AJP-Cell Articles in PresS. Published on June 5, 2002 as DOI 10.1152/ajpcell.00107.2002
1
1
Exposure to UV light Causes Increased Biotinylation of Histones in
2
Jurkat Cells
3
4
DOROTHEA M. PETERS,1 JACOB B. GRIFFIN,1 J. STEVEN STANLEY,2 MARY M.
5
BECK,3 AND JANOS ZEMPLENI1,4
6
Departments of 1Nutritional Science and Dietetics, 3Animal Science, and 4Biochemistry,
7
University of Nebraska at Lincoln, Lincoln, Nebraska, 68583, 2Arkansas Children's
8
Hospital Research Institute, Little Rock, Arkansas, 72202
9
10
Address for reprint requests and other correspondence:
11
J. Zempleni, Department of Nutritional Sciences and Dietetics, University of Nebraska at
12
Lincoln, 316 Ruth Leverton Hall, Lincoln, NE 68583-0806. Phone: (402) 472-3270; fax:
13
(402) 472-1587; email: [email protected]
14
15
Running head: Biotinylation of histones in Jurkat cells
16
Copyright 2002 by the American Physiological Society.
2
1
ABSTRACT
2
3
Biotin in breakdown products of biotinylated carboxylases serves as substrate for
4
biotinylation of histones by biotinidase. Here we determined whether biotinylation of
5
histones might play a role in repair of damaged DNA and in apoptosis. Jurkat cells were
6
exposed to UV light to induce DNA damage. Abundance of thymine dimers increased
7
about three times in response to UV exposure, consistent with DNA damage. Biotin-
8
containing carboxylases were degraded in response to UV exposure, as judged by
9
Western blot analysis and carboxylase activities. Mitochondrial integrity decreased in
10
response to UV exposure (as judged by confocal microscopy), facilitating the release of
11
breakdown products of carboxylases from mitochondria. Biotinylation of histones
12
increased in response to UV exposure; biotinylation of histones did not occur specifically
13
at sites of newly repaired DNA. UV exposure triggered apoptosis, as judged by caspase-3
14
activity and analysis by confocal microscopy. In summary, this study provided evidence
15
that increased biotinylation of histones in DNA-damaged cells might either be a side
16
product of carboxylase degradation or a step during apoptosis.
17
18
3
1
2
Key words: apoptosis; biotin; carboxylases; DNA damage; histones; Jurkat cells
4
1
INTRODUCTION
2
3
In mammals, biotin serves as a covalently bound coenzyme for four biotin-dependent
4
carboxylases: acetyl-CoA carboxylase (E.C. 6.4.1.2); pyruvate carboxylase (E.C.
5
6.4.1.1); propionyl-CoA carboxylase, PCC1 (E.C. 6.4.1.3); and 3-methylcrotonyl-CoA
6
carboxylase (E.C. 6.4.1.4) (27). Degradation of holocarboxylases by proteases leads to
7
the formation of biotinylated peptides. These peptides are further degraded by biotinidase
8
(E.C. 3.5.1.12) to release free biotin, which can then be used for the synthesis of new
9
holocarboxylases (24).
10
Recently, Hymes et al. have proposed an enzymatic mechanism in which biotinidase
11
acts as biotinyl transferase: breakdown products of carboxylases serve as biotinyl donors
12
and histones act as biotinyl acceptors (8). Covalent binding of biotin to histones suggests
13
that biotin might play a role in transcription, replication, or repair of DNA. Based on
14
these pioneering studies, our laboratory has demonstrated that human cells indeed
15
biotinylate histones (22).
16
Histones are the primary proteins that mediate the folding of DNA into chromatin
17
(25). Chromatin consists of a repetitive nucleoprotein complex, the nucleosome. Each
18
(mono)nucleosome consists of DNA that is wrapped around an octamer of core histones
19
(two molecules each of histones H2A, H2B, H3, and H4); nucleosome assembly is
20
completed by incorporation of a linker histone such as histone H1.
21
In vivo, histones are modified posttranslationally by acetylation, methylation,
22
phosphorylation, ubiquitination, poly (ADP-ribosylation), and biotinylation (22, 25).
23
Evidence has been provided that enzymatic poly (ADP-ribosylation) of histones is linked
5
1
to DNA repair mechanisms (1, 4). Induction of DNA damage in mammalian cells by
2
chemical treatment or UV irradiation causes a dramatic increase of poly (ADP-
3
ribosylation) of histones (12). Poly (ADP-ribosylation) of histones catalyzes a transient
4
dissociation from and reassociation of histones with DNA (1). It has been proposed that
5
this mechanism may guide enzymes to sites of DNA repair (1). In addition, poly (ADP-
6
ribosylation) might be involved in nucleosomal unfolding of chromatin in DNA excision
7
repair (1). Also, it has been proposed that poly (ADP-ribosylation) of histone H1 might
8
be involved in early apoptotic events (26).
9
In the present study, we determined whether biotinylation of histones might play a
10
role similar to poly (ADP-ribosylation) of histones in the repair of damaged DNA or in
11
apoptosis. A human T cell line (Jurkat cells) was used to determine whether UV-induced
12
oxidative stress and DNA damage causes breakdown of biotin-dependent carboxylases
13
and increased biotinylation of histones.
14
15
16
MATERIALS AND METHODS
17
18
Cells and culture conditions. Jurkat cells (clone E6-1) were purchased from
19
American Type Culture Collection (Manassas, VA). Cells were cultured in humidified
20
atmosphere (5% CO2) at 37ºC. Culture medium consisted of 90% RPMI-1640 (Hyclone,
21
Logan, UT) and 10% fetal bovine serum (Hyclone), containing 100 I.U./ml penicillin and
22
100 µg/ml streptomycin. Unless otherwise noted, cells were exposed to 79 mW of UV
23
light (254 nm) per 121 cm2 of surface area over a period of 1 h to induce DNA damage;
6
1
at timed intervals, aliquots were collected before, during, and after UV irradiation for the
2
analyses described below.
3
Glutathione reductase. Cellular activity of glutathione reductase may decrease
4
moderately in response to oxidative stress. Activity of glutathione reductase was
5
measured in Jurkat cells before, during, and after exposure to UV light in order to
6
confirm oxidative stress. UV-irradiated Jurkat cells and controls (5 x 106 cells) were
7
centrifuged (250 g for 10 min) and the supernatant was discarded. The cell pellet was
8
suspended in a mixture of 250 µL distilled water, 250 µL detergent (BugBuster,
9
Novagen, Madison, WI), 1 µL of protease inhibitor cocktail (Sigma, St. Louis, MO;
10
catalog number P-8340), and 1 µL of DNase (Benzonase, Novagen). Samples were
11
incubated for 3 min at room temperature to allow for hydrolysis of DNA by DNase.
12
Activity of glutathione reductase in the lysate was measured as described previously (21).
13
The assay is based on monitoring the oxidation of NADPH (decrease of absorbance at
14
340 nm) by glutathione reductase over a period of 15 min.
15
DNA damage. UV light causes DNA damage, e.g., formation of thymine dimers.
16
Here, an ELISA assay was developed on the basis of previously published procedures
17
(13, 14) in order to quantify thymine dimers in DNA. Briefly, 96-well plates (Nunc
18
Maxisorb surface; Fisher, Pittsburgh, PA) were coated with 16 µg of protamine sulfate
19
(in 100 µL of water) per well for 90 min at room temperature; the plates were washed
20
three times with water and were allowed to dry (stable for up to 3 months). DNA was
21
isolated from 8 x 106 UV-exposed Jurkat cells (436 J over 1 h) and normal controls by
22
using a commercially available kit (DNeasy Tissue kit; Qiagen, Valencia, CA). DNA was
23
quantified spectrophotometrically at 260 nm. Fifty nanograms of DNA in 50 µL of
7
1
phosphate-buffered saline were added per protamine sulfate-coated well; plates were
2
dried at 37ºC for 16 h to adsorb DNA to the wells. The plates were washed and
3
unspecific binding sites were blocked as described previously (14). Mouse anti-thymine
4
dimer monoclonal antibody (0.1 ml) was added to each well (clone H3, Sigma; final
5
dilution 1 : 10,000); plates were incubated at 37ºC for 90 min. Plates were washed three
6
times (14) and 0.1 ml of peroxidase-conjugated goat anti-mouse IgG (Sigma; final
7
dilution 1 : 10,000) was added to each well; plates were incubated at 37ºC for 90 min.
8
Plates were washed three times. For color development, 100 µL of TMB (3,3',5,5'-tetra-
9
methylbenzidine) 1-component microwell substrate (Kirkegaard & Perry Laboratories,
10
Gaithersburg, MD; catalog number 50-76-04) were added to each well; plates were
11
incubated for 0.5 h at room temperature. Reaction was terminated by addition of 100 µL
12
of stop solution (Kirkegaard & Perry Laboratories, catalog number 50-85-04) and the
13
absorbance was read at 450 nm in an Emax microwell plate reader (Molecular Devices,
14
Sunnyvale, CA).
15
Propionyl-CoA carboxylase activity. This assay quantifies the binding rate of
16
radioactive bicarbonate to propionyl-CoA, catalyzed by PCC in samples of lysed cells.
17
Aliquots of cell suspension (8 x 106 cells) were centrifuged for 10 min at 2260 g;
18
supernatants were discarded and the cell pellets were suspended in 180 µL of
19
homogenization buffer as described previously (28). After vortexing, 50 µL of 0.5%
20
Triton X-100 were added to lyse the cells. Complete lysis of cells by this procedure was
21
confirmed by light microscopy. PCC activity was quantified as described previously with
22
minor modifications (28). Briefly, lysed Jurkat cells were incubated with propionyl-CoA,
23
[14C]bicarbonate, and cofactors to allow for covalent binding of [14C]bicarbonate to
8
1
propionyl-CoA. After incubation, unbound [14C]bicarbonate was volatilized by addition
2
of perchloric acid and samples were air dried. Finally, samples were suspended in
3
scintillation fluid and the bound [14C]bicarbonate was quantified by liquid scintillation
4
counting.
5
Biotinylation of biotin-dependent carboxylases. Holocarboxylases (as opposed to
6
apocarboxylases) contain covalently bound biotin. Biotin in holocarboxylases can be
7
probed with streptavidin peroxidase, using Western blots of cellular protein extracts. In
8
this study, 6 x 106 Jurkat cells were lysed in 480 µL of detergent (BugBuster, Novagen),
9
immediately followed by the addition of 1.6 µL of protease inhibitor cocktail (Sigma, St.
10
Louis, MO; catalog number P-8340) and 0.5 µL of DNase (Benzonase, Novagen).
11
Samples were incubated for 3 min at room temperature to allow for hydrolysis of DNA
12
by DNase. Next, 65 µL of sample were mixed with 25 µL of loading buffer ("Nupage
13
4x", Invitrogen, Carlsbad, CA), and 10 µL of 1 M dithiothreitol; samples were heated for
14
10 min at 70°C and loaded onto a 3-8% Tris-Acetate gel (Invitrogen). After
15
electrophoresis, proteins were transferred onto polyvinylidene fluoride membranes by
16
electroblotting. Biotin in carboxylases was probed using streptavidin peroxidase in
17
analogy to our previous studies of biotinylated histones (22).
18
Biotinidase activity. Cell suspension containing 10 x 106 Jurkat cells was centrifuged
19
(250 g for 10 min) and the supernatant was discarded. Cell pellets were suspended in 50
20
µL Triton X-100 and 180 µL water to cause lysis. Activity of biotinidase in the lysate
21
was determined by measuring the hydrolysis of N-D-biotinyl-p-aminobenzoic acid as
22
described previously (22).
9
1
Apoptosis. Jurkat cells were monitored for an early marker of apoptosis (activity of
2
caspase-3) by using a commercially available kit ("ApoAlert™ Caspase-3 Colorimetric
3
Assay Kit", Clontech).
4
Confocal laser microscopy. Mitochondria in Jurkat cells were stained using a
5
commercially available kit according to the manufacturer's instructions (MitoTracker
6
CMXRos; Molecular Probes, Eugene, OR). Images of stained mitochondria were taken
7
by Dr. You Zhou at the Beadle Center for Biotechnology at the University of Nebraska-
8
Lincoln by using a confocal laser scanning microscope (BioRad MRC1024ES).
9
Chicken chromatin. Previous studies have revealed that transcriptionally active
10
chromatin has an enhanced sensitivity to DNase compared to transcriptionally inactive
11
chromatin (20); digestion of chromatin by micrococcal nuclease and subsequent salt
12
fractionation of the digest can be used to isolate transcriptionally active and inactive
13
chromatin, respectively. In the present study, nuclei from chicken erythrocytes were used
14
to isolate transcriptionally active chromatin and inactive chromatin using previously
15
published methods (16, 19) with minor modifications. Briefly, 25 ml of blood from adult
16
white leghorn chickens were collected in an equal volume of EDTA-containing saline.
17
Erythrocytes were collected by centrifugation. Nuclei were isolated by washing
18
erythrocytes in detergent-containing buffer. Nuclei were suspended in 20 ml of digestion
19
buffer and incubated with 1000 units of micrococcal nuclease (Takara Ltd., Shiga, Japan)
20
at 37ºC for 20 min; digestion was terminated by addition of 2 ml of 10 mM EGTA and
21
nuclei were collected by centrifugation. Using digested nuclei, chromatin was extracted
22
sequentially with 5 ml of extraction buffer (10 mM Tris, pH 7.0, 2 mM MgCl2, 1%
23
thiodiglycol, 25 mM KCl, 10 mM EGTA, and 30 mM sodium butyrate) containing
10
1
increasing concentrations of NaCl (0.1 and 0.6 M) in analogy to Nickel et al. (16).
2
Previous studies have provided evidence that transcriptionally active chromatin is
3
enriched in the fraction extracted with 0.1 M NaCl; inactive chromatin is enriched in the
4
fraction extracted with 0.6 M NaCl (16, 19, 20). Finally, histones in chromatin extracts
5
were isolated using 1 M HCl (final concentration) as described in our previous studies
6
(22).
7
Note that this procedure of chromatin fractionation has the following limitations. (i)
8
Partitioning of transcriptionally active and inactive chromatin is not complete, i.e., the
9
fraction extracted with 0.1 M NaCl may contain some transcriptionally inactive
10
chromatin and the fraction extracted with 0.6 M NaCl may contain some transcriptionally
11
inactive chromatin (16). (ii) The fraction denoted "transcriptionally active chromatin"
12
may be enriched for both genes that are transcriptionally active and genes that are no
13
longer transcriptionally active but that have been transcribed at earlier states of cellular
14
differentiation. Thus, Rocha et al. have proposed to refer to these regions as
15
"transcriptionally competent chromatin" (20).
16
Biotinylation of histones in UV-treated Jurkat cells. Histones were extracted from
17
Jurkat cell nuclei (4 x 106 cells) by using HCl as described previously (22). Equimolar
18
amounts of histones (as judged by gel densitometry after staining with coomassie blue)
19
were electrophoresed using 16% Tris glycine gels (Invitrogen); biotin in histones was
20
probed using streptavidin peroxidase as described previously (22). Alternatively, biotin in
21
histones was probed using an anti-biotin antibody (22).
22
Previous studies have provided evidence that newly repaired regions of DNA are
23
initially more sensitive than bulk DNA to digestion by nuclease (18). In the present study,
11
1
experiments were conducted to provide evidence as to whether biotinylation of histones
2
occurs specifically at sites of newly repaired DNA. In these experiments Jurkat cells were
3
exposed to UV light for 1 h as described above; controls were not exposed to UV light.
4
One hour after UV exposure, 96 x 106 cells were harvested by centrifugation (250 g for
5
10 min) and nuclei were isolated as described previously (22). Nuclei were digested with
6
micrococcal nuclease as described above for nuclei from chicken erythrocytes with the
7
following minor modification: nuclei from Jurkat cells were suspended in a volume of 1
8
ml of digestion buffer; 50 units of micrococcal nuclease were added. Next, nuclei were
9
collected by centrifugation and chromatin was extracted using 0.25 ml of extraction
10
buffer containing 0.6 M sodium chloride at 4ºC for 20 min. Sixteen microliters of extract
11
were mixed with 4 µL of 5x TBE Hi-density loading buffer (Invitrogen) and
12
electrophoresed under non-denaturing conditions on a 6% TBE DNA retardation gel
13
(Invitrogen). Gels were stained with 0.1 µg/ml ethidium bromide (final concentration) for
14
1 h, followed by visual inspection using a Kodak EDAS 290 Documentation and
15
Analysis System (Rochester, NY), in order to confirm digestion of DNA by nuclease. A
16
second set of gels was used to electroblot histones in chromatin onto polyvinylidene
17
fluoride membranes; biotin in histones was probed with streptavidin peroxidase as
18
described previously (22).
19
Statistics. Homogeneity of variances among groups was tested using Bartlett’s test
20
(10). Variances were homogenous; therefore, data were not transformed before further
21
statistical testing. Significance of differences among groups was tested by one-way
22
ANOVA. Fisher's Protected Least Significant Difference procedure was used for posthoc
23
testing (10). StatView 5.0.1 (SAS Institute; Cary, NC) was used to perform all
12
1
calculations. Differences were considered significant if P < 0.05. Data are expressed as
2
mean ± SD.
3
4
5
RESULTS
6
7
UV-induced oxidative stress and DNA damage. Studies were conducted to confirm
8
that UV exposure of Jurkat cells caused oxidative stress and DNA damage. Previous
9
studies have provided evidence that oxidative stress might be associated with a moderate
10
decrease of glutathione reductase activities (5) and that UV exposure causes the
11
formation of thymine dimers in DNA (6). In the present study, cells were exposed to UV
12
light for 1 h; samples were collected at timed intervals before, during, and after UV
13
exposure. Cells responded to UV exposure with a 15% decrease of activities of
14
glutathione reductase compared to cells before UV exposure (data not shown). This is
15
consistent with the hypothesis that UV exposure of Jurkat cell caused oxidative stress.
16
Note that the coenzyme of glutathione reductase (flavin adenine dinucleotide) is light
17
sensitive; thus, photodegradation of flavin adenine dinucleotide could have contributed to
18
decreased activity of glutathione reductase in response to UV exposure.
19
Next, concentrations of thymine dimers were measured as a marker for UV-induced
20
DNA damage. Thymine dimers were quantified immediately after UV exposure, i.e,
21
when dimer concentrations reach a maximum. The abundance of thymine dimers was
22
significantly greater in UV-exposed cells compared to normal controls: 0.32 ± 0.05
13
1
versus 0.09 ± 0.03 units of absorbance at 450 nm per 50 ng of DNA (P < 0.01),
2
suggesting that UV exposure caused DNA damage.
3
Biotin-dependent carboxylases. Previous studies have suggested that biotin in
4
breakdown products of biotinylated carboxylases serves as substrate for biotinylation of
5
histones (8). We determined whether Jurkat cells respond to UV exposure with increased
6
degradation of biotin-dependent carboxylases, providing substrate for biotinylation of
7
histones. Cells were exposed to UV light for 1 h; cells were collected at timed intervals
8
before, during, and after UV exposure. Cells responded to UV exposure with a
9
progressive degradation of holoenzymes of pyruvate carboxylase, PCC (alpha chain), and
10
3-methylcrotonyl-CoA carboxylase (alpha chain, Fig. 1). Nineteen hours after UV
11
exposure, holocarboxylases were barely detectable in cell extracts. Note that the fourth
12
carboxylase, acetyl-CoA carboxylase, was not detectable in Jurkat cells. The biotinylated
13
alpha chains of PCC and 3-methylcrotonyl-CoA carboxylase have similar molecular
14
weights (80 and 83 kDa, respectively) and migrate as one single band on polyacrylamide
15
gels (Fig. 1).
16
Next, activities of PCC were measured in Jurkat cells at timed intervals before,
17
during, and after exposure to UV light. PCC activities decreased by about 40% during
18
UV exposure, then transiently returned to normal levels, before decreasing again by about
19
35% compared to cells before UV exposure (Fig. 2). This decrease of PCC activity is
20
somewhat less than what was expected based on staining of holocarboxylases with
21
streptavidin peroxidase (Fig. 1); within-day variations may account for the difference
22
between these two experiments. In summary, the findings presented here are consistent
23
with the hypothesis that UV irradiation causes breakdown of holocarboxylases, thereby
14
1
providing substrate for biotinylation of histones. Also, studies outlined below provided
2
evidence for decreased integrity of mitochondria in response to UV exposure, suggesting
3
that breakdown products of holocarboxylases may leak out of mitochondria into
4
cytoplasm.
5
Biotinylation of histones. Jurkat cells were exposed to UV light for 1 h; cells were
6
collected at timed intervals before, during, and after UV exposure. Biotinylation of
7
histones increased in response to UV irradiation (Fig. 3). Apparently, the increased
8
biotinylation of histones was not restricted to a single class of histones but was seen in
9
classes H1, H2A, H2B, H3, and H4. Histones H2A and H2B electrophoresed as one
10
single band. Thus, it remains uncertain whether the probe streptavidin bound to histone
11
H2A, H2B, or both. In parallel experiments, biotin in histones was probed with an
12
antibody to biotin. The antibody bound to histones, providing further evidence that
13
histones were biotinylated (data not shown).
14
Previous studies have suggested that biotinylation of histones is catalyzed by
15
biotinidase (8). In the present study, biotinidase activity was quantified in Jurkat cells at
16
timed intervals before, during, and after UV irradiation. Biotinidase activity was similar
17
for all samples: approximately 40 pmol p-aminobenzoic acid released/(106 cells x 30
18
min). Possible explanations for increased biotinylation of histones despite of constant
19
biotinidase activity are being provided in section "Discussion."
20
Previous studies have provided evidence that newly repaired regions of DNA are
21
initially more sensitive than bulk DNA to digestion by nuclease (18); thus newly repaired
22
DNA is enriched in the mononucleosomal fraction of nuclease-digested chromatin.
23
Theoretically, if biotinylation of histones occurs specifically at sites of DNA damage, the
15
1
abundance of biotinylated histones should be greater in mononucleosomes compared to
2
large fragments of chromatin. In the present study, biotinylation of histones was
3
determined in chromatin from UV-irradiated Jurkat cells and normal controls after
4
digestion with micrococcal nuclease. First, digestion of chromatin by nuclease was
5
confirmed by staining of DNA gels with ethidium bromide (Fig. 4). Next, biotinylated
6
histones in digested chromatin were probed with streptavidin. In both UV-irradiated cells
7
and in normal controls, biotinylated histones were located in large fragments of
8
chromatin (barely visible in Fig. 4) rather than in mononucleosomes. This suggests that
9
biotinylation of histones did not occur specifically at sites of DNA damage or newly
10
repaired DNA. Nevertheless, biotinylation of histones in UV-damaged chromatin might
11
be important for unfolding of higher-order chromatin structures.
12
In analogy, regions of chicken chromatin that contain biotinylated histones are
13
relatively resistant to digestion by micrococcal nuclease. Transcriptionally active
14
chromatin and transcriptionally inactive chromatin was isolated from nuclei in chicken
15
erythrocytes by nuclease digestion and salt fractionation; biotinylated histones were
16
detected exclusively in the fraction containing transcriptionally inactive chromatin (Fig.
17
5). There is precedence for transcriptional silencing of chromatin by covalent
18
modification of histones, e.g., methylation of lysine 9 in histone H3 (2). Taken together,
19
the findings in UV-damaged Jurkat cells and in normal chicken erythrocytes suggest that
20
biotinylation of histones occurs at sites of chromatin that do not contain newly repaired
21
DNA and do not contain transcriptionally active DNA.
22
23
Apoptosis. The present study is consistent with the hypothesis that UV irradiation
caused apoptosis. For example, peak activity of caspase-3 (an early marker of apoptosis)
16
1
was approximately three times greater in UV-irradiated cells compared to cells before
2
UV irradiation (Fig. 6). Note that caspase-3 activity increased later than the observed
3
increase of carboxylase degradation and histone biotinylation (Figs. 1 - 3).
4
Apoptosis in response to UV irradiation was further investigated by using confocal
5
microscopy. Jurkat cells were collected at timed intervals before, during, and after
6
exposure to UV light. Cells were incubated with a dye (MitoTracker) to stain
7
mitochondria; after fixation, cells were evaluated qualitatively by microscopy. UV
8
irradiation caused clustering of mitochondria in cells (Fig. 7, panel B), formation of
9
vacuoles in cells (Fig. 7, panel E), and, eventually, lysis of cells. This decreased
10
mitochondrial and cellular integrity in response to UV exposure parallels the increased
11
biotinylation of histones observed in response to UV treatment (Fig. 3). Theoretically,
12
increased biotinylation of histones might be part of the apoptotic (or necrotic) processes
13
that occur in UV-damaged cells.
14
15
16
DISCUSSION
17
18
This study is consistent with the hypothesis that exposure of Jurkat cells to UV light
19
(254 nm) is associated with increased biotinylation of histones. Evidence has been
20
provided that biotinylation of histones is caused by the following mechanistic sequence:
21
(i) UV irradiation causes oxidative stress and DNA damage, as judged by activity of
22
glutathione reductase and formation of thymine dimers. (ii) Biotinylated carboxylases are
23
degraded, as judged by Western blot analysis of holocarboxylases and cellular activity of
17
1
PCC. (iii) Breakdown products of carboxylases leak from mitochondria into cytoplasm,
2
due to decreased integrity of mitochondria in response to UV irradiation. (iv) Breakdown
3
products of carboxylases diffuse from cytoplasm into the cell nucleus where they are
4
used as substrate for biotinylation of histones. (v) Cells increase biotinylation of histones
5
in response to UV irradiation, as judged by Western blot analysis of histones.
6
This mechanistic sequence is consistent with previous studies by Wolf and
7
coworkers in normal cells and in vitro experiments. These investigators have
8
demonstrated that biotin-dependent carboxylases are degraded by proteases, leading to
9
the release of biotinylated peptides (24). These peptides are hydrolyzed by biotinidase to
10
release biocytin (biotinyl lysine) (3, 24). Biocytin is further hydrolyzed by biotinidase
11
and the biotinyl moiety is transferred to histones for covalent attachment (7-9).
12
Currently it remains uncertain whether increased biotinylation of histones is (i) a
13
mechanism of DNA repair, (ii) a process leading to apoptosis, or (iii) a side product of
14
metabolism that does not play a physiologic role in UV-damaged cells. This study
15
provided evidence that biotinylation of histones does not occur specifically at sites of
16
newly repaired DNA. Rather, biotinylation of histones occurs at sites that have not been
17
repaired (or damaged) and sites that are transcriptionally inactive. These findings, along
18
with the observation that UV-irradiated cells entered apoptosis, suggest that biotinylation
19
of histones might not be associated with DNA repair. Rather, increased biotinylation of
20
histones is an artifact caused by increased availability of substrate (biocytin), or part of a
21
series of events that lead to cell death. Nevertheless, future studies will have to provide
22
conclusive evidence as to whether biotinylation of histones plays a role in DNA repair or
18
1
apoptosis. Both repair and apoptosis are potentially beneficial processes that may prevent
2
tumor initiation by eliminating abnormal DNA.
3
The magnitude by which biotinylation of histones increased in response to UV
4
exposure was quantitatively moderate. However, quantitatively small changes in the level
5
of one modification of a given histone may have important effects regarding other
6
modifications in the same histone molecule. For example, methylation of arginine 3 in
7
histone H4 facilitates subsequent acetylation of histone H4 tails, leading to transcriptional
8
activation (23); numerous other examples for synergistic and antagonistic modifications
9
of histones have been identified (11). By analogy, a moderate increase of biotinylation of
10
11
histones may affect other modifications of histones.
Biotinylation of histones is catalyzed by biotinidase (8). In the present study,
12
evidence has been provided that biotinidase activity remains unchanged if cells are
13
exposed to UV light. How can cells increase biotinylation of histones without increasing
14
the activity of the enzyme that mediates biotinylation? In previous studies we observed
15
that cells responded to proliferation with increased biotinylation of histones in spite of
16
constant expression of the biotinidase gene (22). In these previous studies we offered the
17
following explanations for our findings: biotinylation of histones might be regulated by
18
import of biotinidase into the cell nucleus, or biotinylation of histones might be catalyzed
19
by enzymes other than biotinidase. In the present study, we would like to add another
20
possible explanation: biotinylation of histones is regulated by substrate (biocytin)
21
availability. The following observations are consistent with this hypothesis. The
22
Michaelis-Menten constant of biotinidase for biocytin is approximately 10 µM (17); the
23
concentration of total biotin metabolites (including biocytin) in tissues such as rat liver is
19
1
32 nmol/kg (15), i.e., 300 times less than the Michaelis-Menten constant of biotinidase.
2
Thus, small changes of biocytin concentrations in cells may substantially affect rates of
3
biotinidase-catalyzed reactions. The present study provided evidence that concentrations
4
of biotinylated breakdown products of carboxylases (including biocytin) increase in
5
response to UV irradiation.
6
It remains uncertain why biotinylation of histones at transcriptionally inactive sites
7
of chromatin is favored over sites that are transcriptionally active. Many of the known
8
modifications of histones (e.g., acetylation) require interactions of various regulatory
9
elements such as transcription factors with enzymes that catalyze these modifications.
10
Future studies will have to determine whether transcription factors also direct biotinidase
11
to discrete regions of chromatin.
12
20
1
TEXT FOOTNOTE
2
3
4
5
1Abbreviations used: PCC, propionyl-CoA carboxylase
21
1
ACKNOWLEDGMENTS
2
3
This work was supported by NIH grant DK 60447 and the United States Department
4
of Agriculture/National Research Initiative Competitive Grants Program project award
5
2001-35200-10187. A contribution of the University of Nebraska Agricultural Research
6
Division, Lincoln, NE 68583. Journal Series No. 13614.
7
22
1
REFERENCES
2
3
1.
4
Althaus F. Poly ADP-ribosylation: a histone shuttle mechanism in DNA excision
repair. J Cell Sci 102: 663-670, 1992.
5
2.
Bird A. Methylation talk between histones and DNA. Science 294: 2113-2115, 2001.
6
3.
Cole H, Reynolds TR, Lockyer JM, Buck GA, Denson T, Spence JE, Hymes J,
7
and Wolf B. Human serum biotinidase cDNA cloning, sequence, and
8
characterization. J Biol Chem 269: 6566-6570, 1994.
9
4.
10
11
DNA excision repair. Nature 283: 593-596, 1980.
5.
12
13
6.
Garrett RH, and Grisham CM. Biochemistry. Forth Worth, TX: Saunders College
Publishing, 1995.
7.
16
17
Erden IA, and Kahraman A. The protective effect of flavonol quercetin against
ultraviolet a induced oxidative stress in rats. Toxicology 154: 21-29, 2000.
14
15
Durkacz BW, Omidiji O, Gray DA, and Shall S. (ADP-ribose)n participates in
Hymes J, Fleischhauer K, and Wolf B. Biotinylation of biotinidase following
incubation with biocytin. Clin Chim Acta 233: 39-45, 1995.
8.
Hymes J, Fleischhauer K, and Wolf B. Biotinylation of histones by human serum
18
biotinidase: assessment of biotinyl-transferase activity in sera from normal
19
individuals and children with biotinidase deficiency. Biochem Mol Med 56: 76-83,
20
1995.
21
22
23
9.
Hymes J, and Wolf B. Human biotinidase isn't just for recycling biotin. J Nutr 129:
485S-489S, 1999.
10. SAS Institute. StatView Reference. Cary, NC: SAS Publishing, 1999.
23
1
2
3
4
5
6
7
8
9
10
11. Jenuwein T, and Allis CD. Translating the histone code. Science 293: 1074-1080,
2001.
12. Juarez-Salinas H, Sims JL, and Jacobson MK. Poly(ADP-ribose) levels in
carcinogen-treated cells. Nature 282: 740-741, 1979.
13. Klotz JL, Minami RM, and Teplitz RL. An enzyme-linked immunosorbent assay
for antibodies to native and denatured DNA. J Immunol Methods 29: 155-165, 1979.
14. Leadon SA, and Hanawalt PC. Monoclonal antibody to DNA damage containing
thymine glycol. Mutation Res 112: 191-200, 1983.
15. Lewis B, Rathman S, and McMahon R. Dietary biotin intake modulates the pool
of free and protein-bound biotin in rat liver. J Nutr 131: 2310-2315, 2001.
11
16. Nickel BE, Allis CD, and Davie JR. Ubiquitinated histone H2B is preferentially
12
located in transcriptionally active chromatin. Biochemistry 28: 958-963, 1989.
13
17. Pispa J. Animal biotinidase. Ann Med Exp Biol Fenniae 43: 4-39, 1965.
14
18. Ramanathan B, and Smerdon MJ. Enhanced DNA repair synthesis in
15
hyperacetylated nucleosomes. J Biol Chem 264: 11026-11034, 1989.
16
19. Ridsdale JA, and Davie JR. Chicken erythrocyte polynucleosomes which are
17
soluble at physiological ionic strength and contain linker histones are highly enriched
18
in β-globin sequences. Nucleic Acids Res 15: 1081-1096, 1987.
19
20. Rocha E, Davie JR, van Holde KE, and Weintraub H. Differential salt
20
fractionation of active and inactive genomic domains in chicken erythrocyte. J Biol
21
Chem 259: 8558-8563, 1984.
22
23
21. Sauberlich HE, Judd JH, Nichoalds GE, Broquist HP, and Darby WJ.
Application of the erythrocyte glutathione reductase assay in evaluating riboflavin
24
1
nutritional status in a high school student population. Am J Clin Nutr 25: 756-762,
2
1972.
3
4
5
22. Stanley JS, Griffin JB, and Zempleni J. Biotinylation of histones in human cells:
effects of cell proliferation. Eur J Biochem 268: 5424-5429, 2001.
23. Wang H, Huang Z-Q, Xia L, Feng Q, Erdjument-Bromage H, Strahl BD,
6
Briggs SD, Allis CD, Wong J, Tempst P, and Zhang T. Methylation of histone H4
7
at arginine 3 facilitating transcriptional activation by nuclear hormone receptor.
8
Science 293: 853-857, 2001.
9
10
24. Wolf B, and Heard GS. Biotinidase deficiency. In: Advances in Pediatrics, edited
by Barness L, and Oski F. Chicago, IL: Medical Book Publishers, 1991, p. 1-21.
11
25. Wolffe A. Chromatin. San Diego, CA: Academic Press, 1998.
12
26. Yoon YS, Kim JW, Kang KW, Kim YS, Choi KH, and Joe CO. Poly(ADP-
13
ribosyl)ation of histone H1 correlates with internucleosomal DNA fragmentation
14
during apoptosis. J Biol Chem 271: 9129-9134, 1996.
15
16
27. Zempleni J, and Mock DM. Biotin biochemistry and human requirements. J Nutr
Biochem 10: 128-138, 1999.
17
28. Zempleni J, Trusty TA, and Mock DM. Lipoic acid reduces the activities of
18
biotin-dependent carboxylases in rat liver. J Nutr 127: 1776-1781, 1997.
19
25
1
FIGURE LEGENDS
2
3
FIG. 1. UV exposure affects levels of holocarboxylases in Jurkat cells. Cells were
4
exposed to UV light (254 nm) for 1 h. Aliquots were collected before (0 h), during
5
("hours of UV"), and after ("hours after UV") exposure to UV light and the following
6
holocarboxylases were quantified by Western blot analysis: pyruvate carboxylase (PC),
7
3-methylcrotonyl-CoA carboxylase (MCC), and propionyl-CoA carboxylase (PCC). The
8
biotinylated alpha chains of MCC and PCC have similar molecular weights (83 and 80
9
kDa, respectively) and migrate as one single band.
10
11
26
1
FIG. 2. UV exposure affects activities of propionyl-CoA carboxylase in Jurkat cells. Cells
2
were exposed to UV light (254 nm) for 1 h. Aliquots were collected before (0 h), during
3
("hours of UV"), and after ("hours after UV") exposure to UV light and PCC activity was
4
measured. *,**Significantly different from controls (*P < 0.05; **P < 0.01); n = 8.
5
6
27
1
FIG. 3. UV exposure affects biotinylation of histones in Jurkat cells. Cells were exposed
2
to UV light (254 nm) for 1 h. Aliquots were collected before (0 h), during ("hours of
3
UV"), and after ("hours after UV") exposure to UV light and biotinylation of histones as
4
quantified by probing with streptavidin.
5
6
28
1
FIG. 4. Biotinylation of histones did not occur specifically at sites of newly repaired DNA
2
in Jurkat cells. Cells were exposed to UV light (254 nm) for 1 h; controls were not UV
3
irradiated. One hour after UV exposure, 96 x 106 cells were harvested and nuclei were
4
digested with micrococcal nuclease. Chromatin was extracted and electrophoresed on a
5
non-denaturing 6% TBE gel. Lane a: representative sample of nuclease-digested
6
chromatin after staining with ethidium bromide to verify digestion of chromatin. Lane b:
7
biotinylated histones in digested chromatin from UV-exposed cells were probed with
8
streptavidin. Lane c: same as in b, but cells were not UV irradiated. In both lanes b and c,
9
biotin was associated with large fragments of DNA (barely visible on gels) rather than
10
11
12
with mononucleosomes.
29
1
FIG. 5. Biotinylated histones are enriched in transcriptionally inactive chromatin.
2
Transcriptionally active chromatin and transcriptionally inactive chromatin was isolated
3
from nuclei in chicken erythrocytes. Histones were acid extracted and biotin in histones
4
was probed with streptavidin by Western blot analysis.
5
6
30
1
FIG. 6. UV exposure affects activities of caspase-3 in Jurkat cells. Cells were exposed to
2
UV light (254 nm) for 1 h. Aliquots were collected before, during, and after exposure to
3
UV light and activity of caspase-3 was measured. *P < 0.05 versus zero-hour controls (n
4
= 3).
5
6
31
1
FIG. 7. Effects of UV exposure on mitochondrial and cellular integrity in Jurkat cells.
2
Cells were exposed to UV light (254 nm) for 1 h. Aliquots were collected before, during,
3
and after exposure to UV light; mitochondria were stained and cells were evaluated by
4
microscopy. Panel A: Non-irradiated controls; arrow indicates mitochondria which were
5
evenly distributed. Panel B: Cells after 1 h of UV exposure; arrow indicates clustering of
6
mitochondria in cells. Panel C: Cells 1 h after termination of UV exposure. Panel D:
7
Cells 3 h after termination of UV exposure; arrow indicates clustering of mitochondria.
8
Panel E: Cells 18 h after termination of UV exposure; arrow indicates formation of
9
vacuoles in cells and initiation of cell lysis. Panel F: Cells in which apoptosis was
10
induced by culturing with 50 µM of actinomycin D for 18 h; arrows indicate clustering of
11
mitochondria and lysis of cell membrane.
12