Impact of LBC Fixative Type and Fixation Time on Molecular Analysis of Pancreatic Cancer Cells: A Comparative Study of Cell Morphology, Antigenicity and Nucleic Acids.

Introduction: Liquid-based cytology (LBC) has been widely used since 2000. Next-Generation Sequencing (NGS) analysis of residual specimens in LBC fixative may also be performed for pancreatic cancer in the near future. We examined cell morphology, antigenicity and nucleic acids in pancreatic cancer cells at different fixation times using two types of LBC fixatives.
Methods: PANC-1 cells were fixed in 1 ml CytoRich Red (CR), CytoRich Blue (CB), 95% ethanol (95% AL) or 10% neutral buffered formalin (10% NBF) and evaluated for cell area, antigenicity and nucleic acids with fixation times of 1 hour and 1, 3, 9, and 14 days. Antigenicity was evaluated by immunocytochemical staining for p53 and CK20, and nucleic acid fragmentation was assessed by real-time PCR.
Results: There was no difference in total cell area between 1 hour and 14 day fixation times for the CR group, but the CB group showed cell contraction with 9 days fixation. In immunocytochemical staining, the CR group showed high p53 and CK20 positivity even after 14 days fixation. The CB group had a lower p53 positive rate than the CR group from 1 hour fixation. For nucleic acid fragmentation, Ct values for the CR group increased with fixation time. The CB group had consistently low Ct values.
Conclusion: Different LBC fixatives and fixation time can have varying effects on cell morphology, antigenicity and nucleic acids in pancreatic cancer cells. Therefore, fixative type and fixation time should be considered for molecular testing on residual samples in LBC fixatives. Copyright:

INTRODUCTION

Pancreatic cancer is a refractory cancer with increasing incidence and mortality in Japan and worldwide.[1] The cancer is highly aggressive and its frequent detection late in the disease course can lead to resistance to conventional therapy.[2] Endoscopic Ultra Sound-guided Fine Needle Aspiration (EUS-FNA) is a useful method for detecting malignant tumors of the pancreas and has high specificity, making this approach an indispensable guideline for diagnosis and treatment.[34] Results of recent studies indicated that the diagnostic rate for pancreatic mass can be increased by using liquid-based cytology (LBC) on samples obtained by EUS-FNA.[56789]

LBC methods have been widely used for clinical diagnoses since 2000, especially in the field of gynecology.[101112] The technical advantages of the LBC method over the conventional Papanicolaou smear method include more uniform distribution of cells, dry removal of specimens, reduction of specimen loss, and the ability to prepare multiple specimens.[11] In lung cancer, genetic analysis using the LBC method is also reported to be useful for diagnosis.[131415] In the near future, Next Generation Sequencing (NGS) analysis using residual specimens in LBC fixatives may also be performed clinically for pancreatic cancer.

Some commercially available LBC fixatives contain a small amount of formaldehyde and are known to improve antigen retention. However, the small amount of formaldehyde in LBC fixatives may cause DNA degradation. Thus, effects on antigenicity and nucleic acids must be considered when using LBC fixatives for cytological specimens. Moreover, when performing NGS analysis on residual samples in LBC fixatives, the number of tumor cells should be confirmed. As such, it is important to select a fixative that induces few morphological changes. However, the impact of LBC fixative type and fixation time on cell morphology, antigenicity and nucleic acids in pancreatic cancer cells has not been examined closely to date.

In the current study, we examined the cell area, nuclear and cytoplasmic antigens and DNA extraction or fragmentation in pancreatic cancer cells having a range of fixation times with two types of LBC fixatives. The impact of the different fixatives and fixation times on molecular analysis of pancreatic cancer cells was explored.

MATERIAL AND METHODS

Cell line

The human pancreatic epithelioid carcinoma cell line PANC-1 was purchased from Public Health England (Salisbury, UK). PANC-1 cells were cultured in RPMI-1640 medium (Sigma-Aldrich, Inc., St Louis, MO, USA) supplemented with 10% Fetal bovine serum (FBS; Invitrogen, Thermo Fisher Scientific, Waltham, MA, USA) and antibiotics (100 U/ml penicillin and 100 μg/ml streptomycin; Nacalai Tesque, Inc., Kyoto, Japan) in a humidified atmosphere of 5% CO2 at 37°C. PANC-1 cells are known to have k-ras mutation and p53 expression.

Sample fixation

The medium was aspirated and cells were washed with Dulbecco's phosphate-buffered saline (D-PBS; Nacalai Tesque, Inc., Kyoto, Japan) before detachment with trypsin treatment at 37°C for 3 min. PANC-1 cells (5 × 105) were fixed in 1 ml of CytoRich Red (CR), CytoRich Blue (CB) (Becton, Dickinson and Company, Franklin Lakes, New Jersey), 95% ethanol (95% AL) or 10% neutral buffered formalin (10% NBF) and the treatments were compared. Triplicate samples were evaluated for fixation times of 1 hour, 1 day, 3 days, 9 days, and 14 days and each sample was stored at room temperature.

Papanicolaou staining and cell morphology

After the indicated fixation time, the fixed cultured cells were centrifuged at 600 g for 5 min and the supernatant was removed to obtain the cell sediment, which was used with the BD SurePath method to prepare the specimens. The preparation procedure was as follows: 600 μl of purified water was mixed with the cell sediment, and the resulting cell suspension was dispensed into a dedicated chamber, in which three glass slides for each specimen were prepared. The prepared specimens were fixed in 95% ethanol for 30 minutes and stained with Papanicolaou stain. An Olympus BX53 microscope was used at 100x objective magnification to acquire images for 100 randomly selected cells prepared in each fixative storage solution. Overlapping cells and cells having indistinct boundaries were excluded from the analysis. Images were analyzed using ImageJ software (National Institutes of Health, Bethesda, Maryland). Within ImageJ, images first were converted from red, green, and blue (RGB) color to 8-bit grayscale. Using the Threshold tool in ImageJ, nuclei and cell borders were then isolated according to their nuclear pixel gray value on a scale of 0 to 255, with 0 being black and 255 being white. Once isolated, the area of nuclei and the total area of the cells were measured using the ImageJ measurement tool. The N/C ratio was calculated by dividing the nuclear area by the cell area (nuclear area/cell area).[16]

Immunocytochemical staining

The antibodies for immunocytochemical staining used were p53 (clone: DO-7 mouse monoclonal antibody, dilution factor: 200x, Dako) and CK20 (clone: Ks20.8 mouse monoclonal antibody, dilution factor: 100x, Dako). Immunostaining of p53 and CK20 is useful for pancreatic cancer diagnosis.[171819] Here we evaluated nuclear and cytoplasmic staining to detect p53 and CK20 expression in the PANC-1 cell line. The VENTANA BenchMark GX fully automated immunostaining system (Roche Diagnostics) was used to analyze samples subjected to the following staining procedure. Endogenous peroxidase activity was first blocked by permeation with 70% ethanol with 0.2% H2O2. Antigen activation (95°C, 8 min) was carried out with heat treatment (HIER). The primary antibodies were reacted for 24 minutes for p53 and 32 minutes for CK20, and the color was generated by the polymer method using ultraView DAB (Ventana ultraView DAB Universal Kit, Roche Diagnostics). After color development, the cells were contrast-stained with hematoxylin, dehydrated, permeabilized, and sealed. An Olympus BX53 microscope with a 40x objective was used to acquire images of cells stored in each fixative storage solution. The number of cells in 10 random fields of view was counted. Staining for p53 and CK20 was scored as 3+ if ≥50% of cells were immunostained, 2+ if 10-50% were stained, 1+ if <10%, and 0 (none) if there was no staining. For both immunostains, 0, and 1+ and 2+ were defined as no expression, whereas 3+ was defined as high expression (positive).[20] The positivity rate (number of positive cells/total number of cells) for p53 and CK20 antibodies was calculated. Overlapping cells and cells with unclear boundaries were excluded from the analysis.

DNA extraction

DNA extraction from the fixed cells was performed using the tissue protocol with the QIAamp DNA Mini Kit (Qiagen). The fixed cells were centrifuged at 500 g for 5 minutes to produce a cell pellet, which was vortexed for 15 s in 180 μl buffer ATL and 20 μl proteinase K and incubated for 1 hour at 56°C. Then, 200 μl buffer AL was added and the mixture was vortexed for 15 s before incubation at 70°C for 10 minutes. The cleaning procedure was then performed as described in the manufacturer's protocol. After the washing procedure, the cells were incubated for 1 min in 50 μl of buffer AE at room temperature. Extracted DNA was quantified using a Nano Drop ND-2000 UV-Vis spectrophotometer (Thermo Fisher Scientific) and stored at -20°C.

DNA fragmentation evaluation and PCR amplification

To analyze DNA fragmentation, extracted DNA samples (50 μl) were analyzed by a TaqMan Assay (Thermo Fisher Scientific) using the CFX96 Touch real-time PCR analysis system (Bio-Rad Laboratories). DNA samples were amplified using TaqMan master mix and two different probes (87bp and 256bp). DNA fragmentation was evaluated by calculating ΔCt values and relative ratios based on the number of cycles of amplified DNA, and the ratios were compared for fixation times of 1 hour, 1 day, 3 days, 9 days, and 14 days. The analysis was performed in triplicate. The PCR protocol was denaturation at 95°C for 10 min, followed by 45 cycles of 95°C for 15 s and 60°C for 1 min. ΔCt values were compared by calculating the Long Assay (256 bp)/Short Assay (87 bp) for each sample using Human Control DNA as the reference sample, and then calculating the relative ratio for each sample using the control as 1.0.

Statistical analysis

The Wilcoxon Kruskal–Wallis comparison test was used to compare differences between groups. P values <0.05 were considered statistically significant. Measurement data were expressed as mean ± standard deviation. Statistical analyses were performed using Statcel4 software.

RESULTS

Impact of LBC fixative and fixation time on cell morphology

The total cell area was smaller for the CR and CB group compared to the 95% AL group with 1 hour fixation [Table 1]. However, there was no difference in total cell area between 1 hour and 14 days for the CR group (P = 0.05). The CB group showed cell contraction from 9 days fixation (P < 0.05) [Table 1].

Table 1

Comparison of cell area of PANC-1 cells treated with various fixatives

Fixation time	CytoRich Red		CytoRich Blue		95% AL		10% NBF
	Mean±SD (μm2)	C.V. (%)	Mean±SD (μm2)	C.V. (%)	Mean±SD (μm2)	C.V. (%)	Mean±SD (μm2)	C.V. (%)
1 hour	156.8±40.2	0.26	219.9±123.3	0.56	376.1±112.1	0.30	122.1±36.5	0.30
1 day	148.1±37.8	0.26	186.9±56.4	0.30	295.5±91.9****	0.31	130.3±35.8	0.27
3 days	146.8±36.3	0.25	201.2±77.1	0.38	266.5±91.5***	0.34	123.4±41.5	0.34
9 days	150.1±36.9	0.25	152.5±40.3**	0.26	142.2±32.8**	0.23	127.4±34.3	0.27
14 days	159.4±38.0	0.24	160.1±34.9*	0.22	153.4±34.6*	0.23	127.5±30.9	0.24

**** 1 hour vs. 1 day P<0.05. *** 1 hour vs. 3 days P<0.05. ** 1 hour vs. 9 days P<0.05. * 1 hour vs. 14 days P<0.05

The CB group also showed a contraction in nuclear area from 9 days (P < 0.05) [Table 2], whereas differences for the CR group were only observed for 3 days fixation (P < 0.05). The N/C ratio for the CR group showed no difference as a function of fixation time (P = 0.05), but the CB group showed a decrease in the N/C ratio from 1 day (P < 0.05) [Supplementary Table 1].

Table 2

Comparison of nuclear area of PANC-1 cells treated with various fixatives

Fixation time	CytoRich Red		CytoRich Blue		95% AL		10% NBF
	Mean±SD (μm2)	C.V. (%)	Mean±SD (μm2)	C.V. (%)	Mean±SD (μm2)	C.V. (%)	Mean±SD (μm2)	C.V. (%)
1 hour	95.7±30.1	0.31	123.3±57.1	0.46	246.4±114.9	0.47	81.5±30.4	0.37
1 day	93.3±29.8	0.32	108.7±52.2	0.48	197.5±96.1****	0.49	71.0±24.4	0.34
3 days	81.3±26.7***	0.33	137.2±60.9	0.44	182.7±75.1***	0.41	67.6±26.9***	0.40
9 days	86.0±28.8	0.33	86.2±37.7**	0.44	89.6±36.0**	0.40	70.5±19.8	0.27
14 days	92.3±27.0	0.29	83.0±24.9*	0.30	87.5±22.9*	0.26	69.5±17.7	0.26

**** 1 hour vs. 1 day P<0.05. *** 1 hour vs. 3 days P<0.05. ** 1 hour vs. 9 days P<0.05. * 1 hour vs. 14 days P<0.05

Supplementary Table 1

Comparison of N/C ratio of PANC-1 cells treated with various fixatives

Fixation time	CytoRich Red		CytoRich Blue		95% AL		10% NBF
	Mean±SD (µm2)	C.V. (%)	Mean±SD (µm2)	C.V. (%)	Mean±SD (µm2)	C.V. (%)	Mean±SD (µm2)	C.V. (%)
1 hour	0.59±0.11	0.19	0.64±0.12	0.19	0.76±0.10	0.13	0.57±0.10	0.18
1 day	0.59±0.11	0.19	0.57±0.11****	0.20	0.75±0.11	0.15	0.55±0.10	0.17
3 days	0.58±0.11	0.18	0.56±0.17***	0.30	0.70±0.09***	0.13	0.56±0.10	0.19
9 days	0.56±0.09	0.17	0.52±0.10**	0.19	0.58±0.10**	0.17	0.54±0.08	0.16
14 days	0.55±0.09	0.17	0.53±0.10*	0.19	0.59±0.09*	0.16	0.54±0.08	0.15

****1 hour vs. 1 day P<0.05. *** 1 hour vs. 3 days P<0.05. ** 1 hour vs. 9 days P<0.05. * 1 hour vs. 14 days P<0.05

Evaluation of p53 and CK20 by immunocytochemistry staining

For p53 staining, the intensity for the CR group was maintained even at 14 days fixation, but the CB group showed a slight decrease in staining from 1 hour fixation, and this decreasing trend continued through 14 days [Figure 1]. CK20 staining slightly decreased at 14 days fixation for the CR group compared to 1 hour fixation, but the CB group showed almost no staining at 14 days fixation compared to 1 hour fixation [Figure 2].

Figure 1

Comparison of p53 immunocytochemistry staining for samples with 1 hour and 14 days fixation. Cells fixed with (a) CytoRich Red or (b) CytoRich Blue for 1 hour; and cells fixed with (c) CytoRich Red or (d) CytoRich Blue for 14 days

Figure 2

Comparison of CK20 immunocytochemistry staining for samples with 1 hour and 14 days fixation. Cells fixed with (a) CytoRich Red or (b) CytoRich Blue for 1 hour; and cells fixed with (c) CytoRich Red or (d) CytoRich Blue for 14 days

The CR group showed high positivity for p53 and CK20 staining at 82.7% and 84.8%, respectively, at 14 days fixation [Tables 3 and 4]. Staining intensity for the CR group was independent of fixation time, and both nuclear and cytoplasmic antigenicity were maintained across all fixation times. In contrast, the CB group had a lower p53 positive rate with only 40.3% positive from 1 hour fixation [Table 3]. Moreover, CK20 staining was 93.6% at 1 hour after CB fixation, but the positive rate dropped significantly thereafter [Table 4].

Table 3

p53 positive rate of PANC-1 cells treated with various fixatives

Fixation time	CytoRich Red Mean±SD (%)	CytoRich Blue Mean±SD (%)	95% AL Mean±SD (%)	10% NBF Mean±SD (%)
1 hour	89.5±3.2	40.3±5.9	33.2±3.1	91.8±2.3
1 day	90.5±2.3	33.8±4.4	40.5±8.7	87.3±4.5
3 days	89.1±5.8	30.0±3.3	37.2±8.0	81.2±4.9
9 days	80.9±4.5	28.1±6.7	39.3±8.4	79.6±5.1
14 days	82.7±3.9	61.5±5.7	55.6±12.7	86.5±3.0

Table 4

CK20 positive rate of PACN-1 cells treated with various fixatives

Fixation time	CytoRich Red Mean±SD (%)	CytoRich Blue Mean±SD (%)	95% AL Mean±SD (%)	10% NBF Mean±SD (%)
1 hour	97.9±2.0	93.6±2.1	88.5±4.7	96.1±2.4
1 day	87.3±1.4	1.7±1.8	2.9±1.6	96.6±1.7
3 days	86.9±3.2	2.3±2.6	6.2±4.5	95.5±3.0
9 days	86.5±3.7	2.5±2.4	4.5±2.3	96.2±4.1
14 days	84.8±2.4	2.3±2.7	29.7±12.5	98.0±1.2

Impact of various fixatives and fixation time on nucleic acid yield and quality

In analysis of nucleic acid preservation, the CR group showed a dependence on fixation time, and the amount of nucleic acid extracted gradually decreased as the fixation time increased [Figure 3a]. The CB group showed little change with fixation time, and the amount of extracted nucleic acids was stable through 14 days [Figure 3b]. Meanwhile, the 10% NBF group showed a significant decrease from 1 hour fixation time and showed a lower amount through 14 days. In contrast, the amount of extracted DNA from cells fixed with 95% AL was stable through 14 days. The mean DNA purity (ratio of 260 nm to 280 nm in samples) was acceptable for all groups (1.97, 1.99, 1.78, 1.99, respectively, for the CR, CB, NBF and AL groups).

Figure 3

Nucleic acid yield extracted from 5 × 105 PANC-1 pancreatic cancer cells. PANC-1 cells were fixed in solution containing 95% ethanol, and 10% neutral buffered formalin with (a) CytoRich Red or (b) CytoRich Blue (3 samples per group)

DNA fragmentation evaluation and relative comparison by ΔCt value

The CR and 10% NBF group showed an increase in Ct values with fixation time, whereas the CB and 95% AL groups both had consistently low Ct values [Figure 4a and 4b]. The 10% NBF group showed Ct values for 256 bp up to 9 days after fixation, but no Ct values could be determined at 14 days [Figure 4b].

Figure 4

Real-time PCR amplification in a TaqMan Assay. Amplification with (a) Short Assay (87 bp) and (b) Long Assay (256 bp) was compared and the (c) relative ratio was calculated. Measurements were made triplicate for each sample

The relative ratios of the CB and 95% AL groups fluctuated but did not decrease significantly, indicating little DNA fragmentation during the fixation time. In contrast, the CR and 10% NBF groups showed a decrease in the relative ratio with increasing fixation time, suggesting that DNA fragmentation occurred. In addition, the relative ratio of the 10% NBF groups was much lower than that for the CR groups [Figure 4c].

DISCUSSION

Several studies have reported the usefulness of LBC methods for pancreatic cytology.[5621] However, few have examined the impact of LBC fixative type and fixation time on cell morphology, antigenicity and nucleic acids in pancreatic cancer cells. Fujii et al.[22] reported that immunocytochemical staining of urothelial carcinoma cell lines for p16 and CD44 showed no difference in staining between LBC fixative with and without formaldehyde. In our study, p53 and CK20 staining was better in cells fixed with CR than those fixed with CB [Figures 1 and 2]. These findings suggest that LBC fixative with formaldehyde allows better antigen retention. Our results for LBC fixative without formaldehyde differed from those of Fujii et al.,[22] suggesting that antigen retention could vary depending on the type of antigen and cancer.

Recent studies have reported the usefulness of LBC samples for NGS analysis.[2324] For pancreatic cancer, NGS analysis using LBC specimens was reported to contribute to a possible improvement in the evaluation of pancreatic tumor malignancies and the application of molecular-targeted therapeutic agents.[25] Preservation of samples is a key step to successful NGS analyses. In this study, the nuclear morphology was not significantly different between CR and CB. However, in CB, changes in cell contraction were observed from 9 days [Tables 1 and 2]. This finding suggests that long-term storage with LBC fixative without formaldehyde could affect the determination of tumor burden in samples.

DNA degradation could be caused by crosslinking induced by formaldehyde.[131422] In our study, the amount of DNA extracted from cells with 10% NBF or CR treatment was lower than that for cells treated with 95% AL or CB [Figure 3a and 3b]. Our findings support that DNA degradation could be caused by formaldehyde cross-linking. On the other hand, the Ct value of CR was significantly lower than that for 10% NBF [Figure 4]. CytoRich Red may suppress nucleic acid fragmentation to a greater degree than that by 10% neutral buffered formalin and thus may not affect molecular testing approaches including NGS in pancreatic cancer.

There were some limitations to our study. First, we examined only one pancreatic cancer cell line. In patient specimens obtained with EUS-FNA, blood components can be present, so further investigation of the effects of fixatives in the presence of blood and protein components is needed. Second, in this study, polymerase chain reaction (PCR) analysis of k-ras mutations and all exon analysis by NGS using LBC samples was not performed. Thus, the direct impact of fixative and fixation time on these molecular testing methods requires investigation. Third, this study was limited to investigating the effects of fixatives in pancreatic cancer cells that had short-term storage. Therefore, further investigation of the effects of fixatives after long-term storage is needed.

In summary, we demonstrated that the effect on cell morphology, antigenicity and nucleic acids in pancreatic cancer cells could differ depending on LBC fixative type and fixation time. When performing molecular testing such as immunocytochemical staining and NGS analysis on residual samples in LBC fixatives, fixative type and time after fixation should be taken into consideration.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.