**2.4 Assessment of human MCC on FCM**

Cytotoxicity was assessed by two different FCM cytotoxicity techniques utilizing DIOC18 or mAb staining for target cell labeling (Özdemir, 2007, 2011). The basic strategy of two-color FCM assay involves labeling target cells with a fluorescent membrane dye DIOC18, in addition to staining with PI, to identify dead cells. Alternatively, a new three-color FCM approach called as "flow cytometric mast cell-mediated cytotoxicity assay (FCM-MCMCA) " , which entails target cell marking with specific mAb (CD19) and with AnnV/PI colabeling to identify apoptotic/dead target cells, was used for some tumor cells. Briefly, we performed the following stepwise approach for analyzing the samples:

#### **2.5 Pre-labeling target leukemia cells with DIOC18**

A stock solution was prepared by dissolving DIOC18 (Sigma, St. Louis, MO) in DMSO (2 mg/ml) overnight with agitation. The target cells (106cells/ ml) were incubated with 10µg/ml of DIOC18 (final concentration) in 1 ml of PBS containing %3 FBS for 45 minutes at incubator. Then, target cells were washed with PBS three times to get rid of dye remnant. DIOC18 pre-labeling before coincubation was done for K562, Meg-01, HL-60, DAMI cells and the patient samples (Fig.1C1-E3).

Table 1. The phenotypic characterizations of 4- to 8- week-old human bone marrow- derived

The human malignant B-lymphoblastoid cell lines such as Daudi/ Raji and erythroleukemia cell line K562, known as "LAK- sensitive" and "reference cells" in cytotoxicity studies, were utilized in this study. In several experiments we also utilized LAK- resistant human acute myeloid leukemia (AML) cell lines (HL-60, DAMI and Meg-01) and discarded AML patient samples were used. All cell lines were obtained from ATCC (Manassas, VA) and maintained in RPMI 1640 culture media, supplemented with 10% FBS. Before coincubation, target and effector cell viability were determined by the trypan blue exclusion test; a viability of 90%

Cytotoxicity was assessed by two different FCM cytotoxicity techniques utilizing DIOC18 or mAb staining for target cell labeling (Özdemir, 2007, 2011). The basic strategy of two-color FCM assay involves labeling target cells with a fluorescent membrane dye DIOC18, in addition to staining with PI, to identify dead cells. Alternatively, a new three-color FCM approach called as "flow cytometric mast cell-mediated cytotoxicity assay (FCM-MCMCA) " , which entails target cell marking with specific mAb (CD19) and with AnnV/PI colabeling to identify apoptotic/dead target cells, was used for some tumor cells. Briefly, we

A stock solution was prepared by dissolving DIOC18 (Sigma, St. Louis, MO) in DMSO (2 mg/ml) overnight with agitation. The target cells (106cells/ ml) were incubated with 10µg/ml of DIOC18 (final concentration) in 1 ml of PBS containing %3 FBS for 45 minutes at incubator. Then, target cells were washed with PBS three times to get rid of dye remnant. DIOC18 pre-labeling before coincubation was done for K562, Meg-01, HL-60, DAMI cells

mast cells by flow cytometry in representative samples are shown.

performed the following stepwise approach for analyzing the samples:

**2.5 Pre-labeling target leukemia cells with DIOC18** 

and the patient samples (Fig.1C1-E3).

**2.3 Target lymphoma and leukemia cells** 

**2.4 Assessment of human MCC on FCM** 

was required to proceed.

*Surface Marker Expressions 4th-week 6th-week 8th-week* 

CD14 0 % 0 % 0 % CD15 93 % 3 % 2 % CD19 0 % 0 % 0 % CD33 93 % 25 % 19 % CD34 7 % 1 % 1 % CD38 0 % 0 % 0 % CD44 93 % 85 % 84 % CD45 0 % 0 % 0 % CD49d 76 % 12 % 11 % CD117 (c-kit) 91 % 95 % 98 % HLA-DR 0 % 0 % 0 %

Fig. 1. C1- E3: Target cell (DAMI) death is shown by new flow cytometric assay at 24h and 48h in a representative sample. Identification of effector and target cells in alone/coincubation samples using target cell labeling (DIOC18) and cell size characteristics (FS) is shown in histograms at the first row. Spontaneous death and cytotoxicity evaluation of the cell populations alone or in co-incubation samples are shown with the rectilinear boxes of related histograms at the second and third rows. C1 shows location of effector-mast cell. Since effector cells are not pre-stained with DIOC18, they are DIOC18 negative and in a different area (Region G). C2 and C3 demonstrate the change in viability of effector cell alone population (DIOC18−/PI− viable cells: 94% and 91%; respectively) at 24h/48h. Spontaneous deaths in these effector alone samples are 6% and 9%; respectively. These histograms are obtained after gating on region G of a corresponding sample. D1 shows the target-DAMI cell population from target alone tube. As expected, target cells are very wellmarked with DIOC18 (Region J). D2 and D3 depict the viability of the target population alone (DIOC18+/PI− viable cells: 96% and 95%; respectively) at 24h and 48h. Spontaneous deaths in these samples are 4% and 5%; respectively. These histograms are obtained after gating on region J of a corresponding sample. E1 obtained from an experiment where mast cells are coincubated with DIOC18-positive DAMI cells at 2:1 effector/ target ratio. In this co-incubation sample; effector/target gatings are defined according to the target/effector alone (control) tubes. E2 depicts decrease in viability of previously gated target cells in J region of a corresponding co-incubation sample, 7% of them are stained with PI+ indicating necrotic killing (Region L2). In this representative sample, DAMI cell viability slightly decreased from 96% in the control to 93% after 24h co-incubation. The 4% of spontaneous death increased up to 7% with cytotoxicity mediated by mast cells during 24h co-incubation. E3 shows obvious decrease in viability of previously gated target DAMI cells, as 26% of them are dead. In this representative sample, DAMI cell viability further decreased from 95% to 74% throughout 48h co-incubation. The necrotic populations increased from a corresponding 5% of spontaneous death to 26% kill mediated by mast cell cytotoxicity in the co-incubated sample.

May Mast Cells Have Any Effect in New Modalities of Cancer Treatment? 9

Fig. 1. F1- I2. Target (Daudi) cell kill, caused by mast cell- mediated cytotoxicity, is

cell (target alone sample) in the tube. 1G2 shows spontaneous kill in the target cell population of the control sample. It shows 92% viable (Annexin-/ PI-) target (Daudi) cells

with total 8% spontaneous killing including early apoptotic plus necrotic kill. 1H1 demonstrates co-incubated effector and target cells at 1:1 ratio after 12h (co-incubation sample). It reveals the changes in the target cell population in the quadrant J of the fig. 1G1 in coincubated samples after cytotoxic kill. Compared to fig. 1G1, decrease in size and amount of the target cell population of 1H1 indicates killing due to mast cell –mediated cytotoxicity. Effector cells in this sample are not stained with CD19 and they are in a

different area. However; it is hard to say anything about conjugate formation between target

and effector cells (effector-target doublets) in the co-incubation sample, and it seems probable that it is happening earlier. The doublets probably form and immediately dissolve during the beginning of this process. Thus, there is no reflection of conjugate formation in

fig. 1H1. 1H2 histogram of co-incubation sample was acquired after gating on the population in the quadrant J of fig. 1H1 and reflects the alterations in that co-incubated target population after cytotoxic kill. It depicts obvious decrease in viability of the target cells from 92% to 48%. It shows increased death up to 52 %, which 6 % of killing was necrotic-late apoptotic (Annexin+/PI+) and 46 % of killing was early apoptotic

53% at a higher ratio.

(Annexin+/PI-). Obviously, decline in viability of the target cell population from 92% to 48% strongly reveals significant killing due to human mast cell- mediated cytotoxicity in vitro. Similarly, 1I1 shows further decrease in size and amount of the target cell population of 1G1 and 1H1 due to augmented cytotoxic kill at 2:1 ratio. And 1I2 demonstrates a

decrease in viability of the target cell population from 92% to 37%, but death increased up to

demonstrated by flow cytometric mast cell -mediated cytotoxicity assay (FCM-MCMCA) at 12h in a representative sample. Quadrant gating and determination of populations in fig. 1F1- H2 were defined by control samples consisting of target/effector alone tubes. The histograms in fig. 1G2-H2 were achieved from gating accordingly on target cell populations. 1F1 shows location of effector-mast cell in the histograms of FCM-MCMCA. Effector cells are seen as CD19 negative (effector alone sample). 1F2 demonstrates spontaneous kill in effector cell population in control tube. 1G1 shows target-tumor cells location. As expected, the target cell population is well-marked with CD19 in the histogram and there is no effector

#### **2.6 Coincubation**

A few sets of human BM-derived MCs at 8th weeks of age, without any stimulation (PMA etc.) and human tumor targets were coincubated at certain effector/target ratios (1:1, 2:1, 4:1, 5:1) short-term and long-term. Human MCs were coincubated in vitro with human LAKsensitive (K562, Daudi, Raji cells) and LAK- resistant tumor targets (Meg-01, HL-60, DAMI and patient cells) for the evaluation of human MCC. Tubes were centrifuged at 115xG for 5 minutes and incubated at 37ºC in 5% CO2 for short- (2h) and long-term (up to 48h) period. Although MCC is well-known to take place after long-term incubation (>18h), 2h and 48h coincubation times were selected so as not to miss any possible cytotoxicity. And these experiments were repeated several times with a few sets of MC colonies on different times. Samples were tested in duplicates for reproducibility and reliability.

#### **2.7 Target lymphoma cell labeling with mAb and detection of death after coincubation**

Labeling target cells were done by two different methods in this study. As mentioned earlier, the first one is DIOC18 prelabeling for leukemia samples, which is being done before coincubation. The second technique is based on marking the lymphoma targets such as Daudi and Raji cells after co-incubation with 10µl of mouse anti-human PE conjugated CD19 mAb (Immunotech, Westbrook, ME) for 20 minutes before staining for detection of death. Detection of apoptosis/death in target cells in mAb labeling technique was done by staining with AnnexinV (AnnV) and propidium iodide (PI) (TACS™ AnnexinV-FITC; R&D Systems, Minneapolis, MN) for 20 minutes before FCM acquisition. AnnV+ and/or PI+ events were analyzed from gating on the events in the target population of the control as well as coincubation samples. Ann V+ events represent early apoptotic population, AnnV+/ PI+ dead (late apoptotic or necrotic) cells (Fig.1F1-I2). In the DIOC18 prelabeling method, counterstaining with the nuclear dye PI (R&D Systems, Minneapolis, MN) of target cells was also done for 20 minutes before acquisition to discriminate between live and dead target cells. PI+ events represent dead cell population in FCM (Fig.1C1-E3).

#### **2.8 FCM acquisition and running**

FCM was performed using an EPICS-XL MCL (multicarousel) (Coulter, Miami, FL) equipped with an argon laser (15 mW) source operating at 488 nm. The emission of three fluorochromes was recorded through specific band pass filters: 525 nm for FITC (FL1), 575 nm for PE (FL2), 620 nm for fluorospheres (FL3), and 675 nm for PI (FL4). The instrument was set for 4-color analysis. As the emission spectra of the three different dyes utilized in this bioassay interfere with one another, appropriate electronic compensations were adjusted by running individual cell populations stained with each dye consecutively through the EPICS. Once the compensations had been set, a gating was done on forward scatter (FS) (ordinate) versus log-scale fluorescence of mAb CD19 or DIOC18 (abscissa), to separate target cells from effector cells (Fig.1C1-I2). The FS gating was especially helpful for separating bigger DAMI and Meg-01 target cells from effector MCs. The log-scale fluorescence (CD19 or DIOC18) of abscissa was well enough to separate targets from effectors besides the aid of FS gatings. The different cell populations (target, effector and coincubated cells) were gated on with the help of isotype control samples. To measure target-cell death and apoptosis, CD19-/DIOC18- positive events were gated on, and analysis of log green fluorescence (AnnV) and/or log red fluorescence (PI) was performed depend on the cytotoxicity technique ( Fig.1C1-I2).

A few sets of human BM-derived MCs at 8th weeks of age, without any stimulation (PMA etc.) and human tumor targets were coincubated at certain effector/target ratios (1:1, 2:1, 4:1, 5:1) short-term and long-term. Human MCs were coincubated in vitro with human LAKsensitive (K562, Daudi, Raji cells) and LAK- resistant tumor targets (Meg-01, HL-60, DAMI and patient cells) for the evaluation of human MCC. Tubes were centrifuged at 115xG for 5 minutes and incubated at 37ºC in 5% CO2 for short- (2h) and long-term (up to 48h) period. Although MCC is well-known to take place after long-term incubation (>18h), 2h and 48h coincubation times were selected so as not to miss any possible cytotoxicity. And these experiments were repeated several times with a few sets of MC colonies on different times.

**2.7 Target lymphoma cell labeling with mAb and detection of death after coincubation**  Labeling target cells were done by two different methods in this study. As mentioned earlier, the first one is DIOC18 prelabeling for leukemia samples, which is being done before coincubation. The second technique is based on marking the lymphoma targets such as Daudi and Raji cells after co-incubation with 10µl of mouse anti-human PE conjugated CD19 mAb (Immunotech, Westbrook, ME) for 20 minutes before staining for detection of death. Detection of apoptosis/death in target cells in mAb labeling technique was done by staining with AnnexinV (AnnV) and propidium iodide (PI) (TACS™ AnnexinV-FITC; R&D Systems, Minneapolis, MN) for 20 minutes before FCM acquisition. AnnV+ and/or PI+ events were analyzed from gating on the events in the target population of the control as well as coincubation samples. Ann V+ events represent early apoptotic population, AnnV+/ PI+ dead (late apoptotic or necrotic) cells (Fig.1F1-I2). In the DIOC18 prelabeling method, counterstaining with the nuclear dye PI (R&D Systems, Minneapolis, MN) of target cells was also done for 20 minutes before acquisition to discriminate between live and dead target

FCM was performed using an EPICS-XL MCL (multicarousel) (Coulter, Miami, FL) equipped with an argon laser (15 mW) source operating at 488 nm. The emission of three fluorochromes was recorded through specific band pass filters: 525 nm for FITC (FL1), 575 nm for PE (FL2), 620 nm for fluorospheres (FL3), and 675 nm for PI (FL4). The instrument was set for 4-color analysis. As the emission spectra of the three different dyes utilized in this bioassay interfere with one another, appropriate electronic compensations were adjusted by running individual cell populations stained with each dye consecutively through the EPICS. Once the compensations had been set, a gating was done on forward scatter (FS) (ordinate) versus log-scale fluorescence of mAb CD19 or DIOC18 (abscissa), to separate target cells from effector cells (Fig.1C1-I2). The FS gating was especially helpful for separating bigger DAMI and Meg-01 target cells from effector MCs. The log-scale fluorescence (CD19 or DIOC18) of abscissa was well enough to separate targets from effectors besides the aid of FS gatings. The different cell populations (target, effector and coincubated cells) were gated on with the help of isotype control samples. To measure target-cell death and apoptosis, CD19-/DIOC18- positive events were gated on, and analysis of log green fluorescence (AnnV) and/or log red fluorescence (PI) was performed depend

Samples were tested in duplicates for reproducibility and reliability.

cells. PI+ events represent dead cell population in FCM (Fig.1C1-E3).

**2.8 FCM acquisition and running** 

on the cytotoxicity technique ( Fig.1C1-I2).

**2.6 Coincubation** 

Fig. 1. F1- I2. Target (Daudi) cell kill, caused by mast cell- mediated cytotoxicity, is demonstrated by flow cytometric mast cell -mediated cytotoxicity assay (FCM-MCMCA) at 12h in a representative sample. Quadrant gating and determination of populations in fig. 1F1- H2 were defined by control samples consisting of target/effector alone tubes. The histograms in fig. 1G2-H2 were achieved from gating accordingly on target cell populations. 1F1 shows location of effector-mast cell in the histograms of FCM-MCMCA. Effector cells are seen as CD19 negative (effector alone sample). 1F2 demonstrates spontaneous kill in effector cell population in control tube. 1G1 shows target-tumor cells location. As expected, the target cell population is well-marked with CD19 in the histogram and there is no effector cell (target alone sample) in the tube. 1G2 shows spontaneous kill in the target cell population of the control sample. It shows 92% viable (Annexin-/ PI-) target (Daudi) cells with total 8% spontaneous killing including early apoptotic plus necrotic kill. 1H1 demonstrates co-incubated effector and target cells at 1:1 ratio after 12h (co-incubation sample). It reveals the changes in the target cell population in the quadrant J of the fig. 1G1 in coincubated samples after cytotoxic kill. Compared to fig. 1G1, decrease in size and amount of the target cell population of 1H1 indicates killing due to mast cell –mediated cytotoxicity. Effector cells in this sample are not stained with CD19 and they are in a different area. However; it is hard to say anything about conjugate formation between target and effector cells (effector-target doublets) in the co-incubation sample, and it seems probable that it is happening earlier. The doublets probably form and immediately dissolve during the beginning of this process. Thus, there is no reflection of conjugate formation in fig. 1H1. 1H2 histogram of co-incubation sample was acquired after gating on the population in the quadrant J of fig. 1H1 and reflects the alterations in that co-incubated target population after cytotoxic kill. It depicts obvious decrease in viability of the target cells from 92% to 48%. It shows increased death up to 52 %, which 6 % of killing was necrotic-late apoptotic (Annexin+/PI+) and 46 % of killing was early apoptotic (Annexin+/PI-). Obviously, decline in viability of the target cell population from 92% to 48% strongly reveals significant killing due to human mast cell- mediated cytotoxicity in vitro. Similarly, 1I1 shows further decrease in size and amount of the target cell population of 1G1 and 1H1 due to augmented cytotoxic kill at 2:1 ratio. And 1I2 demonstrates a decrease in viability of the target cell population from 92% to 37%, but death increased up to 53% at a higher ratio.

May Mast Cells Have Any Effect in New Modalities of Cancer Treatment? 11

Student's t-test and/or Wilcoxon test. Correlations were estimated by Pearson correlation coefficients. Significance was considered as ≤0.05. All statistics were done using the

This study demonstrated MCC against tumor cells by using human cells instead of murine effector/target cells. In vitro MCC against various tumor cells at different E:T ratios (1:1-5:1) was measured by FCM-MCMCA and DIOC18 methods on various periods (Table 2A-D).

**Pecentage Based Mast Cell- Mediated Cytotoxicity (%, PC)** 

**LAK-sensitive Death Type 2h 18h 48h** 

**LAK-resistant Death Type 2h 18h 48h** 

**Cell characteristics %, PC on Coincubation Times** 

2±1 3±1 4±2 (3) NS

1±1 2±1 9±1 (3) NS

2±1 1±0 7±1 (3) NS

3±1 5±1 6±2 (3) NS

1±1 2±1 5±1 (6) NS

1±1 5±1 7±1 (9) NS

2±1 5±1 19±2 (9) NS

3±1 5±1 20±2 (9) 0.059

4±1 8±1 18±1 (9) 0.050

3±1 8±1 20±1 (9) NS

3±1 5±1 18±2 (9) NS

3±1 7±1 24±2 (9) NS

3±1 8±1 27±4 (9) **0.031** 

4±1 11±2 39±3 (9) **0.011** 

4±1 13±1 22±2 (9) NS

Statistical Package for Social Sciences (SPSS-13 for Windows; Chicago, IL).

Spontaneous Spontaneous Cytotoxic N= P=

Spontaneous Spontaneous Cytotoxic N= P

Spontaneous Spontaneous Cytotoxic N= P=

Spontaneous Spontaneous Cytotoxic N= P=

Spontaneous Spontaneous Cytotoxic N= P=

denotes the number of experiments were repeated in each series. NS means not significant.

The death/killings (percentage based cytotoxicity) in this table reflect the death at a 5:1 ratio in this study. Spontaneous kill reflects the death in target or effector alone tubes. Cytotoxic death shows the target cell death by only mast cell- mediated cytotoxicity in coincubation sample. The percentage based cytotoxicity was calculated according to the given formula in the methods. All values are given as mean±SEM. P values reflect the significance of differences between spontaneous and mast cell mediated cytotoxic killing. N

Table 2A. In vitro human mast cell -mediated cytotoxicity measured by DIOC18 method

**3. Results** 

Effector (**Mast**) control/alone Target (**K562**) control Coincubation (5:1)

Effector control /alone Target (**DAMI**) control Coincubation (5:1)

Target (**Meg-01**) control Coincubation (5:1)

Effector control

Effector control Target (**HL-60**) control Coincubation (5:1)

Effector control

Target (**Patients`**) control Coincubation (5:1)

against different human tumor cells is shown.

#### **2.9 FCM data analysis**

An average of 10,000 total events and 3,000 target cells were collected per sample. The gating for the target cells was based on the target-alone analysis and kept constant throughout all tubes to avoid exaggeration of the counts due to apoptotic body contamination. Cytotoxicity calculations were based on viable populations in target-alone (control) and co-culture (coincubated) tube analysis results. Viable target-cell percentage was determined, and calculations were based on the control-tube (target alone) values. Apoptotic and/or necrotic death happening in control tubes were identified as spontaneous, but those killed in coincubated tubes were identified as cytotoxic killing. We expressed the MCC as percentage based cytotoxicity:

#### **2.10 Percentage based cytotoxicity (%, PC)**

As mentioned, we defined MCC occurring in co-culture tube (coincubated sample) as percent (%, PC) cytotoxicity. And cytotoxicity calculations were based on viable populations in target-alone (control) and co-culture tube analysis results.

$$\text{PercentCytokivity (PC)} = \frac{\text{[Control-viable cell\%]} - \text{[Coincubation-viable cell\%]}}{\text{[Control-viable cell\%]}}$$

Our in vitro experiments were repeated several times (n=27) with a few sets of MC colonies. Samples were tested in duplicates for reproducibility and reliability (mean coefficients of variation (CV) of 1.1% (95% CI, 0.8- 1.7%); r2=0.93, and p<0.001).

### **2.11 Comparison with chromium- 51Cr release assay (CRA)**

A standard assay was performed as previously described (Özdemir, 2007). Briefly, the assay is a standard 4h CRA using target cells that have been prelabeled with 100 μCi 51Cr (Perkin Elmer, Boston, MA) for 1h. Several concentrations of MCs were added to a fixed number of target cells (5,000) in a round bottom microtiter plate to a total volume of 0.2 ml. Following the 4h incubation, 0.1 ml of the supernatant was carefully harvested and counted on a scintillation counter (Packard, Downers Grove, IL). Maximum release was determined from wells including target cells and 10% sodium dodecyl sulfate in the medium. The PC was calculated using the following equation: (E–S)/(M–S)x100, where E is the experimental counts/minute, S is the spontaneous counts/minute and M is the maximum counts/minute. In the end, when we used some sets of our experimental data in comparing FCM cytotoxicity assays with CRA results, there was a significant correlation for PC (n=58; r= 0.95; P<0.001), similar to our earlier studies (Özdemir, 2007, 2011).

#### **2.12 Cell staining with Wright/Giemsa for light microscopy**

At the end of the coincubation period, the sample was gently mixed and 10µl was used to prepare slides at room temperature without cytospinning. The slides were then stained with Wright/Giemsa. Conjugate formations between some tumor cells and MCs were seen (Fig.1A1-A4).

#### **2.13 Statistical analysis**

Unless otherwise specified, all data are presented as means±SEM. The significance of differences between spontaneous and cytotoxic kill was determined using the paired Student's t-test and/or Wilcoxon test. Correlations were estimated by Pearson correlation coefficients. Significance was considered as ≤0.05. All statistics were done using the Statistical Package for Social Sciences (SPSS-13 for Windows; Chicago, IL).
