count the number of labeled bodies that have those area values. Inside this loop, row positions are stored in elems by filtering with a comparison operation (step 1), then the labels assigned for those positions in   matrix are gathered from   list (step 2) and a loop is performed to calculate averages of intensity signals for every labeled element, storing each value in an auxiliary vector aux (step 3). At the end of the loop, an average of the computed intensity values (step 4) is taken as central value of intensity for all labeled elements of . Additionally, the number of elements of such area is stored in  (i) to generate the Cluster Size distribution (step 5):

  5. where unique () operator returns the values of a list or vector excluding repeated elements and DistA.index(x) operator provide the vector positions in the list DistA where values contained in it are equal to target value x.
      With these variables, the normalized averaged intensity and the percentage of intensity for all given cluster size distribution can be derived just dividing ave_intensity_prof by the maximum or the sum of the distribution, respectively.
  6. The above data calculation scheme can be applied for the whole image dataset covering all tested experimental condition range. Average and standard deviation parameters can be derived from analyzing every batch of images (having k images collected from different biological replica) for every tested experimental condition q. As a result, bar diagrams (see Figure 6A) showing averages and error bars of these batches of images, and distribution curves for sizes and other variables of interest (see Figure 6B) can be obtained. For more details, see a complete set of results in Espeso et al. (2018).

    Figure 6. Typical bar diagram (A) and distribution curves (B) obtained after performing the data analysis step


  1. When developing this protocol, we found several issues related to microscope inspection. Specifically, the use of a microscope equipped with a galvanometric tray for high sensible displacement constrains the maximum amount of weight to operate on it. The described configuration exceeded such parameters, operating out of security thresholds, which caused several problems of stability. To avoid any risk of damage for the microscope, the user should consider using an alternative approach to reduce the weight of the biofilm chamber. We suggest the replacement of cast iron valves coupled to the flow chamber by plastic valves (PVC, PP), separate valves from biofilm chamber by using flexible tubing that allow supporting valve weight out of galvanometric tray or reducing the size of the whole plate to reduce the overall material weight. Additionally, the availability of user-friendly 3D-printing technology offers the possibility to create custom made chambers without the need of using an industrial milling machine, as it was used here. Manufacturing the model with a water-resistant material (i.e., polycarbonate, polypropylene), further treatment with an organic vapor to seal material pores at microscopic level (see Tsuda et al., 2015 for details) and finally covering it with two transparent covers of the same material may be a good alternative.
  2. Air flow should be complemented with a humidity chamber between the air filter and the entrance of the auxiliary tank to avoid loss of liquid by desiccation. To do this, just use a two or four entrance GL-45 cap with a regular 1 L flask filled with 500 ml of autoclaved water, allowing incoming air to bubble through the water column prior to entering into the system. Connect it to the system by simply adding two pieces of silicon tubing ended with Male and female Luer to 1/8" L barb adapter (sterilization will be performed by autoclaving all together).
  3. In the case of small flow cells, we recommend silicon, however for larger biofilm chambers with larger flow rates a stronger fastening would be required (i.e., o-rings combined with bolts and nuts for larger chambers).
  4. The “regionprops” and “Label” operators here mentioned are functions implemented in image software packages (such as MATLAB or Python) that applies algorithms for labeling detected regions in images considered as “connected”, and measure typical parameters of interest (i.e., centroid, moments, etc.). Depending on the chosen programming language, invoking such functions and provided outputs may vary. Detailed information of the algorithms used by these functions may be found in Haralick and Shapiro, 1992; Jain et al., 1995; Fiorio and Gustedt, 1996; Wu et al., 2005.


In order to prepare the minimal medium, prepare the following stocks separately:

  1. 10x M9 salts stock (for 500 ml of solution)
    42.5 g Na2HPO4·2H2O
    15 g KH2PO4
    2.5 g NaCl
    5 g NH4Cl
    500 ml H2O
  2. Glycerol 20% (w/v) (or any other desired carbon source, for 100 ml)
    23.53 ml 85% (w/v) Glycerol
    76.47 ml H2O
  3. 1 M MgSO4 (for 100 ml of solution)
    12.03 g MgSO4
    100 ml H2O
    Autoclave and store the stocks at room temperature
  4. M9 minimal medium supplemented with 0.2% (w/v) glycerol (1 L)
    100 ml M9 salts 10x
    10 ml Glycerol 20% (w/v)
    2 ml MgSO4 1 M
    888 ml sterile H2O
    Note: All ingredients must be mixed under sterile conditions (if the water is previously autoclaved), or mixed and filtered using a 0.45 μm PES filter unit to avoid contaminations. It is recommended to prepare it when needed to prevent contamination and minimize antibiotic degradation.


This work was funded by the HELIOS Project of the Spanish Ministry of Economy and Competitiveness BIO 2015-66960-C3-2-R (MINECO/FEDER); the ARISYS (ERC-2012-ADG-322797), EmPowerPutida (EU-H2020-BIOTEC-2014-2015-6335536), MADONNA (H2020-FET-OPEN-RIA-2017-1-766975), BioRoboost (H2020-NMBP-BIO-CSA-2018), and SYNBIO4FLAV (H2020-NMBP/0500) Contracts of the European Union and the S2017/BMD-3691 InGEMICS-CM funded by the Comunidad de Madrid (Spain) and the European Structural and Investment Funds.

Competing interests

Authors declare no conflict of interest.


  1. Espeso, D. R., Martinez-Garcia, E., Carpio, A. and de Lorenzo, V. (2018). Dynamics of Pseudomonas putida biofilms in an upscale experimental framework. J Ind Microbiol Biotechnol 45(10): 899-911.
  2. Fiorio, C. and Gustedt, J. (1996). Two linear time Union-Find strategies for image processing. Theor Comp Sci 154: 165-181.
  3. Haralick, R. M. and Shapiro, L. G. (1992). Computer and robot vision. Volume I. Addison-Wesley Longman Publishing Co., Inc. Boston. ISBN: 0201569434.
  4. Jain, R., Kasturi, R. and Schunck, B. G. (1995). Machine Vision. McGraw-Hill, Inc., ISBN: 0-07-032018-7.
  5. Tsuda, S., Jaffery, H., Doran, D., Hezwani, M., Robbins, P. J., Yoshida, M. and Cronin, L. (2015). Customizable 3D printed 'plug and play' millifluidic devices for programmable fluidics. PLoS One 10(11): e0141640.
  6. Wu, K., Otoo, E.J., and Shoshani, A. (2005). Optimizing connected component labeling algorithms. Medical Imaging: Image Processing.





在该协议中,我们描述了用于组装,测试和操作工业类装置以研究平台菌株 Pseudomonas putida mt-2的生物膜扩散模式的程序。该装置的设计和操作程序是专门选择的,以考虑在使用工业材料和设备时通常发现的约束和特征,为给定的研究目的提供可行的解决方案。

关键字:生物膜, 定制设备, 流体系统, 扩散模式, 成像分析, 恶臭假单胞菌


  1. 实验室耗材
    1. 0.4μm孔径腈过滤器(Millipore,目录号:SLHA033SS)
    2. 0.45μmPES过滤器单元(VWR,目录号:514-0335)
    3. 显微镜载玻片(VWR,目录号:16004-422)
    4. 显微镜盖板(VWR,目录号:48393-026)
    5. 铝箔
    6. 标记笔

  2. 管道
    1. 1 x 25 m 5.5 mm内径柔性聚氨酯管(Legris,目录号:1025U0805)
    2. 1 x硅胶管,25 m,2 mm(内径)x 4 mm(od),(VWR,目录号:228-0704)
    3. 1 x硅胶管,4 m,4 mm(内径)x 6 mm(外径),(VWR,目录号:228-0709)

  3. 阀门和夹子
    1. 8 x F-F¼'BSPP球阀(RS,RS Pro,目录号:733-5202)
    2. 1 x ISMATEC 2停止编码管,标识:1.65 mm(蓝 - 蓝,目录号:SC0018),垫子:Tygon R-3607(Cole Parmer,目录号:EW-96460-38)
    3. 2 x软管夹(VWR,目录号:21716-102)

  4. 连接器和配件
    1. 1 x 8 mm-8 mm-8 mm Tee Legris(RS,目录号:445-4495)
    2. 14 x Legris气动直螺纹到管适配器,R 1/4“公头,推入8 mm(RS,目录号:182-4784)
    3. 3个SMC气动直螺纹到管适配器,Rc 1/8“母头,8英寸推入式(RS,目录号:771-5002)
    4. 2个SMC气动弯头螺纹到管适配器,推入8 mm 1/4“螺纹(RS,目录号:771-6030)
    5. 2 x Legris气动弯头螺纹转接件,R 1/4“公x R 1/4”公(RS,产品目录号:367-5792)
    6. 2 x Legris Brass 1/4“in.BSPT公头螺纹接头(RS,目录号:231-5045)
    7. 1个Legris气动三通螺纹管接头,1/4“。x 8 mm x 8 mm,30 bar(RS,目录号:225-0617)
    8. 1个带倒钩的配件,直式连接器,天然PP,1/8“内径; 10 / pk(Cole-Parmer,目录号:EW-51518-05)
    9. 带刺T形连接器,聚丙烯,1/8“; 25 / pk(Cole-Parmer,目录号:SI-50623-66)
    10. 1 x ADCF公Luer至1/8“L倒钩适配器,聚丙烯,25 / pk(Cole-Parmer,目录号:SI-30800-24)
    11. 1 x ADCF母鲁尔至1/8“L倒钩适配器,聚丙烯,25 / pk(Cole-Parmer,目录号:SI-30800-08)

  5. 饰品
    1. 电工胶带,19 mm x 20m(RS,目录号:511-4306)
    2. 插座延长线(RS,目录号:122-1105)
    3. 1 x透明聚碳酸酯PC板,1.25 m x 610 mm x 6 mm(RS,目录号:681-665)
    4. 1 x丁腈橡胶O形圈线,直径2 mm(RS,目录号:138-1600)
    5. 1 x白色PTFE胶带12 mm x 12 m x 0.075 mm(RS,目录号:512-238)
    6. 1 x盘头光亮镀锌钢机螺丝袋,M3,20 mm(RS,目录号:560-625)
    7. 1 x不锈钢,六角螺母袋,M3(RS,目录号:189-563)
    8. 1 x不锈钢平垫圈,0.5 mm厚度,M3(RS,目录号:189-620)
    9. 1 x氰基丙烯酸酯胶(RS,目录号:533-478)

  6. 化学制品
    1. 庆大霉素(Sigma-Aldrich,目录号:PHR1077-1G)
    2. 绝对乙醇(Sigma-Aldrich,目录号:1009832500)
    3. NaClO(VWR,目录号:470302-586)
    4. 甘油85%(Sigma-Aldrich,目录号:1040942500)
    5. Na 2 HPO 4 ·2H 2 O(Sigma-Aldrich,目录号:71643-250G)
    6. KH 2 PO 4 (Sigma-Aldrich,目录号:P5655-500G)
    7. NaCl(Sigma-Aldrich,目录号:S3014-1KG)
    8. NH 4 Cl(Sigma-Aldrich,目录号:A9434-500G)
    9. MgSO 4 (Sigma-Aldrich,目录号:M2643-500G)
    10. 10x M9盐料(参见食谱)
    11. 20%甘油(见食谱)
    12. 1 M MgSO 4 (见食谱)
    13. M9基本培养基补充0.2%(w / v)甘油(见食谱)


  1. 2 x 2 L Pyrex ®窄口锥形瓶(Sigma-Aldrich,目录号:CLS49802L)
  2. 3 x 1 L Pyrex ®窄口锥形瓶(Sigma-Aldrich,目录号:CLS49801L)
  3. 1x Duran TM 4端口组装HPLC螺帽GL-45(Fisher Scientific,目录号:10740834)
  4. 1 x 1 L带螺帽的实验室瓶,DURAN ® GL-45(VWR,目录号:215-1517)
  5. 1 x 5 L BRAND ® PP烧杯,带有喷口,低压型(Sigma-Aldrich,目录号:BR87826-4EA)
  6. 1 x 16 mm长磁力搅拌棒(VWR,目录号:442-0366)
  7. 1 x 250 x 250 mm磁力搅拌器(VWR,目录号:444-0572)
  8. 轨道振动筛
  9. 1个流量计(奥尔堡P-150 mm,1/4“FNTP接头,目录号:P11A3-BB0)
  10. 2 x 12VDC泵(EKWD,EK-DCP 2.2,目录号:3831109862506)
  11. 3 x 220V AC / 12V DC适配器(RS Online,目录号:737-8149)
  12. 1 x 19-37.8升水族箱空气泵(Hagen,A842 Aquaclear Air pump 10,目录号:15561108447)
  13. 1 x Ismatec IPC-N ISM936D IPC-N(ISM936D)低速数字蠕动泵,带有Click-'-Go-Cartridge; 8通道,230V(Cole-Parmer,目录号:EW-78001-16)
  14. 1个可调节扳手(RS,目录号:469-7018) 
  15. 1台EMCO Concept Mill 105铣床
  16. 1x荧光显微镜,具有以下特点:
    1. 光源和滤光片,可以观察GFP和德克萨斯红荧光染料
    2. 带EMCCD芯片的单色数码相机(最低要求)
    3. 电动托盘,用于自动XY位移和自动拼接扫描采集
    4. 自动Z自动对焦


  1. 科学编程语言(即,MATLAB)或通用编程语言(即,Python,C +)能够与支持基本统计功能的科学计算包一起使用( ie ,均值,标准差),数组/结构操作(数组排序,长度)和元素选择运算符(唯一元素)
  2. 所选程序语言的图像处理库(即。,Python的scikit-image,MATLAB的图像处理工具箱)
  3. 包含适当库的图像处理软件(ImageJ或等效物),用于打开具有原生格式(即,*。lif)的显微镜文件,或能够执行常规格式文件存储的显微镜相关软件



  1. 流通池/生物膜室设计与制造

    图1.定制的生物膜室蓝图显示(A)侧视图,(B)前视图和(C)俯视图。根据该蓝图,使用铣床雕刻聚碳酸酯板。当前最先进的制造技术允许使用其他技术(如3D打印)在不同类型的材料中创建此设计或任何其他概念( ie ,PP,PC, 等。)与实验目的一致。尺寸以mm表示。

  2. 装配和设置流体回路
    1. 按照图2(上)所示的分布组织设备,同时考虑设备所在房间的可用空间。图2(下)示出了组装好的装置的斜视图和俯视图。通过这样做,您将意识到需要多少管来连接不同的元件以及如何放置不同的元件以便于所有组装,操作,检查和维护任务。


    2. 带有阀门,流量计和流通池的螺旋管连接器,如图3所示,在外螺纹中使用PTFE带。在不破坏胶带的情况下牢固地拧紧螺纹。然后,切割所需长度的管件并按图示连接它们。


    3. 组装定制O形圈以用于密封生物膜室盖:切割400 mm的2 mm丁腈橡胶O形环绳,并使用一滴氰基丙烯酸酯胶将两端连接起来。等待至少15分钟让胶完全干燥。
    4. 用盖子盖住流动池/生物膜室并使用最合适的系统密封(更多细节见注3)。
    5. 检查所有泵和空气压缩机是否按照回路中设计的流向连接,并标记生物膜室的流入和流出部件(即,盖上的标记用笔,贴纸, etc。)有一个参考并知道它内部流动的方向(因为腔室的几何对称性)。
    6. 准备离心泵的电气连接。手动连接离心泵和AC / DC连接器。为此,请检查AC / DC转换器是否未连接到电网并切断电线的极端以使其空闲。分开正极和负极导线并剥离它们的极端,以便在AC / DC转换器和泵之间进行连接。将来自离心泵和转换器的相同极性的电线连接在一起,扭转它们。用电隔离胶带覆盖去皮部分。或者,如果用户具有压接工具(24AWG),则可以使用对接线接头连接器来连接电线。
    7. 将所有电气设备连接到插座延长线。所有电线必须安全地放置在水接触之外,从而检查电线是否处于升高位置。使用任何可用的方法(胶带悬挂它们,升高的支撑将它们与水面分开)以避免在泄漏或意外溢出时与任何液体接触。
    8. 建议执行泄漏测试以检测错误/不良/失败的连接:
      1. 首先关闭组装回路的所有阀门。
      2. 用蒸馏水填充2L Erlenmeyer烧瓶:这将用作整个实验的培养容器(在图3中标记为BR-1)。打开阀门V-1并通过手动填充蒸馏水灌注离心泵(P-1,P-2),将管道与培养容器(BR-1)连接,以尽可能多地吹扫空气。如果使用蠕动泵泵送水,则不需要最后一步。
      3. 以最大容量(如果有)和再循环管路(V-1,V-3,V-4,V-6)中的所有阀门打开流量计FM-1。打开泵P-1,P-2,检查液体是否完全充满电路。
      4. 让液体流动最少30分钟,并通过检查螺纹连接器中存在的小水滴或电路任何元件下方的水溢出,定期检查没有泄漏。如果检测到泄漏,请停止流动并首先尝试拧紧连接,干燥区域并重新启动流量以查看水是否仍然泄漏。如果泄漏仍然存在,请关闭电流,通过辅助管路2清洗水(参见图2)并卸下受影响的部件。如果损坏,请断开螺钉并更换PTFE胶带。添加额外的PTFE胶带层以在需要时增加隔离。
      5. 泄漏测试结束后,关闭阀门V-1,V-3,V-4和V-6,并手动清空培养皿。

  3. 最初的细菌培养和库存
    1. 在实验开始前一天(第0天),接种50ml Erlenmeyer烧瓶,其中含有10ml补充有0.2%(w / v)甘油作为碳源的M9基本培养基和0.5μg/ ml庆大霉素。直接加入-80°C的 P冷冻甘油原液。 putida mt-2。
    2. 在设定为170rpm的定轨振荡器中于30℃孵育过夜。
    3. 在实验开始当天(第1天),测量培养物的OD 600 并将其稀释在1L新鲜M9 + 0.2%(w / v)甘油中以达到OD 600 0.0025。
    4. 在启动装置之前,准备一个4 L的M9 + 0.2%(w / v)甘油补充了庆大霉素(0.5μg/ ml)的无菌原液。

  4. 启动
    1. 在开始实验之前,必须对流体回路进行灭菌。为了做到这一点:
      1. 使用带刺倒刺的直连接器连接对应于营养/废物蠕动泵送管线PP-1的ISMATEC管道和2 mm硅胶管部件。还连接2毫米管线,空气将通过公鲁尔至1/8“L倒钩适配器将空气传导至主烧瓶。使用铝箔分别覆盖两个管线的极端,并对它们进行高压灭菌。
      2. 将500ml 1L Erlenmeyer烧瓶FL-3和700ml BR-1容器装入70%乙醇(v / v)中。在BR-1中引入磁力搅拌棒并将FL-3连接到辅助管线1.检查循环回路中的阀门V2,V-3,V-4和V-6是否打开,V-1和V-5是关闭,使流量计也以最大容量打开,并打开离心泵。将储存的乙醇在FL-3中转移至BR-1,直至大部分烧瓶为空,然后关闭泵P-1,P-2。这将用乙醇填充辅助管线1。关闭V-2并打开V-1。然后再次打开泵P-1和P-2,从而在60分钟内完成系统的灭菌。
      3. 停止离心泵P-1,P-2,打开阀门V-5,关闭V-6并打开P-1和P-2,让乙醇离开系统。必须通过让泵尽可能地清空培养容器BR-1来进行乙醇的去除。当泵仍然工作时,避免其全部空,因为这会在循环管路中引入空气并导致泵中的启动问题。 
      4. 一旦BR-1完全排空(并且没有空气仍然进入回路,如上所述),则断开P-1和P-2,并用1L无菌的另一个更换1L Erlenmeyer烧瓶FL-3。水。再次连接P-1,P-2,用大约100毫升的水清洗再循环管线中剩余的乙醇。关闭V-5并打开V-6让淡水进入主水箱并在1小时内清洗乙醇。
      5. 重复先前描述的乙醇 - 水替代操作,但使用1.2L新鲜M9培养基补充0.2%(w / v)甘油代替系统中的水,最后在BR-1中留下1L新鲜培养基。 
    2. 此时将调整显微镜设置。为图像选择12位或16位分辨率,以获得更高的灰度值灵敏度。选择通道来检测研究菌株产生的荧光信号(这里我们推荐GFP,因为它的稳定性和性能)。取10μl饱和过夜培养物,用移液管尖端将其扩散到显微镜载玻片上。将盖玻片放在样品上并用显微镜检查。调整显微镜的增益和激光强度以获得非饱和图像。保持这些设置,因为它们将通过实验保持不变 
    3. 关闭泵P-1,P-2和PP-1,关闭所有阀门,拆卸生物膜室BC-1并按照图像采集部分中详述的步骤将其放置在显微镜下。在生物膜室中收集图像(如果检测到任何信号,则它将由材料和/或液体介质自发荧光引起)。这些图像将用于设置背景噪声阈值。
    4. 取下铝箔,用蠕动泵PP-1连接营养液FL-1和废液罐FL-2,用BR-1连接各自的输出。将管道放入烧瓶内时,避免管道与任何表面接触,以防止污染。将ISMATEC停止编码管放入泵盒中,检查它们是否正确固定。
    5. 将4 mm硅胶管件连接到气泵的输出口。添加一个三通并在90°分支处连接一个带夹具CL-1的硅胶管,用于调节空气流入。然后将注射器过滤器的鲁尔侧连接到湿度箱HC-1。最后,通过在烧瓶中引入管的自由侧将HC-1的输出管与BR-1连接。有关其他注释,请参见注释2。
    6. 以150rpm开启磁力搅拌器,并连接空气泵。使用夹具CL-1调节进入BR-1的空气流量,以避免形成过多的泡沫。
    7. 以期望的目标流速连接蠕动泵。

  5. 系统的运作
    1. 每天检查营养素库存是否为空。必要时在无菌环境中手动重新填充。 
    2. 生物膜通过再循环管线(管道和设备)的内表面的扩散增加了要克服的流体动力学阻力,并且可能通过部分堵塞现象部分地产生瓶颈。尽管事实上流量计在实验开始时设定在所需的流速,但这可以在再循环管线中产生流速下降。每天检查流量计FM-1给出的流量值是实验选择的流量值,并根据需要进行调整。 
    3. 查看由于连接松动而在操作过程中可能出现的任何泄漏。仔细检查接近关节的意外溢出物或液滴。如果有的话,停止泵并使用可调扳手拧紧接头。用乙醇70%(v / v)润湿泄漏,干燥并重新连接系统以检查问题是否已解决。
    4. 实验结束后,如前所述,使用辅助管线1和2,用无菌水和次氯酸钠(10%[w / v]的NaClO)的混合物替换培养物。让系统将漂白剂再循环1小时以对再循环回路进行灭菌。清洗回路并在1小时内用蒸馏水重新加注,以清洗系统的漂白剂。
    5. 拆卸回路,处理油管和生物膜腔,并存放备件和泵。
    6. 对每个生物复制品和每个测试的实验条件重复所有先前的步骤。

  6. 图像采集过程
    1. 在所需的时间,停止离心泵P-1,P-2和蠕动泵PP-1,关闭阀门V-1,V-6并打开阀门V-2,V-5。在无菌条件下,在500ml Erlenmeyer烧瓶中制备300ml新鲜M9培养基+ 0.2%(w / v)甘油,并将其连接到FL-3中。将FM-1的设定流量降低当前值的50%,然后打开离心泵P-1,P-2。这将用清洁的营养培养基清洗流动池室,减少浮游细菌在收集图像时产生的非特异性亮度。当引入400毫升液体时,关闭P-1,P-2,避免回路中的空气进入,并关闭阀门V-2,V-5。
    2. 关闭阀门V-3和V-4并使用可调扳手将它们与电路断开。用70%(v / v)的乙醇填充烧杯并浸入两个弯头螺纹 - 管接头,以避免外部污染。如果在实验开始时将生物膜室放在显微镜上,请跳过此步骤。
    3. 将生物膜室置于显微镜下,并通过观察其不规则性使用明场来定位材料的底部。如果腔室材料光滑透明,则可能难以聚焦样品底部(没有视觉参考来指导用户在哪里找到底部位置):用记号笔在底部做一个小点在检查样品时,腔室的外表面创建视觉参考以跟随显微镜。
    4. 可以通过挑选随机快照或通过合并通过n×m图像的规则网格空间分布的图像来执行合成图像(即,参见图4)。如果需要进行区域演化,请选择生物膜室的区域来收集它并用标记物进行识别。否则,在腔室中选择一个随机点。
    5. 获取图像后,将生物膜室重新连接到其原始位置,检查室的方向是否与流动方向匹配,重置FL-1中的初始流速并重新连接设备以继续实验(适用时)。



实验结束后,一组图像将被视为n×n矩阵,其中将对这些矩阵的每个元素执行减法,计数和平均操作。 报告显微镜在特定空间位置捕获的GFP荧光相关强度值。因此,较大的强度值与每个像素中较大量的检测到的GFP直接相关。


  1. 使用显微镜软件或任何其他图像软件(即,ImageJ),将图像保存为* .tiff文件。
  2. 使用所选的图像分析软件将* .tiff文件转换为灰度图像。
  3. 对于每组实验条件q,打开t = 0图像()并计算图像中检测到的强度的平均值(”AVG“运算符),这将是生物膜室和营养培养基的背景信号。将结果舍入到最接近的整数(“round”运算符)以获得一个整数,以便在从强度曲线中折扣其效果时避免使用十进制数。
  4. 对于每个可用图像,执行以下循环:
    1. 通过减去自动荧光值来消除样品的自发荧光贡献 (从t = 0样本计算)到图像的所有像素。
    2. 在减法后为所有具有负值的像素指定值0。
    3. 计算基础部分图像,定义为与处理后的图像大小相同的矩阵,其值均等于所有像素图像中登记的强度值的平均值()。从概念上讲,它将是在图像中检测到的恒定值强度的一部分,其与腔室底部的均匀生物膜层的存在和GFP在生物膜细胞内的积累有关。 是一个具有等于AVG的同质信号值的矩阵()。
    4. 过滤细胞内GFP积累产生的噪音值和生物膜孔内浮游细胞的存在。这里,可以根据用户的偏好应用不同的标准。在我们的例子中,噪声计算如下:


    5. 基底层信号和噪声贡献之间的差异将是覆盖图像的真实生物膜信号的估计()。
    6. 将所有负值降至零
    7. 现在,计算波动部分;  ,再次将所有负值降至零。

图5.灰度图像的滤波和形态数据提取。框图显示了提取相关形态参数的计算方案( 和  变量)用于进一步的统计分析。

  1. 复制  进入一个名为,其中将通过将小于所选阈值 th 的所有值降至零来创建二进制掩码,否则将值设置为1。选择等于1的阈值。
  2. 使用图像处理软件中包括的标记算法标记二进制图像中的连接像素,并使用类似于区域提取的算法计算在标签中检测到的每个连接区域的像素区域。请注意,此运算符的结果是包含对检测到的实体的不同度量的列表,然后需要选择“area”属性。有关其他信息,请参见注释。
  3. 仅通过对标记物体进行选择和计数来计算感兴趣的参数。这里评估了四个指标:生物膜层的平均强度(p1),平均簇大小(p2),簇占据的平均面积百分比(p3)和簇数(p4)。
    其中 phys_ratio 是用于将正方形像素单位转换为μm 2 的因子,一旦在使用显微镜采集图像时选择了物镜和变焦,就会给出因子n 2 是图像中要覆盖的像素的最大可能区域(形成图像的像素总数),length()运算符提供给定列表的元素数。
  4. 对于任何感兴趣的图片,以像素为单位计算聚类大小分布()标记对象基于其占用区域。为此,请获取区域分布中出现的区域的唯一值(以像素为单位)()和micron 2 (“”style)然后遍历循环以计算具有这些区域值的已标记实体的数量。在此循环中,行位置存储在 elems 中,通过比较操作(步骤1)进行过滤,然后为 中的那些位置分配标签。  矩阵从 收集。此外,此类区域的元素数量存储在  (i)生成群集大小分布(步骤5):

  5. 其中unique()运算符返回列表或向量的值,不包括重复元素,DistA.index(x)运算符提供列表DistA中的向量位置,其中包含的值等于目标值x。
  6. 上述数据计算方案可应用于覆盖所有测试实验条件范围的整个图像数据集。对于每个测试的实验条件q,可以从分析每批图像(具有从不同生物复制品收集的k个图像)导出平均和标准偏差参数。结果,可以获得显示这些批次图像的平均值和误差条的条形图(参见图6A),以及尺寸和其他感兴趣的变量的分布曲线(参见图6B)。有关详细信息,请参阅Espeso等人的完整结果集。 (2018)。


  1. 在制定该协议时,我们发现了与显微镜检查有关的几个问题。具体而言,使用配备有用于高显灵位移的检流盘的显微镜限制了在其上操作的最大重量。所描述的配置超出了这些参数,在安全阈值之外操作,这导致了几个稳定性问题。为避免显微镜损坏的任何风险,用户应考虑使用另一种方法来减轻生物膜腔的重量。我们建议用塑料阀门(PVC,PP)更换连接到流动室的铸铁阀门,通过使用柔性管道将生物膜腔室中的单独阀门更换,从而允许将阀门重量从测量托盘中移出或减小整个板材的尺寸以减少整体材料重量。此外,用户友好的3D打印技术的可用性提供了创建定制腔室的可能性,而无需使用工业铣床,如此处所使用的那样。使用防水材料(即,聚碳酸酯,聚丙烯)制造模型,使用有机蒸汽进一步处理以在微观水平密封材料孔(详见Tsuda等,2015),最后用两个相同材料的透明盖子覆盖它可能是一个很好的选择。
  2. 空气流量应辅助空气过滤器和辅助水箱入口之间的湿度箱,以避免干燥造成液体流失。要做到这一点,只需使用一个两个或四个入口GL-45盖子和一个装有500毫升高压灭菌水的普通1升烧瓶,允许进入的空气在进入系统之前通过水柱鼓泡。通过简单地将两片硅管连接到公母和鲁尔接头到1/8“L倒钩接头(将通过高压灭菌一起进行灭菌),将其连接到系统。
  3. 在小流通池的情况下,我们建议使用硅,但是对于流速较大的较大生物膜腔室,需要更强的固定(即,O形环与更大腔室的螺栓和螺母结合)。
  4. 这里提到的“regionprops”和“Label”操作符是在图像软件包(例如MATLAB或Python)中实现的函数,它们应用算法来标记被视为“连接”的图像中的检测区域,并测量感兴趣的典型参数(即,质心,时刻,等。)。根据所选的编程语言,调用这些函数和提供的输出可能会有所不同。这些函数使用的算法的详细信息可以在Haralick和Shapiro,1992; Jain等,1995; Fiorio和Gustedt,1996年;吴等人,2005年。



  1. 10x M9盐原料(500 ml溶液)
    42.5 g Na 2 HPO 4 ·2H 2 O
    15克KH 2 PO 4
    5克NH 4 Cl
    500毫升H 2 O.
  2. 甘油20%(w / v)(或任何其他所需碳源,100 ml)
    23.53毫升85%(w / v)甘油
    76.47ml H 2 O.
  3. 1M MgSO 4 (对于100ml溶液)
    12.03g MgSO 4
    100毫升H 2 O
  4. M9基本培养基补充0.2%(w / v)甘油(1L)
    10毫升甘油20%(w / v)
    2毫升MgSO 4 1 M
    888ml无菌H 2 O


这项工作由西班牙经济和竞争力部BIO 2015-66960-C3-2-R(MINECO / FEDER)的HELIOS项目资助; ARISYS(ERC-2012-ADG-322797),EmPowerPutida(EU-H2020-BIOTEC-2014-2015-6335536),MADONNA(H2020-FET-OPEN-RIA-2017-1-766975),BioRoboost(H2020-NMBP-) BIO-CSA-2018)和SYNBIO4FLAV(H2020-NMBP / 0500)欧洲联盟和由马德里共和国(西班牙)和欧洲结构和投资基金资助的S2017 / BMD-3691 InGEMICS-CM合同。




  1. Espeso,D.R.,Martinez-Garcia,E.,Carpio,A。和de Lorenzo,V。(2018)。 高级实验框架中 Pseudomonas putida 生物膜的动态。 J Ind Microbiol Biotechnol 45(10):899-911。
  2. Fiorio,C。和Gustedt,J。(1996)。 两种线性时间联合查找图像处理策略。 Theor Comp Sci 154:165-181。
  3. Haralick,R。M.和Shapiro,L。G.(1992)。 计算机和机器人愿景。第一卷Addison-Wesley Longman Publishing Co.,Inc 波士顿 ISBN:0201569434。
  4. Jain,R.,Kasturi,R。和Schunck,B。G.(1995)。 机器视觉。 McGraw-Hill,Inc。,ISBN:0-07-032018-7。
  5. Tsuda,S.,Jaffery,H.,Doran,D.,Hezwani,M.,Robbins,P.J。,Yoshida,M。和Cronin,L。(2015)。 可自定义3D打印'即插即用'可用于可编程流体的毫流量设备。 PLoS One 10(11):e0141640。
  6. Wu,K.,Otoo,E.J。和Shoshani,A。(2005)。 优化连接组件标签算法。 医学影像:图像处理。
  • English
  • 中文翻译
免责声明 × 为了向广大用户提供经翻译的内容, 采用人工翻译与计算机翻译结合的技术翻译了本文章。基于计算机的翻译质量再高,也不及 100% 的人工翻译的质量。为此,我们始终建议用户参考原始英文版本。 Bio-protocol., LLC对翻译版本的准确性不承担任何责任。
Copyright: © 2019 The Authors; exclusive licensee Bio-protocol LLC.
引用:Espeso, D. R., Martínez-García, E. and Lorenzo, V. d. (2019). Assembly of a Custom-made Device to Study Spreading Patterns of Pseudomonas putida Biofilms. Bio-protocol 9(10): e3238. DOI: 10.21769/BioProtoc.3238.