We would like to share our newly-developed procedure for cleaning the FACSAria cuvette flow cell with the community of FACSAria users, as we believe that it can be handy and helpful. It has been stated, for example by Howard M. Shapiro, that any irregularities of the flow cell's wall or adherent particles may produce turbulence and any protein residue can affect the optical quality of the cuvette (1). Any contamination of the cuvette flow cell in flow cytometers, for example by proteins, antibodies or some fluorescent dyes, can therefore represent a serious problem that may introduce turbulences and/or increase background fluorescence and cause light scattering. This may logically result in higher CV and lower precision.
We have encountered similar complications recently when using Hoechst 33342 and propidium iodide staining in our experiments. As the samples stained with Hoechst 33342 were measured in large quantities on a daily basis and the propidium iodide was added shortly prior to measurement for viability staining, the potential for flow cell contamination was relatively high. Indeed, it was manifested as increased Bright Bead %Robust CV (%rCV) (a population CV calculated based on median instead of mean) (2), especially in channels related to UV or violet lasers, subsequently leading to problems passing the CS&T procedure. Long-lasting washings with sodium hypochlorite (BD CLEAN), BD RINSE (both BD Biosciences) or potassium hydroxide containing Contrad 70 (up to 15%; Decon Labs, King of Prussia, PA) solutions run as samples produced some improvement temporarily but were rather time-consuming. There are also other approaches employing for example 0.1 N sodium hydroxide followed by CoulterCleanse solution and distilled deionized water (Mario Roederer, Purdue Cytometry Mailing List, August 28, 2001; https://lists.purdue.edu/pipermail/cytometry/2001-August/02 0245.html). However, Mario Roederer himself and also Howard Shapiro advise making sure whether there is any part of the instrument's fluidic system that will not stand up to the sodium hydroxide solution (1). According to industrial standards of cleaning and our previous experience, we presumed that turbulent convection within the flow cell should produce more effective washing in comparison to linear flow. The following cleaning procedure has proved to be simple and effective in maintaining the condition of the BD FACSAria cuvette flow cell, and can be easily executed on a daily basis.
Protocol (a schematic illustration is presented in Fig. 1E):
3Execute “Clean flow cell” procedure2 with 15% Contrad 70 solution.
4Turn the stream on without any nozzle and activate any dummy sample from the sample list but insert in an empty tube.
5Set the highest flow rate value. Generation of bubbles in the cuvette of the instrument should be registered within 30–60 s. The empty tube can be run as long as the bubbles are created and driven out through the cuvette, or for at least 3–5 min.
6Shut the stream off; execute the “clean flow cell” procedure again with BD CLEAN solution instead of Contrad 70 solution.
7Repeat the step 4.
8Set the highest flow rate value. Generation of bubbles in the cuvette of the instrument should be registered within 30–60 s. Run the empty tube for at least 3–5 min.
9Turn the stream off, insert the correct nozzle and turn the stream on again. Let the stream run for at least 3–5 min to clean the flow cell. Meanwhile execute the “Sample line backflush” procedure and let it run for at least 1–2 min to remove the residuals of cleaning solutions from the sample line.
10Turn the stream off and on again to remove any bubbles that might be left in the flow cell. “CS&T” procedure3 can be run now.
Bubbles that are created by air entering the flow cell instead of the sample are driven through the flow cell generating strong turbulent convection that helps to clean the walls of the cuvette. If the generation of bubbles is not sufficient, the empty tube can be unloaded and reloaded again. Similarly, sodium hypochlorite is not a detergent, so problems with insufficient generation of bubbles when running the empty tube as a sample after BD CLEAN solution cleaning may occur. Repeated unloading and reloading of the sample will help to generate more bubbles.
As laser illumination of the cuvette during the cleaning procedure may lead to undesired fixing of the residues to the cuvette wall, it is recommended to avoid this for example by keeping the laser-shutter closed during the whole cleaning procedure. FACSAria users can achieve this by keeping the flow cell access door (hood) open. Moreover, it is said that there is a short delay in laser-shutter closing when switching off the stream on the FACSAria instrument (personal communication with BD technicians, advisers and product specialists). Sample line clog can destabilize the laminar flow and that can be the moment when the turbulent flow brings the sample residues into direct contact with the cuvette flow cell wall, where they can be fixed to the wall by the laser. This cannot be prevented in the case of a clog, but FACSAria users can eliminate short illuminations of the “naked” cuvette at regular stream switch-off by opening the hood first.
Despite the recommendations of the sorters manufacturer (BD) (3), there are some doubts in the community about the usage of potassium hydroxide containing solutions (including Contrad 70) for cleaning cuvette flow cells. According to personal communications with multiple users, we have to warn those interested about a theoretical risk that some procedures using highly concentrated potassium/sodium hydroxide solutions and/or prolonged exposure can affect the lifespan of the cuvette flow cell. Our proposed protocol exposes the cuvette to 15% Contrad 70 solution only for a few seconds, followed by extensive washing with sheath fluid and turbulent convection. However, as the efficiency of the presented protocol is based on turbulent convection in the flow cell and principally should not depend only on the constitution of the cleaning solution, we support our concept with a simple experiment using 0.1% sodium dodecyl sulphate (SDS) instead of 15% Contrad 70.
A newly-installed cuvette flow cell (BD FACSAria Instrument, Dr. Šinkorová Lab) was contaminated on purpose using the following procedure: fixed blood leukocytes (prepared by blood lysis with BD FACS Lysing Solution) were stained with propidium iodide (10 mg/ml), the suspension was sucked into the cuvette using the “Clean flow cell” procedure and the cuvette was irradiated with lasers for 1 min (near UV 377 nm–7 mW; sapphire 488 nm–20 mW, HeNe 633 nm–20 mW). The instrument was cleaned according to the above-mentioned protocol using 0.1% SDS instead of 15% Contrad 70. The CS&T procedure was executed before the contamination (initial), after the contamination (contaminated), after cleaning with 0.1% SDS according to steps 3–5, 9, and 10 (0.1% SDS), and after the cleaning with CLEAN solution according to steps 6–10 (CLEAN). As the fluorescent channels related to blue 488 nm laser were affected in the largest extent, only these are presented (Fig. 1F). It is evident that even cleaning with 0.1% SDS was significantly effective and removed the contamination.
Results and Discussion
The results of CS&T procedures executed between February and July 2012 in our laboratory presented as variations in Bright Bead %Robust CV (%rCV) for UV and violet lasers can be seen in Figures 1A–1D. Although some increases were visible in the case of blue-488 nm and red-635 nm lasers as well, these were not as dramatic and were kept quite well under control using standard long-lasting cleaning procedures. Logical elimination identified contamination with fluorescent dyes (Hoechst 33342 and propidium iodide), as we registered the rise in %rCV only in corresponding channels and we were able to lower these %rCV values with extensive washing. The values of %rCV, especially the Hoechst Blue and Hoechst Red channels (Figs. 1A and 1B), peaked frequently as we were trying to clean the cuvette flow cell using various procedures based on long-lasting (30 min or more) cleanings with 15% Contrad 70, BD RINSE, BD CLEAN or ethanol or their combinations, and washings with laminar flow. Some of these procedures (e.g., 15% Contrad 70/BD RINSE – 30 min each) successfully lowered the %rCV value (e.g., CS&T 26th to 29th); however, these were time-consuming as they took over an hour. An attempt to shorten the procedure resulted in failure (e.g., CS&T 30th to 37th). When we started using propidium iodide for viability measurements more extensively (approx. CS&T 31st), we registered dramatically increased %rCVs also in the violet-405 nm laser related channels, especially in the Pacific Blue, AmCyan and Qdot605 (Fig. 1C). Another procedure recommended by our technician based on BD CLEAN solution preheated to 50°C and the “Clean flow cell” procedure repeated 3 × 10 min was effective only in the case of parameters related to the violet-405 nm laser (CS&T 35th and 36th) (Fig. 1C) and led to some leakage of the solution and salt deposit accumulation around the screw at the entry of the sample line into the cuvette flow cell. Although perfectly laminar flow is elementary for flow cytometer operation, turbulent flow is one of the essential principles of the “clean-in-place” procedures that are applied in industrial cleaning, especially for tubing and pipelines, and is considerably more effective than laminar flow (4). In laminar flow, the velocity of a fluid is highest along the axis, but there is actually a thin boundary layer of water that does not move at the walls of the tube. As a result a so-called parabolic profile of flow velocities is created (1). Cleaning the flow cell with laminar flow is therefore less effective as the velocity of sheath fluid in close proximity to the wall actually reaches zero. Turbulent convection offers higher efficiency as the velocity of fluid can hit the walls with much higher power. Indeed, we had previously established our own procedures for effective cleaning of our BD FACSCalibur instrument based on turbulent convection (unpublished results); we therefore continued trying to find a way of achieving the same results without interfering with the FACSAria's fluidics. We went through various optimization steps as we finally established the above-mentioned protocol. Starting from the 47th CS&T in the series we saw low and stable %rCV values in all affected channels (Figs. 1B and 1D), and slight improvement in all the other channels as well. We also started using this procedure at the end of the measurements before the “Fluidics shut down” procedure.
In response to the above-mentioned doubts on the usage of potassium hydroxide-containing solutions, we successfully tested our proposal using 0.1% SDS solution instead of 15% Contrad 70.
We have established a quick and easy procedure for FACSAria cuvette flow cell cleaning effective even in case of heavy contamination by Hoechst 33342 and propidium iodide dyes. This procedure helps to maintain low and over-time stable %rCV. There is the possibility of extending the individual steps or of adding another cleaning solution in order to optimize the protocol according to the type of contamination. Since there is some empirical experience and doubts that Contrad 70 or other potassium hydroxide-containing cleaning solution can be aggressive towards the flow cell, we also tested 0.1% SDS or BD RINSE solutions instead and we found them able to remove contamination or at least to maintain the purity of the cuvette flow cell. We can recommend establishing a protocol individually for every lab according to the samples that are analyzed/sorted using milder approaches first and turning to stronger cleaning solutions when the instrument's performance gets affected. However, it has to be emphasized that the presented protocol exposes the cuvette flow cell directly to diluted Contrad 70 only for few seconds, and it also omits exposure of the nozzle sealing that can be negatively affected as well. This is negligible in comparison to some previously recommended protocols that used even undiluted Contrad 70 for several minutes, or its 15% dilution as the “cleaning solution” for shutdown procedure (3). Ultimately, to our knowledge there is no direct evidence that makes Contrad 70 responsible for cuvette flow cell damage.
Although sodium hypochlorite-containing solutions (e.g., FACSClean) are aggressive, the tubing and cuvette of FACSAria are fairly resistant and therefore these solutions should be safe, as they are, for example, the recommended ones for “Prepare for aseptic sort” procedure (5). Nevertheless, it is important to note that sodium hypochlorite can be rather aggressive towards the rubber washers, so even if the presented procedure can also be run with the nozzle in place, this approach should be used carefully and executed only when needed and it should not be used with Contrad 70 solution. In any case, the nozzle can also be cleaned using other, standard methods. It is also important to remember that the BD CLEAN solution containing sodium hypochlorite is aggressive towards any type of steel, so the closed loop nozzle must not be used in combination with BD CLEAN solution or other sodium hypochlorite-containing solution.
We believe that this procedure can also be adapted to other types of cuvette-based sorters. If this procedure or a modification is successfully adopted by FACSAria users, we recommend that the manufacturer implement it into new versions of the software, as it only requires integration of a simple step-by-step wizard such as those which are already part of the FACSDiva software. We also believe that development of new cleaning procedures based on turbulent convection can help to improve cleaning efficiency and cuvette life-time in multiple types of flow cytometers. It has to be kept in mind that, however, construction of the cuvette flow cell in FACSAria cell sorters is inverted in contrast to most other cytometers, so it allows accumulation of foam in the cone of the flow cell and consequent releases in pulses causing strong turbulent convection. For this reason, a different approach will be needed for cleaning other flow cytometers.
BD FACSAria II SORP (Institute of Biology and Ecology, Faculty of Science, P. J. Šafárik University in Košice):
UV laser (355 nm, 60 mW)—(A) Hoechst Red (DM635LP-670 LP) and (B) Hoechst Blue (405/50); Violet laser (405 nm, 100 mW)—(A) Qdot800 (DM750LP-780/60), (B) Qdot700 (DM685LP-710/50), (C) Qdot655 (DM630LP-660/20), (D) Qdot605 (DM600LP-610/20), (E) AmCyan (DM505LP-525/50), and (F) Pacific Blue (450/50); Blue laser (488, 100 mW)—(A) PE-Cy7 (DM750LP-780/60), (B) PerCP-Cy5-5 (DM685LP-710/50), (C) PerCP (DM670LP-685/35), (D) PE-Texas Red (DM600LP-610/20), (E) PE (DM550LP-575/25), (F) FITC (DM505LP-525/50), and (G) SSC (488/10); Red laser (640 nm, 40 mW)—(A) APC-Cy7 (DM750LP-780/60), (B) Alexa Fluor 700 (DM685LP-710/50), and (C) APC (670/14).
BD FACSAria II (Department of Radiobiology, Faculty of Military Health Sciences, University of Defence, Hradec Králové):
Near UV laser (375 nm, 7 mW)—(A) Alexa Fluor 430 (DM502LP-530/30) and (B) Hoechst Blue (450/40); Blue laser (488 nm, 20 mW)—(A) PE-Cy7 (DM735LP-780/60), (B) PerCP-Cy5-5 (DM655LP-695/40), (C) PE-Texas Red (DM610LP-616/32), (D) PE (DM556LP-585/42), (E) FITC (DM502LP-530/30), and (F) SSC (488/10); Red laser (633 nm, 18 mW)—(A) APC-Cy7 (DM735LP-780/60) and (B) APC (660/20).
The authors are grateful to Andrew J. Billingham for the proofreading.
1Shapiro H. Practical Flow Cytometry, 4th ed.Hoboken, NJ: Wiley; 2003. p 168, 179, 444.
Closed loop nozzle: Nozzle connected to the airline of the sorter for routine cleaning of the cuvette. It also seals the cuvette after shut down to prevent its drying.
Clean flow cell procedure: Procedure for washing the cuvette with cleaning solution in the whole volume of the cuvette. The cleaning solution is directly sucked into the flow cell.
CS&T procedure: Cytometer setup and tracking (CS&T) procedure is run daily or whenever the cytometer configuration is changed in order to set and track cytometer parameters (laser delays, PMT voltages, cytometer performance, etc.).