Tissue Dissociation Can Change Surface Markers: Protecting Flow Cytometry Phenotypes

## Introduction

You digest a tumour or a spleen, stain the suspension, and a CD marker reads weaker than before. Is that a real drop in expression, or did the enzyme cut the epitope off? You cannot tell from the digested, stained sample alone, so the default reading is that a post-digestion drop is a method artifact until a control separates the two. Enzymatic digestion demonstrably reduces the level of detectable surface molecules on immune cells (Autengruber, 2012; Abuzakouk, 1996; Ford, 1996), the effect is epitope- and clone-specific rather than uniform, and only a comparator differing in the digestion step alone can separate trimming from biology: an enzyme-free or mechanically dissociated aliquot, a second clone against a different epitope, or an orthogonal readout. This is a tissue dissociation surface markers flow cytometry problem, and epitope loss is not the only artifact: the same enzymes change which cells survive and what state they arrive in.

## 1. First Principle: A Flow Cytometry Phenotype Is an Epitope Measurement, Not a Protein Census

### 1.1 What a surface stain actually measures

A stain reports how much conjugated antibody stayed bound to one epitope on one cell that survived to cross the laser. A positive event needs all four: the protein present, the epitope that clone binds intact and accessible, the conjugate in range, and the cell alive and single. The optics behind that number sit in the [flow cytometry principle](https://www.bosterbio.com/protocol-and-troubleshooting/flow-cytometry-principle) walkthrough. So "CD4 is lower" means "less anti-CD4 of this clone bound per cell here", which a preparation step can change without protein leaving the cell.

### 1.2 What dissociation actually does to a tissue

Releasing single cells from solid tissue means mincing plus a warm enzymatic digest, commonly a collagenase and DNase mixture, sometimes with dispase, trypsin or papain, typically for 30 to 60 minutes and usually at 37 degrees Celsius; the practical version is in the [flow cytometry sample preparation](https://www.bosterbio.com/protocol-and-troubleshooting/flow-cytometry-sample-preparation) guide. Those proteases are not selective for extracellular matrix; they cleave accessible peptide bonds, and cell-surface proteins are accessible. That is why trypsin was known to strip cell-surface antigens decades ago (Fakhri, 1975) and why preparation method is a first-class variable in current solid-tissue guidance (Reichard, 2018; Cossarizza, 2021).

## 2. Is the Weaker CD Signal Real Biology or Enzymatic Cleavage?

### 2.1 Why the digested sample alone cannot answer this

Both explanations predict lower per-cell fluorescence, and nothing inside the digested tube distinguishes them: the sample carries no record of the epitope before the enzyme reached it.

Consider an illustrative pair of aliquots from one tissue. In the enzyme-free or mechanically released aliquot the marker-positive population sits about 10-fold above the negative peak; in the enzymatically digested aliquot of the same tissue the same population sits about 3-fold above the negative peak, roughly a 3-fold loss of per-cell signal. Those three figures are illustrative, not measured. The percentage of cells scored positive is essentially unchanged between the two aliquots, so the population did not disappear and only its per-cell signal moved, which is the pattern that points at epitope trimming; a real down-regulation would show the same drop in the aliquot that never met the enzyme.

### 2.2 The routes by which a marker loses signal during dissociation

Four routes produce a dimmer marker. The first is direct proteolysis of the epitope, the loss of detectable lymphocyte and macrophage surface molecules after collagenase or dispase digestion (Abuzakouk, 1996; Ford, 1996; Autengruber, 2012). The second is cell-driven shedding: live cells release ectodomains through the ADAM10 and ADAM17 sheddases, with CD62L a well-described substrate (Le Gall, 2009; Yang, 2011); that route exists in any warm suspension and is not a measurement of a dissociation enzyme. The third is cell loss rather than marker loss, since fragile populations are under-recovered. The fourth is measurement drift: dead cells, debris, autofluorescence and Fc-receptor binding raise the background, so separation falls with the epitope intact.

### 2.3 What a genuine expression change looks like by comparison

A real change behaves like biology: it survives in the aliquot that never met the enzyme, reproduces with a second clone, tracks the dose and timing of the treatment, and agrees with an orthogonal readout. It also changes the shape of the data, often shifting the percentage of positive cells, whereas trimming lowers the whole positive population and leaves its frequency roughly unchanged.

## 3. The Two Kinds of Damage a Digest Causes

### 3.1 Epitope damage

Protease activity removes or alters the structure the clone recognises, so the antibody binds less or not at all, and the magnitude depends on the enzyme, its activity, the digestion time and temperature, not on anything the cell decided (Autengruber, 2012; Abuzakouk, 1996). It is read at the end of the digest, so later recovery in culture cannot help a phenotype you stain immediately.

### 3.2 Cell-state damage

The second kind of damage is to the cells. Single-cell sequencing work has repeatedly shown that warm enzymatic dissociation induces a stress and immediate-early gene program in the released cells (van den Brink, 2017; O'Flanagan, 2019; Denisenko, 2020), that this is a recognised class of artifact with its own mitigation literature (Machado, 2021), and that it can be measured directly with dissociation-time labelling (Neuschulz, 2022). Time from excision to processing is its own variable (Massoni-Badosa, 2020). That is transcriptome-level evidence about cell state, not evidence that an epitope was cleaved.

## 4. Why the Same Digest Hits Some Markers and Not Others

Sensitivity to a digest is a property of the epitope, not the marker. Accessibility, distance from the membrane, glycosylation and the presence of a cleavable site differ between epitopes on one protein, so two clones against the same antigen can behave very differently. Enzymatic digestion reshapes immune-cell analysis tissue by tissue, shown in mouse reproductive mucosa with a focus on gamma delta T cells (Skulska, 2019), a mucosal result rather than a tumour or spleen one. Not every difference is the enzyme, though. A conjugate titrated on blood can sit outside its range on digested tissue where autofluorescence and debris are higher, so titration and clone choice are worth repeating per tissue, as the [flow cytometry optimization](https://www.bosterbio.com/protocol-and-troubleshooting/flow-cytometry-optimization) guidance sets out. Recovery bias compounds this: disaggregation methods optimised to preserve the phenotype and function of immune cells from human lung tumours exist because a careless digest changes which cells you recover (Quatromoni, 2014).

## 5. Separating Cleavage From Real Down-Regulation: The Control Set

Any tissue dissociation surface markers flow cytometry workflow needs a short control set. First, an aliquot of the same tissue released without the enzyme, mechanically or with a non-enzymatic reagent; non-enzymatic dissociation has been used that way to quantify PD-L1 and PD-1 on tumour and immune cells (Chargin, 2016). Second, a second clone against a different epitope: agreement argues for biology, disagreement for trimming. Third, an orthogonal readout on the same specimen, such as in-situ staining or a mass cytometry panel (Leelatian, 2016). Fourth, bracket the digest with a short and a long incubation, since a marker that falls with digestion time names its own cause.

The rest of the control set has nothing to do with cleavage and is skipped as often. A viability dye and a doublet gate decide which events are eligible, an Fc-receptor block and an isotype or fluorescence-minus-one control decide where the negative sits, and a stock of [isotype control antibodies](https://www.bosterbio.com/products/isotype-control-antibodies.html) makes that judgement reproducible between batches. Record cell yield and time from excision to digest as data, not logistics.

## 6. Epitope Loss vs Real Down-Regulation: Different Questions, Different Purposes

Row 1 - What it describes

Epitope Loss During Dissociation: a measurement artifact; the protein is still on the cell, but the structure your clone binds was damaged during processing.

Real Biological Down-Regulation: a change in the cell, which presents fewer copies of the protein at its surface.

Row 2 - What causes it

Epitope Loss During Dissociation: protease and sheddase activity, digestion time, temperature and handling, all after the tissue left the animal.

Real Biological Down-Regulation: signalling in the intact tissue, acting through transcription, trafficking or internalisation, before you touched the sample.

Row 3 - How you test for it

Epitope Loss During Dissociation: compare an aliquot differing only in the digestion step, or a second clone against a different epitope.

Real Biological Down-Regulation: confirm it persists in the enzyme-free comparator and tracks treatment dose or timing.

Row 4 - The question it answers

Epitope Loss During Dissociation: did my method survive the tissue.

Real Biological Down-Regulation: did the cells change.

## 7. Dissociation Scenarios

In This Article

  1. Mouse spleen, thymus or lymph node. Soft lymphoid tissue can often be released mechanically with little or no enzyme, so the enzyme-free comparator is cheap and should always be run; red-cell lysis and time on ice still move viability and recovery without touching an epitope.
  2. A solid tumour, mouse or human. An enzymatic digest cannot be avoided, so the artifact must be characterised once and held constant: one enzyme formulation, one lot where possible, one incubation time, one tissue mass per tube. Expect recovery bias as well as epitope loss, since cell types release at different rates.
  3. A comparison across treatment groups, batches or processing days. The digest must be identical across arms: a shared artifact does not cancel, and an unequal one becomes a group difference that looks biological. Randomise processing order and record enzyme lot and digestion time.
  4. A workflow coupling phenotyping to sorting or single-cell sequencing. The digest that trims epitopes also drives a stress program in the cells you are about to sequence, which is why cold-active protease digestion (Adam, 2017; O'Flanagan, 2019), nuclei-based workflows (Denisenko, 2020) and in-situ fixation (Machado, 2017) exist as mitigations.
  5. Reporting a post-digestion intensity drop as reduced expression with no comparator that differs only in the digestion step. Without that aliquot the claim cannot be falsified.
  6. Assuming every clone against a marker behaves the same. Sensitivity belongs to the epitope the clone binds, so no marker is digest-sensitive in general, only with a particular clone.
  7. Letting the digest drift between groups or days. Enzyme lot and activity, digestion time, temperature, tissue mass and agitation all move the result; unfixed, processing becomes a hidden variable.
  8. Blaming the enzyme for losses that are really gating, viability, Fc-receptor background or spreading. Those failures look like a dim marker too, and the [flow cytometry troubleshooting](https://www.bosterbio.com/protocol-and-troubleshooting/flow-cytometry-troubleshooting) checklist rules them out faster than another protocol change.
  9. An epitope that no releasing digest leaves intact. Some phenotypes are not measurable in a suspension at all, and the honest answer is an orthogonal readout: an [immunohistochemistry protocol](https://www.bosterbio.com/protocol-and-troubleshooting/ihc-protocol) on sections, imaging, or a nuclei-based assay that never asks the surface to survive.
  10. The sample itself. Ischaemia time, handling and cell-type-specific fragility decide which cells survive at all, and a gentler enzyme cannot restore a population that never reached the tube.
  11. The antibody and the panel. An unvalidated clone, high autofluorescence and low event counts on a rare population are not dissociation problems; starting from [knockout-validated antibodies](https://www.bosterbio.com/ko-kd-validated-antibodies) removes specificity from the suspect list before the tissue is cut.
  12. Autengruber A, 2012, European Journal of Microbiology and Immunology. Impact of enzymatic tissue disintegration on the level of surface molecule expression and immune cell function. DOI 10.1556/eujmi.2.2012.2.3 (PMID 24672679)
  13. Abuzakouk M, 1996, Journal of Immunological Methods. Collagenase and Dispase enzymes disrupt lymphocyte surface molecules. DOI 10.1016/0022-1759(96)00038-5 (PMID 8765174)
  14. Ford AL, 1996, Journal of Immunological Methods. Tissue digestion with dispase substantially reduces lymphocyte and macrophage cell-surface antigen expression. DOI 10.1016/0022-1759(96)00067-1 (PMID 8690942)
  15. Fakhri O, 1975, Cellular Immunology. The effect of trypsin on cell surface antigens. DOI 10.1016/0008-8749(75)90021-0 (PMID 46186)
  16. Skulska K, 2019, Journal of Immunological Methods. Impact of tissue enzymatic digestion on analysis of immune cells in mouse reproductive mucosa with a focus on gamma delta T cells. DOI 10.1016/j.jim.2019.112665 (PMID 31525366)
  17. Quatromoni JG, 2014, Journal of Leukocyte Biology. An optimized disaggregation method for human lung tumors that preserves the phenotype and function of the immune cells. DOI 10.1189/jlb.5ta0814-373 (PMID 25359999)
  18. Chargin A, 2016, Cancer Immunology, Immunotherapy. Quantification of PD-L1 and PD-1 expression on tumor and immune cells in non-small cell lung cancer (NSCLC) using non-enzymatic tissue dissociation and flow cytometry. DOI 10.1007/s00262-016-1889-3 (PMID 27565980)
  19. Reichard A, 2018, Cytometry Part A. Best Practices for Preparing a Single Cell Suspension from Solid Tissues for Flow Cytometry. DOI 10.1002/cyto.a.23690 (PMID 30523671)
  20. Leelatian N, 2016, Cytometry Part B Clinical Cytometry. Single cell analysis of human tissues and solid tumors with mass cytometry. DOI 10.1002/cyto.b.21481 (PMID 27598832)
  21. Cossarizza A, 2021, European Journal of Immunology. Guidelines for the use of flow cytometry and cell sorting in immunological studies (third edition). DOI 10.1002/eji.202170126 (PMID 34910301)
  22. van den Brink SC, 2017, Nature Methods. Single-cell sequencing reveals dissociation-induced gene expression in tissue subpopulations. DOI 10.1038/nmeth.4437 (PMID 28960196)
  23. Adam M, 2017, Development. Psychrophilic proteases dramatically reduce single cell RNA-seq artifacts: a molecular atlas of kidney development. DOI 10.1242/dev.151142 (PMID 28851704)
  24. O'Flanagan CH, 2019, Genome Biology. Dissociation of solid tumor tissues with cold active protease for single-cell RNA-seq minimizes conserved collagenase-associated stress responses. DOI 10.1186/s13059-019-1830-0 (PMID 31623682)
  25. Denisenko E, 2020, Genome Biology. Systematic assessment of tissue dissociation and storage biases in single-cell and single-nucleus RNA-seq workflows. DOI 10.1186/s13059-020-02048-6 (PMID 32487174)
  26. Massoni-Badosa R, 2020, Genome Biology. Sampling time-dependent artifacts in single-cell genomics studies. DOI 10.1186/s13059-020-02032-0 (PMID 32393363)
  27. Machado L, 2021, Trends in Cell Biology. Stress relief: emerging methods to mitigate dissociation-induced artefacts. DOI 10.1016/j.tcb.2021.05.004 (PMID 34074577)
  28. Machado L, 2017, Cell Reports. In Situ Fixation Redefines Quiescence and Early Activation of Skeletal Muscle Stem Cells. DOI 10.1016/j.celrep.2017.10.080 (PMID 29141227)
  29. Neuschulz A, 2022, Molecular Systems Biology. A single-cell RNA labeling strategy for measuring stress response upon tissue dissociation. DOI 10.15252/msb.202211147 (PMID 36573354)
  30. Le Gall SM, 2009, Molecular Biology of the Cell. ADAMs 10 and 17 represent differentially regulated components of a general shedding machinery for membrane proteins such as transforming growth factor alpha, L-selectin, and tumor necrosis factor alpha. DOI 10.1091/mbc.e08-11-1135 (PMID 19158376)
  31. Yang S, 2011, PLOS ONE. The shedding of CD62L (L-selectin) regulates the acquisition of lytic activity in human tumor reactive T lymphocytes. DOI 10.1371/journal.pone.0022560 (PMID 21829468)

1. Mouse spleen, thymus or lymph node. Soft lymphoid tissue can often be released mechanically with little or no enzyme, so the enzyme-free comparator is cheap and should always be run; red-cell lysis and time on ice still move viability and recovery without touching an epitope.

2. A solid tumour, mouse or human. An enzymatic digest cannot be avoided, so the artifact must be characterised once and held constant: one enzyme formulation, one lot where possible, one incubation time, one tissue mass per tube. Expect recovery bias as well as epitope loss, since cell types release at different rates.

3. A comparison across treatment groups, batches or processing days. The digest must be identical across arms: a shared artifact does not cancel, and an unequal one becomes a group difference that looks biological. Randomise processing order and record enzyme lot and digestion time.

4. A workflow coupling phenotyping to sorting or single-cell sequencing. The digest that trims epitopes also drives a stress program in the cells you are about to sequence, which is why cold-active protease digestion (Adam, 2017; O'Flanagan, 2019), nuclei-based workflows (Denisenko, 2020) and in-situ fixation (Machado, 2017) exist as mitigations.

## 8. Common Mistakes When Reading a Post-Dissociation Phenotype

1. Reporting a post-digestion intensity drop as reduced expression with no comparator that differs only in the digestion step. Without that aliquot the claim cannot be falsified.

2. Assuming every clone against a marker behaves the same. Sensitivity belongs to the epitope the clone binds, so no marker is digest-sensitive in general, only with a particular clone.

3. Letting the digest drift between groups or days. Enzyme lot and activity, digestion time, temperature, tissue mass and agitation all move the result; unfixed, processing becomes a hidden variable.

4. Blaming the enzyme for losses that are really gating, viability, Fc-receptor background or spreading. Those failures look like a dim marker too, and the [flow cytometry troubleshooting](https://www.bosterbio.com/protocol-and-troubleshooting/flow-cytometry-troubleshooting) checklist rules them out faster than another protocol change.

## 9. What a Gentler Protocol Cannot Fix

1. An epitope that no releasing digest leaves intact. Some phenotypes are not measurable in a suspension at all, and the honest answer is an orthogonal readout: an [immunohistochemistry protocol](https://www.bosterbio.com/protocol-and-troubleshooting/ihc-protocol) on sections, imaging, or a nuclei-based assay that never asks the surface to survive.

2. The sample itself. Ischaemia time, handling and cell-type-specific fragility decide which cells survive at all, and a gentler enzyme cannot restore a population that never reached the tube.

3. The antibody and the panel. An unvalidated clone, high autofluorescence and low event counts on a rare population are not dissociation problems; starting from [knockout-validated antibodies](https://www.bosterbio.com/ko-kd-validated-antibodies) removes specificity from the suspect list before the tissue is cut.

## Conclusion

A weaker CD signal after digestion is a result about your method until proved otherwise. Resolve it with a comparator that differs only in the digestion step, a second clone against a different epitope, or an orthogonal readout, and keep the digest constant across every arm. Any tissue dissociation surface markers flow cytometry experiment also makes claims about recovery, viability and cell state that the same enzymes influence.

Protecting a phenotype starts with reagents whose behaviour you can predict. Well-validated flow cytometry antibodies and detection reagents, with [documented validation evidence](https://www.bosterbio.com/antibodies-validation-information) behind each clone, keep the surviving question about the tissue and the enzyme rather than the antibody.

## References

1. Autengruber A, 2012, European Journal of Microbiology and Immunology. Impact of enzymatic tissue disintegration on the level of surface molecule expression and immune cell function. DOI 10.1556/eujmi.2.2012.2.3 (PMID 24672679)

2. Abuzakouk M, 1996, Journal of Immunological Methods. Collagenase and Dispase enzymes disrupt lymphocyte surface molecules. DOI 10.1016/0022-1759(96)00038-5 (PMID 8765174)

3. Ford AL, 1996, Journal of Immunological Methods. Tissue digestion with dispase substantially reduces lymphocyte and macrophage cell-surface antigen expression. DOI 10.1016/0022-1759(96)00067-1 (PMID 8690942)

4. Fakhri O, 1975, Cellular Immunology. The effect of trypsin on cell surface antigens. DOI 10.1016/0008-8749(75)90021-0 (PMID 46186)

5. Skulska K, 2019, Journal of Immunological Methods. Impact of tissue enzymatic digestion on analysis of immune cells in mouse reproductive mucosa with a focus on gamma delta T cells. DOI 10.1016/j.jim.2019.112665 (PMID 31525366)

6. Quatromoni JG, 2014, Journal of Leukocyte Biology. An optimized disaggregation method for human lung tumors that preserves the phenotype and function of the immune cells. DOI 10.1189/jlb.5ta0814-373 (PMID 25359999)

7. Chargin A, 2016, Cancer Immunology, Immunotherapy. Quantification of PD-L1 and PD-1 expression on tumor and immune cells in non-small cell lung cancer (NSCLC) using non-enzymatic tissue dissociation and flow cytometry. DOI 10.1007/s00262-016-1889-3 (PMID 27565980)

8. Reichard A, 2018, Cytometry Part A. Best Practices for Preparing a Single Cell Suspension from Solid Tissues for Flow Cytometry. DOI 10.1002/cyto.a.23690 (PMID 30523671)

9. Leelatian N, 2016, Cytometry Part B Clinical Cytometry. Single cell analysis of human tissues and solid tumors with mass cytometry. DOI 10.1002/cyto.b.21481 (PMID 27598832)

10. Cossarizza A, 2021, European Journal of Immunology. Guidelines for the use of flow cytometry and cell sorting in immunological studies (third edition). DOI 10.1002/eji.202170126 (PMID 34910301)

11. van den Brink SC, 2017, Nature Methods. Single-cell sequencing reveals dissociation-induced gene expression in tissue subpopulations. DOI 10.1038/nmeth.4437 (PMID 28960196)

12. Adam M, 2017, Development. Psychrophilic proteases dramatically reduce single cell RNA-seq artifacts: a molecular atlas of kidney development. DOI 10.1242/dev.151142 (PMID 28851704)

13. O'Flanagan CH, 2019, Genome Biology. Dissociation of solid tumor tissues with cold active protease for single-cell RNA-seq minimizes conserved collagenase-associated stress responses. DOI 10.1186/s13059-019-1830-0 (PMID 31623682)

14. Denisenko E, 2020, Genome Biology. Systematic assessment of tissue dissociation and storage biases in single-cell and single-nucleus RNA-seq workflows. DOI 10.1186/s13059-020-02048-6 (PMID 32487174)

15. Massoni-Badosa R, 2020, Genome Biology. Sampling time-dependent artifacts in single-cell genomics studies. DOI 10.1186/s13059-020-02032-0 (PMID 32393363)

16. Machado L, 2021, Trends in Cell Biology. Stress relief: emerging methods to mitigate dissociation-induced artefacts. DOI 10.1016/j.tcb.2021.05.004 (PMID 34074577)

17. Machado L, 2017, Cell Reports. In Situ Fixation Redefines Quiescence and Early Activation of Skeletal Muscle Stem Cells. DOI 10.1016/j.celrep.2017.10.080 (PMID 29141227)

18. Neuschulz A, 2022, Molecular Systems Biology. A single-cell RNA labeling strategy for measuring stress response upon tissue dissociation. DOI 10.15252/msb.202211147 (PMID 36573354)

19. Le Gall SM, 2009, Molecular Biology of the Cell. ADAMs 10 and 17 represent differentially regulated components of a general shedding machinery for membrane proteins such as transforming growth factor alpha, L-selectin, and tumor necrosis factor alpha. DOI 10.1091/mbc.e08-11-1135 (PMID 19158376)

20. Yang S, 2011, PLOS ONE. The shedding of CD62L (L-selectin) regulates the acquisition of lytic activity in human tumor reactive T lymphocytes. DOI 10.1371/journal.pone.0022560 (PMID 21829468)