© 1991 Oxford University Press
Nucleic Acids Research, Vol. 19, No. 8 1745
A novel DNA nucleotide in Trypanosoma brucei only
present in the mammalian phase of the life-cycle
Janet Gommers-Ampt, Jan Lutgerink1* and Piet Borst
Division of Molecular Biology, The Netherlands Cancer Institute, Plesmanlaan 121, 1066 CX
Amsterdam and 1 Division of Chemical Carcinogenesis, The Netherlands Cancer Institute, Plesmanlaan
121, 1066 CX Amsterdam, The Netherlands
Received February 13, 1991; Revised and Accepted March 18, 1991
ABSTRACT
INTRODUCTION
The existence of an unusual form of DNA modification
In the bloodstream form of the African trypanosome
Trypanosoma brucei has been Inferred from partial
resistance to cleavage of nuclear DNA with Pstl and
Pvull (Bernards et al, 1984; Pays et al, 1984). This
putative modification Is correlated with the shut-off of
telomeric Variant-specific Surface Glycoprotein (VSG)
gene expression sites (ESs). The modification only
affects inactive VSG genes with a telomeric location,
and it is absent in procydic (insect form) trypanosomes
In which no VSG is made at all. Previous attempts to
detect unusual nucleosldes In T.brucel DNA were
unsuccessful, but we now report the detection of two
unusual nucleotides, called pdJ and pdV, in T.brucel
DNA, using the ^P-postlabeling technique. Nucleotide
pdV was present In both bloodstream form and
procycllc T.brucel DNA and co-migrated in two different
two-dimensional thin layer chromatography (2D-TLC)
systems with hydroxymethyldeoxyurldine 5'-monophosphate (pHOMedU). In contrast, nucleotide pdJ was
exclusively present In bloodstream form trypanosomal
DNA. Levels of pdJ were higher in DNA enriched for
telomeric sequences than In total genomlc DNA and
pdJ was also detected In other Kinetoplastlda species
exhibiting antlgenic variation. Postlabeling and 2D-TLC
analyses showed base J to be different from the known
eukaryotic unusual DNA bases 5-methylcytoslne,
N6-methyladenine and hydroxymethyluracll, and also
from (glucosylated) hydroxymethylcytoslne, uracil, aputrescinylthymlne, 5-dlhydroxypentyluracil and N6carbamoylmethyladenlne. We conclude that pdJ is a
novel eukaryotic DNA nucleotide and that K is probably
responsible for the partial resistance to cleavage by
Pvull and Pstl of Inactive telomeric VSG genes. It may
therefore be Involved in the regulation of ES activity
In bloodstream form trypanosomes.
The African trypanosome Trypanosoma brucei is able to evade
the immune response of its mammalian host by continuously
changing its surface coat, consisting of a single protein species,
the Variant-specific Surface Glycoprotein (VSG). Expressed VSG
genes are invariably located near the ends of chromosomes, where
they are part of a large multicistronic transcription unit, the
telomeric Expression Site (ES). The trypanosome can change its
coat either by replacing the VSG gene in an active expression
site or by switching from one ES to another one (reviewed in
1). Previous work has shown that the transcriptional shut-off of
an ES in bloodstream form trypanosomes was accompanied by
a partial modification of Pstl and Pvull sites (2,3) and possibly
Hindin and SphI sites (3) in and around the inactivated VSG
gene. The degree of modification varied with the length of the
telomere. Trypanosome telomeres grow by addition of GGGATT units (4,5) at a rate of 1 - 2 hexamer units per cell division
(6,7) and they contract by die occasional loss of large telomeric
segments (6). The degree of modification of a given restriction
site was found to increase wim the length of the adjacent telomere
from 1 - 5 % in a telomere of 2 kb to >20% in a telomere >20
kb (2). Modification was strictly limited to telomeric VSG genes
in bloodstream form trypanosomes. The (always) silent
chromosome-internal VSG genes of bloodstream trypanosomes
were not modified; procyclic (insect-form) trypanosomes, in
which all VSG gene ESs are shut off, contained no modified sites
at all (3).
1
Previous attempts to detect the putative unusual nucleoside
responsible for the partial cleavage of restriction sites, were
unsuccessful. In enzymic hydrolysates of trypanosome
bloodstream form DNA no unusual nucleosides were found using
HPLC analysis (8). As modification affects only a small fraction
of total trypanosome DNA, the possibilities remained that a
modified nucleoside was present but at a level below the detection
limit of the HPLC method used (0.1 % of total nucleoside), or
that it was undetectable due to comigration with one of the normal
nucleosides. We have therefore turned to a more sensitive
Present address: Department of Human Biology BMC, State University of Limburg, Beeldsnijdersdreef 101, 6200 MD Maastricht, The Netherlands
1746 Nucleic Acids Research, Vol. 19, No. 8
method, ^-postlabeling followed by 2-dimensional thin layer
chromatography (2D-TLC) for separation of the labeled
nucleotides (9). This paper describes the detection of two unusual
nucleotides in trypanosome DNA and analyses their possible
relation to the partial cleavage of PstI and PvuII sites in silent
telomeric VSG genes.
MATERIALS AND METHODS
Trypanosomes
The trypanosomes used belong to strain 427 of T.brucei brucei.
Trypanosome variant 221a (MiTat 1.2a) is described by Cross
(10). Variants 118a (MiTat 1.5a) and 118a' are described by
Michels et al (11,12). Variant 1.8 expresses the VSG gene 1.8
and is a relapse of 221a (unpublished result). Trypanosomes were
grown in Sprague Dawley rats and blood was collected from
animals with a high parasitaemia by cardiac puncture.
Trypanosomes were separated from blood cells by DEAEcellulose chromatography according to Fairlamb et al (13).
Procyclic culture form trypanosomes, recently isolated from the
insect vector, were grown in the semi-defined medium described
by Brun and Schonenberger (14).
Isolation of DNA
Total trypanosome DNA was isolated as described by Bernards
et al. (15) and resuspended in 2 mM Tris-HCl, pH 7.6 . Bacillus
subtilis bacteriophage HI DNA was isolated essentially as
described by Arwert and Venema (16). Bacteriophage HI and
its B. subtilis host 0G1 were kindly provided by Dr. S. Bron
(Dept. of Genetics, University of Groningen, Haren, The
Netherlands).
Southern blotting and hybridization
Digested DNA was transferred to nitrocellulose by standard
procedures (17). The 0.7 kb Hinfl-PvuII fragment of TgBl .1006
(probe 3 from 18) was labeled by nick translation (19). Blots
were hybridized as described (20) and washed in 0.3 xSSC, 0.1 %
SDS at 65°C.
Pulsed-field gradient gel electrophoresis (PFGE) and isolation
of chromosomal DNA
Trypanosomes were prepared as described by Van der Ploeg et
al (21). In initial experiments mini-chromosomal DNA was
separated from larger chromosomal DNA in 1 % agarose gels
in 1XTBE (16) for 24 h at 14°C at 10 V/cm with a pulse
frequency of 60 s (fig. 1A). For the isolation of both minichromosomal and large chromosomal DNA separations were
performed in 1.5% agarose gels for 12 h at 14°C at 10 V/cm
with a pulse frequency of 30 s. DNA was visualized with ethidium
bromide. Mini-chromosomes and large chromosomes (all other
chromosomes that migrated out of the slot) were cut out and DNA
was recovered from the gel slices by PFGE in dialysis tubes.
The electro eluted DNA was concentrated by n-butanol
extractions prior to extensive dialysis against 10 mM Tris-HCl,
pH 8.0, and 0.1 mM EDTA. After dialysis, the DNA was
precipitated with an equal volume of isopropanol in 0.3 M sodium
acetate, pH 5.2, and resuspended in 2 mM Tris-HCl, pH 7.6.
32
P-postlabeUng analysis
The method was essentially as described by Gupta et al (22),
with a few modifications. 1 /xg DNA was hydrolysed to
deoxyribonucleoside 3'-monophosphates (dNp's) with 0.2 U
micrococcal endonuclease (Sigma) and 0.01 U spleen
phosphodiesterase (Cooper) for 3h at 37 °C in a total volume of
10 /J of 20 mM sodium succinate, pH 6.0, containing 10 mM
CaCl2. Both enzymes were dialysed against deionized water at
4°C for 15 h before use. Subsequent ^P-labeling of the dNp's
with T4 polynucleotide kinase (PNK) and [y-^PJATP were
done as described (22), with the exception that T4 PNK (diluted
to 2 U//J) was obtained from Boehringer (3'-phosphatase free).
32
P-labeled deoxyribonucleoside 5',3'-biphosphates (pdNp's)
were converted into deoxyribonucleoside 5'-monophosphates
(pdN's) by incubation with 1 U nuclease PI (Boehringer) per
pmole pdNp in 60 mM sodium acetate, pH 5.0, containing 0.1
mM ZnCl2 for lh at 37°C. If the labeled pdN's were separated
by 2D-TLC under conditions A (see next paragraph), the
unreacted [ T 3 2 P ] A T P was removed by adding 0.025 U apyrase
(Sigma, grade 1) in 10 mM N,N-bis-(2-hydroxyethyl)glycine,
pH 9.6, to the reaction mixture, followed by incubation for
30 min. at 37°C. This was done because under these
chromatographic conditions ATP migrates close to the nucleotides
of interest.
Two-dimensional thin layer chromatography
Labeled nucleotides were separated by two-dimensional thin layer
chromatography (2D-TLQ either using unmodified cellulose
sheets (Merck) or polyethyleneimine-impregnated sheets
(Polygram eel 300 PEI, Macherey Nagel). Conditions A: Samples
spotted onto unmodified cellulose were developed in the first
dimension with isobutyric acid:H2O:NH4OH (60:20:l,v/v),
slightly modified from Dawid et al. (23) and in the second
dimension with saturated (NH^SO^isopropanol:! M sodium
acetate (80:2:18,v/v) (24). Conditions B: PEI sheets were soaked
for 20 min. in 0.1 M ammonium formate, pH 3.5 (prepared by
titration of formic acid with concentrated ammonia) and
subsequently dried prior to application of the sample. The first
dimension (Dl) was developed with 0.3 M ammonium formate,
pH 3.5. After development in Dl, the sheets were dried and
washed in methanol to remove the ammonium formate.
Development in the second dimension was with saturated
(NH^SC^ adjusted to pH 3.5 with sulfuric acid. After
autoradiography, spots containing labeled nucleotides were
excised from the sheets and counted by Cerenkov assay. The ratio
of unusual nucleotide relative to the total of normal
deoxyribonucleotides was calculated from the percentage of
unusual nucleotide relative to 2(pdC + pdT), as (pdG + pdA)
was lower than (pdC + pdT) possibly due to depurination.
Treatment of nucleotides with trifluoroacetk add (TFA) and
hydrazine
DNA hydrolysates (dNp's) were lyophilized and subsequently
treated either with 80% v/v TFA in H2O for 1 h at room
temperature or with 62% v/v hydrazine in Hfi for 4 h at 60 c C.
The samples were then lyophilized and resuspended in 10 /d of
H2O.
Synthesis of 5-hydroxymethvldeoxycvtidme 3'-monopbosphate
(HOMedCp)
HOMedCp was synthesized and purified as described by Alegria
(25).
Nucleic Acids Research, Vol. 19, No. 8 1747
RESULTS
Partial restriction enzyme digestion in a mini-chromosomal
telomeric VSG gene
In previous work (2,3) only telomeres containing a VSG gene
expression she (ES) were studied for the presence of the proposed
base modifications. Whether other telomeres were modified as
well, remained unknown. The detection of the modified
nucleotide would be complicated if modification were restricted
to telomeres with an inactive ES, because less than 10% of all
telomeres are supposed to contain an ES (about 240 telomeres
per nucleus (21,26) and at most 20 ESs (27,28)). We therefore
tested whether telomeric VSG genes not located in a potential
ES are also modified. Such VSG genes are mainly present in
the about 100 mini-chromosomes of 50 — 150 kb. Minichromosomes contain VSG genes (21), but lack ES specific
sequences, such as VSG promoter and Expression Site Associated
Gene (ESAG) sequences (28), and therefore cannot contain a
functional ES. A diagnostic PvuII site in mini-chromosomal VSG
gene 1.1006 (18,21), with a telomeric location (not shown), was
tested for partial cleavage in isolated mini-chromosomal DNA
of bloodstream form trypanosomes. Mini-chromosomal DNA
from procyclic culture form trypanosomes that should be
completely cleaved, served as a control.
Mini-chromosomal DNA was isolated from Pulsed Field
Gradient (PFG) gels (Fig. 1A) and digested with both Hinfl and
various amounts of PvuII. In the procyclic culture form DNA,
increasing amounts of PvuII resulted in a complete digestion of
the 1.6 kb Hinfl fragment, yielding the 0.7 kb Hinfl-PvuII
fragment (Fig. IB lanes 1-4) expected from the VSG 1.1006
gene map (Fig. IB). In contrast, PvuII overdigestion of
bloodstream form DNA of variant 118a left about 22% (as
determined by scanning of the autoradiogram) of the 1.6 kb Hinfl
fragment uncleaved (Fig. IB lanes 5—8), indicating the presence
of base modifications. In three additional experiments with minichromosomal DNA of variant 221a, 15%, 10% and 15% of the
1.6 kb Hinfl fragment remained uncleaved. The lower level of
modification in 221a could be due to the rather short length of
the telomere downstream of the VSG 1.1006 gene ( < 14 kb in
variant 221a, not shown). Bernards et al. (2) have shown for
another telomere that the level of modification increased with
the length of the array of telomeric repeats, modification being
below 20% for short telomeres. From these results we conclude
that modification of inactive telomeres is not restricted to
telomeres harbouring an ES.
purified mini-chromosomal DNA is enriched in telomeric
sequences and hence (presumably) in modified nucleotide.
Figures 2A and 2B show the postlabeled nucleotides derived
from both bloodstream form variant 221a and procyclic culture
form mini-chromosomal DNA, each separated under two
different 2D-TLC conditions. In mini-chromosomal DNA of both
life cycle stages spots representing the four normal
deoxyribonucleotides, indicated as pdG, pdA, pdT, pdC and small
amounts of four ribonucleotides, pG, pA, pU, pC are visible.
The ribonucleotides are derived from the reagents used in the
postlabelings procedure, as shown in a postlabeling experiment
— slot
— compr.
} Interm.
(up to 700 kb)
— mini
(50-150 kb)
B
Procyclic
Bloodstream
culture form form
1 2 3 4
5 6 7 8
kb
0.7
Hf
Detection of an unusual nucleotide specific for bloodstream
form trypanosomes
Since our attempts to visualize a modified nucleoside in total
genomic DNA of T. brucei using HPLC analysis were
unsuccessful (8), we switched to the more sensitive 32Ppostlabeling technique, developed by Randerath and coworkers
(9). In this technique DNA is enzymically hydrolysed to
deoxyribonucleoside 3'-monophosphates (dNp's), which are
subsequently labeled with [7-32P]ATP and polynucleotide
kinase. The resulting deoxyribonucleoside 5',3'-biphosphates
(pdNp's) are converted into deoxyribonucleoside
5'-monophosphates (pdN's) and separated by two dimensional
thin layer chromatography (2D-TLQ. In addition, we used minichromosomal DNA rather than total genomic DNA for the
analysis. Because the mini-chromosomes contain only about 10%
of the total T. brucei DNA but about 80% of the telomeres,
I
Pv
|_
Hf
END
1.6 _
0.7 _
probe.
Figure 1. Partial cleavage of a PvuII site in mini-chromosomal telomeric VSG
gene 1.1006 of bloodstream form variant 118a. Panel A shows the separation
of mini-chromosomes from larger chromosomes of variant 118a in an ethidiumstained 1 % agarose gel, after Pulsed Field Gradient (PFG) gel electrophoresis
according to methods. Positions of slot, compression zone (compr.), intermediate
chromosomes (interm.) and mini-chromosomes (mini) are indicated. Panel B.
Autoradiogram of a Southern blot containing a Hinfl-PvuII double digest of
procyclic culture form (lanes 1 —4) and bloodstream form variant 118a (lanes
5 - 8 ) genomic DNA. 0.1 /ig DNA was digested with 10 U Hinfl (lanes 1,5),
followed by 0.1 U PvuII (lanes 2,6), 1 U PvuII (lanes 3,7), 10 U PvuE (lanes
4,8). The 0.7 kb Hinfl-PvuII fragment of TgBl. 1006 was used as a probe. The
map beneath the blot shows the 1.1006 gene and indicates the position and length
(in kb) of the hybridizing fragments (Hf: Hinfl; Pv: PvuII) and the probe.
1748 Nucleic Acids Research, Vol. 19, No. 8
TaWe I. Modified nucleotide pdJ is specific for bloodstream form trypanosomes
and enriched in mini-chromosomal DNA
Procyclic culture form
Bloodstream form
-DNA
POA
PA
pdJ
%pdV
mini-chromosomes:
bloodstream form 221a
bloodstream form 118a
average(5)
procyclic culture form
0.50; 0.45; 0.27
0.29; 0.30
0.36
<0.0007; <0.002
0.08; 0.14; 0.09
0.12
total DNA:
bloodstream form 221a
bloodstream form 1.8
bloodstream form 118a
average(5)
procyclic culture form
0.09; 0.07
0.07; 0.04
0.04
0.06
<0.0002; <0.0003
n.q.
n.q.
n.q.
large chromosomal DNA:
bloodstream form 118a
0.10
0.03
0.02
n.q.
pdC
Percentages unusual nucleotide relative to the total of normal deoxyribonucleotides
were determined as described in methods. The average levels of pdJ were calculated
from the number of experiments indicated between brackets. All data are derived
from separate postlabeling experiments. Levels of pdJ and pdV presented here
are minimal levels, as postlabeling of pdJ and pdV may be less efficient than
that of the normal deoxyribonucleotides (see also discussion),
n.q.: detected but not quantified.
< : not detected, below indicated detection level.
pdV
•
pU
pG
B
DNA%
Bloodstream form
Procyclic culture form
pdJ
Bloodstream form
Procyclic culture form
Figure 2. Detection of a modified nucleotide specific for bloodstream form
trypanosomes. The autoradiograms show 2D-TLC separations of 32P-postlabeled
nucleotides derived from mini-chromosomal DNA of bloodstream form 221a and
procyclic culture form T.brucei and from the reagents used for post labeling
(-DNA). Panel A: TLC conditions as described under A in methods. Panel B:
TLC conditions as described under B in methods. Panel C: Longer exposures
of autoradiograms shown in panel B. Only the area indicated with dashed lines
(panel B) is shown. The arrow and the arrow with circle indicate pdV and pdJ
respectively. pdN: deoxyribonucleoside 5'-monophosphate; pN: ribonucleoside
5'-monophosphate. The smear in panel B is from free inorganic phosphate. Dl
and D2 indicate direction of chromatography in the first and second dimension
respectively.
without added DNA (fig.2A, -DNA). In bloodstream form
DNA we detect two additional spots, indicated with arrows.
These two spots were not detected in a postlabeling experiment
with RNA isolated from variant 221a (not shown), nor in an
experiment without DNA (fig.2A, —DNA). Moreover, it is
highly unlikely that these spots are the result of incomplete
digestion of DNA during the postlabeling procedure, since such
intermediate products do not migrate from the origin in the first
dimension under the chromatography conditions used in B. We
therefore conclude that the two additional spots in bloodstream
form DNA are deoxyribonucleotides, and refer to them as pdJ
and pdV. Nucleotide pdJ is absent from procyclic culture form
DNA, as can be clearly seen in fig. 2C, which presents longer
exposures of the part indicated with the dashed lines in the
chromatograms of panel B. Nucleotide pdV is present in DNA
of both life cycle stages, but more prominently in the bloodstream
form than in the procyclic culture form DNA (fig. 2C).
Percentages of the nucleotides pdJ and pdV compared to total
deoxyribonucleotides, as determined from several independent
experiments by Cerenkov counting of the excised spots, are
presented in table I. We conclude that we have detected an
unusual nucleotide pdJ that is exclusively present in bloodstream
form DNA.
Modified nucleotide pdJ is enriched in mini-chromosomes
Mini-chromosomes are 7 to 8 times enriched in telomeric
sequences as compared to total genomic DNA. To investigate
whether the bloodstream form specific pdJ is associated with
telomeres, we compared levels of pdJ found in minichromosomes with those obtained with total genomic DNA.
The results in Table I show that pdJ is 6 times enriched in
mini-chromosomal DNA, suggesting that pdJ has a telomeric
location. Since mini-chromosomes are not only enriched in
telomeres, but also in 177-bp Alul repeats (29), another
explanation for the enrichment of pdJ could be that these repeats
are highly modified. We therefore tested whether pdJ was
associated with these repeats. T. brucei DNA was digested with
Alul and the resulting 177-bp fragments (identified by
hybridization with cloned repeats) were isolated from a
preparative 1,5% agarose gel. The levels of pdJ found in these
purified 177-bp repeats were equal to those in total DNA (not
shown). The low levels of modification in these repeats is
probably due to modification of telomere-proximal repeats and/or
Nucleic Acids Research, Vol. 19, No. 8 1749
Table O. Modified bases of which the corresponding nucleotides were investigated
for comigration with pdJ and pdV via postlabeling and 2D-TLC analysis of the
DNA types specified in the second column
Modified base
DNA sources) used
5-methylcytosine
calf thyxnus
rat white blood cells
E.coli plasmid
E.coli plasmid
E.coli phages T2, T4
B.subtUis phage HI
hydroxymethylcytidine 3'-monophosphate,
chemically synthesized (see methods)
E.coli phages T2, T4
deoxyuridine 3'-monophosphate
P.addovirans phage <
B.subtilis phage SP15
E.coli phage Mu
N*-methyladenine
5-hydroxymethyluracil
5-hydroxymethylcytosine
(non, mono or diglucosylated)
uracil
a-putrescinylthymine
5-dihydroxypentyluracil
N*-carbamoylmethyladenine
contamination with telomere-derived fragments. Clearly,
however, modification of 177-bp repeats cannot account for the
enrichment of pdJ in mini-chromosomes. We therefore conclude
that the enrichment of pdJ in mini-chromosomes indicates that
pdJ is associated with telomeres. The presence of pdJ in large
chromosomal DNA (table I) shows that pdJ is not restricted to
mini-chromosomes only.
Nucleotide pdV was detected in total DNA of both bloodstream
form and procyclic culture form trypanosomes but not quantified.
However, its amount was estimated from four independent
experiments with bloodstream form total DNA. In all four
experiments the amount of pdV was between half and equal that
of pdJ. Therefore, pdV appears to be somewhat enriched in minichromosomal DNA.
Nucleotide pdJ is present in other species exhibiting antigenk
variation
Postlabeling and 2D-TLC analyses were performed on DNA of
all species belonging to the trypanozoon group. pdJ was detected
in all these species, i.e. T.brucei brucei, T.brucei gambiense,
T.brucei rhodesiense, T.equiperdum and T.evansi. In Crithidia
fasciadata, lacking antigenic variation, pdJ was not detectable.
This indicates that pdJ is either specific for species having
antigenic variation, or that our analysis is not sensitive enough
to detect pdJ in a species without mini-chromosomes, like
C.fasciadata (21). We have deferred a more detailed study of
pdJ in other Kinetoplastida until more sensitive and quantitative
analytical procedures for pdJ will be available.
pdJ may be a novel pyrimidine nucleotide
We tried to identify the structure of pdJ and pdV by comparing
their chromatographic properties with those of modified
nucleotides previously identified in DNA from a variety of
sources (30—32). Table n gives a schematic overview of the
modified bases investigated. In eukaryotic DNAs three unusual
bases have been found thusfar (33), 5-methylcytosine,
N^methyladenine and hydroxymethyluracil. In ^P-postlabeling
analyses of calf thymus, E.coli and phage HI DNAs, we found
the chromatographic behavior of p5mdC, pNfodA and
pHOMedU to be different from that of pdJ (not shown).
Therefore we conclude that J is a novel eukaryotic modified base.
To test whether J is identical to one of the other unusual bases
listed in table n , the corresponding DNAs were analysed as
described above. With the exception of <£W14 DNA, each DNA
yielded a specific extra spot, but none of these co-migrated with
TFA
Hydrazfn*
Figure 3. Modified base J has the chemical properties of a pyrimidine nudeotide.
The autoradiograms show 32P-labded digests of variant 118a mini-chromosomal
DNA chromatographed as described under B in methods, except that ammonium
formate with a pH of 3.2 instead of 3.5 was used, causing slightly less migration
of pdJ in Dl. Digests were either untreated (left panel), or treated with
trifluoroacetic acid (TFA; middle panel), or treated with hydrazine (right panel)
according to methods. The arrow with a circle marks pdJ. For the explanation
of other spots see fig.2B. The spot below pdT is probably deoxyinosine
5'-monophosphate, resulting from deamination of pdA.
H1
Bloodstream form + H1
Figure 4. Comparison of pdV with pHOMedU in 2D-TLC. The autoradiograms
show two-dimensional separations of MP-labeled digests ot Badllus subtihs phage
HI DNA (left panel) and of a 200:1 mixture of variant 221a and phage HI
respectively (right panel). TLC conditions were as described under A in methods.
The arrowhead and the arrow indicate pHOMedU and pdV respectively. For the
explanation of other spots see fig. 2A.
pdJ. Only 5'-a-putrescinyl dTMP was not detected, but this
nucleotide is known to be resistant to several enzymes (34) and
therefore probably not released from the DNA by the enzymes
used, unlike pdJ. pdJ may therefore be different from all known
nucleotides.
To analyse whether J is a purine or a pyrimidine base, a
hydrolysate made from bloodstream form variant 118a minichromosomal DNA was treated with either trifluoroacetic acid
(TFA) or hydrazine prior to labeling of the nucleotides. 2D-TLC
separations of the resulting nucleotides are shown in fig. 3.
Treatment of dNp's with TFA leads to hydrolysis of the Nglycosidic bonds of purine deoxyribonucleotides, but not of
pyrimidine deoxyribonucleotides (35). The middle panel shows
that in the TFA treated sample dJp, like dCp and dTp, but unlike
dAp and dGp, has been labeled as efficiently as in the untreated
sample (left panel). In contrast, hydrazine treatment of the dNp's
prior to labeling, selectively destroys the pyrimidine rings in dCp
and, to a lower extent, in dTp (36). Therefore, only purine
deoxyribonucleotides and part of dTp will subsequently be
labeled. The absence of pdJ in the right panel indicates that pdJ,
like dCp and part of dTp, has been destroyed by hydrazine,
whereas dAp and dGp are unaffected. The combination of
resistance to TFA treatment and sensitivity to hydrazinolysis
shows that dJp has the chemical properties of a pyrimidine
deoxyribonucleotide.
1750 Nucleic Acids Research, Vol. 19, No. 8
pdV co-migrates in two chromatographic systems with
hydroxymethyldeoxyuridme 5'-monophosphate (pHOMedU)
Fig. 4, left panel, shows a 2D-TLC separation (conditions A)
of postlabeled nucleotides derived from phage HI DNA, in which
100% of thymine is replaced by hydroxymethyluracil (16).
Comparison of this nucleotide map with that of T. brucei DNA
(fig.2A) shows that the position of pHOMedU is similar to that
of pdV. This is also the case when chromatographic conditions
B were used (not shown). Comigration of pHOMedU and pdV
was confirmed in a mixing experiment (fig.4 right panel).
Hydrolysates of both phage HI DNA and bloodstream form
variant 221a mini-chromosomal DNA were postlabeled and a
mixture was made with equal amounts of pHOMedU and pdV.
This mixture was analysed on TLC sheets under conditions A.
Fig. 4 right panel shows that pHOMedU and pdV comigrate.
DISCUSSION
We have detected two unusual nucleotides in DNA of
bloodstream form trypanosomes, using the sensitive 32 Ppostlabeling technique developed by Randerath et al. (9,22). One
of these nucleotides, pdJ, has the characteristics expected of the
putative unusual nucleotide responsible for the partial cutting of
PstI and PvuII sites in and around silent telomeric VSG genes:
pdJ is present in bloodstream form DNA in low amounts; it is
not detectable in procyclic trypanosomes that lack partial digestion
(ref.3 and fig. IB); it is enriched in mini-chromosomes that contain
most of the silent telomeric VSG genes in the nucleus (21); and
it is not enriched in the 177-bp satellite DNA, also mainly present
in mini-chromosomes (21). So far, we have found pdJ in all
species belonging to the trypanozoon group, all known to undergo
antigenic variation, but not in Crithidia fasciculata. It has yet
to be established whether pdJ is indeed restricted to those
Kinetoplastida exhibiting antigenic variation. pdJ was not detected
in DNA from rodent, bovine and bacterial sources, nor from
Caenorhabditis elegans, Drosophila melanogaster, and all phages
mentioned in table II (unpublished observations). pdJ cannot be
derived from contaminating kinetoplast DNA QcDNA), since
purified kDNA (37) contained no detectable pdJ (not shown).
pdJ does not co-migrate with any of the three known unusual
nucleotides in eukaryotic DNA (33), 5-MedCMP,
N6-MedAMP, and 5-HOMedUMP. It also did not co-migrate
with any of the bacteriophage unusual nucleotides available to
us. As pdJ is only detected in some DNA sources and not in
others it cannot be an artifact of the method of detection. It is
also highly unlikely that pdJ is not a DNA nucleotide. The
polynucleotide kinase labeling procedure used is highly specific
for 3' nucleotides and we have not detected pdJ in trypanosome
RNA. Moreover, we have recently succeeded in obtaining pure
dJp by chromatography on two subsequent HPLC columns. This
dJp has a typical nucleotide absorbance spectrum excluding the
possibility that pdJ is derived from an unusual contaminant. We
therefore conclude that pdJ is a novel unusual nucleotide in
eukaryotic DNA, possibly even a novel DNA nucleotide
altogether. Treatment with TFA and hydrazine shows that dip
is likely to be a pyrimidine nucleotide.
As several modified nucleotides are partially resistant to the
type of enzymes used in the postlabeling procedure (34,38), the
levels of pdJ presented here may be underestimated. In enzyme
titration studies we indeed found that pdJp reacts less effectively
with nuclease PI than the standard deoxyribonucleoside
5',3'-biphosphates (pdNp's). We have also obtained at least twofold higher yields of column-purified dJp than expected on the
basis of 2D-TLC of post-labeled nucleotides (unpublished results).
The pdJ levels listed in Table I are therefore minimal estimates.
Besides pdJ, we have detected a second nucleotide, pdV, in
trypanosome DNA hydrolysates. pdV comigrates with
5-HOMedUMP in two different chromatographic systems and
we conclude that it is identical to this unusual nucleotide. Low
levels of pHOMedU have been found in DNA samples of many
sources and it is a known product of oxidative damage of DNA
(39). The pHOMedU we find both in procyclic DNA and in
bloodstream form DNA is possibly also the result of DNA
damage, in vivo or after isolation. An explanation for the
increased level of pHOMedU in bloodstream form DNA (table
I) could be that pdJ is unstable and yields pHOMedU on
degradation. We have not observed an inverse relation between
pdJ and pHOMedU levels, however. Another possibility is that
pHOMedU is the precursor of pdJ at the polynucleotide level,
as HOMeU is the precursor of some thymine analogues (30).
In higher eukaryotes the modified DNA base 5-methylcytosine
is involved in the regulation of gene expression (reviewed in
40—42). Although we have found a nucleotide co-migrating with
5-MedCMP in preliminary experiments (unpublished
observations), it was present in both bloodstream form and
procyclic DNA, levels were low (<0.03%) and equal in both
life cycle stages and not higher in mini-chromosomal DNA.
Therefore, unlike pdJ, this nucleotide does not correlate with the
partial digestion of inactive bloodstream form telomeric VSG
genes.
Even though pdJ has the testable properties of the unusual
nucleotide postulated to prevent cleavage of PstI and PvuII sites,
this association does not prove identity. This proof requires
identification of the chemical nature of pdJ and further
experiments based on this identity, i.e. specific chemical
modification of DNA and experiments with antibodies against
pdJ. Attempts to establish the structure of purified dJp by mass
spectrometry are in progress.
ACKNOWLEDGEMENTS
We thank Joost Zomerdijk, Paul Shiels, Carsten Lincke, Jaap
Smit, Lisette Eijdems, Alfred Schinkel, Ronald Plasterk and Leo
den Engelse for critical reading of the manuscript, Pirn van Dijk
for sharing unpublished results, and Francesca Fase-Fowler,
Dr.P. Van de Putte (Dept. of Molecular Genetics, Univ. of
Leiden, Leiden, The Netherlands), Dr.R. Warren (Dept. of
Microbiology, Univ. of British Columbia, Vancouver, Canada),
Dr.M. Ehrlich (Dept. of Biochemistry, Tulane University
Medical Center, New Orleans, Louisiana, USA) and Dr.S. Bron
(Dept. of Genetics, Univ. of Groningen, Haren, The Netherlands)
for kindly providing phage DNAs or bacteriophages. This work
was supported by grants from the Netherlands Foundation for
Chemical Research (SON), with financial support of the
Netherlands Organization for Scientific Research (NWO), and
from the Dutch Cancer Foundation (J.L.).
REFERENCES
1. Boret.P. (1986) Ann. Rev. Biochcm., 55, 701-732.
2. Bernards.A., Van Haiten-Loosbroek.N. and Borst.P. (1984) Nucleic Acids
Res., 12, 4153-4170.
3. Pays,E., Delauw.M.F., Laurent.M. and Steinert.M. (1984) Nucleic Acids
Res., 12, 5235-5247.
Nucleic Acids Research, Vol. 19, No. 8 1751
4. Van dcr Plocg.L.H.T., LiuAY.C. and Borst.P. (1984) Cell, 36, 459-468.
5. Blackburn.E.H. and Challoner.P.B. (1984) Cell, 36, 447-457.
6. BernardsA, Michels.P.A.M., Uncke.C.R. and Borst.P. (1983) Nature,
303, 592-597.
7. Pays.E., Laurent.M. DelinteJC., Van Meirvenne.N. and Steinert.M. (1983)
Nucleic Acids Res. 11, 8137-8147.
8. Crozaner.M., DeBrij,R-J., DenEngelse.L., Johnson,PJ. andBorst,P. (1988)
Mol. Biochem. Parasitol., 31, 127-132.
9. Reddy.M.V., Gupta.R.C. and Randerath.K. (1981) Anal. Biochem., 117,
271-279.
10. Cross.G.A.M. (1975) Parasitology, 71, 393-418.
11. Michels.P.A.M., LiuAY.C., BemardsA, Sloof.P., Vander Bijl.M.M.W.,
SchinkeLA.H., Menke.H.H., Borst.P., Veeneman.G.H., Tromp.M.C. and
Van BoomJ.H. (1983) J. Mol. Biol., 166, 537-556.
12. Michels.P.A.M., Van der Ploeg.L.H.T., Liu.A.Y.C. and Borst.P. (1984)
EMBOJ., 3, 1345-1351.
13. Fairlamb.A.H., Weislogel.P.O., HoeijmakersJ.H.J. and Borst.P. (1978)
J. CeU. Biol., 76, 293-309.
14. Brun,R. and Schdnenberger.M. (1979) Acta Trop., 36, 289-292.
15. Bemards.A., Van der Ploeg.L.H.T., Carlos.A. Frasch.C. Borst.P.,
BoothroydJ.C, Coleman.S. and Cross.G.A.M. (1981) CeU, 27,497-505.
16. Arwert.F. and Venema.G. (1974) Molec. gen. Genet. 128, 55-72.
17. MnnintU T Fritsch.E.F. and Sambrook^l. (1982) Molecular Cloning: a
Laboratory Manual.- Cold Spring Harbor Laboratory Press, Cold Spring
Harbor, N.Y.
18. De Lange.T., KooterJ.M., Michels.P. and Borst.P. (1983) Nucleic Acids
Res., 11, 8149-8165.
19. Rigby.P.WJ., Dieckman.M., Rhodes.C. and Berg.P. (1977) J. Mol. Biol.,
113,237-251.
20. Jefrreys.A.J. and Flavell.R.A. (1977) Cell, 12, 429-439.
21. Van der Ploeg.L.H.T., Schwartz.D.C, Cantor.C.R. and Borst.P. (1984)
Cell, 37, 77-84.
22. Gupta.R.C., Reddy.M.V. and Randerath.K. (1982) Carcinogenesis, 3,
1081-1092.
23. Dawid.I.B., Brown.D.D. and Reeder.R.H. (1970) J. Mol. Biol., 51,
341-360.
24. Markham.R. and SmithJ.D. (1951) Biochem. J., 49, 401-406.
25. Alegria.A.H. (1967) BBA, 149, 317-324.
26. GottesdienerJC., Garcia-AnoverosJ., Lee,M.G-S. and Van der Ploeg.L.H.T.
(1990) Mol. Cell. Biol., 10, 6079-6083.
27. Cully,D.F., Ip,H.S. and Cross.G.A.M. (1985) Cell, 42, 173-182.
28. ZomerdiM.C.B.M., Ouelette.M., Ten Asbroek.A.L.M.A., Kieft.R.,
Bommer.A.M.M., Clayton.C.E. and Borst,P. (1990) EMBO J. 9,
2791-2801.
29. Sloof.P., Menke.H.H., Caspers.M.P.M. and Borst.P. (1983) Nucleic Acids
Res., 11, 3889-3901.
30. Warren.R.A.J. (1980) Ann. Rev. Microbiol., 34, 137-158.
31. Swinton.D., Hattman.S., Crain.PP.F., Cheng,C-S., Smith.D.L. and
McCloskeyJ.A. (1983) Proc. Nad. Acad. Sci. USA, 80, 7400-7404.
32. Hattman.S. (1970) Virology, 42, 359-367.
33. Rae.P.M.M., and Steele.R.E. (1978) Biosystems, 10, 37-53.
34. Kropinski.A.M.B., Bose.RJ. and Warren.R.A.J. (1973) Biochemistry, 12,
151-157.
35. LutgerinkJ.L., De GraafA and Westra.G. (1991) Manuscript in preparation.
36. TemperliA, TQrler.H., Rust.P., Danon.A. and Chargaff.E. (1964) BBA,
91, 462-476.
37. Borst.P. and Hoeijmakers.J.HJ. (1979) Plasmid, 2, 2 0 - 4 0 .
38. MarmurJ., Brandon.C, Neubort,S., Eluiich.M., Mandel.M., Konvkka^J.
(1972) Nature New Biol., 239, 68-70.
39. Freakel.K., CummingsA, SolomonJ., Cadet^l., Steinberg^.J. and
Teebor.G.W. (1985) Biochemistry, 24, 4527-4533.
40. Bird.A.P. (1986) Nature, 321, 209-213.
41. Cedar.H.W. (1988) Cell, 53, 3 - 4 .
42. Dynan/W.S. (1989) Trends Genet., 5, 35-36.