Journal of Fluorescence (2018) 28:1393–1404

https://doi.org/10.1007/s10895-018-2306-4

ORIGINAL ARTICLE

The Detailed Comparison of Cell Death Detected by Annexin V-PI

Counterstain Using Fluorescence Microscope, Flow Cytometry
and Automated Cell Counter in Mammalian and Microalgae Cells
Emine Koç 1 & Selcen Çelik-Uzuner 2 & Uğur Uzuner 2 & Ramazan Çakmak 1

Received: 31 July 2018 / Accepted: 14 October 2018 / Published online: 21 October 2018
# Springer Science+Business Media, LLC, part of Springer Nature 2018

Abstract
The evaluation of cell wellness is an important task for molecular biology research. This mainly comprises the
assessment for morphology and viability of culturing cells. Annexin V-Propidium iodide counterstaining has been
currently one of the common and easy methods to discriminate apoptotic and necrotic cell profiles. The method is
operated by fluorescence-based detection of counterstain via laser beam-employed instruments including flow
cytometer, fluorescence microscope and automated cell counter. The detection is primarily conducted based on the
same principle; however the efficiency of instruments may vary. Here we evaluated the efficiency of those instruments
for the clear-cut detection of cell death through various mammalian and microalgae cell lines. To the best of our
knowledge, this is the first study revealing comparative analyses of apoptotic and necrotic cells in mammalian and
microalgae cells using Annexin V-PI counterstain detected by flow cytometer, fluorescence microscope and automated
cell counter. Fluorescence microscope and cell counter instruments were also tested and compared for the traditional
trypan blue-based cell viability detection performance. For these, cell death was induced by UV-irradiation and/or bee
venom for mammalian (pancreatic cancer, metastatic breast cancer and mouse fibroblasts) and microalgae cells
(Chlorella vulgaris), respectfully. Findings postulated that automated cell counter and fluorescence microscopy re-
vealed similar patterns for the detection by both counterstain and trypan blue in mammalian cells. Interestingly, flow
cytometry did provide an accurate and significant detection for only one mammalian cell line when UV-treatment was
followed by routine Annexin V-Propidium iodide counterstaining. Unlike, only flow cytometry revealed a significant
change in the detection of death of microalgae cells by Annexin V-Propidium iodide method, but both Annexin and
conventional trypan blue methods were not applicable for the automated cell counter and microscopic detections for
microalgae cells. The related outputs propose that the obtaining reliable quantitation strongly depends on cell type and
instruments used. These suggest the necessity of optimization and validation endeavors before any cell death detection
initiative. The analytical outcomes present insights into detailed assessment of cell death detection of eukaryotic cells
and provide a direction to researchers to consider.

Keywords Cell death . Annexin V-PI counterstain . Trypan blue . Flow cytometer . Fluorescence microscope . Automated cell
counter

Introduction

* Selcen Çelik-Uzuner
selcen.celik@ktu.edu.tr
1
Department of Molecular Biotechnology, Faculty of Science,
Karadeniz Technical University, 61080 Trabzon, Turkey
2
Department of Molecular Biology and Genetics, Faculty of Science,
Karadeniz Technical University, 61080 Trabzon, Turkey
The assessment of cell viability is one of the major appli-
cations used in molecular biology research. This plays an
important role for both routinely verify the wellness of
culturing cells and the experimental evaluation of the cy-
totoxic effects of potential drug candidates on cells. There
is a range of methods to detect the ratio of dying cells in a
cell population. Of which, trypan blue dye is quite useful
1394
for the exclusion of live cells from dead ones under light
microscopes. This commonly used method is typically
open to some subjectivity when practice the counting
manually. The most important restrictions of manual
counting are; (i) only limited number of cells are counted
and this may not properly cover the actual specimen, (ii)
personal artifacts may lead to false positive counting of
debris or dirt as cells, (iii) long-lasting counting may ad-
versely affect the whole experimental process. On the other
hand, modern cell counters are able to minimize these limita-
tions as they can automatically count cells and calculate the
ratios without artifacts. Sizes of particles that are not cells can
be identified, differentiated and excluded from the counting
values by automated machines.
Another common method is based on the staining of
Annexin-V that is a specific marker for apoptosis. This
method uses fluorophores that bind to a marker protein,
and requires expensive instruments including fluorescence
microscope and flow cytometer. Another high standard
staining process is performed by a combinatory use of
Annexin V with propidium iodide (PI) which is a DNA
intercalating counterstain and commonly preferred to dis-
criminate apoptotic and necrotic cells in a population. In
principle, Annexin V binds to phosphatidylserine local-
ized in the outer layer of membrane of apoptotic cells,
and PI goes into the cell and intercalates to DNA when
cell membrane is disrupted. Cells can be analyzed by this
method using flow cytometer, cell counter and/or fluores-
cence microscope. Annexin V-positive cells are specified
as early apoptotic, PI-positive cells are denoted as necrot-
ic, and positive cells for both staining are symbolized as
late apoptotic in a population. The related kit has been
marketed by different manufacturers and the protocols re-
ferred are almost identical. Here, we aimed to compara-
tively evaluate the effectiveness of flow cytometer, auto-
mated cell counter and fluorescence microscope on the
quantitative detection of cell death in mammalian (both
cancer and normal cells) and microalgae cells. Typical cell
death process was induced by UV exposure and the cells
were then separated into three tubes for the quantitative
detection through those three instruments. The results re-
vealed that automated cell counter and fluorescence mi-
croscope provided the detection of induced cell death with
similar significance from all cells examined, yet flow
cytometer unexpectedly delivered a significant difference
for only one cell line. Instead, the best cell death detection
and related quantitation in microalgae cells was identified
through flow cytometry. The observations suggest that
Annexin V – PI test may not be suitable for flow
cytometry-based cell death detection of mammalian cells,
but fluorescence microscope and automated cell counter.
Nevertheless, it appears more sensitive to measure cell
death of microalgae.
J Fluoresc (2018) 28:1393–1404
MMaterials and Methods
1. Cell culture: Mammalian cells AR42J (ECACC, Cat.
No. 93100618), MDA-MB231 (ATCC, Cat. No. HTB-
26) and NIH3T3 (ATCC, Cat. No. CRL1658) cells
were cultured in RPMI (Sigma, Cat No. R8758) or
DMEM (Wisent, Cat. No. 319–005-CL) media includ-
ing 10% fetal bovine serum (Cat. No. 12103C) (bovine
calf serum, Sigma, Cat No. 12133C, for NIH3T3) and
1% Penicillin-Streptomycin (Wisent, Cat. No. 450–
201-EL). Cells were incubated in a 37 °C incubator,
humidified with 5% CO2. Cells used were at passages
12–19. Microalgae cells (CV211/11 J) were incubated
in TAP medium.
2. Inducement of cell death: In mammalian cells, cell death
was induced by two designs of UV treatment by UV lamb
(253.7 nm) targeting chronic and acute cytotoxic effect of
UV; (i) Model 1: 1 min UV treatment was followed by
24 h post-incubation, or (ii) Model 2: 10 min UV treat-
ment was followed by 1 h post-incubation. The cells were
exposed to UV light from the same distance on dishes
without lid. Microalgae cells were treated with UV for
15 or 45 min followed by 4 h or 24 h post-incubation
respectively. Bee venom at concentrations of 100, 300,
375 or 500 mg/ml for 24 h or 6 days was also used to
induce cell death in microalgae cells.
3. Trypan Blue staining: Cells after treatments were
trypsinized or directly centrifuged and the cell suspen-
sions in 1xPBS were incubated with trypan blue (0.4%)
(Sigma, Cat. No. T8154) (1:1 volume) for 5 min, and
assessed using a hemocytometer.
4. Annexin V – Propidium Iodide staining: After treat-
ments, the mammalian cells were trypsinized, yet
microalgae cells were directly centrifuged before
performing Annexin V-PI staining protocol (BD, Cat.
No. 556547). The protocol briefly includes (i) washing
cells with cold 1xPBS once, (ii) suspension of 1 × 106
cells with 1xbinding buffer (100 μl), (iii) addition of
Annexin V (5 μl) and PI (5 μl), and finally, (iv) comple-
tion the volume to 500 μl with 1xbinding buffer. Cells
were kept in dark till assessment and were analyzed by
flow cytometer within one hour. Cells at FSC-SSC (for-
ward scatter-side scatter) plot were gated according to
untreated cells and percentages felt into four regions were
calculated.
5. Instruments: Trypan blue staining was assessed by
brightfield tool of fluorescence microscope (Zeiss
AxioVert, Germany) and automated cell counter
(Countess FLII, Thermoscientific). Annexin V-PI staining
was measured by fluorescence microscope, cell counter
and flow cytometer (BD Accuri). Some experiments in-
cluded serial measurement of cells loaded on the specific
hemocytometer of cell counter (0.5, 1, 3 and 5 min) to
J Fluoresc (2018) 28:1393–1404 1395

identify whether the counter was able to calculate the
percentages consistently for the same sample.
6. Statistics: Cell viabilities by trypan blue method and
apoptotic/necrotic cells by Annexin V-PI method were
measured as percentages, and percentages were arcsine-
transformed followed by comparison with UNIANOVA
test of SPPS software (Version 13). Multiple significant
levels as p < 0.05 (*), p < 0.01 (**), p < 0.001 (***) and
p < 0.0001 (****) were used for detailed comparisons. All
experiments were performed as independent triplicates.
Results
Cell Viability Detection Experiments via Trypan Blue
Staining in Mammalian Cells
Cell viabilities after treatments were calculated by both man-
ual counting using a hemocytometer and automated cell
counter. For MDA-MB231 cells, manual counting did not
reveal a significant change in cell death after UV treatment
(p > 0.05) (Fig. 1a and c), whereas automated cell counter was
able to significantly identify the induced cell death in both UV
models as compared to untreated cells (p < 0.0001) (Fig. 1b
and d). Serial measurements with automated cell counter pro-
vided similar results in all cases (Fig. 1b and d). In automated
counting, the standard error of the mean of total cell number
(between independent repeats) was also at acceptable level,
suggesting that automated cell counter could reveal more ac-
curate results among the same measurements (Fig. 1e). In
contrast to MDA-MB231 cells, both counting approaches re-
vealed a significant decrease in AR42J cells after 1 min UV
followed by 24 h post incubation (Model 1) (p < 0.01 in man-
ual, p < 0.0001 in automated) (Fig. 2a and b) and 10 min UV
followed by 1 h post incubation (Model 2) (p < 0.05 in man-
ual, p < 0.0001 in automated) (Fig. 2c and d) compared to
untreated control cells. In model 1, cell viability detection with
automated cell counter after serial measurements was differed
from each other (p < 0.05) (Fig. 2b), whereas the only

Fig. 1 Detection of cell viability

of MDA-MBD231 cells using
trypan blue exclusion method
after UV treatment. Cells were
treated with UV for 1 min or
10 min followed by incubation for
24 h (Model 1, a-b) or 1 h (Model
2, c-d), respectively. Control cells
were untreated. Cells then were
incubated with trypan blue
followed by the analyses of dead
and live cell numbers by
conventional hemacytometer (a,
c), and automated cell counter (b-
d). Measurements were done in
series (at 0.5, 1, 3 and 5 min) for
cell counter (b-d). The total cell
numbers calculated are shown in
(e). Experiments were performed
as independent triplicates. Error
bars represent +/− standard error
of the means of replicates. **** =
p < 0.0001
1396 J Fluoresc (2018) 28:1393–1404

Fig. 2 Detection of cell viability

of AR42J cells using trypan blue
exclusion method after UV
treatment. Cells were treated with
UV for 1 min or 10 min followed
by incubation for 24 h (Model 1,
a-b) or 1 h (Model 2, c-d),
respectively. Control cells were
untreated. Cells then were
incubated with trypan blue
followed by the analyses of dead
and live cell numbers by
conventional hemacytometer (a,
c), and automated cell counter (b-
d). Measurements were done in
series for cell counter (at 0.5, 1, 3
and 5 min) (b-d). The total cell
numbers calculated are shown in
(e). Experiments were performed
as independent triplicates. Error
bars represent +/− standard error
of the means of replicates. *
p < 0.05, ** p < 0.01, **** =
p < 0.0001

difference was found between measurements at 0.5 and 5 min
in untreated cells of model 2 (p < 0.05), but not in UV-treated
cells (Fig. 2d). Total cell count was detected as similar along
with all replicates (Fig. 2e). The pattern of NIH-3T3 cells was
also similar to MDA-MB231 and AR42J cells (Fig. 3a-e). UV
induced cell death in both model, but it was only detectable by
automated cell counter in model 2 (Fig. 3d). As in MDA-
MB231, serial measurements provided similar results (Fig.
3b and d). In microalgae, the detection of cells stained with
or without trypan blue was not successful regardless the in-
strument used.
Cell Death Detection Experiments via Annexin V –
Propidium Iodide Staining in Mammalian Cells
After UV exposure of cells with definite time intervals, the
identification of cells based on their staining profiles with
Annexin V and PI were examined using flow cytometer,
fluorescence microscope and automated cell counter. For this
evaluation by flow cytometer, in forward-side scatter plot,
cells were grouped as stained with both; late apoptotic (UR
region), stained with Annexin-V alone; early apoptotic (LR
region), stained with PI alone (UL region), not stained with
any; live cells (LL region) (Fig. 4a-c). A similar analysis was
performed for the fluorescence microscope observation of
cells as cells were stained with Annexin-V (green) or PI
(red), and late apoptotic cells were evaluated as merging im-
ages captured using green and red filter (yellow) (Fig. 5).
In MDA-MB231 cells, the significance levels of increased
cell death after UV exposure were identified by the statistical
analyses of measurements from automated cell counter
(Fig. 6a-d), fluorescence microscope (Fig. 6b-e) and flow
cytometer (Fig. 6c-f) in both models. The analyses for auto-
mated cell counter showed that the proportions of early apo-
ptotic cells in both models were not found as significantly
increased in comparison with untreated cells (Fig. 6a-d). On
J Fluoresc (2018) 28:1393–1404 1397

Fig. 3 Detection of cell viability

of NIH3T3 cells using trypan blue
exclusion method after UV
treatment. Cells were treated with
UV for 1 min or 10 min followed
by incubation for 24 h (Model 1,
a-b) or 1 h (Model 2, c-d),
respectively. Control cells were
untreated. Cells then were
incubated with trypan blue
followed by the analyses of dead
and live cell numbers by
conventional hemacytometer (a,
c), and automated cell counter (b-
d). Measurements were done in
series for cell counter (at 0.5, 1, 3
and 5 min) (b-d). The total cell
numbers calculated are shown in
(e). Experiments were performed
as independent triplicates. Error
bars represent +/− standard error
of the means of replicates. *
p < 0.05, **** = p < 0.0001

the other hand, the analyses with fluorescence microscope
revealed a difference after UV in model 1 only as compared
to untreated control cells (p < 0.05) (Fig. 6e), whereas flow
cytometer provided the most significant and accurate compar-
ison in both models (p < 0.0001) (Fig. 6c-f). When automated
cell counter and fluorescence microscope utilized, the tenden-
cy of cells in terms of the types of induced cell death was
found quite similar in both models (Fig. 6a-b and Fig. 6d-e).
Nevertheless, interestingly in AR42J (Fig. 7c and f) and
NIH3T3 (Figs. 8C and 7F) cells, the flow cytometric measure-
ments did not display substantial decrease in cell death after
different UV treatments. Automated cell counter was identi-
fied as the most promising tool as it was able to quantify the
changes in cell death levels, especially in necrotic cell death
following the UV treatment of both AR42J (p < 0.0001 in
model 1, p < 0.05 in model 2) (Fig. 7a and d) and NIH3T3
(p < 0.0001 in model 1, p < 0.001 in model 2) (Fig. 8a and d)
cells. The analysis results from fluorescence microscope
revealed that the long time exposure to UV (model 2) signif-
icantly induced necrosis and late apoptosis in AR42J cells
(Fig. 7e) (p < 0.01 in necrotic cells, p ≤ 0.05 in late apoptotic
cells), but only necrosis in NIH3T3 cells (p ≤ 0.01) (Fig. 8e).
The significance levels detected for each cell line examined
and the related models by these three instruments are given in
Table 1.
Identification of Cell Death by Trypan Blue
and Annexin V-PI Method in Microalgae Cells
The microalgae cells stained with trypan blue after UV treat-
ment for 45 min followed by 24 h post-incubation could not
be detectable with microscope (Fig. 9a) and also by automated
cell counter. Yet the analysis by Annexin-PI counterstain
method, automated cell counter was successful to detect im-
portant changes in UV-induced cell death as necrotic
(p < 0.0001), late (p ≤ 0.001) and early apoptotic cells
1398 J Fluoresc (2018) 28:1393–1404

Fig. 4 Representative flow

cytometry dot plots of cells for
Annexin V-PI counterstain. Cell
death was induced in MDA-
MBD231 (upper), AR42J
(middle) and NIH3T3 (lower)
cells after UV treatment for 1 min
followed by 24 h incubation.
Untreated cells (left panels) and
UV-treated cells (right panels) are
shown. X and Y axes represent
green (FL1-H) and red (FL3-H)
fluorescence channels and show
Annexin V and PI staining
respectively. Cells were gated
according to untreated cells for
analyses. Each plot shows cells
positive for Annexin V only
(region LR), positive for PI only
(region UL), positive for both
(region UR), and negative for
both (region LL). Regions LR,
UL, UR and LL represent early
apoptotic, necrotic, late apoptotic
and live cells in the population

(p < 0.01) (Fig. 9b). Bee venom also induced cell death de-
tected by flow cytometry (Fig. 10), but not by automated cell
counter (data not shown). Gated cells (Fig. 10a) were ana-
lyzed, and only the proportion of early apoptotic cells in-
creased after bee venom (500 μg/ml) as compared to untreated
cells (p < 0.01) (Fig. 10b-d). But, interestingly untreated cells
had more late apoptotic cells than each dose (p < 0.01 and
p < 0.001 for 300 and 500 μg/ml, respectively) (Fig. 10b).
There could be couple of reasons: 1) microalgae cells experi-
enced were at late log phase that may induce cell death in
untreated cells, 2) microalgae cells have a thick cell wall
which may slow down the diffusion speed of bee venom to-
wards intracellular environment, therefore bee venom may not
induce cellular pathways involved in late apoptosis process, 3)
microalgae cells may have an extra and undefined resistance
or tolerance system against bee venom-induced late apoptotic
pathway. Similarly, microalgae cells have resistance against
heavy metal accumulation within the cells.
Discussion
The detection of cell death is an important leverage to reveal
the potential cytotoxic effects of novel drug candidates, natu-
ral products or synthetic compounds on therapeutic ap-
proaches of diseases. Cytotoxicity assays mainly include the
methods based on the detection of metabolic and enzymatic
activities, membrane integrity and changes on mitochondria
J Fluoresc (2018) 28:1393–1404
Fig. 5 Representative microscopy images of MDA-MBD231 cells for
Annexin V-PI counterstain. MDA-MBD231 cells were treated with UV
for 1 min, and analyzed for Annexin V-PI after 24 h. Cells were counted
for co-staining of Annexin V (green filter) and PI (red filter). Untreated
Fig. 6 Detection of cell death profiles of MDA-MBD231 cells using
Annexin V-PI counterstain method after UV treatment. Cells were
treated UV for 1 min or 10 min followed by incubation for 24 h (Model
1, a-b-c) or 1 h (Model 2, d-e-f), respectively. Control cells were
untreated. Cells then were proceed with Annexin V-PI staining protocol
followed by the analyses of cells by automated cell counter (a, d),
1399
(upper panel) and treated (lower panel) cells are shown for brightfield,
Annexin V, PI and merged of Annexin V-PI. Cells stained for Annexin V
are green, cells stained for PI are red, and cells stained for both are merged
and shown as yellow. Images were captured using an ×40 objective
fluorescence microscope (b, e) and flow cytometry (c-f). Cell
percentages are calculated according to LR (Annexin V), UL (PI) and
UR (both) (see Fig. 4) regions and arcsine-transformed followed by
UNIANOVA test. Experiments were performed as independent
triplicates. Error bars represent +/− standard error of the means of
replicates. * p < 0.05, ** p < 0.01, *** p < 0.001, **** = p < 0.0001
1400
Fig. 7 Detection of cell death profiles of AR42J cells using Annexin V-PI
counterstain method after UV treatment. Cells were treated UV for 1 min
or 10 min followed by incubation for 24 h (Model 1, a-b-c) or 1 h (Model
2, d-e-f), respectively. Control cells were untreated. Cells then were
proceed with Annexin V-PI staining protocol followed by the analyses
of cells by automated cell counter (a, d), fluorescence microscope (b, e)
and DNA [1]. Counterstaining of Annexin V and PI based on
the disruption of membrane integrity has been one of the most
common methods used to detect cell death [2–7]. This method
is not recommended for the use of adherent cells by the man-
ufacturer as trypsinization of cells can disturb the intact cell
membrane of live cells leading to false positive events; how-
ever there is a number of studies that successfully used the
counterstain for adherent cells [3, 8–10]. The key point ap-
pears to have a minimum incubation time with trypsin as
possible before proceeding with the protocol.
Automated cell counter, fluorescence microscope and flow
cytometer work similarly for the fluorescence-based detection
of cells through different fluorescence filters allowing the si-
multaneous detection of multiple parameters (such as Annexin
V at green channel and PI at red channel). Flow cytometer and
fluorescence microscopy can provide quantitative data as be-
ing eligible to calculate staining amounts represented by mean
fluorescence intensities. However, cell counter does not
J Fluoresc (2018) 28:1393–1404
and flow cytometry (c-f). Cell percentages are calculated according to LR
(Annexin V), UL (PI) and UR (both) (see Fig. 4) regions and arcsine-
transformed followed by UNIANOVA test. Experiments were performed
as independent triplicates. Error bars represent +/− standard error of the
means of replicates. * p < 0.05, ** p < 0.01, *** p < 0.001, **** =
p < 0.0001
present the amount of staining but the existence of staining.
These instruments theoretically should provide a similar pat-
tern of analyses for the presence of Annexin V and/or PI stain-
ing of cells in a population. Differentially, fluorescence micro-
scope does not directly provide an automatic analysis of cells,
but require an extra image-analysis tool for merging of two
images captured using various fluorescence channels and
manually counting cells based on staining profiles with or
without both Annexin V and PI. Automated cell counter and
flow cytometer are eligible to detect cells using two fluores-
cence channels at the same time. But in this study, cell counter
appeared as more efficient in order to identify the level of
difference in cell viability between UV-treated and untreated
control cells rather than manual counting. In addition, both
cell counter and microscopy measurements provided
cooperated results with each other, but flow cytometer did
not. Different gating strategies on FSC-SSC dot plots have
been also implemented for the analyses, but still a significant
J Fluoresc (2018) 28:1393–1404 1401

Fig. 8 Detection of cell death profiles of NIH3T3 cells using Annexin V-
PI counterstain method after UV treatment. Cells were treated UV for
1 min or 10 min followed by incubation for 24 h (Model 1, a-b-c) or 1 h
(Model 2, d-e-f), respectively. Control cells were untreated. Cells then
were proceed with Annexin V-PI staining protocol followed by the
analyses of cells by automated cell counter (a, d), fluorescence
microscope (b, e) and flow cytometry (c-f). Cell percentages are
calculated according to LR (Annexin V), UL (PI) and UR (both) (see
Fig. 4) regions and arcsine-transformed followed by UNIANOVA test.
Experiments were performed as independent triplicates. Error bars
represent +/− standard error of the means of replicates. * p < 0.05, **
p < 0.01, *** p < 0.001, **** = p < 0.0001

difference in the comparison of percentages of apoptotic/ study, for only MDA-MB231 cells, cytometer revealed signif-
necrotic cells between untreated and treated cells were not icant increases in early apoptotic, late apoptotic and necrotic
identified. The accuracy of Annexin V-PI test detected by flow cells through both UV treatment models. A significant differ-
cytometer highly depends on the cell type examined. In our ence in the cell death profiles of MDA-MB231 cells compared

Table 1 Significance levels (p values) detected between untreated and treated cells (Annexin V-PI staining)

Cell Model Automated Cell Counter Fluorescence Microscope Flow cytometer

Early Late Necrotic Early Late Necrotic Early Late Necrotic

apoptotic apoptotic apoptotic apoptotic apoptotic apoptotic

MDA-MB231 1 (UV) NS† 0.008 0.000 NS† 0.014 0.027 0.000 0.000 0.000
2 (UV) NS† 0.001 0.000 0.021 0.014 0.002 0.000 0.000 0.000
AR42J 1 (UV) 0.006 0.000 0.000 NS† NS† NS† NS† NS† NS†
2 (UV) NS† NS† 0.038 NS† 0.05 0.004 NS† NS† NS†
NIH3T3 1 (UV) NS† 0.000 0.000 NS† NS† 0.031 NS† NS† NS†
2 (UV) NS† 0.004 0.001 NS† NS† 0.01 NS† 0.05 NS†
Chlorella vulgaris UV 0.006 0.001 0.000 ND* ND* ND* NS† NS† NS†
Bee Venom NS† NS† NS† ND* ND* ND* NS† 0.01 NS†

ND*: not detectable

NS†: not significant
1402
Fig. 9 Detection of cell death by trypan blue method and Annexin V-PI
method in microalgae cells after UV treatment. Chlorella vulgaris cells
were stained with trypan blue after UV treatment for 45 min followed by
24 h post-incubation. But trypan blue stained cells was not be detectable
with microscope (a). Automated cell counter was however successful to
detect important changes in UV-induced cell death as necrotic, late and
early apoptotic cells using Annexin-PI counterstain method (b).
Experiments were performed as independent triplicates. Error bars
represent +/− standard error of the means of replicates. ** p < 0.01, ***
p < 0.001, **** = p < 0.0001
to untreated cells has also been reported previously [3].
Nevertheless, in three groups of cells, unlike cytometer, both
counter and fluorescence microscope provided more signifi-
cant comparisons of late apoptotic/necrotic cells between un-
treated and UV-treated cells. The identification of the differ-
ence of early apoptotic cells for both treated and untreated
cells by either counter or microscope was however not in a
significant trend. NIH3T3 cells were previously treated with
acrylamide followed by the flow cytometric detection of
Annexin V staining profiles of cells, and found to have signif-
icant levels of apoptosis compared to untreated counterparts
[11]. AR42J cells were also treated with deoxycholic acid [12]
which is a bile acid and reported to have a cytotoxic activity
[13], and with Exentin-4 which is a peptide stimulating insulin
secretion [14]. Both agents were able to induce cell death in a
dose and time-dependent manner. In our design, UV could not
J Fluoresc (2018) 28:1393–1404
be adequate to induce cell death, the use of genotoxic agents
such as Doxorubicin and Cisplatin might be helpful to induce
a significant amount of cell death in AR42J and NIH3T3 cells.
Therefore, it is suggested that the choice of agent strongly
depends on the cell type.
Flow cytometry is the most sensitive instruments compared
to counters and microscopy as it detects different cell popula-
tions concisely and also discriminates dirt and cell debris
along with cell size and granularity (in FSC-SSC dot plots),
and that measures mean fluorescence intensity of a cell popu-
lation without bias. However, in some cells, the detected
amount of fluorescence may include self cellular autofluores-
cence [15] as some compounds within the live cells have been
known to emit itself in the fluorescence channels depending
on the cell conditions [16]. The majority of auto-fluorescent
components are excited at 488 nm and detected by green
channel (FL-1) [17], and some organelles, including mito-
chondria and lysosomes, may lead to background staining
[18]. There are some exceptional methods developed to ex-
clude cellular autofluorescence while analyzing data by flow
cytometry [15, 16] and by microscope [16, 18, 19]. In the
protocol used, Annexin V is tagged with fluorescein isocya-
nate (FITC) and detected by FL-1. The possible interference
of those components may lead to current insignificant results
obtained by flow cytometry. In the current work, autofluores-
cence was automatically excluded since gating strategy was
defined according to FSC-SSC plots of untreated cells.
The quantification tests also revealed that both microscope
and cell counter were failed to differentiate the trypan blue-
stained microalgae cells after bee venom and UV treatments.
One possible explanation for this is that the operated micro-
scope resolution may not be sufficient to visualize microalgae
(cell diameter ~4-10 μm). The maximum magnification was
obtained using a 40x objective that was probably not enough.
Another possibility could be that the instruments used were
not suitable for the detection of trypan blue-stained for small
size microalgae cells. As counter revealed all the cells from
treated and untreated populations as dead, it may detect cells
as false positive for trypan blue staining since algae’s own
greenish color may be interfered with trypan blue. On the
other hand, in microalgae cells, inducement of cell death
was not as accessible as in mammalian cells. Microalgae cells
contain cell wall structures as well as the membrane so that
cells have an additional barrier against external agents
[20–22], and therefore been more resistant than the cells with
membrane only. Bee venom was found to be significant in late
apoptotic cells by only flow cytometer, but not by counter.
Flow cytometer has been commonly used to detect the lipid
content of various microalgae through Nile Red staining
[23–25]. The advantage of flow cytometry for gating makes
it more applicable for small cells in size to be detected.
Microalgae cells have been commonly used in biotechnology
J Fluoresc (2018) 28:1393–1404 1403

Fig. 10 Flow cytometric

detection of cell death by trypan
blue method and Annexin V-PI
method in microalgae cells after
treatment with bee venom.
Chlorella vulgaris cells were
treated with bee venom for
6 days. After incubation, gated
cells (a) were analyzed, and only
the proportion of early apoptotic
cells increased after bee venom
(500 μg/ml) as compared to
untreated cells (b-d). Untreated
cells had more number of late
apoptotic cells than each dose (B).
Experiments were performed as
independent triplicates. Error bars
represent +/− standard error of the
means of replicates. * p < 0.05, **
p < 0.01

research as they can grow easy and produce a large number of
metabolites. This suggests that the accurate detection(s) for
wellness of microalgae cells is crucial to proceed biotechno-
logical production of molecules.
It should be also noted that combined approaches can in-
crease the sensitivity of these instruments. For instance, imag-
ing flow cytometry (IFC) is designed as to include advantages
of both flow cytometry and fluorescent microscopy as it al-
lows fluorescence profiles of cells as well as morphological
changes in cells after various environmental changes [26].
Further evaluation with these kinds of hybrid technologies
can reveal more detailed and accurate outcomes of patterns
of apoptotic and necrotic ones in a cell population.
recommended that researchers should be aware of considering
the type of cells they use to discuss the outcomes of Annexin
V-PI staining. Fluorescence microscope and automated cell
counter examined here appear not suitable for the analyses
of microalgae cells after Annexin V-PI, but flow cytometer.
Acknowledgements A part of this study was supported by the equipment
grant (Project ID: FAY-2016-5755) of Karadeniz Technical University,
Scientific Research Coordination Unit, to Dr. Uğur Uzuner. Authors
would like to thank to Prof Sevgi Kolayli (Karadeniz Technical
University, Department of Biochemistry) for the kindly gift of Annexin/
PI detection kit, and to Prof Figen Celep Eyupoglu (Karadeniz Technical
University, Department of Medical Biology) for providing a batch of
MDA-MB231 cells.

Author Contribution EK performed all experiments with mammalian

cells, RC performed microalgae experiments. SCU designed the study
Conclusions and wrote the manuscript. UU edited the manuscript.

This study suggests automated cell counters and fluorescence Compliance with Ethical Standards
microscopes are more suitable instruments for the detection of
Annexin V-PI staining than flow cytometer. It is Competing Interest Authors declare no competing interest exists.
1404
References
1. Alfraidi A, Fagundes TC (2018) Cell-based cytotoxicity methods.
JAN 1(1):3–12
2. Huang X, Chen L, Liu WJ, Qiao Q, Wu K, Wen J, Huang CH, Tang
R, Zhang XZ (2015) Involvement of oxidative stress and cytoskel-
etal disruption in microcystin-induced apoptosis in CIK cells.
Aquat Toxicol 165:41–50
3. Xiao K, Li YP, Wang C, Ahmad S, Vu M, Kuma K, Cheng YQ,
Lam KS (2015) Disulfide cross-linked micelles of novel HDAC
inhibitor thailandepsin a for the treatment of breast cancer.
Biomaterials 67:183–193
4. Shan X, Tian LL, Zhang YM, Wang XQ, Yan Q, Liu JW (2015)
Ginsenoside Rg3 suppresses FUT4 expression through inhibiting
NF-kappa B/p65 signaling pathway to promote melanoma cell
death. Int J Oncol 47(2):701–709
5. Deng ZT, Xiao Y, Pan M, Li F, Duan WL, Meng L, Liu X, Yan F,
Zheng HR (2016) Hyperthermia-triggered drug delivery from
iRGD-modified temperature-sensitive liposomes enhances the
anti-tumor efficacy using high intensity focused ultrasound. J
Control Release 243:333–341
6. Hwang JH, Park H, Choi DW, Nam KT, Lim KM (2018)
Investigation of dermal toxicity of ionic liquids in monolayer-
cultured skin cells and 3D reconstructed human skin models.
Toxicol in Vitro 46:194–202
7. Yao RS, Han DY, Sun XY, Xie Y, Wu QY, Fu CL, Yao Y, Li HJ, Li
ZY, Xu KL (2018) Scriptaid inhibits cell survival, cell cycle, and
promotes apoptosis in multiple myeloma via epigenetic regulation
of p21. Exp Hematol 60:63–72
8. Svystonyuk DA, Ngu JMC, Mewhort HEM, Lipon BD, Teng GT,
Guzzardi DG, Malik G, Belke DD, Fedak PWM (2015) Fibroblast
growth factor-2 regulates human cardiac myofibroblast-mediated
extracellular matrix remodeling. J Transl Med 13:147
9. Sevastre B, Sarpataki O, Stan RL, Taulescu M, Sevastre-Berghian
AC, Olah NK, Furtuna F, Hanganu D, Hangan AC, Cenariu M,
Baldea I (2017) Anticancer activity of euonymus Europaeus fruits
extract on human melanoma cells. Farmacia 65(1):56–62
10. Turkington RC, Longley DB, Allen WL, Stevenson L, McLaughlin
K, Dunne PD, Blayney JK, Salto-Tellez M, Van Schaeybroeck S,
Johnston PG (2014) Fibroblast growth factor receptor 4 (FGFR4): a
targetable regulator of drug resistance in colorectal cancer. Cell
Death Dis 5:e1046
11. Sahinturk V, Kacar S, Vejselova D, Kutlu HM (2018) Acrylamide
exerts its cytotoxicity in NIH/3T3 fibroblast cells by apoptosis.
Toxicol Ind Health 34(7):481–489
12. Zhang G, Zhang J, Shang D, Qi B, Chen H (2015) Deoxycholic
acid inhibited proliferation and induced apoptosis and necrosis by
regulating the activity of transcription factors in rat pancreatic aci-
nar cell line AR42J. In Vitro Cell Dev Biol Anim 51(8):851–856
J Fluoresc (2018) 28:1393–1404
13. Sousa T, Castro RE, Pinto SN, Coutinho A, Lucas SD,
Moreira R, Rodrigues CMP, Prieto M, Fernandes F (2015)
Deoxycholic acid modulates cell death signaling through
changes in mitochondrial membrane properties. J Lipid Res
56(11):2158–2171
14. Li Z, Zhu S, Huang L, Shang M, Yu C, Zhu H, Han D, Huang H,
Yu X, Li X (2018) Exendin-4 impairs the autophagic flux to
induce apoptosis in pancreatic acinar AR42J cells by down-
regulating LAMP-2. Biochem Biophys Res Commun 496(2):
294–301
15. Celik-Uzuner S, Peters L, O'Neill C (2016) Quenching of cellular
autofluorescence is necessary for specific detection of DNA meth-
ylation by flow cytometry compared to microscopy-based analysis.
FEBS J 283:249–250
16. Aubin JE (1979) Autofluorescence of viable cultured mammalian-
cells. J Histochem Cytochem 27(1):36–43
17. Billinton N, Knight AW (2001) Seeing the wood through the trees:
a review of techniques for distinguishing green fluorescent protein
from endogenous autofluorescence. Anal Biochem 291(2):175–197
18. Andersson H, Baechi T, Hoechl M, Richter C (1998)
Autofluorescence of living cells. J Microsc (Oxford) 191:1–7
19. Nawa Y, Inami W, Miyake A, Ono A, Kawata Y, Lin S, Terakawa S
(2014) Dynamic autofluorescence imaging of intracellular compo-
nents inside living cells using direct electron beam excitation.
Biomed Opt Express 5(2):378–386
20. Gerken HG, Donohoe B, Knoshaug EP (2013) Enzymatic cell wall
degradation of Chlorella vulgaris and other microalgae for biofuels
production. Planta 237(1):239–253
21. Yap BHJ, Crawford SA, Dagastine RR, Scales PJ, Martin GJO
(2016) Nitrogen deprivation of microalgae: effect on cell size, cell
wall thickness, cell strength, and resistance to mechanical disrup-
tion. J Ind Microbiol Biotechnol 43(12):1671–1680
22. Safi C, Ursu AV, Laroche C, Zebiba B, Merah O, Pontalier PY,
Vaca-Garcia C (2014) Aqueous extraction of proteins from
microalgae: effect of different cell disruption methods. Algal Res
3:61–65
23. da Silva TL, Reis A, Medeiros R, Oliveira AC, Gouveia L (2009)
Oil production towards biofuel from autotrophic microalgae
Semicontinuous cultivations Monitorized by flow cytometry.
Appl Biochem Biotechnol 159(2):568–578
24. Montero MF, Aristizabal M, Reina GG (2011) Isolation of high-
lipid content strains of the marine microalga Tetraselmis suecica for
biodiesel production by flow cytometry and single-cell sorting. J
Appl Phycol 23(6):1053–1057
25. Pereira H, Schulze PSC, Schuler LM, Santos T, Barreira L, Varela J
(2018) Fluorescence activated cell-sorting principles and applica-
tions in microalgal biotechnology. Algal Res 30:113–120
26. Barteneva NS, Fasler-Kan E, Vorobjev IA (2012) Imaging flow
cytometry: coping with heterogeneity in biological systems. J
Histochem Cytochem 60(10):723–733