Abstract
Most shared resource flow cytometry facilities do not permit analysis of radioactive samples. We are investigating low-dose molecular targeted radionuclide therapy (MTRT) as an immunomodulator in combination with in situ tumor vaccines and need to analyze radioactive samples from MTRT-treated mice using flow cytometry. Further, the sudden shutdown of core facilities in response to the COVID-19 pandemic has created an unprecedented work stoppage. In these and other research settings, a robust and reliable means of cryopreservation of immune samples is required. We evaluated different fixation and cryopreservation protocols of disaggregated tumor cells with the aim of identifying a protocol for subsequent flow cytometry of the thawed sample, which most accurately reflects the flow cytometric analysis of the tumor immune microenvironment of a freshly disaggregated and analyzed sample. Cohorts of C57BL/6 mice bearing B78 melanoma tumors were evaluated using dual lymphoid and myeloid immunophenotyping panels involving fixation and cryopreservation at three distinct points during the workflow. Results demonstrate that freezing samples after all staining and fixation are completed most accurately matches the results from noncryopreserved equivalent samples. We observed that cryopreservation of living, unfixed cells introduces a nonuniform alteration to PD1 expression. We confirm the utility of our cryopreservation protocol by comparing tumors treated with in situ tumor vaccines, analyzing both fresh and cryopreserved tumor samples with similar results. Last, we use this cryopreservation protocol with radioactive specimens to demonstrate potentially beneficial effector cell changes to the tumor immune microenvironment following administration of a novel MTRT in a dose- and time-dependent manner.
Introduction
Continued preclinical and clinical progress in the field of tumor immunology hinges on the ability to analyze accurately and reliably the immune tumor microenvironment (TME) in response to treatment. Flow cytometry is a powerful tool for the interrogation of different immune cell subsets, markers of activation, signaling pathway phosphorylation states, transcription factors, cytokines, and others in complex populations of cells (1). Effective, rigorous flow cytometric analyses can be hindered by several barriers that disrupt experiment workflow. Unexpected instrument breakdown, maintenance, or overcrowding of shared resource facilities can delay analysis of samples following tissue harvest or delay planned harvest days, which may have detrimental effects on time-sensitive experiments. An extreme example our working group has encountered is the sudden shutdown of our shared resource flow cytometry core because of the global COVID-19 pandemic. There are also instances in which same-day acquisition of samples by flow cytometry is not logistically possible, for instance if samples are collected at multiple sites and need to be transported to other buildings, institutions, or countries for analysis in a central laboratory. Another critical example is that most shared resource facilities prohibit analysis of radioactive samples for safety reasons. Dedicated cytometers for analysis of radioactive material are expensive and difficult to maintain and may not benefit from the expertise and efficiencies afforded by the shared resource facility structure. In these instances, the ability to cryopreserve harvested samples for later flow cytometric analysis would be of great benefit.
Although cryopreservation is a commonplace laboratory procedure with many different applications and protocol variations, it is generally not recommended in flow cytometry if possible (2–6). It is known that cells undergoing cryopreservation exhibit a stress response to the freezing process, which is one reason why many prefer flow cytometry analyses be performed on freshly obtained cells (7). One could consider avoiding the biological stress of cryopreservation by staining and fixing samples prior to freezing. However fluorophore-conjugated Abs, especially those using tandem dye fluorophores, are generally not considered stable through freeze/thaw cycles because of the risk of degradation (8–10). Furthermore, fixed cells may have a substantial degree of cross-linked structural rigidity (11). Although it has not been clearly demonstrated and likely depends on the fixative used, there is a risk that crystals formed during the cryopreservation process may damage these rigid cells and impair the recovery of intact cells following cryopreservation. Several studies have described the changes that can occur to detected frequencies of various immune populations in human PBMCs following cryopreservation and concluded that although possible, it is preferable to stain and analyze fresh cells without cryopreservation (4, 5, 12, 13). Pinto and colleagues (14) have demonstrated that it is feasible to surface stain and fix human PBMCs prior to subsequent cryopreservation and observed comparable immune cell frequencies to those of freshly analyzed samples. However, it is unclear how this translates to immunophenotyping in the murine TME and how the cryopreservation methodology might influence detection of markers such as transcription factors that require cell membrane permeabilization (perm).
External beam radiation therapy (RT) is known to have direct tumor cell–killing effects and is used in clinic to treat cancer patients (15). Recently, much interest has arisen in the effect of radiation on the immune system and the TME. Low to moderate doses of radiation (8–12 Gy) have been shown to elicit a type I IFN response through the activation of the stimulator of IFN genes pathway (16, 17). Studies have also demonstrated a rise in MHC class I concurrent with infiltrating CD4+ and CD8+ T cells as well as increased proinflammatory cytokines in the TME following 2–5 Gy doses of radiation (18–20). Radiation can also have a direct cytotoxic effect on certain immune cell lineages, especially lymphocytes, with almost 90% of lymphocytes driven to apoptosis within 8 h of a 3 Gy dose in one study (21). This may have the beneficial effect of temporarily clearing the tumor of suppressive lineages like CD4+FOXP3+ T regulatory (Tregs) cells (22, 23). Taken together, these findings suggest that low to moderate doses of radiation have the potential to temporarily immunomodulate a suppressive TME into a more favorable effector environment capable of responding to subsequent immunotherapy. This principle was proven effective in demonstrating a synergy between 12 Gy local RT and intratumorally (IT) injected hu14.18-IL-2 immunocytokine (IC) (an anti-disialoganglioside D2 [anti-GD2] Ab fused to IL-2) in treating GD2+ murine tumors (24). This form of immunotherapy is one of several strategies that aim to induce an in situ vaccine response using the tumor’s own Ags. This is driven by targeted immunogenic cell death, Ag processing, and presentation, which then induce recruitment and sensitization of anti-tumor effector cells capable of recognizing multiple different Ags selectively expressed by the tumor (25).
In preclinical studies, we have begun to test treatment approaches using systemically administered molecular targeted radionuclide therapy (MTRT) to deliver low to moderate doses of immunomodulatory RT to all tumor sites. This approach holds promise at least in part because the efficacy of in situ vaccines may be suppressed in the setting of distant, nonradiated metastases (26). In this sense, systemic low-dose MTRT may modulate the immune susceptibility of the collective TME for all tumor sites and thereby facilitate greater systemic propagation of an in situ vaccine delivered at a single tumor site (27). Alkylphosphocholine analogues can be used to deliver MTRT and have been shown to be preferentially retained in over 50 mammalian tumor types regardless of anatomic location (28). [90Y]NM600 is an alkylphosphocholine analogue that has been characterized in murine and canine models and can be used to safely deliver 2–5 Gy to all tumor sites without causing bone marrow suppression or systemic lymphopenia (29). Although preclinical anti-tumor efficacy has been demonstrated following combination MTRT and checkpoint blockade immunotherapy (27), the mechanisms by which this form of low-dose MTRT alters the TME to provide this immunotherapeutic synergy has not been clearly described. Addressing these issues requires careful flow cytometry analyses of serially obtained tumor samples from mice undergoing curative combination therapy with MTRT and immunotherapy. However, this form of flow cytometry requires either a dedicated flow cytometer for radioactive samples or the ability to cryopreserve these samples long enough to allow radiation to decay to background levels prior to analysis.
In this study, we address this methodological challenge by evaluating the impact of introducing a cryopreservation step to our flow cytometry workflow for dissociated murine tumor and spleen samples. Our goal was to identify a cryopreservation strategy/workflow that would give flow cytometry results most similar to those of freshly obtained and analyzed samples. We demonstrate that timing of cryopreservation relative to surface staining, fixation, perm, and internal target staining has a strong effect on immunophenotyping results. Our data show that freezing after all staining and fixation is completed, rather than before or during these sequential steps, yields flow cytometry results that are most concordant with those obtained from freshly analyzed, noncryopreserved aliquots of these samples for most immune populations of interest to our research group. We demonstrate that this protocol of freezing cells after all staining is completed can be used to evaluate the impact of an in situ vaccine and describe the changes to the TME as a function of time and dose of radioactive [90Y]NM600.
Materials and Methods
Syngeneic tumor cell line
B78-D14 (B78) murine melanoma is a cell line derived from B16-F10 melanoma, as previously described (27). This cell line was generously provided by Ralph Reisfeld at the Scripps Research Institute (La Jolla, CA). Cells were cultured in RPMI 1640 (Mediatech, Manassas, VA) supplemented with 10% FBS, 100 U/ml penicillin, 100 μg/ml streptomycin, and 2 mM l-glutamine. As previously described, B78 cells have been engineered to express GD2 and GD3 synthase under the control of 400 μg/ml G418 and 50 μg/ml hygromycin and were selected over 3 d in selection media prior to being frozen down as laboratory stock (31, 32). Cells were kept below 90% confluence and were used within seven passages of thaw from the common bank. Cells were confirmed mycoplasma negative by PCR within 6 mo of use.
Animal studies and tumor models
Animals used in this study were housed and cared for using an approved protocol reviewed by the University of Wisconsin—Madison Institutional Animal Care and Use Committee. Human PBMCs from volunteer donors (used for experiments outlined in Supplemental Figs. 1–3) were collected according to a protocol approved by the University of Wisconsin Institutional Review Board for human subjects research. Female 6–8-wk-old C57BL/6 mice were ordered from Taconic Biosciences (Rensselaer, NY) and allowed to acclimate in our animal facility for at least 1 wk following arrival. Tumors were engrafted by injecting 2 × 106 B78 cells in 100 μL of sterile PBS intradermally in the right flank using a 26-G needle. Of note, the flank was shaved 24–48 h prior to implantation to ensure consistency of injection and to ensure resolution of any irritation and inflammation resulting from shaving. Tumors were monitored twice weekly and developed over 6 wk. Tumor volume was measured using calipers and approximated as (width2 × length)/2. Mice receiving MTRT or immunotherapy were randomized to begin treatment when tumor volumes reached 90–180 mm3.
External beam radiation, molecularly targeted radionuclide preparation, and administration
External beam RT was delivered using an X-RAD 320 system (Precision X-Ray, North Branford, CT). Mice were immobilized using custom lead jigs and surgical tape such that only the dorsal right flank was exposed, with the rest of the mouse (including the contralateral flank and spleen) shielded. Radiation settings were beam strength 320 Kv/12.5 mA, beam conditioning filter 2, platform height 36 cm, and treatment duration 331 s, which has been measured to reliably deliver 11 ± 1 Gy to the tumor by thermoluminescent dosimetry (data not shown). 90YCl3 was purchased from PerkinElmer (Waltham, MA). The vehicle used for MTRT, 2-(trimethylammonio)ethyl(18-(4-(2-(4,7,10-tris(carboxymethyl)-1,4,7,10-tetraazacyclododecan-1-yl)acetamido)phenyl)octadecyl) phosphate (NM600), was kindly provided by Archeus Technologies (Madison, WI). The radiolabeling and characterization of [90Y]NM600 and its positron emission tomography–detectible counterpart [86Y]NM600 has been described elsewhere (29, 33). Briefly, 5–10 mCi of 90YCl3 was buffered in 0.1 M NaOAc (pH 5.5), and 10–15 nmol/mCi of NM600 was added to the mixture. The reaction was placed at 90°C for 30 min under constant shaking at 500 rpm. The reaction mixture was loaded into a hydrophilic-lipophilic balance solid phase extraction cartridge (Waters) and washed with 5 mL of H2O, and [90Y]NM600 was eluted in 2 mL of absolute ethanol. Nitrogen steam was used to evaporate the eluate, and [90Y]NM600 was reconstituted in normal saline containing 0.4% v/v Tween 20 and sodium ascorbate (0.5% w/v). Mice receiving RT were radiated on treatment day 1, and mice receiving MTRT were injected with 50 or 250 μCi of [90Y]NM600 preparation by tail vein injection on treatment day 1.
Immunotherapy preparation and administration
IC (hu14.18-IL-2) was provided in lyophilized form (4 mg/vial) by APEIRON Biologics (Vienna, Austria). It was reconstituted by adding 8 mL of sterile PBS for a working concentration of 0.5 mg/ml. For mice being treated with RT+IC, 100 μL of the 0.5 mg/ml IC solution was injected IT daily on treatment days 6 through 10, for a total dose of 250 μg per mouse in five doses. Injections were through a 30-G needle, and care was taken to inject slowly when administering IC via IT injection. In addition, the drug was not drawn up into the syringe from the vial through the 30-G needle when preparing IC as to avoid shear-induced denaturation of the IC.
Tumor and spleen tissue harvest
At the time of harvest, mice were euthanized by CO2 asphyxiation, and the tumor was dissected out. The tumor was cut into ∼5 mm fragments and added to GentleMACS C tubes (Miltenyi Biotec, Bergish Gladbach, Germany) containing 2.5 ml of RPMI 1640 + 10% FBS, 100 U/ml penicillin, 100 μg/ml streptomycin, and 2 mM l-glutamine. Then, 100 μL of DNAse I solution in RPMI 1640 (2.5 mg/ml, Sigma-Aldrich, St. Louis, MO) and 100 μL collagenase IV solution in RPMI 1640 (25 mg/ml, Life Technologies, Grand Island, NY) were then added, and the samples were disaggregated using a Miltenyi GentleMACS Octo Dissociator (Miltenyi Biotec) using the preset dissociation protocol 37C_m_TDK1 for mouse tumor dissociations. Sample dissociates were filtered through a 70-μm cell strainer, washed with 10 mL of cold PBS, and kept on ice until aliquoted into flow cytometry tubes. Tumors analyzed for changes in immune microenvironment were dissociated individually. For comparing different cryopreservation protocols, n = 5 tumor-bearing mice were harvested, individually dissociated, strained, and pooled into one common reference population. When required, spleens were dissected out and transferred into a sterile petri dish with 1 ml of ice-cold sterile PBS. Spleens were physically ground on the plate into the PBS and transferred into a 5 ml Eppendorf tube. RBCs in the spleen dissociates were lysed by adding 3 mL of 1× RBC lysis buffer for 10 min per the manufacturer’s instructions (BioLegend, San Diego, CA). They were then filtered through a 70-μm cell strainer, and single-cell suspensions from five separate spleens were combined to create a common reference population.
Flow cytometry Abs
We optimized two Ab panels for analysis of myeloid and lymphoid immune cells. See Table I for the complete list of Abs, including fluorophore, clone, company, and optimized volume of stock added per sample, Table II for the two full immunophenotypes used in separate panels, and Fig. 1 for the corresponding gating strategy. These complete Ab panels were used in the comparisons of cryopreservation techniques as seen in Figs. 2 and 3. In Figs. 4 and 5, single staining panels consisting of a combination of targets from the lymphoid and myeloid panels were used for efficiency. See Supplemental Table I for a list of Abs used in Figs. 4 and 5 and in testing single-color consistency corresponding to Supplemental Fig. 2.
Sample preparation, staining, and cryopreservation
Single-cell suspensions from either tumor or spleen were counted, and 3 × 106 cells were aliquoted to labeled flow tubes. Aliquoted samples were stained with 0.5 μL GhostRed780 stock (Tonbo Biosciences, San Diego, CA) diluted in 50 μL of PBS per sample and light protected at 4°C for 30 min. Samples were washed with flow buffer (PBS + 2% FBS) and then Fc blocked by adding 0.25 μL of stock TruStain FcX Plus anti-mouse CD16/32 Ab (BioLegend, San Diego, CA) diluted in 50 μL flow buffer per sample at room temperature and light protected for 10 min. Samples were then stained using 50 μL of an Ab master mix (prepared up to 24 h before) for the corresponding surface markers and light protected at 4°C for 30 min. Master mix volumes were calculated given a 50 μL per sample target volume, with optimal volumes of stock Ab (in Table I, Supplemental Table I) included in each 50-μL test volume. In calculating the required volumes of both flow buffer and Abs for all master mixes, the total number of samples for volume calculation purposes was increased by 10% to account for volume loss because of pipetting and any bubbles generated while mixing (e.g., 11 samples worth of master mix was prepared to stain 10 samples). After surface staining, samples were again washed with flow buffer, then fixed and permeabilized using the eBioscience FOXP3 Fixation/perm Kit at room temperature, and light protected for 30 min according to manufacturer’s instructions (Thermo Fisher Scientific, Waltham, MA). Of note, samples were vortexed immediately prior to and following the addition of fixative to prevent the formation of large cell aggregates. After washing with 1× perm buffer from the Fixation/perm kit, samples were stained for intracellular targets (e.g., FOXP3) by adding 50 μL of perm buffer containing the optimal Ab volume per sample and incubated (with light protection) at room temperature for 30 min. For consistency of staining and sample reproducibility, all samples were fixed, even if no intracellular targets were in the Ab staining panel.
Modified preparation techniques were conducted according to the schematic outlined in Fig. 2 and discussed in the Results. When cryopreservation was performed, samples were washed and resuspended in 0.5 mL FBS, then transferred to cryotubes containing 0.5mL of FBS + 20% DMSO (final concentration 1 mL of FBS + 10% DMSO), and frozen in isopropyl alcohol bath containers in a −80°C freezer in a custom wooden box with lead and plexiglass interior lining. Radioactive samples were stored until 30 d postinjection of activity when radioactivity is at background (∼11.25 half-lives for [90Y]). Proper shielding was confirmed via survey with Geiger counters calibrated by the University of Wisconsin Department of Environment, Health, and Safety. Samples were thawed at 37°C and transferred into flow cytometry tubes containing 3 mL of flow buffer. Tubes were then pelleted and resuspended in 200 μL of flow buffer.
Flow cytometry
Previously radioactive samples were confirmed to be at background levels via survey with a calibrated Geiger counter. Samples were acquired on an Attune NxT flow cytometer (Thermo Fisher Scientific) with manufacturer-provided acquisition software. This instrument was maintained by the University of Wisconsin Carbone Cancer Center Flow Cytometry Laboratory, which performs daily quality control checks and instrument calibration using Attune Performance Tracking Beads (Thermo Fisher Scientific, catalog 4449754). This cytometer was equipped with the following excitation lasers: 488 nm blue laser (BL), 561 nm yellow laser (YL), 405 m violet laser (VL), and 633 nm red laser (RL). The cytometer was equipped with the following channel/bandpass filter combinations: BL1 (530/30), BL2 (590/40), BL3 (695/40), YL1 (585/16), YL2 (620/15), YL3 (695/40), YL4 (780/60), VL1 (440/50), VL2 (512/25), VL3 (603/48), VL4 (710/50), RL1 (670/14), RL2 (720/30), and RL3 (780/60). Of note, staining panels that used Brilliant Violet (BV) 711 used the 710/50 filter on VL4, and staining panels that used BV785 required substituting the 710/50 filter on VL4 with a 780/60 filter. To ensure the validity of comparisons made on different days, rainbow fluorescent beads (Spherotech, Lake Forest, IL) were acquired after the first day of sample acquisition and used to align voltages to calibrate equivalent fluorescence intensities on later analysis dates. All flow cytometry experiments included fluorescence minus one controls used for setting gates. Data were analyzed using the FCS Express 7 software (De Novo Software, Pasadena, CA) platform.
Statistical analysis
All data presenting results for analyses of identical replicate samples include the individual sample values as well as mean ± SEM, except where otherwise noted. All significance tests were determined by p values at the α = 0.05 level. Comparisons between the “Control” staining technique to each cryopreservation technique were done using two-sided, two-sample t tests. Each method was compared with the Control staining technique based on the mean difference from the Control mean. Comparisons in Fig. 4 were made using linear models assessing outcome association with cryopreservation method, treatment, or preparation–treatment interaction. Comparisons in Fig. 5 were made using linear models to assess the effect of time, dose of treatment, and time–dose interaction on the measured immune parameters. For both analyses, F tests were conducted to evaluate covariate significance. If the interaction term was determined insignificant, it was removed from the model. If significance was found in treatment effect by F tests, then pairwise differences between treatment categories were reported. The p values were not corrected for multiple comparisons. Statistical analyses were done using R and Prism (GraphPad Software, San Diego, CA).
Results
Optimization of Ab staining panels and gating strategy
After gating out debris and doublets using both forward scatter and side scatter area versus height, viable cells are identified in all staining panels by their low expression of fixable viability dye GhostRed780 (Fig. 1A). In the live cell gate, tumor cells (GD2+) are distinguished from hematopoietic-derived (CD45+) cells. In the lymphoid staining panel, live CD45+ cells are then divided by expression of CD3 (Fig. 1A). Within the CD3+ population, the identification of NKT cells, CD4+ T cells, and CD8+ T cell subpopulations can be made. The CD3− population then contains the NK cell and CD19+ B cell populations. In our hands, NK cells were best identified by plotting NK1.1 against side scatter, although the population can be identified using other parameters such as NK1.1 versus CD45 or CD3. Among CD4+ T cells, coexpression of CD25 and the transcription factor FOXP3 clearly identifies the T (Treg) cell subset (34, 35). An exclusion gate then distinguishes the Treg cell from the non-Treg CD4+ T cells, which can include both naive and effector helper T cells. Programmed cell death receptor 1 (PD1) was included in the adaptive panel to quantify activation/exhaustion status of PD1-expressing immune cells including T, B, and NK cells (36). Similarly, CD103 was included in this panel to identify tissue-resident lymphocytes, specifically CD8+ and CD4+ memory cell populations (Table I) (37, 38).
Representative gating strategy for the evaluation of immune parameters in dissociated tumor or splenic tissue. Tumors or spleens from C57BL/6 mice were harvested and dissociated as outlined in Materials and Methods. The gating strategy outlined in (A) shows the lymphoid immune panel used to identify the following populations from a sample after excluding debris and gating on single cells: total live cells, GD2+ tumor cells, CD45+, CD3+, CD4+, Treg cells, Th cells, CD8+, CD19+, NK, and NKT cell populations. Populations of T, B, and NK cells were also evaluated for expression of the CD103 memory marker and PD1 activation/exhaustion marker (the representative gating strategy presented here shows PD1 and CD103 evaluation for the T cell subsets [CD8+, CD4+ Treg cells, and CD4+ non-Treg cells] only). The gating strategy outlined in (B) shows the myeloid immune panel used to identify the following populations in a sample [after excluding debris and gating on single cells as in (A)]: total live cells, CD45+, neutrophils/MDSCs, macrophages, cDC1, cDC2, and moDCs. Macrophages were also subcategorized into M1-like and M2-like based on high or low expression of MHCII, respectively. Gates were set using corresponding fluorescence minus one controls. All assay replicates were pooled for this representative gating strategy for visualization of rare populations.
Representative gating strategy for the evaluation of immune parameters in dissociated tumor or splenic tissue. Tumors or spleens from C57BL/6 mice were harvested and dissociated as outlined in Materials and Methods. The gating strategy outlined in (A) shows the lymphoid immune panel used to identify the following populations from a sample after excluding debris and gating on single cells: total live cells, GD2+ tumor cells, CD45+, CD3+, CD4+, Treg cells, Th cells, CD8+, CD19+, NK, and NKT cell populations. Populations of T, B, and NK cells were also evaluated for expression of the CD103 memory marker and PD1 activation/exhaustion marker (the representative gating strategy presented here shows PD1 and CD103 evaluation for the T cell subsets [CD8+, CD4+ Treg cells, and CD4+ non-Treg cells] only). The gating strategy outlined in (B) shows the myeloid immune panel used to identify the following populations in a sample [after excluding debris and gating on single cells as in (A)]: total live cells, CD45+, neutrophils/MDSCs, macrophages, cDC1, cDC2, and moDCs. Macrophages were also subcategorized into M1-like and M2-like based on high or low expression of MHCII, respectively. Gates were set using corresponding fluorescence minus one controls. All assay replicates were pooled for this representative gating strategy for visualization of rare populations.
In a separate panel focused on myeloid cells (Fig. 1B), tumor and splenic tissues were again gated to remove debris, isolate single cells, and isolate live CD45+ cells. Neutrophils and/or myeloid-derived suppressor cells (MDSCs) are first identified by their coexpression of CD11b and Ly-6G. Second, macrophages were identified from the Ly-6G–negative population based on coexpression of F4/80 and CD64 (39). The use of a second macrophage identifier like CD64 is critical for downstream differentiation of type 2 classical dendritic cells (cDC2s), as some of these dendritic cells also express F4/80, and removal of the more autofluorescent macrophage population helps to eliminate contamination in other gates (40, 41). MHC class II (MHCII) can then be used to further subcategorize the macrophages into the more active, MHCIIhi, M1-like macrophages and more anti-inflammatory, MHCIIlo, M2-like macrophages. Although this panel is certainly not enough to fully delineate the phenotypes of these macrophage lineages, the panel can easily be adjusted by including markers such as CD206 and CD80/86 to more completely phenotype the macrophage population (42) if desired. The neutrophil-negative, macrophage-negative population is then subdivided based on expression of CD11b and Ly6C and separates into four distinct populations. Within the CD11b−Ly6C− population, the CD11c+MHCII+ double positive cells are likely to be type 1 classical dendritic cells (cDC1s) based on their confirmatory coexpression of CD103 and XCR1 (40, 43). Conversely, within the CD11b+Ly6C+ population, the CD11c+MHCII+ double positive cells are likely to be cDC2s based on their CD103– and XCR1–double negative status. The CD11b+Ly6C− population that expresses both CD11c and MHCII is largely monocyte-derived dendritic cells (moDCs) based on its confirmatory expression of F4/80 and absence of CD103 (44, 45). The CD11b−Ly6C+ population may consist of granulocyte precursors or could represent Ly6C+ T cells in the tumor (46). Further delineation of these populations could be achieved using a pooled lineage marker for CD3, CD19, and NK1.1 and confirming populations to be lineage negative (40), although this was not done in this study. Exact immunophenotypes for these above-described populations are outlined in Table II.
Flow cytometry panels used to evaluate different cryopreservation protocols in (Fig. 1
. | Target . | Fluorophore . | Company (Catalog) . | Clone . | Volume per Test (μL) . |
---|---|---|---|---|---|
Lymphoid panel | CD25 | BB515 | BD Biosciences (564424) | PC61 | 1.5 |
CD103 | PE | BioLegend (121406) | 2E7 | 1.5 | |
NK1.1 | PE-CF594 | BD Biosciences (562864) | PK136 | 1.2 | |
CD19 | PE-Cy5 | BioLegend (115510) | 6D5 | 0.5 | |
FOXP3 | PE-Cy7 | Invitrogen (25-5773-82) | FJK-16s | 1.4 | |
PD-1 | V450 | Tonbo Biosciences (75-9981-U100) | RMP1-30 | 1.2 | |
CD45 | BV510 | BioLegend (103137) | 30-F11 | 1 | |
CD3 | BV605 | BioLegend (100351) | 145-2C11 | 1.2 | |
CD4 | BV785 | BioLegend (100453) | GK1.5 | 1 | |
GD2 | Allophycocyanin | BioLegend (357306) | 14G2a | 1 | |
CD8 | Allophycocyanin-R700 | BD Biosciences (564983) | 53-6.7 | 1 | |
LIVE/DEAD | GhostRed 780 | Tonbo Biosciences(13-0865-T100) | — | 0.5μL | |
Myeloid panel | CD45 | FITC | Tonbo Biosciences (35-0451-U500) | 30-F11 | 1 |
CD103 | PE | BioLegend (121406) | 2E7 | 1.5 | |
MHCII | PE-Dazzle594 | BioLegend (107648) | M5/114.15.2 | 1.2 | |
F4/80 | PE-Cy5 | Invitrogen (15-4801-82) | BM8 | 1 | |
CD64 | PE-Cy7 | BioLegend (139314) | X54-5/7.1 | 1.2 | |
CD11b | V450 | BD Biosciences (560455) | M1/70 | 1.5 | |
XCR1 | BV510 | BioLegend (148218) | ZET | 3 | |
Ly6C | BV605 | BD Biosciences (563011) | AL-21 | 2 | |
CD24 | BV711 | BD Biosciences (563450) | M1/69 | 1.2 | |
CD11c | Allophycocyanin | BioLegend (117310) | N418 | 1.2 | |
Ly6G | AF700 | BD Biosciences (561236) | 1A8 | 1.2 | |
LIVE/DEAD | GhostRed 780 | Tonbo Biosciences (13-0865-T100) | — | 0.5μL |
. | Target . | Fluorophore . | Company (Catalog) . | Clone . | Volume per Test (μL) . |
---|---|---|---|---|---|
Lymphoid panel | CD25 | BB515 | BD Biosciences (564424) | PC61 | 1.5 |
CD103 | PE | BioLegend (121406) | 2E7 | 1.5 | |
NK1.1 | PE-CF594 | BD Biosciences (562864) | PK136 | 1.2 | |
CD19 | PE-Cy5 | BioLegend (115510) | 6D5 | 0.5 | |
FOXP3 | PE-Cy7 | Invitrogen (25-5773-82) | FJK-16s | 1.4 | |
PD-1 | V450 | Tonbo Biosciences (75-9981-U100) | RMP1-30 | 1.2 | |
CD45 | BV510 | BioLegend (103137) | 30-F11 | 1 | |
CD3 | BV605 | BioLegend (100351) | 145-2C11 | 1.2 | |
CD4 | BV785 | BioLegend (100453) | GK1.5 | 1 | |
GD2 | Allophycocyanin | BioLegend (357306) | 14G2a | 1 | |
CD8 | Allophycocyanin-R700 | BD Biosciences (564983) | 53-6.7 | 1 | |
LIVE/DEAD | GhostRed 780 | Tonbo Biosciences(13-0865-T100) | — | 0.5μL | |
Myeloid panel | CD45 | FITC | Tonbo Biosciences (35-0451-U500) | 30-F11 | 1 |
CD103 | PE | BioLegend (121406) | 2E7 | 1.5 | |
MHCII | PE-Dazzle594 | BioLegend (107648) | M5/114.15.2 | 1.2 | |
F4/80 | PE-Cy5 | Invitrogen (15-4801-82) | BM8 | 1 | |
CD64 | PE-Cy7 | BioLegend (139314) | X54-5/7.1 | 1.2 | |
CD11b | V450 | BD Biosciences (560455) | M1/70 | 1.5 | |
XCR1 | BV510 | BioLegend (148218) | ZET | 3 | |
Ly6C | BV605 | BD Biosciences (563011) | AL-21 | 2 | |
CD24 | BV711 | BD Biosciences (563450) | M1/69 | 1.2 | |
CD11c | Allophycocyanin | BioLegend (117310) | N418 | 1.2 | |
Ly6G | AF700 | BD Biosciences (561236) | 1A8 | 1.2 | |
LIVE/DEAD | GhostRed 780 | Tonbo Biosciences (13-0865-T100) | — | 0.5μL |
Cryopreservation following staining and fixation is most concordant with freshly stained samples for many immune cell populations
After optimizing our comprehensive immunophenotyping panels, we sought to determine how cryopreservation affects immunophenotyping outcome based on when during the staining process cryopreservation is performed. Based on literature review (14, 47–50), we identified three points in the process of sample preparation and staining to test the introduction of a cryopreservation step (Fig. 2A, Tables I and II). The first point, labeled “Before” in (Fig. 2A, freezes the cells after dissociation but before any staining. This minimizes handling time following harvest and cryopreserves the cells in the “freshest” possible state. Because freeze/thaw cycles are not recommended for either fixed samples or for the integrity of fluorophore-labeled Abs (8–10), we reasoned the Before method may minimize the impact of cryopreservation on the quality of fixation and staining. The potential downside is that living cells experiencing a freeze/thaw cycle may alter their metabolism, surface expression of various markers, and viability and therefore become a confounding factor in subsequent analyses (14). The second point, labeled “During” in (Fig. 2A, freezes the cells after surface staining but prior to fixation and perm. Thus, any change in surface marker expression because of cryopreservation would not be reflected in these samples. The cells are also still alive, meaning they are known to be able to survive cryopreservation without substantial damage or rupture of the cells. The During method does freeze the fluorophore-labeled Abs bound to the cells, which is a potential risk for signal degradation, particularly in tandem dyes (8–10). The third point, labeled “After” in (Fig. 2A, freezes the cells after all surface staining, fixation/perm, and internal staining. This method most accurately captures the “fresh” state of the TME at the time of harvest for the entire staining process and avoids any expression changes that may be due to prolonged processing time and the stress of the cryopreservation process. It does, however, carry the risk of damaging the more rigid fixed cells as well as destroying the fluorophore labels through both fixation and freezing in the process.
Comparison of sample staining and cryopreservation techniques across comprehensive immunophenotyping panels. (A) As described in the text, tumor and spleen samples were dissociated and resuspended. These single-cell suspensions are stained with fixable viability dye, Fc blocked, surface target stained, fixed and permeabilized, internal target stained, and analyzed, with a wash and pelleting step in between each. Results of these analyses on freshly obtained, never cryopreserved samples are shown in black (Control). For replicate samples, three separate times in this workflow of staining–fixation–perm–staining were tested for cryopreservation. These included cryopreservation prior to all staining (Before, red); after surface staining but before fixation (During, green); or after all fixation and staining (After, blue). (B) Five C57BL/6 mice bearing B78 syngeneic melanoma tumors had tumors (shown in this study) and spleens (shown in Supplemental Fig. 1) harvested as described in Materials and Methods and pooled to create a homogeneous starting population of tumor or spleen dissociates. Each population was analyzed using the lymphoid and myeloid phenotyping panels described in (Fig. 1 and Tables I and II, using each of the four cryopreservation protocols: Control (not cryopreserved) or cryopreserved Before, During, or After. The 23 separate graphs show the mean and 95% confidence interval of n = 7 assay replicates in each parameter listed above each graph for each cryopreservation protocol in the dissociated tumor cell preparations. Filled red, green, or blue symbols indicate a significant (p < 0.05) difference on a t test when compared with the Control (black) protocol, and open red, green or blue symbols indicate nonsignificant (p > 0.05) differences when compared with the Control protocol.
Comparison of sample staining and cryopreservation techniques across comprehensive immunophenotyping panels. (A) As described in the text, tumor and spleen samples were dissociated and resuspended. These single-cell suspensions are stained with fixable viability dye, Fc blocked, surface target stained, fixed and permeabilized, internal target stained, and analyzed, with a wash and pelleting step in between each. Results of these analyses on freshly obtained, never cryopreserved samples are shown in black (Control). For replicate samples, three separate times in this workflow of staining–fixation–perm–staining were tested for cryopreservation. These included cryopreservation prior to all staining (Before, red); after surface staining but before fixation (During, green); or after all fixation and staining (After, blue). (B) Five C57BL/6 mice bearing B78 syngeneic melanoma tumors had tumors (shown in this study) and spleens (shown in Supplemental Fig. 1) harvested as described in Materials and Methods and pooled to create a homogeneous starting population of tumor or spleen dissociates. Each population was analyzed using the lymphoid and myeloid phenotyping panels described in (Fig. 1 and Tables I and II, using each of the four cryopreservation protocols: Control (not cryopreserved) or cryopreserved Before, During, or After. The 23 separate graphs show the mean and 95% confidence interval of n = 7 assay replicates in each parameter listed above each graph for each cryopreservation protocol in the dissociated tumor cell preparations. Filled red, green, or blue symbols indicate a significant (p < 0.05) difference on a t test when compared with the Control (black) protocol, and open red, green or blue symbols indicate nonsignificant (p > 0.05) differences when compared with the Control protocol.
Gating immunophenotypes used in comparing staining protocols
. | Population . | Gating Phenotype (after Cells/Single/Live) . |
---|---|---|
Lymphoid Panel | CD45+ | CD45+ |
Tumor cells | CD45−/GD2+ | |
NKT cells | CD45+/CD3+/NK1.1+ | |
CD8+ T cells | CD45+/CD3+/NK1.1−/CD8+ | |
CD4+ T cells | CD45+/CD3+/NK1.1−/CD4+ | |
Treg cells | CD45+/CD3+/NK1.1−/CD4+/CD25+/foxp3+ | |
B cells | CD45+/CD3−/CD19+ | |
NK cells | CD45+/CD3−/NK1.1+ | |
Myeloid Panel | Neutrophils | CD45+/CD11b+/Ly6G+ |
Macrophages | CD45+/Ly6G−/F4/80+CD64+ | |
M1-like macrophages | CD45+/Ly6G−/F4/80+CD64+/MHCIIhi | |
M2-like macrophages | CD45+/Ly6G−/F4/80+CD64+/MHCIIlo | |
Classical type 1 DCs | CD45+/Ly6G−/F4/80−CD64−/CD11b−/Ly6C−/CD11c+/MHCII+/CD103+/XCR1+ | |
Classical type 2 DCs | CD45+/Ly6G−/F4/80−CD64−/CD11b+/Ly6C+/CD11c+/MHCII+/CD103−/XCR1− | |
moDCs | CD45+/Ly6G−/F4/80−CD64−/CD11b+/Ly6C−/CD11c+/MHCII+/F4/80+/CD103− |
. | Population . | Gating Phenotype (after Cells/Single/Live) . |
---|---|---|
Lymphoid Panel | CD45+ | CD45+ |
Tumor cells | CD45−/GD2+ | |
NKT cells | CD45+/CD3+/NK1.1+ | |
CD8+ T cells | CD45+/CD3+/NK1.1−/CD8+ | |
CD4+ T cells | CD45+/CD3+/NK1.1−/CD4+ | |
Treg cells | CD45+/CD3+/NK1.1−/CD4+/CD25+/foxp3+ | |
B cells | CD45+/CD3−/CD19+ | |
NK cells | CD45+/CD3−/NK1.1+ | |
Myeloid Panel | Neutrophils | CD45+/CD11b+/Ly6G+ |
Macrophages | CD45+/Ly6G−/F4/80+CD64+ | |
M1-like macrophages | CD45+/Ly6G−/F4/80+CD64+/MHCIIhi | |
M2-like macrophages | CD45+/Ly6G−/F4/80+CD64+/MHCIIlo | |
Classical type 1 DCs | CD45+/Ly6G−/F4/80−CD64−/CD11b−/Ly6C−/CD11c+/MHCII+/CD103+/XCR1+ | |
Classical type 2 DCs | CD45+/Ly6G−/F4/80−CD64−/CD11b+/Ly6C+/CD11c+/MHCII+/CD103−/XCR1− | |
moDCs | CD45+/Ly6G−/F4/80−CD64−/CD11b+/Ly6C−/CD11c+/MHCII+/F4/80+/CD103− |
All samples were gated on all cells, followed by single cells identified by both forward and side scatter area versus height and live cells by viability dye exclusion.
To compare these four staining techniques (termed Control, Before, During, and After throughout this manuscript) across both lymphoid and myeloid panels, common pools of dissociated tumors and RBC-lysed spleens were made. We measured 27 distinct parameters for both tumor (Figs. 2B, 3) and spleen (Fig. 3, Supplemental Fig. 1) cell preparations and compared the performance of the three different cryopreservation protocols to the fresh control and to each other. To test whether the different cryopreservation methods had different effects on flow cytometry results, we compared the Before, During, and After conditions to each other using Kruskal–Wallis tests for differences in all 27 immune parameters for both tumor and spleen cell preparations (Table III). Significant (p < 0.05) differences were observed for all 27 parameters in the spleen samples and in 25 out of 27 parameters in the tumor samples, indicating that the cryopreservation timing did impact most immunophenotyping results.
Sequence of cryopreservation nonuniformly alters PD1 expression in T cells, B cells, and NK cells. Five C57BL/6 mice bearing B78 syngeneic melanoma tumors had tumors and spleens harvested as described in Materials and Methods and pooled to create a homogeneous starting population of both tumor- and spleen-dissociated cells. Each population was analyzed using the lymphoid and myeloid phenotyping panels described in (Fig. 1 and Tables I and II, using each of the four cryopreservation protocols (Control, Before, During, and After). MFI of PD1 staining was measured for CD8+ T cells (A and E), CD4+ helper (gated as CD4+, non-Treg) T cells (B and F), NK1.1+ NK cells (C and G), and CD19+ B cells (D and H) in both tumor (A–D) and spleen (E–H) samples. Statistical comparisons were made using two-sample t tests, comparing each staining condition to the nonfrozen control method. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. ns, not significant.
Sequence of cryopreservation nonuniformly alters PD1 expression in T cells, B cells, and NK cells. Five C57BL/6 mice bearing B78 syngeneic melanoma tumors had tumors and spleens harvested as described in Materials and Methods and pooled to create a homogeneous starting population of both tumor- and spleen-dissociated cells. Each population was analyzed using the lymphoid and myeloid phenotyping panels described in (Fig. 1 and Tables I and II, using each of the four cryopreservation protocols (Control, Before, During, and After). MFI of PD1 staining was measured for CD8+ T cells (A and E), CD4+ helper (gated as CD4+, non-Treg) T cells (B and F), NK1.1+ NK cells (C and G), and CD19+ B cells (D and H) in both tumor (A–D) and spleen (E–H) samples. Statistical comparisons were made using two-sample t tests, comparing each staining condition to the nonfrozen control method. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. ns, not significant.
Statistical differences between cryopreservation methods are detected in almost all immune parameters
Parameter . | Spleen . | Tumor . |
---|---|---|
Total events | <0.001 | 0.018 |
Total live events | <0.001 | <0.001 |
GD2 (% of live) | — | <0.001 |
CD45 (% of live) | <0.001 | <0.001 |
CD19 (% of live) | <0.001 | <0.001 |
NK1.1 (% of live) | <0.001 | <0.001 |
CD3 (% of live) | <0.001 | <0.001 |
NKT (% of live) | <0.001 | 0.002 |
CD8 (% of live) | <0.001 | <0.001 |
CD4 (% of live) | <0.001 | <0.001 |
Treg cells (% of CD4) | <0.001 | <0.001 |
Non-Treg cells (% of CD4) | <0.001 | <0.001 |
CD4+PD1+ (% of non-Treg nells) | <0.001 | <0.001 |
CD4+CD103+ (% of non-Treg cells) | <0.001 | 0.010 |
CD8+PD1+ (% of CD8) | <0.001 | <0.001 |
CD8+CD103+ (% of CD8) | <0.001 | <0.001 |
Neutrophils/MDSCs (% of live) | 0.011 | <0.001 |
Macrophages (% of live) | 0.006 | 0.002 |
M1-like (% of macrophages) | 0.004 | <0.001 |
M2-like (% of macrophages) | 0.005 | <0.001 |
cDC1 (% of live) | 0.001 | 0.363 |
cDC2 (% of live) | <0.001 | 0.558 |
moDCs (% of live) | <0.001 | 0.007 |
CD4+ non-Tregc cells PD1 MFI | <0.001 | <0.001 |
CD8+ PD1 MFI | <0.001 | 0.003 |
NK1.1+ PD1 MFI | <0.001 | <0.001 |
CD19+ PD1 MFI | <0.001 | <0.001 |
Parameter . | Spleen . | Tumor . |
---|---|---|
Total events | <0.001 | 0.018 |
Total live events | <0.001 | <0.001 |
GD2 (% of live) | — | <0.001 |
CD45 (% of live) | <0.001 | <0.001 |
CD19 (% of live) | <0.001 | <0.001 |
NK1.1 (% of live) | <0.001 | <0.001 |
CD3 (% of live) | <0.001 | <0.001 |
NKT (% of live) | <0.001 | 0.002 |
CD8 (% of live) | <0.001 | <0.001 |
CD4 (% of live) | <0.001 | <0.001 |
Treg cells (% of CD4) | <0.001 | <0.001 |
Non-Treg cells (% of CD4) | <0.001 | <0.001 |
CD4+PD1+ (% of non-Treg nells) | <0.001 | <0.001 |
CD4+CD103+ (% of non-Treg cells) | <0.001 | 0.010 |
CD8+PD1+ (% of CD8) | <0.001 | <0.001 |
CD8+CD103+ (% of CD8) | <0.001 | <0.001 |
Neutrophils/MDSCs (% of live) | 0.011 | <0.001 |
Macrophages (% of live) | 0.006 | 0.002 |
M1-like (% of macrophages) | 0.004 | <0.001 |
M2-like (% of macrophages) | 0.005 | <0.001 |
cDC1 (% of live) | 0.001 | 0.363 |
cDC2 (% of live) | <0.001 | 0.558 |
moDCs (% of live) | <0.001 | 0.007 |
CD4+ non-Tregc cells PD1 MFI | <0.001 | <0.001 |
CD8+ PD1 MFI | <0.001 | 0.003 |
NK1.1+ PD1 MFI | <0.001 | <0.001 |
CD19+ PD1 MFI | <0.001 | <0.001 |
The performance of both lymphoid and myeloid staining panels in tumor and spleen were compared with each other by testing for differences within the three cryopreservation protocols (excluding the noncryopreserved Control group) using Kruskal–Wallis tests. Note that GD2 was not tested in splenic tissue as it is a tumor-specific marker. Significant differences were detected within the three cryopreservation protocols for all parameters except for cDC1 (percentage of live cells) in the tumor and cDC2 (percentage of live cells) in the tumor.
To determine which of the three cryopreservation protocols most closely resembled the noncryopreserved control, we used a difference of means approach to determine the cryopreservation method that was closest to the reference Control method for all parameters. Our results demonstrate that cryopreservation after all staining and fixation (the After condition) had the smallest difference of means compared with the Control in 18 out of 27 parameters for tumors and 17 out of 26 parameters for spleen, including total live cells, total CD45+ cells, B cells, NK cells, T cells, NKT cells, Treg cells, macrophages, and others. The Before condition had the smallest difference compared with the Control in 6 out of 27 parameters for tumor (total events, neutrophils, cDC1 and cDC2 cells, CD8 PD1 median fluorescence intensity (MFI), and CD103 proportion among CD8 cells) and 2 out of 26 parameters for spleen (NK cells and total CD3+ cells). The During condition had the smallest difference compared with the Control in 3 out of 27 parameters in the tumor (moDCs, CD4eff PD1 MFI, and CD19+ B cell PD1 MFI) and 7 out of 26 parameters in the spleen (including neutrophils, macrophages, and cDC1 dendritic cells). In addition, Student t tests could not reject the null hypothesis of equivalence (p > 0.05) when directly comparing the After and Control groups in 13 out of 27 parameters in the tumor and 5 out of 26 parameters in the spleen; in other words, for these parameters, these data could not prove the After and Control groups were different, even though this does not prove they were equivalent. Although unable to statistically establish noninferiority or equivalence, these results do support the conclusion that the After cryopreservation protocol yielded results very similar to the Control method for these nonsignificant comparisons. Finally, the total number of cells analyzed (recovered events) were similar among the cryopreservation protocols; however, the degree of live cell recovery was substantially lower in the Before staining group compared with the Control, During, or After groups. Taken together, these results suggest that cryopreservation after staining and fixation is a feasible approach to staining samples that require longer-term storage. In the setting of tumor dissociates, the After protocol (cryopreservation after all staining and fixation) yields immunophenotyping results most concordant with freshly analyzed samples for overall cell recovery for almost all lymphocyte populations and some myeloid populations including macrophages. The Before staining condition yielded immunophenotyping results most concordant with freshly analyzed samples for neutrophil and dendritic cell populations.
It is known that PD1 is expressed on several immune cells, in particular T cells in response to TCR stimulation (51). Thus, it is used as a gauge of both T cell activation and exhaustion, which is commonly evaluated in the context of the immune TME. In comparing the degree of PD1 expression measured on CD8+ T cells and CD4+ non–Treg cells, a substantial difference was observed based on staining method (Fig. 3). MFI of PD1 on CD4+ non–Treg cells was almost double in the Before staining condition in both tumor and spleen compared with the freshly stained control (Fig. 3B, 3D). CD8+ T cells in the spleen exhibited the same doubling effect using the Before staining condition (Fig. 3C). However, in the tumor, where resident T cells are known to be exhausted and have high expression of PD1 (52), no such difference in PD1 expression was detected (Fig. 3A). In both CD19+ B cells (Fig. 3E, 3F) and NK1.1+ NK cells (Fig. 3G, 3H), the Before staining condition resulted in increased PD1 expression in both the tumor and the spleen, again by an almost 2-fold difference. These results suggest that the process of cryopreservation and thawing can either increase expression of PD1 on cells that are not already highly expressing PD1 or preferentially eliminate cells with low PD1 expression and that the After cryopreservation method results in the least change to population PD1 levels compared with Control.
To assess the broader applicability of the After cryopreservation technique toward constructing different staining panels with different fluorophores, we conducted a single-color staining study with C57BL/6 splenocytes using the Control and After staining protocols (Supplemental Fig. 2, Supplemental Table I). Single-color stains for CD45, CD4, CD8, CD19, and NK1.1, each with four different fluorophores (consisting of both standard and tandem dyes), were largely unaffected by fluorophore choice or by the After cryopreservation technique. Yield of CD11b+ cells was lower following the After cryopreservation technique compared with Control; however, the degree of change was similar regardless of fluorophore used. To demonstrate applicability of the After technique across strains and species, a simple four-color panel of CD45, CD4, CD8, and CD11b was tested in C57BL/6 mouse splenocytes, BALB/c mouse splenocytes, and human PBMCs (Supplemental Fig. 3). Results suggest that the C57BL/6 and BALB/c splenocytes were both able to similarly withstand the After cryopreservation protocol and had results very similar to the Control stained samples. The human PBMCs had percentages of CD8, CD4, and CD11b cells that were minimally affected by the After cryopreservation protocol as well. Taken together, these studies suggest that the After cryopreservation protocol is applicable across several different fluorophores and tissue types and can likely be adapted for use in a variety of applications.
The After cryopreservation technique yields similar conclusions as noncryopreserved samples when analyzing the immune effects of an in situ tumor vaccine
It is known that immune cells can have different physical and biological properties depending on their activation state. For example, activated lymphocytes are more resistant to ionizing radiation and genotoxic anticancer drugs than naive lymphocytes (53). To study the potential for a differential impact on immunophenotyping results based on the activation state of the immune TME, we compared the Control and After cryopreservation techniques (as depicted in (Fig. 2A) by flow cytometry analyses for tumor samples from mice receiving a previously published in situ vaccine tumor immunotherapy regimen (24). C57BL/6 mice bearing ∼5 wk syngeneic B78 flank melanoma tumors were treated with either PBS sham, 12 Gy RT only, or 12Gy RT + intratumoral hu14.18-IL-2 IC as previously described (24). Tumors were harvested, dissociated, and processed according to either the Control (fresh) or After (cryopreserved) staining protocols (from (Fig. 2A) using a hybrid lymphoid/myeloid flow cytometry staining panel (Supplemental Table I). Eight immune parameters were assessed using linear regression models for association with treatment (PBS, RT, or RT+IC), preparation condition (After or Control protocol), or treatment–preparation interaction (Fig. 4). No interaction between treatment and preparation technique was detected for any of the immune parameters, suggesting that the preparation technique used did not affect the impact of treatment. In addition, no significant effect of preparation was detected in the model for all cell populations other than NK cells, which had a significant p value of 0.046 (Fig. 4D). This suggests that NK cell content is slightly different in the cryopreserved preparation compared with fresh, but the nonsignificant treatment–preparation interaction suggests this difference does not change with treatment. Our analyses, based on modeling of treatment effect, identified significantly greater percentages for all seven distinct immune populations (CD45+ cells, CD4+ T cells, NK1.1+ NK cells, CD8+ T cells, CD4+CD25+ Treg cells, CD19+ B cells, and CD11b+ myeloid cells), expressed as a percentage of live cells, in the RT+IC immunotherapy group compared with PBS control (Fig. 4A, 4C–H). Although immunosuppressive Treg cells were higher in the RT+IC group, the CD8/Treg cell ratio was also substantially greater in the RT+IC group (Fig. 4B). Taken together, these data suggest that the After cryopreservation technique largely reflects the same flow cytometry results in the immunotherapy-treated TME as obtained using the Control technique and allow for similar conclusions to be drawn about the effect of an in situ vaccine.
Comparison of Control and After preparation methods for flow cytometric analyses of immune cells in the TME following immunotherapy with RT and IT-IC. C57BL/6 mice bearing ∼5 wk syngeneic B78 melanoma flank tumors were treated using a previously published in situ vaccine consisting of 12 Gy external beam radiation (RT) on treatment day 1, followed by five daily intratumoral injections of 50 μg hu14.18-IL-2 IC on treatment days 6–10. Staining method is indicated by the Cryopreservation row with (−) for Control and (+) for the After protocol. Staining was done using a hybrid lymphoid/myeloid panel of Abs. Immune populations in (A) and (C)–(H) are expressed as a percentage relative to all single, live cells in the sample, and (B) depicts the CD8+ T cell:Treg cell ratio. Data are presented with individual mouse tumors as data points, with bars representing mean ± SEM for each treatment and staining condition combination. Statistical comparisons were made using linear models assessing preparation protocol, treatment effect, and interaction. All parameters were determined to have an insignificant interaction on F test. The p values corresponding to Control versus After preparation protocol differences are presented in the bottom left of each corresponding graph. Data are representative of two independent biological replicates. Pairwise differences between PBS, RT, and RT+IC treatment groups derived from the linear model are indicated as *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.
Comparison of Control and After preparation methods for flow cytometric analyses of immune cells in the TME following immunotherapy with RT and IT-IC. C57BL/6 mice bearing ∼5 wk syngeneic B78 melanoma flank tumors were treated using a previously published in situ vaccine consisting of 12 Gy external beam radiation (RT) on treatment day 1, followed by five daily intratumoral injections of 50 μg hu14.18-IL-2 IC on treatment days 6–10. Staining method is indicated by the Cryopreservation row with (−) for Control and (+) for the After protocol. Staining was done using a hybrid lymphoid/myeloid panel of Abs. Immune populations in (A) and (C)–(H) are expressed as a percentage relative to all single, live cells in the sample, and (B) depicts the CD8+ T cell:Treg cell ratio. Data are presented with individual mouse tumors as data points, with bars representing mean ± SEM for each treatment and staining condition combination. Statistical comparisons were made using linear models assessing preparation protocol, treatment effect, and interaction. All parameters were determined to have an insignificant interaction on F test. The p values corresponding to Control versus After preparation protocol differences are presented in the bottom left of each corresponding graph. Data are representative of two independent biological replicates. Pairwise differences between PBS, RT, and RT+IC treatment groups derived from the linear model are indicated as *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.
Use of the After cryopreservation preparation technique enables demonstration of alterations to the TME following [90Y]NM600 MTRT
To demonstrate the use of the After cryopreservation protocol in analyzing radioactive tissue and to explore the time- and dose-dependent effects of [90Y]NM600 MTRT on the immune TME, we conducted a time course study in C57BL/6 mice bearing ∼5 wk syngeneic B78 flank tumors (Fig. 5). On treatment day 0, mice were given a tail vein injection of either PBS control, 50 μCi of [90Y]NM600, or 250 μCi of [90Y]NM600. Tumors were harvested weekly on treatment days 1, 7, 14, and 21 following injection, dissociated, and stained for analysis according to the After cryopreservation protocol using a hybrid innate/adaptive staining panel (Supplemental Table I). Immune parameters were assessed using linear regression models for association with time following injection, dose of treatment (PBS, 50 μCi, or 250 μCi), or time–dose interaction. Results demonstrate a significant (p < 0.05) effect of time for all parameters and significant time–dose interaction for all parameters except for NK cell content (Fig. 5B), the fraction of NK cells expressing CD11b (Fig. 5C), and neutrophil/MDSC content (Fig. 5E). This demonstrates that MTRT had a time- and dose-dependent effect on most immune populations.
Use of the After cryopreservation method to analyze radioactive tumor samples over time following MTRT. Mice bearing ∼5 wk B78 syngeneic melanoma tumors were treated with tail vein injections of either PBS control, 50 μCi, or 250 μCi of [90Y]NM600 MTRT. n = 4 tumors per group were harvested on days 1, 7, 14, and 21 postinjection, dissociated, and stained using a hybrid lymphoid/myeloid panel using the After cryopreservation protocol described in (Fig. 2. After 30 d stored behind lead shielding at −80°C, samples were confirmed to be at background radioactivity, thawed, and analyzed by flow cytometry. The above immune parameters depicted in (A), (B), (D), (E), (H), and (I) were calculated and presented as a percentage of cells of interest relative to all single, live cells. (C) Depicts the percentage of NK cells that are CD11b+, and (G) depicts the CD8 T cell:Treg cell ratio. Because of the very low frequency of CD19+ B cells, (F) presents data as natural log transformed. These data shown in this study are representative of two independent experiments. Data were analyzed using linear models to assess the effects of time, treatment condition, and time–treatment interaction on the above immune parameters. F statistics for time–dose interaction are presented as p values in the bottom left corner of each graph. No corrections for multiple comparisons were conducted. Pairwise analyses from the models were conducted within each time point only and are depicted as *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. ns, not significant.
Use of the After cryopreservation method to analyze radioactive tumor samples over time following MTRT. Mice bearing ∼5 wk B78 syngeneic melanoma tumors were treated with tail vein injections of either PBS control, 50 μCi, or 250 μCi of [90Y]NM600 MTRT. n = 4 tumors per group were harvested on days 1, 7, 14, and 21 postinjection, dissociated, and stained using a hybrid lymphoid/myeloid panel using the After cryopreservation protocol described in (Fig. 2. After 30 d stored behind lead shielding at −80°C, samples were confirmed to be at background radioactivity, thawed, and analyzed by flow cytometry. The above immune parameters depicted in (A), (B), (D), (E), (H), and (I) were calculated and presented as a percentage of cells of interest relative to all single, live cells. (C) Depicts the percentage of NK cells that are CD11b+, and (G) depicts the CD8 T cell:Treg cell ratio. Because of the very low frequency of CD19+ B cells, (F) presents data as natural log transformed. These data shown in this study are representative of two independent experiments. Data were analyzed using linear models to assess the effects of time, treatment condition, and time–treatment interaction on the above immune parameters. F statistics for time–dose interaction are presented as p values in the bottom left corner of each graph. No corrections for multiple comparisons were conducted. Pairwise analyses from the models were conducted within each time point only and are depicted as *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. ns, not significant.
The 50 μCi dose resulted in increased infiltration of CD11bhi myeloid cells on days 1 and 7 (Fig. 5D), and a mildly decreased infiltration of CD19+ B cells on day 7 compared with PBS (Fig. 5F). The 50 μCi dose also briefly increased the CD8+ T cell:Treg cell ratio on day 1 (Fig. 5G) because of a decrease in Treg cell content (Fig. 5I). The larger 250 μCi dose resulted in a sustained decreased in CD45+ cells (Fig. 5A), NK cells (Fig. 5B), CD19 B cells (Fig. 5F), and Treg cells (Fig. 5I) compared with both the PBS control and the 50 μCi dose on days 7, 14, and 21. The higher dose also resulted in a relative increase in CD11bhi myeloid cells on day 1 but a relative decrease on day 14 compared with PBS control (Fig. 5D). In addition to substantially decreased Treg cells, the higher MTRT dose also resulted in a decrease in CD8+ T cells on days 7 and 21 compared with control (Fig. 5H); however, the CD8 T cell:Treg cell ratio was significantly greater on days 7, 14, and 21 in the higher dose group compared with either PBS control or the lower 50 μCi dose group (Fig. 5G). Last, although total NK cells were decreased in the higher dose group on days 7, 14, and 21 (Fig. 5B), the fraction of NK cells expressing CD11b was higher in the 250 μCi dose group compared with PBS control for days 1, 7, and 14 (Fig. 5C). Together, these data demonstrate the changes in the immune TME following systemic administration of [90Y]NM600 as a function of both MTRT dose and time, with prominent, MTRT dose–dependent, long-lasting decreases of several immune populations, which are now able to be analyzed using the After cryopreservation protocol.
Discussion
In this study, we describe a cryopreservation and immunophenotyping approach for analyzing dissociated cell preparations from murine spleens and tumors requiring cryopreservation prior to flow cytometry analyses. We used two panels of phenotyping mAbs to evaluate the impact of three separate cryopreservation strategies that differed by when the cryopreservation step is performed in the staining/processing sequence. We compared cryopreservation done Before, During, and After staining/processing (Fig. 2A). Our data show that cryopreserving samples after all surface staining, fixation/perm, and internal staining (the After protocol) resulted in the least impact of cryopreservation on immunophenotyping results among all lymphocyte and macrophage populations when compared with freshly obtained (noncryopreserved) spleen and tumor cell preparations. This cryopreservation approach was then compared head-to-head with a no-cryopreservation protocol for tumors in mice receiving combination immunotherapy; analyses of the fresh and the After cryopreserved tumor samples demonstrated the same TME phenotype changes in response to an in situ vaccine. Last, this flow cytometry approach was used to characterize time- and dose-dependent changes to the TME following [90Y]NM600 injection in radioactive tissues requiring cryopreservation (to allow the radioactivity to decay to baseline levels to enable analysis using our core facility’s flow cytometer).
Cryopreservation of PBMCs is a common practice for multicenter studies to reduce variability and delays in processing and has been used heavily in the clinical study of HIV treatments (54). One study found that cryopreservation of PBMCs prior to staining did not substantially change the frequency of detected CD4 and CD8 T cell subsets but did notice changes in markers of activation, including CD62L and CCR5 (5). Another study concluded that cryopreservation of human PBMCs had a relatively small impact on the detected frequencies of terminally differentiated, but not naive, CD4+ and CD8+ T cells following cryopreservation for up to 12 mo (55). However, other more recent studies have described a drop in detected T and NK subsets following cryopreservation for up to 6 wk followed by staining (56, 57). This is consistent with our study, which shows that the frequency of NK1.1+, CD3+, CD4+, and CD8+ T cells are substantially reduced with the Before and During cryopreservation methods, but these decreased frequencies are substantially less (or not) evident with the After cryopreservation method (Fig. 2B, Supplemental Fig. 1). We also observed a significant decrease in CD19+ B cell frequency in both tumor and spleen, particularly following the Before method, which is consistent with published observations with cryopreserved PBMCs in humans (13). This decrease in B cells was substantially less evident when using the After cryopreservation method (Fig. 2B, Supplemental Fig. 1), which may be of interest to groups studying B cell–directed cancer immunotherapies.
Studies of human Treg cells, defined by surface marker expression of CD4, CD25, and lack of CD127, suggest a decrease in the frequency of detectable Treg cells following the cryopreservation of PBMCs (48, 58). However, van Pul and colleagues (12) detected a relative increase in the number of CD4+CD25+ cells in cryopreserved breast sentinel lymph node dissociates. This latter result is consistent with our observation that the relative number of CD4+CD25+FOXP3+ cells is substantially higher in the Before and During cryopreservation methods for both spleen and tumor but not for the After cryopreservation method (Fig. 2B, Supplemental Fig. 1). One possible explanation for the differences reported for human PBMC and solid tissue Treg cells (4, 12, 48) may be that PBMCs from liquid tissue and splenocytes extracted from solid tissue may react to cryopreservation differently. It is also important to note that our method for defining Treg cells includes the intracellular transcription factor FOXP3 in addition to surface markers. Nevertheless, van Pul and colleagues (12) speculate that the cryopreservation process may stress these populations such that regulatory immunosuppressive processes are activated. Indeed, we observed that lymphocytes (CD4, CD8, NK, and B cells) analyzed with the Before method had higher expression of PD1 compared with control (Fig. 3). This pattern is observed in tumor and splenic CD4+ T cells, B cells, and NK cells and in splenic (but not tumor) CD8+ T cells. However, in the tumor, which has high baseline expression of PD1 on tumor-infiltrating CD8+ T cells, no further increase over control was seen for PD1 levels using the Before cryopreservation method. This alteration may be due to induction of stress responses following the cryopreservation and thawing process or may represent selective survival of PD-1–expressing cells. It is certainly reasonable that exhausted PD-1+ lymphocytes may be less metabolically active and therefore less affected by cryopreservation and the subsequent thawing. In either scenario, this suggests that cryopreservation prior to staining may nonuniformly alter PD1 expression in a given population depending on its baseline expression and may therefore bias the resulting conclusions. As most multicenter clinical trials use a version of the Before cryopreservation method, studying the potential magnitude of this bias on other markers of activation and exhaustion may prove to be helpful.
The above-described changes in T cell, NK cell, and B cell frequencies and PD1 expression following cryopreservation highlight the significant stress that cells can be subjected to during the cryopreservation and thawing process. Although there are multiple published protocols that propose methods of mitigating the impact of cryopreservation on surface marker expression, including resting the cells in culture conditions prior to staining (35), it is likely that any protocol that subjects live, metabolically active cells to cryopreservation prior to staining risks nonuniformly altering an immune population and biasing conclusions. In particular, cryopreserving live cells and then thawing those cells that remained viable through the cryopreservation process might cause an increase in stress-related markers (like PD1) that would be apparent when the anti-PD1 staining is done after both the freezing and thawing. Alternatively, cryopreserving a viable cell population prior to fixation may result in selective cell survival that alters population-level expression of various markers, including PD-1 as demonstrated in this study. Our study demonstrates that, contrary to some of our initial assumptions, staining and fixing an immune cell population prior to cryopreservation reliably produces flow cytometry results most concordant with that of noncryopreserved samples and avoids the risk of such nonuniform biases. This principle has been demonstrated in ex vivo–stimulated splenocytes by Alice, Zebertavage, and colleagues (59, 60), in which splenocytes were surface stained, fixed, frozen for 4 h, and then stained for internal markers in a study of IFN-γ production. We hypothesize that the nonuniform “stress” of cryopreservation would be less likely to change the amount of stress molecules on the cells detected by fluorescent mAbs when the Abs have been applied and the cells have been fixed prior to cryopreservation, as is the case for the After method used in this study.
Our comparison between the Before, During, and After cryopreservation protocols demonstrated that the fluorophores themselves were able to withstand fixation and freeze/thaw cycles more robustly than previously hypothesized. Although repeated freeze/thaw cycles certainly are detrimental and do accelerate fluorophore degradation (61), a single freeze/thaw cycle in the context of the After protocol does not seem to degrade fluorophore signal to a degree that would interfere with detection of immune populations. This of course requires appropriate unstained and fluorescence minus one controls, which have experienced the same fixation and cryopreservation conditions, to be able to confidently capture immune populations. We have used multiple panels during the development and implementation of this protocol and have used the After cryopreservation technique with many common, commercially available fluorophores without issue (data not shown). Further, our direct comparison of staining CD45, CD4, CD8, CD19, CD11b, and NK1.1 with multiple different fluorophore-conjugated Abs demonstrates preservation of robust signal following cryopreservation for all tested Ab–fluorophore combinations (Supplemental Fig. 2, Supplemental Table I). Our protocol demonstrates similar resilience between tissue types as well; in addition to B78 melanoma and C57BL/6 splenocytes, we demonstrate similar staining following After cryopreservation in BALB/c splenocytes and human PBMCs (Supplemental Fig. 3). We additionally have experience staining and analyzing various immunotherapies in A/J, HHDII-DR1 (transgenic mice expressing human HLA-A2 and HLA-DR1) (62), and FVB mice as well as adoptively transferred OT-1 T cells (data not shown). These mice have ranged in ages from 6 to 52 wk of age without appreciable changes in the ability to resolve immune populations following the After cryopreservation protocol.
Another key variable when considering cryopreservation stability is the duration of cryopreservation. Most published protocols that cryopreserve live cells prior to analysis similar to the Before protocol note that, if done correctly, cells can be cryopreserved for years at a time and still be viable following thaw. Paredes and colleagues (47) directly observed that immunostained human PBMC samples frozen at −20°C following external staining and fixation could be reliably analyzed at 5, 15, and 30 d postcryopreservation for CD3, CD4, CD8, and CD11b. They also demonstrated that intracellular staining for IL-8 was unchanged in LPS-stimulated myeloid cells after 30 d of cryopreservation. Our observations extend these previously published findings to include murine tumor disaggregates and splenocytes, which are stable following staining fixation and cryopreservation for 30 d. In other applications of this After protocol, we have reliably stained and stored samples for up to 120 d at −80°C while preserving detection of lymphoid and myeloid immune populations (data not shown). It is important to note, however, that careful development and implementation of this protocol will depend on the specific Abs and fluorophores being used as well as the treatment conditions, sample preparation, and sample storage conditions of the facility. We would highly recommend testing the specific staining panel and cryopreservation protocol prior to implementation into critical studies.
Although the After cryopreservation protocol yielded results most concordant with the Control protocol for the majority of the immune parameters tested, certain myeloid populations were more accurately captured using either the Before or During method. This includes CD11b+Ly6G+ neutrophils in both spleen and tumor, dendritic cells in the tumor, and NK cells in the spleen (Fig. 2, Supplemental Fig. 1). Our analysis of cells treated with RT+IC also demonstrated a significant cryopreservation effect on NK cells using the After cryopreservation technique (Fig. 4). This is consistent with our single-color stain of CD11b in Supplemental Fig. 2, which did experience the most change following cryopreservation with the After protocol. Our research group focuses mostly on T cells in the tumor, and as such, these differences were largely acceptable. However, for applications focused on dendritic cell or NK cell biology and therapy, consideration of the Before or During conditions is warranted. As is the case with most cryopreservation-based flow cytometry protocols, analysis of freshly obtained samples is always preferable. However, when necessary, any of these three cryopreservation protocols can be considered based on the immune populations of interest and the tissue being studied. As before, we recommend testing the specific staining panel and tissue of interest using all three cryopreservation techniques prior to selecting a cell preparation technique for a given experiment.
Our data redemonstrate evidence of enhanced immune cell activation following combination RT+IC. In a previous study by Morris and colleagues (24), immunohistochemical analysis was used to demonstrate increased concentration of both CD8+ T cells and NK cells in B78 melanoma tumors treated with combined RT+IC compared with either alone. Our results confirm these prior findings of increased infiltrates of CD8+ T cells and NK cells. With these observations, we extend the characterization of the immune infiltrate following RT+IC in situ vaccine by demonstrating increased CD45+ cells, CD4+ T cells, CD19+ B cells, and CD11b+ myeloid cells and an increased CD8/Treg cell ratio. We also observe an increase in CD4+CD25+FOXP3+ Treg cells, which could limit the efficacy of treatment, although the enhanced CD8/Treg cell ratio supports a predominantly anti-tumor effector response. Importantly, comparing the effect of the cryopreservation technique (Control versus After) in the linear models demonstrated no difference in outcome based on technique for all parameters except for NK cells, which had a significant impact of cryopreservation compared with fresh samples. However, there was not a significant interaction between the treatment effect and the preparation effect for the difference in NK results, suggesting that this difference we observe in NK results between Control versus After should not affect comparisons between a treatment group and control group evaluated with the same technique. Nevertheless, further studies should clarify the change in detected NK cell populations, possibly using other NK markers that complement NK1.1, following treatment with RT+IC. It is also feasible to consider other cryopreservation protocols such as the Before protocol in the setting of specifically studying NK cells. However, the lack of significant treatment–preparation interaction in all tested parameters using the After protocol implies that cryopreservation does not affect activated immune cells differently than inactive ones in the immune subpopulations tested in (Fig. 4. Thus, our After cryopreservation technique is a reasonable approach to interrogate changes to the immune TME following immunotherapy.
Our cryopreservation protocol was used to evaluate samples from mice treated with [90Y]NM600 MTRT, which required 25+ days of lead-shielded storage to enable sufficient half-lives of the 90Y to expire to allow the radioactivity of the mouse tissues to decay to background levels. This allowed us to do the flow cytometry analyses of these samples, which were no longer radioactive, using our cancer center’s shared core service flow cytometry equipment (which does not allow analyses of any radioactive samples). This analysis successfully demonstrated phenotypic changes to the cells in the TME in a time- and dose-dependent manner following the administration of [90Y]NM600. The analysis did detect a time-dependent effect in most cell populations for all treatment groups, including the PBS-untreated control (data in (Fig. 5, statistical analysis not shown). These time-dependent alterations in the immune TME of these untreated tumors may be explained by the constantly changing size of the tumors over time. Tumor size at the time of treatment is known to have an impact on the efficacy of immunotherapies in similar preclinical settings (63). Thus, it is reasonable to expect some variation in immune populations in the control PBS tumors that are growing over time. However, the model results additionally demonstrated that the effect of dose changes with time for almost all of the measured parameters (as indicated by significant treatment–time interaction by F tests), which confirms that, in addition to baseline time-dependent changes without treatment, the MTRT agent does cause alterations to the immune TME in a dose-dependent manner.
The use of the After cryopreservation protocol has allowed for more detailed analysis of dose effect in the immune TME than has been previously described. Compared with untreated controls on the same day posttreatment, the 50 μCi dose did not alter most immune populations substantially, save for CD11bhi myeloid cells and Treg cells on postinjection day 1 (Fig. 5). This caused a small but statistically significant increase in the CD8/Treg cell ratio relative to the untreated control tumor. More substantial phenotypic changes were demonstrated with the higher 250 μCi dose, including a decrease in the number of CD45+ immune cells, NK cells, B cells, and Treg cells on days 7, 14, and 21 following injection compared with control. In addition to these decreases in cell numbers for some parameters, there is evidence that the proportions of certain immune populations were also altered following the higher dose. For instance, the fraction of NK cells expressing CD11b, which are considered “mature” NK cells, rose in the 250 μCi group following treatment (64). In addition, although CD8+ T cells were decreased on days 7 and 21, Treg cells were more substantially decreased on all posttreatment days, resulting in greater CD8/Treg cell ratios in the 250 μCi group on days 7, 14, and 21. These observations may indicate different radiosensitivities in different immune populations (for example, Treg cells versus CD8+ T cells). Importantly, these changes observed at 250 μCi were generally not observed at 50 μCi, which emphasizes the dose-dependent nature of MTRT. Additionally, changes following a 50 μCi injection were not detected after day 7, but the changes from the 250 μCi injection lasted longer, in some cases through day 21. This timing effect may be important in future studies when considering combining MTRT with other immunotherapies. Further investigation of the kinetic changes, dose response, and sensitivity of tumor type to MTRT are warranted to follow up on these findings. Clearly, there is a complex dynamic in the tumor in response to MTRT, and our After cryopreservation technique provides an opportunity to study these effects in detail.
This cryopreservation strategy may also be helpful in the several other situations in which flow cytometric analyses need to be performed at a delayed time, requiring cryopreservation of samples. Many immune flow cytometric analyses in animal models receiving cancer immunotherapy require analyses on samples from randomized groups of mice collected over several weeks; time is needed to implant tumors, allow them to grow to treatment size, administer treatment, and then wait until the predetermined time points posttreatment to harvest and analyze the tissues. During this window when mice are “committed” to the experiment, a loss in access to flow cytometers would result in a lost experiment. The ability to stain and cryopreserve the tumor samples for prolonged periods is a critical reserve option to preserve the experiment. In our institution’s case, the University of Wisconsin Carbone Cancer Center Flow Cytometry Laboratory was abruptly shut down for a 14-d quarantine following a positive contact with a COVID-19 patient. Using this cryopreservation method, experiments requiring flow cytometry were able to be stained and cryopreserved until the laboratory was properly disinfected, institutional policies could be drafted and followed, and adequate personal protective equipment was available to use safely, enabling completion of the planned flow cytometry analyses.
Acknowledgements
We thank all the individuals who contributed to the planning, experimentation, and analysis of this work with special thanks to Alexander L. Rakhmilevich, Amy K. Erbe, and Jacquelyn A. Hank for thoughtful discussion of experimental design and results. We thank Steve Kovo for the design and creation of the artwork in (Fig. 2A (website at www.stevekovo.com).
Footnotes
This work was supported by Midwest Athletes Against Childhood Cancer, Stand Up 2 Cancer, the St. Baldrick’s Foundation, the Crawdaddy Foundation, and the University of Wisconsin Carbone Cancer Center. This research was also supported in part by National Institutes of Health Grants TR002373, U54-CA232568, R35-CA197078, 5K08CA241319, 1DP5OD024576, U01-CA233102, P30 CA014520, P01 CA250972, F30CA228315 (to P.M.C.), and TL1 TR002375 (to P.M.C.). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.
The online version of this article contains supplemental material.
Abbreviations used in this article
- B78
B78-D14
- BL
blue laser
- cDC1
type 1 classical dendritic cell
- cDC2
type 2 classical dendritic cell
- GD2
disialoganglioside D2
- IC
immunocytokine
- IT
intratumorally
- MDSC
myeloid-derived suppressor cell
- MFI
median fluorescence intensity
- MHCII
MHC class II
- moDC
monocyte-derived dendritic cell
- MTRT
molecular targeted radionuclide therapy
- PD1
programmed cell death receptor 1
- perm
permeabilization
- RL
red laser
- RT
radiation therapy
- TME
tumor microenvironment
- Treg
T regulatory
- VL
violet laser
- YL
yellow laser
References
Disclosures
P.M.C., R.H., R.B.P., J.W., Z.S.M., and P.M.S. hold patents related to NM600. J.W. and Z.S.M. are members of the Scientific Advisory Board of Archeus Technologies. The other co-authors have no financial disclosures.