Question: How can I reduce background from unbound CMOs when using the 3’ CellPlex kit for Cell Multiplexing?
Answer: Below, we list several tips and best practices for obtaining optimal signal-to-noise ratios in 3' CellPlex data. These tips and best practices are also highlighted in our Technical Note on Cell Multiplexing and our Demonstrated Protocol for CMO labeling.
For further information on how poor signal-to-noise ratios can impact CMO tag assignment and other metrics in the CellPlex Web Summary file, please see: How does high background from unbound CMOs affect Cell Multiplexing metrics in 3’ CellPlex experiments?
Tips and Best Practices for CellPlex experiments:
1) Start with high-quality cells/nuclei
Sample quality is critical for a successful CellPlex experiment. Poor quality samples may lead to high background noise in CellPlex data.
Cells: Input single-cell suspensions should have >80% (ideally >90%) viability. Samples with viability below 80% may be used if flow sorting is performed to increase viability to >90%. Very fragile sample types may not be suitable for the 3’ CellPlex assay, as they may not withstand the additional handling steps associated with CMO labeling and washing. For these sample types, it may be more appropriate to run the standard Single Cell 3’ Gene Expression assay without Cell Multiplexing.
Nuclei: It is critical to start with high-quality single nuclei suspensions that do not show signs of nuclear membrane blebbing when examined under a 40x or 60x microscope. The CellPlex assay has only been validated on nuclei isolated using protocol CG000124. Nuclei isolated using alternative protocols or nuclei isolated from frozen tissues have not been tested for compatibility with the 3’ CellPlex assay. When working with a new tissue type, perform pilot experiments to assess nuclei quality after CMO labeling and washing. If nuclei clumping or low nuclei recovery are observed, optimization of upstream nuclei isolation protocols (eg. lysis time) is recommended. See: Which nuclei isolation protocols are supported for use with the 3' CellPlex Kit for Cell Multiplexing?
2) Remove debris before or after CMO labeling
If your sample contains debris consisting of cell/nuclei fragments, the CMO lipids may bind to the debris. If debris is not removed before loading the cells/nuclei onto the 10x Genomics chip, this can lead to high background noise levels in the Gene Expression data and CellPlex data.
We recommend fully removing debris before loading the sample on the 10x chip. An excellent method for removing debris is to perform flow sorting, which can be performed on cells or nuclei after CMO labeling, as described here. During in-house tests, we found that using FACS to remove debris and dead/dying cells improved signal-to-noise ratios in CellPlex data and improved Gene Expression data.
3) Keep CMO labeled cells/nuclei on ice and load as soon as possible after pooling
Keep cells/nuclei on ice after each wash step and after completion of the CMO labeling and washing protocol. Do not let CMO labeled cells/nuclei sit at room temperature, as this may increase background in 3’ CellPlex data. If performing flow sorting, ensure that the sample chamber and collection chamber of the FACS instrument are chilled.
Pool cells/nuclei as soon as possible after CMO labeling and washing (within 30 minutes). After pooling, load cells/nuclei onto the 10x Genomics chip as soon as possible (within 30 minutes). It is critical to work efficiently and avoid letting cells/nuclei sit for extended periods, as this can decrease cell viability or nuclei quality and can lead to increased background noise in 3’ CellPlex data.
4) Fully remove the supernatant after each wash step
Remove as much supernatant as possible after each wash step post CMO labeling. If the supernatant is not completely removed, unbound CMOs remaining in the sample will lead to high background noise levels in the final CellPlex data.
We recommend leaving no more than 10 ul of supernatant after each wash step, as illustrated here:
5) Do not reduce the number or volume of wash steps
We recommend performing three total wash steps post CMO labeling. This includes adding 1.9 ml wash buffer immediately after CMO incubation (wash 1), followed by two additional wash steps with 2 ml wash buffer (wash 2 & 3).
If flow sorting is performed, it is possible to reduce the number of wash steps prior to flow sorting (one wash step instead of three), given that the flow sorting step itself acts as a very efficient wash step. Please refer to Protocol 3 and Protocol 4 in the CMO labeling Demonstrated Protocol for guidance on reducing the number of wash steps prior to flow sorting.
If flow sorting is not performed, we do not recommend performing fewer wash steps or reducing the volume of wash buffer, as this may lead to insufficient removal of background from unbound CMOs.
6) Use at least 100,000 cells/nuclei as input into CMO labeling, or 500,000 if possible
When working with low cell/nuclei numbers, a pellet may not be visible after centrifugation, making it challenging to fully remove the supernatant after each wash step. We do not recommend or support using fewer than 100,000 cells/nuclei as input into CMO labeling*. Furthermore, if the sample is not limited, we recommend using at least 500,000 cells/nuclei as input into CMO labeling. See: How many cells do I need for CMO labeling when using the 3' CellPlex Kit for Cell Multiplexing?
* The recommended lower limit of 100,000 cells/nuclei applies to our tube-based CMO labeling protocol CG000391. Lower cell inputs can be used with our plate-based CMO labeling protocol CG000426.
7) Use a swinging bucket centrifuge
It may be preferable to use a centrifuge with a swinging bucket rotor, especially when working with nuclei or lower cell numbers. Using a swinging bucket rotor, the cell/nuclei pellet will be more compact and easier to see. This will make it easier to fully remove the supernatant after each wash step.
8) Select a wash buffer that maintains viability for your sample type
As outlined in our Demonstrated Protocol, the 3’ CellPlex assay was validated using the following buffers to wash cells/nuclei after CMO labeling:
|Wash buffer post CMO labeling||Sample type|
|PBS + 1% BSA||PBMCs, cell lines, dissociated tumor cells|
|NbActiv-1 + 1% BSA||Dissociated brain tissue|
|PBS + 10% FBS||Cells with <80% viability|
|PBS + 1% BSA + RNase Inhibitor (0.2 U/μl)||Nuclei|
If using BSA, we do not recommend decreasing the percentage of BSA in the wash buffer below 1%. During in-house experiments, we found that using 1% BSA, compared to various other wash buffers tested (including 0.04% BSA), produces the best signal-to-noise ratios in CellPlex data and also minimizes cell/nuclei loss and aggregation.
For sensitive cell types, 10% FBS can be used instead of 1% BSA, to maximize cell viability.
For sensitive cell types, a cell culture media appropriate for the sample type can also be used instead of PBS. For example, for the dissociated brain tissue mentioned above, we used the neuronal cell culture media NbActiv-1.
See: Which buffers can be used for the wash steps post CMO labeling when using the 3’ CellPlex kit for Cell Multiplexing?
9) Before CMO labeling, suspend cells/nuclei in the appropriate buffer
For in-house experiments, cells/nuclei were suspended in the following buffers before CMO labeling:
- PBS + 0.04% BSA -- for cells
- PBS + 10% FBS -- for cells of lower viability
- PBS + 1% BSA + RNase Inhibitor -- for nuclei
If you use alternative cell/nuclei isolation protocols that include different resuspension buffers, we recommend washing and resuspending the cells/nuclei in our recommended buffers before CMO labeling. This will avoid introducing components that may interfere with the CMO labeling reaction. For example, if detergents are used during nuclei isolation, wash steps should be performed with PBS + 1% BSA + RNase Inhibitor to remove detergents before CMO labeling.
Products: Single Cell Gene Expression, CellPlex