Journal of Immunologwal Methods, 139 (1991) 271-279
271
© 1991 ElsexaerScience Pubhshers B V 0022-1759/91/$03 50 ADONIS 0022175991001841 JIM05939
A rapid and simple method for measuring thymocyte apoptosis by propidium iodide staining and flow cytometry I. N i c o l e t t i 1, G . M i g l i o r a t i 2, M.C. Pagliacci 1, F. G r i g n a n i 1 a n d C. R i c c a r d i 2 1 Istttuto dt Chmca Medwa 1, and ' lstttuto dl Farmacologta, Perugla Umverstty School of Medicine, 06100 Perugla, Italy
(Recewed 19 October 1990, rewsed received 14 January 1991, accepted 19 February 1991)
Corticosteroids, calcium ionophores and anti-CD3 monoclonal antibodies kill mouse thymocytes incubated in vitro. Cell death is preceded by extensive D N A fragmentation into ohgonucleosomal subumts. Tins type of cell death (apoptosis), which physiologically occurs m the intrathymic process of immune cell selection, is usually evaluated by either electrophoretic or colorimetric methods which measure D N A fragmentation in the nuclear extracts. These techniques are unable to determine the percentage of apoptotic nuclei or recognize the apoptotic cells in a heterogeneous cell population. We have developed a flow cytometric method for measuring the percentage of apoptotic nuclei after propidium iodide staining in hypotonic buffer and have compared it with the classical colonmetric and electrophoretic techniques using dexamethasone (DEX)-treated mouse thymocytes. Apoptotic nuclei appeared as a broad hypodiploxd D N A peak which was easily discriminable from the narrow peak of thymocytes with normal (diploid) D N A content in the red fluorescence channels. When the DEX-induced apoptosis was inhibited by either low-temperature ( 4 ° C ) incubation or cyclohexlmide treatment, no hypodiploid D N A peak appeared. Similarly, thymocyte death induced by s o d m m azide, a substance with cell-kilhng activity through non-apoptotic mechanisms, &d not result in any variation in the normal D N A peak. The flow cytometrlc data showed an excellent correlatton with the results obtained with both electrophoret~c and colorimetric methods. This new rapid, simple and reproducible method should prove useful for assessing apoptos~s of specific cell populations in heterogeneous tissues such as bone marrow, thymus and lymph nodes. Key words Apoptosls, Thymocyte, DNA fragmentation, Flow cytometry, Dexamethasone
Introduction Apoptosis is a c o m m o n form of cell death xn eukariotes. The process is operative during embryogenesis, in tumor regression and in the
Correspondence to I. Ntcoletti, IsUtuto dt Chmca Med~ca1, Umversit/t dl Perugia, Pohchmco Monteluce, 06100 Perugaa, Italy
elirmnation of self-reactive T lymphocytes m the thymus (Wyllie et al., 1980). Apoptotic cell death can be induced by glucocorticoid treatment (Cohen and Duke, 1984), antibodies to C D 3 / T cell receptor complex (Shi et al., 1989; Smith et al., 1989), exposure to Ca 2+ lonophores (Kazaki et al., 1989; McConkey et al., 1989) or "y-irradiauon of mouse thymocytes (Selllns and Cohen, 1987), anti-APO monoclonal antibody treatment of tumor cells (Trauth et al., 1989) and growth factor deprivation
272
of both hematopoietlc and lymphoid cells in vitro (Williams et al., 1990). The death of cells undergoing apoptosls is preceded by chromatin cleavage at the hnker regions between nucleosomes by specific endonucleases, which results m extenswe fragmentation of the D N A into oligonucleosomal subunits (Wyllie et al., 1980). The D N A fragments can be demonstrated by agarose gel electophoresis and the percentage of D N A fragmentation is usually measured by colorimetric methods after separating intact from fragmented D N A by ultracentrifugation (Selhns and Cohen, 1987). The main flaw in these methods, since the D N A fragmentation assays are performed with the whole nuclear extract, is their inabdlty to either directly evaluate the percentage of fragmented nuclei or recognize the apoptotic cells in a heterogeneous cell population. We have developed a rapid, simple and reproducible method for measuring the percentage of apoptotic cells by flow-cytometric analysis of isolated nuclei stained with propidium iodide (PI) in hypotonic buffer and, using mouse thymocytes, demonstrated that the flow cytometric data correlate excellently with the classical D N A fragmentation assays.
Materials and methods
Cell suspenstons Thymocytes were obtained from 2-4-week-old C 3 H / H e N mice purchased from Charles River (Corsico, Milan). The animals were killed by cervical dislocation and the thymus teased in R P M I 1660 medium. The cell suspension was washed, filtered and adjusted to a concentration of 1.5 × 10 6 cell/ml in R P M I 1660 m e d m m supplemented wlth 5% FCS and 10 m M Hepes buffer. Aliquots of 2 ml thymocytes were incubated at either 37°C or 4°C with DEX (at concentrations from 10 - H to 10 - 6 M). At pre-established times, the cells were centrifuged at 200 × g for 10 nun, washed, and processed (see below). DNA labehng techmque and flow cytometrtc analysts Two different methods were evaluated for D N A labeling.
(1) the 200 × g centnfuged cell pellet was fixed in 2 ml cold 70% ethanol at 4°C for 60 min. The cells were then centrifuged, washed m 1 ml PBS and resuspended m 0.5 ml PBS. To a 0.5 ml cell sample 0.5 ml RNAse (Type I-A, Sigma, St. Lores, MO, U.S.A., 1 m g / m l in PBS) was added, followed after gentle mixing by 1 ml propldlum iodide (PI, Sigma, 100 # g / m l m PBS) solution. The mixed cells were incubated in the dark at room temperature for 15 man and kept at 4°C m the dark until measured. (2) The second method was essentially that described by Fried et al. (1978). The 200 × g centrifuged cell pellet was gently resuspended in 1.5 ml hypotonlc fluorochrome solution (PI 50 /~g/ml in 0.1% sodium citrate plus 0.1% Triton X-100, Sigma), in 12 × 75 polypropylene tubes (Becton Dickinson, Lincoln Park, N J, U.S.A.). The tubes were placed at 4°C in the dark overmght before the flow-cytometric analysis. Since this procedure gave lower coefficients of variation of D N A peaks and better discrimination of subdiploid and diploid cells than the classical method of PI staining after ethanol fixation and RNA-ase treatment (Fig. 2), it was used in the whole study. The PI fluorescence of individual nuclei was measured using a FACScan flow cytometer (Becton and Dickinson, Mountain View, CA, U.S.A.). The nuclei traversed the light beam of a 488 nm Argon laser. A 560 nm dichrolc nurror ( D M 570) and a 600 nm band pass filter (bandwidth 35 nm) were used for collecting the red fluorescence due to PI staining of D N A and the data were registered on a logarithmic scale. The forward scatter (FSC) and side scatter (SSC) of particles were simultaneously measured. All the data were recorded in a Hewlett Packard (HP 9000, model 310) computer using specific FACScan research software (Becton Dickinson). Cell debris were excluded from analysis by appropriately raising the FSC threshold. The correct threshold value was selected experimentally. A suspension of 10 /~m diameter fluorescent polystyrene microspheres ( D N A check, Coulter Corporation, Hialeah, FL, U.S.A.) was run and the FSC peak was adjusted around channel 600 of a 1024 channel linear scale. The red fluorescence peak was set at channel 250 in the logarithmic mode. The FSC threshold was then progressively increased until the unwanted
273 signals in the 1-200 channels of the red fluorescence were < 1%. The resulting FSC threshold was utihzed for the analysis of thymocyte nuclei stained with PI hypotonic solution. The residual cell debris ( < 3%) had a very low D N A red fluorescence emission (channels 1-5) and low side scatter (SSC) signal. Apoptotic cell nuclei, which were easily distinguishable from debris by the high SSC value due to the condensation of nuclear chromatm, emitted fluorescence in channels 10250. As there was no overlap between apoptotic nucle~ and debris, the small percentage of residual low-fluorescent detritus was eliminated by gating at channel 10 of the red PI fluorescence scale. The flow rate was set at about 200 nuclei × s-1 and at least 10 4 cells of each sample were analysed. All measurements were done under the same instrument settings.
/~g/ml) was added to each sample and D N A was extracted twice first with phenol plus chloroform and then with chloroform, and recovered by centrifugation after overnight precipitation at - 2 0 ° C in 2 vols. of ethanol in the presence of 0.3 M sodium acetate. Pellets were air dried, dissolved m 10 m M Tris-HC1, 1 m M E D T A ( p H 8) at 4°C and the D N A concentrauon determined by absorbance at 260 nm with a Perkln Elmer spectrofluorimeter. 20 ~g D N A were mixed with 25% Ficoll 400 ( P h a r m a o a , Uppsala, Sweden), 0.25% bromophenol blue and 0.25% xylene cyanol (loading buffer), before loading into the wells of a 1% agarose gel. Electophoresis was carried out in TBE buffer (2 m M EDTA, 89 m M boric acid, 89 m M Tris, p H 8.4) and D N A visualized by ethidium bromide staining.
Colorlmetrlc DN.4-fragmentatton assay and DNA electrophoresis
Results
DEX-induced apoptosis of mouse thymocytes was measured in selected experiments by both the above described flow cytometric method and a classical colorimetric D N A fragmentation assay (Sellins and Cohen, 1987). Briefly, cell suspensions were divided in two aliquots and the resulting cell pellets were either resuspended in PI fluorochrome solution for flow cytometric analysis or dissolved in hypotonic lysing buffer (Tris 10 mM, E D T A 1 mM, Triton X-100 0.2%). Intact and fragmented D N A were separated by centrifugation at 13,000 × g for 10 min and measured by the diphenylamine reaction (Burton, 1956). Percent fragmentation was calculated as the ratio of D N A in the 13,000 x g supernatants to the total D N A (supernatants plus pellets). For D N A electrophoresis, 4 )< 106 mouse thymocytes were incubated overnight at 37°C in medium alone, in 10 - 7 M DEX-supplemented medium, in medium containing 10 - 7 M D E X plus the apoptosis inhibitor cycloheximide (50/~g/ml), or m medium plus 0.1% sodium azide. The cells were washed, centrifuged at 200 × g for 10 min and dissolved in hypotonic lysing buffer (100 m M NaC1, 10 m M Tris, 1 m M EDTA, 1% SDS, 200 /xg/ml protemase K, p H 7.5). Samples were extracted once with phenol plus chloroform ( 1 : 1 , v / v ) and then with chloroform. RNAse A (100
The flow cytometrw pattern of dexamethasone-mduced thymocyte apoptosts Fig. 1 shows the appearance by fluorescence microscopy of ethanol (70%) fixed, PI-stained thymocyte nuclei after 24 h incubation in either medium alone or medium plus 10 - 7 M D E X at 37°C. The majority of steroid-treated thymocytes displayed a clear reduction m nuclear diameter and a complete derangement in chromatin structure. Small areas with highly condensed chromatin were visible. Flow cytometric analysis of ethanol fixed, PIstained cells from parallel cultures revealed a typical diploid D N A peak in R P M I 1660 incubated thymocytes and a minor one in the hypodiploid range; DEX-treated cells manifested an impressive reduction in the percentage of nuclei with diploid D N A content, while the hypodiploid D N A peak was greatly enhanced (Fig. 2, left panels). A better separation between the two peaks was obtained when thymocytes were stained in hypotonic PI solution according to Fried et al. (1978) and the logarithmic mode of PI-fluorescence acquisition utilized during FACScan analysis (Fig. 2, right panels). To demonstrate that the subdiploid peak was really due to apoptotic cells, mouse thymocytes were analyzed in experimental conditions known
274 to m o d i f y t h e a p o p t o t t c p r o c e s s . S i n c e D E X - i n d u c e d D N A f r a g m e n t a t i o n is a t e m p e r a t u r e d e p e n d e n t p h e n o m e n o n ( O j e d a et al., 1990), t h e e f f e c t o f i n c u b a t i o n at b o t h 3 7 ° C a n d 4 ° C o n
DEX action was evaluated. When mouse thymocytes w e r e m a i n t a i n e d for 24 h at 4 ° C t h e D N A h i s t o g r a m s o f cells i n c u b a t e d e i t h e r w i t h o r w i t h o u t 10 - 7 M D E X w e r e a l m o s t s u p e r l m p o s a b l e a n d
O Fig. 1. Fluorescence rmcroscopy (1250 × ) appearance of ethanol-fixed, PI-staaned thymocyte nuclei after 24 h incubation m medium alone (A) or in medium plus 10 -7 M DEX (B) The majority of DEX-treated nuclei show a marked reduction m their diameter and an exadent condensation of chromatin
275
over 90% of nuclei showed a diploid D N A peak (F~gs. 3A and 3B). In parallel cultures incubated at 37°C, the same DEX concentration led to a marked reduction of diploid peak and the appearance of a large number of cells with sub&ploid D N A content (Figs. 3C and 3D). The active process of D N A fragmentation induced by glucocorticoids in thymocytes is known to be lnh~blted by R N A and protein synthesis mhibitors such as D-actmomycin and cyclohex~mide (Cohen and Duke, 1984). The addition of cyclohex~m~de to the incubation m e & u m fully pre-
2 H
vented the appearance of the DEX-lnduced hypodiploid D N A peak and the percentage of diploid nuclei was mrmlar m thymocytes incubated for 24 h at 37°C in medium alone or in medium containmg DEX plus cycloheximide (Fig. 3E). The possibility that the appearance of the sub&plo~d D N A peak was simply related to the presence of fragments of dead cells, was ruled out by evaluating the effect of sodium azade, a substance which induces cell death through a non-apoptotic mechamsm (Wyllie et al., 1984), on mouse thymocytes. Incubation of thymocytes with 0.1% sodium
MEDIUM
07%
1 i ~IIFN I
n I Iil~l~]
lll~.n
I
n i IHIII
I~ ~
I
24 H MEDIUM
178~ I
-I
1
't 1
4
4
t9 Q
, , llrH, I , * ,,NI~ 1 t~ I t~ 2
I ,q,,ll I i i rl,iH I IB a t~ a
1
-I
24 H DEX 10-TM
-1 1 -I
! -1
,,,,
, , , , I , , , , i , , , , [ , , r ,
I 23%
III , ,',~i~, I
" 7 - ~ , . q ~ , , ,
/
.......
t~ ~
DNA CONTENT (PI FLUORESCENCE) Fig 2 DNA fluorescence flow cytometnc profiles of Pl-stamed thymocytes after 2 h and 24 h mcubauon m medium alone or in medium plus 10-7 M DEX. Left panels: D N A tustograms (hnear scale) after 70% ethanol hxatlon, RNAse treatment and PI staining Rtght panels DNA histograms of parallel cultures dtrectly stained with hypotomc PI-fluorochrome solution. This procedure and the data acqulsmon m loganthmac mode perrmt a better resoluuon of the two D N A peaks
276
Compartson of flow cytometmc wtth colonmetrtc DNA fragmentatton assays
azide markedly enhanced the number of dead cells at 24 h ( > 90% by trypan blue exclusion) but did not cause modify the diploid D N A peak of PIstained nuclei (Fig. 3F). Similarly, rapid complement lysts caused a > 95% thymocyte death (trypan blue exclusion) but did not induce any hypodiploid peak in the D N A histogram (data not shown). The absence or presence of sub&ploid D N A peaks paralleled the absence or presence of a ladder of ohgonucleosome length D N A degradation in agarose gel electophoresis when thymocytes were incubated in medium alone, or medium w i t h 10 - 7 M DEX, DEX plus cycloheximide or sodium azide (Fig. 4).
~o+:
~ ....
~@@ , , P~@@ ,
The reliability of the method we propose to measure apoptosis was compared wtth the clasmcal colorimetric D N A fragmentation assay descrtbed by Sellins and Cohen (1987). Fig. 5 gives the percentage of DEX-mduced D N A fragmentation in mouse thymocytes, as evaluated by the colorimetrtc assay, m a dose response experiment. There was a clear dose related increase in the D N A fragmentation of DEX-treated cells, while cycloheximlde strongly inhibited both spontaneous and DEX-mduced chromatm breakdown.
F@@ , ~ ~@@ , , I t eBee'~
MEDIUM 4°C
t _]' DEX 4°C
A
97%1'
IQ~4
B
96%
t
t'
I
1
L
4
-1 -I
1 IBI
%e ¢l
"+t=
Ie . . . .
r~)
le+ , , P@@
lm 4
| e ~1
, , IS~a0 ,
lelialP i I F e e
I
MEDIUM 37°C
C
72%
.
toQ
' ''""'1
Im. . . .
ta l
........
I IB •
'""q
IB •
' '''"q
tg 4
,rap+,, 1,0•, ,iFp~p, impGpi r team
_3 ' DEX 37°C
'I
O
I z
28% , , ,, ,ir~.; z
lea
te l
U,h. I 1o 2
'lllill'l I1111111 le ~ IB 4
le B
|e I
IBa
tm~
|B 4
I
DEX+C'YCLOHEXIH. ~4% 'I
16% , , ,.,,,~ tee
~ , ,+,h, i let
Ie2
¢,,,i
E
_~ Na-AZIDE
93~
1 J q
F
, ,,,,,,, t e Ii
lilt
te ~
te ~
t~ 4
DNA CONTENT (PI FLUORESCENCE) Fig. 3 D N A fluorescence t u s t o g r a m s of P I - s t a l n e d t h y m o c y t e s after 24 h i n c u b a t i o n u n d e r v a n o u s e x p e r i m e n t a l c o n d i t i o n s A m e d i u m alone at 4 ° C , B m e d i u m plus 10 7 M D E X at 4 ° C , C" m e d i u m alone at 37°C; D m e d m m plus 10 - 7 M D E X at 4 ° C , E as D plus cycloherdnude; F. m e d m m plus 0.1% s o d i u m azade.
277
(/~ PT-
Lid V" l::l:: .~
~-"
::~ ~ r'~ LLI
~-" i~
X LLI
I 0
-1>-
"----4 X iii
~ -IX lJJ
LLI r'~
N < (l~
Fig 4 Agarose gel electrophoresls of DNA extracted from mouse thymocytes after 24 h incubation using various experimental con&t~ons The DNA ladder of DEX-treated thymocytes greatly exceeds that of controls Cyclohexamadecomcubatlon almost completelyinhibited DNA fragmentaUon No DNA fragments were seen m sodium azlde treated thymocytes
Fig. 6 shows the flow cytometric pattern of PI-stained nuclei obtained from parallel cultures. The percentage of hypodiplold thymocytes was almost supenmposable on the percentage of D N A fragmentation as measured with the colorimetnc assay and the correlation between the data of the two methods was impressive (r = 0.946).
fact that events known to inhibit apoptotic cell death, such as 4°C incubation or cyclohexlrmde treatment, also prevented the appearence of the hypodiploid peak, clearly demonstrate that apoptotic cells can be identified by our method. As far as we are aware, no other cytofluorometnc method employs PI for analysing D N A fragmentation. Ojeda et al. (1990) described an acridine-orange flow cytometric technique for assessing thymocyte apoptosis. However, the problems inherent in the use of acridine-orange in flow cytometry (Shapiro, 1988), and the large CV of the peaks and the poor separation between hypo&plold and diploid D N A peaks which these investigators reported, make this assay of questionable value for precise measurement of D N A fragmentation. The method we propose appears to offer a number of advantages over the previously published procedures. It gives a reliable and reproduclble estimate of apoptosis, as shown by the fact that the percentage of hypodiplold cells revealed by flow cytometry m both time-course and doseresponse experiments correlated excellently with the results of the classical D N A fragmentation assay. It is also more specific than colorimetric methods since it permits unequivocal assessment
11111
.~, 8O Z
o F-
Discussion
Flow cytometric analysis of cells after PI staremg is widely utilized for evaluating cell death (PI exclusion), determining ploidy m tumor samples and measuring cell cycle parameters of cultured cells (Shapiro, 1988). The new method we propose for the cytometric analys~s of PI-stained thymocyte nuclei permits simple, quantitative and reproducible measurement of apoptos~s. The reduced D N A content of apoptotlc nuclei resulted m a unequivocal hypodiploid D N A peak in the red fluorescence channels. The clear relauonstnp between DEX treatment and the percentage of hypo&ploid nuclei, and the
-'
271
© 1991 ElsexaerScience Pubhshers B V 0022-1759/91/$03 50 ADONIS 0022175991001841 JIM05939
A rapid and simple method for measuring thymocyte apoptosis by propidium iodide staining and flow cytometry I. N i c o l e t t i 1, G . M i g l i o r a t i 2, M.C. Pagliacci 1, F. G r i g n a n i 1 a n d C. R i c c a r d i 2 1 Istttuto dt Chmca Medwa 1, and ' lstttuto dl Farmacologta, Perugla Umverstty School of Medicine, 06100 Perugla, Italy
(Recewed 19 October 1990, rewsed received 14 January 1991, accepted 19 February 1991)
Corticosteroids, calcium ionophores and anti-CD3 monoclonal antibodies kill mouse thymocytes incubated in vitro. Cell death is preceded by extensive D N A fragmentation into ohgonucleosomal subumts. Tins type of cell death (apoptosis), which physiologically occurs m the intrathymic process of immune cell selection, is usually evaluated by either electrophoretic or colorimetric methods which measure D N A fragmentation in the nuclear extracts. These techniques are unable to determine the percentage of apoptotic nuclei or recognize the apoptotic cells in a heterogeneous cell population. We have developed a flow cytometric method for measuring the percentage of apoptotic nuclei after propidium iodide staining in hypotonic buffer and have compared it with the classical colonmetric and electrophoretic techniques using dexamethasone (DEX)-treated mouse thymocytes. Apoptotic nuclei appeared as a broad hypodiploxd D N A peak which was easily discriminable from the narrow peak of thymocytes with normal (diploid) D N A content in the red fluorescence channels. When the DEX-induced apoptosis was inhibited by either low-temperature ( 4 ° C ) incubation or cyclohexlmide treatment, no hypodiploid D N A peak appeared. Similarly, thymocyte death induced by s o d m m azide, a substance with cell-kilhng activity through non-apoptotic mechanisms, &d not result in any variation in the normal D N A peak. The flow cytometrlc data showed an excellent correlatton with the results obtained with both electrophoret~c and colorimetric methods. This new rapid, simple and reproducible method should prove useful for assessing apoptos~s of specific cell populations in heterogeneous tissues such as bone marrow, thymus and lymph nodes. Key words Apoptosls, Thymocyte, DNA fragmentation, Flow cytometry, Dexamethasone
Introduction Apoptosis is a c o m m o n form of cell death xn eukariotes. The process is operative during embryogenesis, in tumor regression and in the
Correspondence to I. Ntcoletti, IsUtuto dt Chmca Med~ca1, Umversit/t dl Perugia, Pohchmco Monteluce, 06100 Perugaa, Italy
elirmnation of self-reactive T lymphocytes m the thymus (Wyllie et al., 1980). Apoptotic cell death can be induced by glucocorticoid treatment (Cohen and Duke, 1984), antibodies to C D 3 / T cell receptor complex (Shi et al., 1989; Smith et al., 1989), exposure to Ca 2+ lonophores (Kazaki et al., 1989; McConkey et al., 1989) or "y-irradiauon of mouse thymocytes (Selllns and Cohen, 1987), anti-APO monoclonal antibody treatment of tumor cells (Trauth et al., 1989) and growth factor deprivation
272
of both hematopoietlc and lymphoid cells in vitro (Williams et al., 1990). The death of cells undergoing apoptosls is preceded by chromatin cleavage at the hnker regions between nucleosomes by specific endonucleases, which results m extenswe fragmentation of the D N A into oligonucleosomal subunits (Wyllie et al., 1980). The D N A fragments can be demonstrated by agarose gel electophoresis and the percentage of D N A fragmentation is usually measured by colorimetric methods after separating intact from fragmented D N A by ultracentrifugation (Selhns and Cohen, 1987). The main flaw in these methods, since the D N A fragmentation assays are performed with the whole nuclear extract, is their inabdlty to either directly evaluate the percentage of fragmented nuclei or recognize the apoptotic cells in a heterogeneous cell population. We have developed a rapid, simple and reproducible method for measuring the percentage of apoptotic cells by flow-cytometric analysis of isolated nuclei stained with propidium iodide (PI) in hypotonic buffer and, using mouse thymocytes, demonstrated that the flow cytometric data correlate excellently with the classical D N A fragmentation assays.
Materials and methods
Cell suspenstons Thymocytes were obtained from 2-4-week-old C 3 H / H e N mice purchased from Charles River (Corsico, Milan). The animals were killed by cervical dislocation and the thymus teased in R P M I 1660 medium. The cell suspension was washed, filtered and adjusted to a concentration of 1.5 × 10 6 cell/ml in R P M I 1660 m e d m m supplemented wlth 5% FCS and 10 m M Hepes buffer. Aliquots of 2 ml thymocytes were incubated at either 37°C or 4°C with DEX (at concentrations from 10 - H to 10 - 6 M). At pre-established times, the cells were centrifuged at 200 × g for 10 nun, washed, and processed (see below). DNA labehng techmque and flow cytometrtc analysts Two different methods were evaluated for D N A labeling.
(1) the 200 × g centnfuged cell pellet was fixed in 2 ml cold 70% ethanol at 4°C for 60 min. The cells were then centrifuged, washed m 1 ml PBS and resuspended m 0.5 ml PBS. To a 0.5 ml cell sample 0.5 ml RNAse (Type I-A, Sigma, St. Lores, MO, U.S.A., 1 m g / m l in PBS) was added, followed after gentle mixing by 1 ml propldlum iodide (PI, Sigma, 100 # g / m l m PBS) solution. The mixed cells were incubated in the dark at room temperature for 15 man and kept at 4°C m the dark until measured. (2) The second method was essentially that described by Fried et al. (1978). The 200 × g centrifuged cell pellet was gently resuspended in 1.5 ml hypotonlc fluorochrome solution (PI 50 /~g/ml in 0.1% sodium citrate plus 0.1% Triton X-100, Sigma), in 12 × 75 polypropylene tubes (Becton Dickinson, Lincoln Park, N J, U.S.A.). The tubes were placed at 4°C in the dark overmght before the flow-cytometric analysis. Since this procedure gave lower coefficients of variation of D N A peaks and better discrimination of subdiploid and diploid cells than the classical method of PI staining after ethanol fixation and RNA-ase treatment (Fig. 2), it was used in the whole study. The PI fluorescence of individual nuclei was measured using a FACScan flow cytometer (Becton and Dickinson, Mountain View, CA, U.S.A.). The nuclei traversed the light beam of a 488 nm Argon laser. A 560 nm dichrolc nurror ( D M 570) and a 600 nm band pass filter (bandwidth 35 nm) were used for collecting the red fluorescence due to PI staining of D N A and the data were registered on a logarithmic scale. The forward scatter (FSC) and side scatter (SSC) of particles were simultaneously measured. All the data were recorded in a Hewlett Packard (HP 9000, model 310) computer using specific FACScan research software (Becton Dickinson). Cell debris were excluded from analysis by appropriately raising the FSC threshold. The correct threshold value was selected experimentally. A suspension of 10 /~m diameter fluorescent polystyrene microspheres ( D N A check, Coulter Corporation, Hialeah, FL, U.S.A.) was run and the FSC peak was adjusted around channel 600 of a 1024 channel linear scale. The red fluorescence peak was set at channel 250 in the logarithmic mode. The FSC threshold was then progressively increased until the unwanted
273 signals in the 1-200 channels of the red fluorescence were < 1%. The resulting FSC threshold was utihzed for the analysis of thymocyte nuclei stained with PI hypotonic solution. The residual cell debris ( < 3%) had a very low D N A red fluorescence emission (channels 1-5) and low side scatter (SSC) signal. Apoptotic cell nuclei, which were easily distinguishable from debris by the high SSC value due to the condensation of nuclear chromatm, emitted fluorescence in channels 10250. As there was no overlap between apoptotic nucle~ and debris, the small percentage of residual low-fluorescent detritus was eliminated by gating at channel 10 of the red PI fluorescence scale. The flow rate was set at about 200 nuclei × s-1 and at least 10 4 cells of each sample were analysed. All measurements were done under the same instrument settings.
/~g/ml) was added to each sample and D N A was extracted twice first with phenol plus chloroform and then with chloroform, and recovered by centrifugation after overnight precipitation at - 2 0 ° C in 2 vols. of ethanol in the presence of 0.3 M sodium acetate. Pellets were air dried, dissolved m 10 m M Tris-HC1, 1 m M E D T A ( p H 8) at 4°C and the D N A concentrauon determined by absorbance at 260 nm with a Perkln Elmer spectrofluorimeter. 20 ~g D N A were mixed with 25% Ficoll 400 ( P h a r m a o a , Uppsala, Sweden), 0.25% bromophenol blue and 0.25% xylene cyanol (loading buffer), before loading into the wells of a 1% agarose gel. Electophoresis was carried out in TBE buffer (2 m M EDTA, 89 m M boric acid, 89 m M Tris, p H 8.4) and D N A visualized by ethidium bromide staining.
Colorlmetrlc DN.4-fragmentatton assay and DNA electrophoresis
Results
DEX-induced apoptosis of mouse thymocytes was measured in selected experiments by both the above described flow cytometric method and a classical colorimetric D N A fragmentation assay (Sellins and Cohen, 1987). Briefly, cell suspensions were divided in two aliquots and the resulting cell pellets were either resuspended in PI fluorochrome solution for flow cytometric analysis or dissolved in hypotonic lysing buffer (Tris 10 mM, E D T A 1 mM, Triton X-100 0.2%). Intact and fragmented D N A were separated by centrifugation at 13,000 × g for 10 min and measured by the diphenylamine reaction (Burton, 1956). Percent fragmentation was calculated as the ratio of D N A in the 13,000 x g supernatants to the total D N A (supernatants plus pellets). For D N A electrophoresis, 4 )< 106 mouse thymocytes were incubated overnight at 37°C in medium alone, in 10 - 7 M DEX-supplemented medium, in medium containing 10 - 7 M D E X plus the apoptosis inhibitor cycloheximide (50/~g/ml), or m medium plus 0.1% sodium azide. The cells were washed, centrifuged at 200 × g for 10 min and dissolved in hypotonic lysing buffer (100 m M NaC1, 10 m M Tris, 1 m M EDTA, 1% SDS, 200 /xg/ml protemase K, p H 7.5). Samples were extracted once with phenol plus chloroform ( 1 : 1 , v / v ) and then with chloroform. RNAse A (100
The flow cytometrw pattern of dexamethasone-mduced thymocyte apoptosts Fig. 1 shows the appearance by fluorescence microscopy of ethanol (70%) fixed, PI-stained thymocyte nuclei after 24 h incubation in either medium alone or medium plus 10 - 7 M D E X at 37°C. The majority of steroid-treated thymocytes displayed a clear reduction m nuclear diameter and a complete derangement in chromatin structure. Small areas with highly condensed chromatin were visible. Flow cytometric analysis of ethanol fixed, PIstained cells from parallel cultures revealed a typical diploid D N A peak in R P M I 1660 incubated thymocytes and a minor one in the hypodiploid range; DEX-treated cells manifested an impressive reduction in the percentage of nuclei with diploid D N A content, while the hypodiploid D N A peak was greatly enhanced (Fig. 2, left panels). A better separation between the two peaks was obtained when thymocytes were stained in hypotonic PI solution according to Fried et al. (1978) and the logarithmic mode of PI-fluorescence acquisition utilized during FACScan analysis (Fig. 2, right panels). To demonstrate that the subdiploid peak was really due to apoptotic cells, mouse thymocytes were analyzed in experimental conditions known
274 to m o d i f y t h e a p o p t o t t c p r o c e s s . S i n c e D E X - i n d u c e d D N A f r a g m e n t a t i o n is a t e m p e r a t u r e d e p e n d e n t p h e n o m e n o n ( O j e d a et al., 1990), t h e e f f e c t o f i n c u b a t i o n at b o t h 3 7 ° C a n d 4 ° C o n
DEX action was evaluated. When mouse thymocytes w e r e m a i n t a i n e d for 24 h at 4 ° C t h e D N A h i s t o g r a m s o f cells i n c u b a t e d e i t h e r w i t h o r w i t h o u t 10 - 7 M D E X w e r e a l m o s t s u p e r l m p o s a b l e a n d
O Fig. 1. Fluorescence rmcroscopy (1250 × ) appearance of ethanol-fixed, PI-staaned thymocyte nuclei after 24 h incubation m medium alone (A) or in medium plus 10 -7 M DEX (B) The majority of DEX-treated nuclei show a marked reduction m their diameter and an exadent condensation of chromatin
275
over 90% of nuclei showed a diploid D N A peak (F~gs. 3A and 3B). In parallel cultures incubated at 37°C, the same DEX concentration led to a marked reduction of diploid peak and the appearance of a large number of cells with sub&ploid D N A content (Figs. 3C and 3D). The active process of D N A fragmentation induced by glucocorticoids in thymocytes is known to be lnh~blted by R N A and protein synthesis mhibitors such as D-actmomycin and cyclohex~mide (Cohen and Duke, 1984). The addition of cyclohex~m~de to the incubation m e & u m fully pre-
2 H
vented the appearance of the DEX-lnduced hypodiploid D N A peak and the percentage of diploid nuclei was mrmlar m thymocytes incubated for 24 h at 37°C in medium alone or in medium containmg DEX plus cycloheximide (Fig. 3E). The possibility that the appearance of the sub&plo~d D N A peak was simply related to the presence of fragments of dead cells, was ruled out by evaluating the effect of sodium azade, a substance which induces cell death through a non-apoptotic mechamsm (Wyllie et al., 1984), on mouse thymocytes. Incubation of thymocytes with 0.1% sodium
MEDIUM
07%
1 i ~IIFN I
n I Iil~l~]
lll~.n
I
n i IHIII
I~ ~
I
24 H MEDIUM
178~ I
-I
1
't 1
4
4
t9 Q
, , llrH, I , * ,,NI~ 1 t~ I t~ 2
I ,q,,ll I i i rl,iH I IB a t~ a
1
-I
24 H DEX 10-TM
-1 1 -I
! -1
,,,,
, , , , I , , , , i , , , , [ , , r ,
I 23%
III , ,',~i~, I
" 7 - ~ , . q ~ , , ,
/
.......
t~ ~
DNA CONTENT (PI FLUORESCENCE) Fig 2 DNA fluorescence flow cytometnc profiles of Pl-stamed thymocytes after 2 h and 24 h mcubauon m medium alone or in medium plus 10-7 M DEX. Left panels: D N A tustograms (hnear scale) after 70% ethanol hxatlon, RNAse treatment and PI staining Rtght panels DNA histograms of parallel cultures dtrectly stained with hypotomc PI-fluorochrome solution. This procedure and the data acqulsmon m loganthmac mode perrmt a better resoluuon of the two D N A peaks
276
Compartson of flow cytometmc wtth colonmetrtc DNA fragmentatton assays
azide markedly enhanced the number of dead cells at 24 h ( > 90% by trypan blue exclusion) but did not cause modify the diploid D N A peak of PIstained nuclei (Fig. 3F). Similarly, rapid complement lysts caused a > 95% thymocyte death (trypan blue exclusion) but did not induce any hypodiploid peak in the D N A histogram (data not shown). The absence or presence of sub&ploid D N A peaks paralleled the absence or presence of a ladder of ohgonucleosome length D N A degradation in agarose gel electophoresis when thymocytes were incubated in medium alone, or medium w i t h 10 - 7 M DEX, DEX plus cycloheximide or sodium azide (Fig. 4).
~o+:
~ ....
~@@ , , P~@@ ,
The reliability of the method we propose to measure apoptosis was compared wtth the clasmcal colorimetric D N A fragmentation assay descrtbed by Sellins and Cohen (1987). Fig. 5 gives the percentage of DEX-mduced D N A fragmentation in mouse thymocytes, as evaluated by the colorimetrtc assay, m a dose response experiment. There was a clear dose related increase in the D N A fragmentation of DEX-treated cells, while cycloheximlde strongly inhibited both spontaneous and DEX-mduced chromatm breakdown.
F@@ , ~ ~@@ , , I t eBee'~
MEDIUM 4°C
t _]' DEX 4°C
A
97%1'
IQ~4
B
96%
t
t'
I
1
L
4
-1 -I
1 IBI
%e ¢l
"+t=
Ie . . . .
r~)
le+ , , P@@
lm 4
| e ~1
, , IS~a0 ,
lelialP i I F e e
I
MEDIUM 37°C
C
72%
.
toQ
' ''""'1
Im. . . .
ta l
........
I IB •
'""q
IB •
' '''"q
tg 4
,rap+,, 1,0•, ,iFp~p, impGpi r team
_3 ' DEX 37°C
'I
O
I z
28% , , ,, ,ir~.; z
lea
te l
U,h. I 1o 2
'lllill'l I1111111 le ~ IB 4
le B
|e I
IBa
tm~
|B 4
I
DEX+C'YCLOHEXIH. ~4% 'I
16% , , ,.,,,~ tee
~ , ,+,h, i let
Ie2
¢,,,i
E
_~ Na-AZIDE
93~
1 J q
F
, ,,,,,,, t e Ii
lilt
te ~
te ~
t~ 4
DNA CONTENT (PI FLUORESCENCE) Fig. 3 D N A fluorescence t u s t o g r a m s of P I - s t a l n e d t h y m o c y t e s after 24 h i n c u b a t i o n u n d e r v a n o u s e x p e r i m e n t a l c o n d i t i o n s A m e d i u m alone at 4 ° C , B m e d i u m plus 10 7 M D E X at 4 ° C , C" m e d i u m alone at 37°C; D m e d m m plus 10 - 7 M D E X at 4 ° C , E as D plus cycloherdnude; F. m e d m m plus 0.1% s o d i u m azade.
277
(/~ PT-
Lid V" l::l:: .~
~-"
::~ ~ r'~ LLI
~-" i~
X LLI
I 0
-1>-
"----4 X iii
~ -IX lJJ
LLI r'~
N < (l~
Fig 4 Agarose gel electrophoresls of DNA extracted from mouse thymocytes after 24 h incubation using various experimental con&t~ons The DNA ladder of DEX-treated thymocytes greatly exceeds that of controls Cyclohexamadecomcubatlon almost completelyinhibited DNA fragmentaUon No DNA fragments were seen m sodium azlde treated thymocytes
Fig. 6 shows the flow cytometric pattern of PI-stained nuclei obtained from parallel cultures. The percentage of hypodiplold thymocytes was almost supenmposable on the percentage of D N A fragmentation as measured with the colorimetnc assay and the correlation between the data of the two methods was impressive (r = 0.946).
fact that events known to inhibit apoptotic cell death, such as 4°C incubation or cyclohexlrmde treatment, also prevented the appearence of the hypodiploid peak, clearly demonstrate that apoptotic cells can be identified by our method. As far as we are aware, no other cytofluorometnc method employs PI for analysing D N A fragmentation. Ojeda et al. (1990) described an acridine-orange flow cytometric technique for assessing thymocyte apoptosis. However, the problems inherent in the use of acridine-orange in flow cytometry (Shapiro, 1988), and the large CV of the peaks and the poor separation between hypo&plold and diploid D N A peaks which these investigators reported, make this assay of questionable value for precise measurement of D N A fragmentation. The method we propose appears to offer a number of advantages over the previously published procedures. It gives a reliable and reproduclble estimate of apoptosis, as shown by the fact that the percentage of hypodiplold cells revealed by flow cytometry m both time-course and doseresponse experiments correlated excellently with the results of the classical D N A fragmentation assay. It is also more specific than colorimetric methods since it permits unequivocal assessment
11111
.~, 8O Z
o F-
Discussion
Flow cytometric analysis of cells after PI staremg is widely utilized for evaluating cell death (PI exclusion), determining ploidy m tumor samples and measuring cell cycle parameters of cultured cells (Shapiro, 1988). The new method we propose for the cytometric analys~s of PI-stained thymocyte nuclei permits simple, quantitative and reproducible measurement of apoptos~s. The reduced D N A content of apoptotlc nuclei resulted m a unequivocal hypodiploid D N A peak in the red fluorescence channels. The clear relauonstnp between DEX treatment and the percentage of hypo&ploid nuclei, and the
-'
