Download The trimethoprim-resistant dihydrofolate reductase associated with

volume 9 Number 31981
Nucleic A c i d s Research
Characterization of a R plasmid-associated, trimethoprim-resistant dihydrofolate reductase and
determination of the nucleotide sequence of the reductase gene
J.Wemer Zolg1 and Urs J.Hanggi
Institut fiir Physiologische Chemie, Physikalische Biochemie und Zellbiologie der Univeisitat,
Goethestr. 33, 8000 Miinchen 2, GFR
Received 17 November 1980
ABSTRACT
The trimethoprim-resistant dihydrofolate reductase associated
with the R plasmid R388 was isolated from strains that overproduce the enzyme. It was purified to apparent homogeneity by
affinity chromatography and two consecutive gel filtration
steps under native and denaturing conditions. The purified
enzyme is composed of four identical subunits with molecular
weights of 8300. A 11OO bp long DNA segment which confers
resistance to trimethoprim was sequenced. The structural gene was
identified on the plasmid DNA by comparing the amino acid composition of the deduced proteins with that of the purified enzyme.
The gene is 234 bp long and codes for 78 amino acids. No homology can be found between the deduced amino acid sequence of the
R388 dihydrofolate reductase and those of other prokaryotic or
eukaryotic dihydrofolate reductases. However, it differs in only
17 positions from the enzyme associated with the trimethoprimresistance plasmid R67.
INTRODUCTION
Trimethoprim prevents the bacterial production of tetrahydrofolate coenzymes by inhibiting the enzyme dihydrofolate reductase (2). The resulting deficiency of tetrahydrofolates leads to
impaired protein, RNA, and DNA synthesis and eventually to cell
death (3). The production of tetrahydrofolates is not inhibited
in bacteria which harbor trimethoprim-resistance plasmids. These
cells contain additional dihydrofolate reductases that are insensitive to the drug (4-6). Gene dosage experiments and investigations on the production of the plasmid-associated enzymes in
minicells have shown that they are encoded on the plasmid DNA
(7-9) .
Several trimethoprim-resistance plasmids have been mapped by
in vitro recombinant DNA techniques (7-11). These studies have
shown that the genes for the resistant dihydrofolate reductases
V) IRL Press Limited. 1 Falconberg Court, London W 1 V 5FG. U.K.
697
Nucleic Acids Research
of the plasmids R483 and R67 reside on DNA fragments of about
41OO bp and 24OO bp length (9) and that of the plasrnid R388 on a
DNA segment of less than 1200 bp length (7). In the present
study the nucleotide sequence of the latter DNA segment was
determined. The resulting amino acid sequences were compared to
the amino acid composition of the isolated enzyme. The comparison
made it possible to identify the structural gene on the DNA segment, to analyze the organization of the regulatory elements, and
to deduce the amino acid sequence of the enzyme.
MATERIALS AND METHODS
Chemicals and enzymes. Acrylamide, 2 times crystallized,was
obtained from Serva, Heidelberg. Piperidine, dimethylsulfate,
and guanidinium chloride were from E. Merck, Darmstadt, hydrazine from Roth, Karlsruhe, lyophilyzed alkaline phosphatase
(calf intestine), T4 polynucleotide kinase, and NADPH from
Boehringer, Mannheim, and 3-(4,5-dimethylthiazolyl-2-)-2,5-diphenyltetrazolium bromide from Sigma, Miinchen. Matrex Gel Red A
was purchased from Amicon, Witten. All other chemicals used were
as described (7) . Dihydrofolate was prepared according to (12) .
Purification of dihydrofolate reductase. E.coli C carrying
PWZ82O or pWZ7O3 (7) was grown in a fermenter in L-broth to a
density of 2-5.10 cells/ml. About 6O g of cells were broken up
in a French-press in 2OO-3OO ml of 1O mM Tris-HCl, pH 8.0 containing 1 mM mercaptoethanol and 1 mM EDTA (TME-8 buffer). After
high speed centrifugation (45 min at 75OOO x g, O C) solid
ammonium sulfate was added first to 2 M, then to 3.6 M. The 2 M
precipitate was discarded. The 3.6 M ammonium sulfate precipitate
was dissolved in 100-200 ml of TME-8 and applied to a Sephadex
G-7 5 column (10x60 cm) equilibrated with TME-8. 20 ml fractions
were collected. The fractions containing the trimethoprimresistant dihydrofolate reductase were pooled and precipitated
by adding ammonium sulfate to 3.6 M, the pellet was dissolved in
50 ml of 10 mM Tris-HCl, pH 9.0 containing 1 mM mercaptoethanol,
1 mM EDTA, and 1O mM NaCl (TMENa-9 buffer), dialyzed against 3
times 1 1 of TMENa-9, and applied to a Matrex Gel Red A column
(2.5x30 cm) equilibrated with TMENa-9. The column was eluted
with the same buffer without using a gradient. The two trimetho-
698
Nucleic Acids Research
prim resistant activity peaks (see Results) were pooled and concentrated to 10 ml. 7.6 g of guanidinium chloride and 1O ul of
concentrated mercaptoethanol were added, and the mixture was held
at 100 C for 5-10 min. After cooling it was applied to a Sephadex
G-150 column (3.5x30 cm) which previously was equilibrated with
10 fold concentrated TMENa-9 containing 6 M urea. Elution was
with the same buffer mixture. Fractions of 5 ml were collected.
The fractions containing dihydrofolate reductase were freed of
urea by gel filtration in Sephadex G-2 5 (3.5x30 cm) equilibrated
with 1O mM Na-phosphate, pH 6.8 containing 1 mM mercaptoethanol.
The enzyme pool was concentrated to 0.5 ml by flash evaporation
and rechromatographed on a small column (1.5x15 cm) of Sephadex
G-2 5 in the phosphate buffer mentioned above. The most active
fractions were pooled and stored at -20 C.
Dihydrofolate reductase activity was assayed according to
Ref. 13. 2 nM trimethoprim was included in the assay mixture whenever the trimethoprim-resistant activity had to be distinguished
from the trimethoprim-sensitive E.coli reductase activity.
Molecular weight determinations. The subunit molecular
weight of the purified R388 dihydrofolate reductase was determined by electrophoresis on 7% and 10% polyacrylamide gels containing 0.2% SDS (14). The gels were calibrated with insulin
chain B (Mr 34OO), aprotinin (Mr 6500), cytochrorae C (Mr 12 5O0),
and soybean trypsin inhibitor (M 215OO) (Combithek I, Boehringer).
The molecular weight of the purified native enzyme was determined
by gel filtration in Sephadex G-200 superfine. The column (1x100
cm) was equilibrated with 1O mM Tris-HCl, pH 8.O, containing
10 mM mercaptoethanol and 1 mM EDTA. It was calibrated with
bovine serum albumin (M 68OOO), ovalbumin (M 4500O), chymotrypsinogen A (M 25000), and cytochrome C (M 12500) (Combithek
II, Boehringer).
Gel electrophoresis. Non-denaturing polyacrylamide gels were
prepared and stained for proteins or dihydrofolate reductase activity according to Ref. 15. The conditions used for separating DNA
fragments on agarose gels were as described (7). DNA fragments
were isolated from agarose by the glass powder adsorption method
(16) . The sizes of the fragments were determined by coelectrophoresis with pBR322 DNA cleaved with either Sau3A or Hpall•
699
Nucleic Acids Research
Restriction enzyme analysis. DNA was isolated and cleaved
with restriction enzymes as previously described (7) . The
restriction nucleases BamHl, EcoRl, Sau3A and Haelll were kindly
donated by T. Igo-Kemenes. Alul, Pstl, Hhal, Hinfl, Sau96l, and
TagI were gifts of R.E. Streeck. Hinfl was prepared by chromatography on DEAE-cellulose and heparin-agarose, Sau96l and TagI
according to Ref. 17. The latter enzyme was used in a buffer
containing 6 mM Tris, pH 7.4, 6 mM MgCl2» and 0.6 mM dithiothreitol at 50 C. Hpall was prepared by H. Feldmann, and Hindu
was obtained from Boehringer, Mannheim.
Terminal labelling and seguence analysis. 2O-5O ug of DNA
weBB cleaved with restriction nucleases and the resulting mixture of fragments was treated with phosphatase prior to the
separation of the fragments by electrophoresis in polyacrylamide
gels. The fragments were recovered from the gels by adding 2
volumes of 1 M NaCl to the gel slices and by passing the gelNaCl mixture through a syringe. The resulting paste was incubated overnight at 37 c and centrifuged. The supernatant was
filtered through Whatman 3MM paper. The isolated fragments were
denatured, labelled at their 5' ends with polynucleotide kinase,
renatured, and cleaved according to Maxam and Gilbert (18) with
minor modifications (19). The reactions specific for C, C+T,
A > C, and G were performed according to Ref. 18, and that for
A+G according to Ref. 20. Separation of the partial fragments
was on 8% or 2O% polyacrylamide gels in 8.3 M urea. Autoradiography was at -70 C with and without intensifying screens
(SE6, Cawo, Schrobenhausen). Methylated cytosines were detected
according to Ref. 21.
RESULTS
Purification and characterization of the R388 dihydrofolate
reductase. In a previous study we had found that dihydrofolate
reductase is stable under denaturing conditions and that the
level of the enzyme is considerably increased in bacteria which
carry multiple copies of the reductase gene (7). These two observations were used as part of a novel purification scheme.
Starting from cell extracts of bacteria with increased dihydrofolate levels the enzyme was partially purified by ultracentri-
700
Nucleic Acids Research
fugation, ammonium sulfate precipitation and gel filtration on
Sephadex G-7 5 as previously described (7) . Further purification
was achieved by chromatography on triacyl dye agarose (Procion
Red HE 3B, Amicon Matrex Gel Red A) which selectively adsorbs
NADP -dependent enzymes (22). Preliminary experiments had shown
that the R388 dihydrofolate reductase was retained by the gel.
In order to purify the enzyme on a preparative scale, pH and
buffer compositions were chosen which allowed the separation of
the bulk of the non-dihydrofolate reductase proteins from the
enzyme without using a salt gradient (Fig. 1A) . 1.5 U of enzyme
were retained per ml of affinity gel. Chromosomal dihydrofolate
reductase is strongly absorbed and is only eluted at high salt
concentrationsThe R388 dihydrofolate reductase displayed a two-peak elution pattern on the Procion Red affinity column (Fig. 1A). However, since no difference between the two fractions could be
detected on SDS gels, both were combined. To remove remaining
05
-
SO
*
•
_
•
400
1/
•
02
>
01
E
i
:
100
500
U/ml
1
\
/
/
" 11
1
/
B
E
<
<
-
300
200
ACIiv • y
a
10
t]
-
too
..••••••
*
A
"*—
IS
E
0
l»>
200
Elulion v o l u m e , ml
2S0
100
SO
100
150
200
250
300
Etulion volume,ml
Figure 1. (A). Affinity chromatography of R388 dihydrofolate
reductase. The concentrated and dialyzed Sephadex G-7 5 pool was
chromatographed on Procion Red agarose as described in Materials
and Methods. Aliquots of the eluate were assayed for absorbance
at 280 nm ( A — A ) and for activity (•--•). The bar indicates
the fractions which were pooled. (B).Gel filtration under denaturing conditions. The enzyme fractions of the affinity gel
were denatured and chromatographed on Sephadex G-150 in the
presence of 6 M urea as described in Materials and Methods.
Aliquots of 1-1O nl of the eluate were assayed for activity
without removing the urea.
701
Nucleic Acids Research
non-reductase contaminants, which in their native conformation
had copurified with the native R388 dihydrofolate reductase on
the first Sephadex G-7 5 column, the pooled enzyme fractions were
denatured by boiling in 6 M guanidinium chloride and chromatagraphed in Sephadex G-150 in the presence of 6 M urea (Fig. IB) .
The R388 dihydrofolate reductase disaggregated under these
conditions and was well separated from the larger, non-dihydrofolate reductase proteins. The fractions containing dihydrofolate reductase were freed of urea by gel filtration on Sephadex G-5O, concentrated, and rechromatographed on a small column
of Sephadex G-50. The resulting enzyme was more than 95% pure as
judged by electrophoresis on SDS gels and the overall yield was
between 15-2o% in the different preparations. The specific activity was about 1.5 U/mg of protein (Table 1 ) .
The molecular weight of the purified, catalytically active
enzyme was estimated by gel filtration on Sephadex G-2OO. R388
dihydrofolate reductase eluted as a single protein and activity
peak in a position corresponding to a molecular weight of about
3600O. A single protein band was also seen on SDS gels. However, its molecular weight was only about 8400. Since the band
showed weak dihydrofolate reductase activity after prolonged
staining for the enzyme, it was concluded that the native enzyme
is composed of four identical subunits with molecular weights of
8400. A similar subunit structure has been found for the R67 dihydrofolate reductase (23).
Table 1. Purification of R388 dihydrofolate reductase
Fraction
Volume
Protein
Acticoncentration vity
Total
activity
Specific
activity
ml
Crude extract
3OO
(NH4)2SO4 precipitate 17O
Sephadex G-7 5 pool 26O
Procion Red agarose 1OO
Sephadex G-5O
5
702
10
0.77
0.42
0.93
157
430
136O
Nucleic Acids Research
The amino acid composition of the purified enzyme is shown
in Table 2. The calculated molecular weight (8O2O) is in close
agreement with the value determined from SDS gels.
Identification of the coding sequence of the dihydrofolate
reductase gene. Our previous analysis of the trimethoprim-resistance gene of R388 has shown that the plasmid-induced dihydrofolate reductase is encoded on a DNA segment of less than
12OO bp length (7) . Further cloning experiments (not shown)
suggested that the entire segment might be essential for proper
expression of the gene. Therefore the DNA was cleaved with
various restriction nucleases and sequenced by the Maxam and
Gilbert technique. The restriction map and the sequence strategy
Table 2. Amino acid composition of R388 dihydrofolate reductase.
About O.6 nmoles of purified enzyme were hydrolyzed for 30 min
at 160 in 5.7 N HCl as described by Wachter and Werhahn (24) .
Amino acid
Cys
Asx
Asp
Asn
Met
Thr
c)
d)
Ser
Glx
nmoles found/nmol of
enzyme a '
suggested theoretical
content °>
content
1
O.98
4.O6
—
O.32
2.98
6.55
10.02
-
Glu
Gin
Pro
Gly
Ala
Val
He
Leu
Tyr
Phe
His
Lys
Arg
e
Trp '
2
1
—
2
1
1
3
7
5
6
5
8
11
8
1
5
3
2
1
1
3
3
2
3
3
2
4
—
1
3
7
10
4
9
11
6
1
5
4.06
8.9O
1O.68
6.21
1.19
5.O7
2.38
1.93
l.OO
3.23
2.56
1.58
3
M : 8O2O
M ; 8266
r
a) mean of two determinations; b) deduced from the nucleotide
sequence shown in Fig. 3 and 4; c) Cystein was determined as
cysteic acid; d) Methionine was determined as methionine sulfone;
e) Tryptophan was determined according to Beaven and Holiday (2 5)
r
703
Nucleic Acids Research
are shown in Fig. 2. The sequence, 112 5 nucleotides long,
was determined, about 80% of which was confirmed by sequencing
both DNA strands. The whole nucleotide sequence is shown in
Fig. 3. The GC-content is 60% between bases 1 and 400, 50%
between bases 400 and 500, and 63% in the remaining part of
the DNA. The overall GC-content is 56%. Three methylated cytosines were found at positions 234, 333, and 959.
Eight open reading frames could be detected by locating the
positions of the start and stop codons in the nucleotide sequence.
Three have no stop codons within the sequenced part. The lengths
of the proteins which are encoded in the open reading frames
vary between 43 and 177 amino acids or more. In order to find
the coding sequence of the dihydrofolate reductase gene, the
amino acid compositions of all proteins that could be derived
from the nucleotide sequence were compared to the composition
of the isolated enzyme. Only a single protein had the composition,
the length, and the characteristic His/Ile/Met/Cys (1:1:1:1),
Leu/Ile (5:1), and Tyr/Trp (3:2) ratios of the isolated enzyme
(cf. Table 2). Therefore it was concluded that this protein is
the dihydrofolate reductase and that the corresponding nucleo-
I — i w ~ V If i f—I "WV\ >t i n n
Figure 2. Restriction map of the DNA segment carrying
the R388 dihydrofolate reductase gene and strategy
used for determining the nucleotide sequence. The
numbers refer to the cleavage sites on the DNA sequence
shown in Fig. 3. The directions in which the DNA fragments were sequenced and the distances are indicated by
arrows. The reductase structural gene is delineated by
the bar.
704
Nucleic Acids Research
GGTCCCGATC CTTGGAGCCC TTGCCCTCCC GCACGATGAT CGTGCCGTGA
TCGAAATCCA GATCCTTGAC CCGCAGTTGC AAACCCTCAC TGATCCGCAT
100
GCCCGTTCCA TACAGAAGCT GGGCGAACAA ACGATGCTCG CCTTCCAGAA
AACCGAGGAT GCGAACCACT TCATCCGGGG TCAGCACCAC CGGCAAGCGC
200
CGCGACGGCC GAGGTCTTCC GATCTCCTGA AGCCAGGGCA GATCCGTGCA
CAGCACCTTG CCGTAGAAGA ACAGCAAGGC CGCCAATGCC TGACGATGCG
300
TGGAGACCGA AACCTTGCGC TCGTTCGCCA GCCAGGACAG AAATGCCTCG
ACTTCGCTGC TGCCCAAGGT TGCCGGGTGA CGCACACCGT GGAAACGGAT
400
GAAGGCACGA ACCCAGTTGA CATAAGCCTG TTCGGTTCGT AAACTGTAAT
GCAAGTAGCG TATGCGCTCA CGCAACTGGT CCAGAACCTT GACCGAACGC
500
AGCGGTGGTA ACGGCGCAGT GGCGGTTTTC ATGGCTTGTT ATGACTGTTT
TTTTGTACAG TCTAGCCTCG GGCATCCAAG CTAGCTAAGC GCGTTACGCC
600
GTGGGTCGAT GTTTGATGTT ATGGAACAGC AACGATGTTA CGCAGCAGGG
TAGTCGCCCT AAAACAAAGT TAGGCAGCCG TTGTGCTGGT GCTTTCTAGT
700
AGTTGTTGTG GGGTAGGCAG TCAGAGTTCG ATTTGCTTGT CGCCATAATA
GATTCACAAG AAGGATTCGA CATGGGTCAA AGTAGCGATG AAGCCAACGC
800
TCCCGTTGCA GGGCAGTTTG CGCTTCCCCT GAGTGCCACC TTTGGCTTAG
GGGATCGCGT ACGCAAGAAA TCTGGTGCCG CTTGGCAGGG TCAAGTC6TC
900
GGTTGGTATT GCACAAAACT CACTCCTGAA GGCTATGCGG TCGAGTCCGA
ATCCCACCCA GGCTCAGTGC AAATTTATCC TGTGGCTGCA CTTGAACGTG
1 000
TGGCCTAACA ATTCGCTCAG GACGGTTCGC CGCCCGCTGA GCTTTGTCGT
TAGGCGTCAT GGGCTGACGC TCACAATCGA ACCAAGCGGC GAGAGTCGCG
1100
GGCACTGCGA CTGCTGTGGG AATTC
Figure 3. Nucleotide sequence of the R388 DNA segment carrying
the dihydrofolate reductase gene. The coding sequence for the
R388 dihydrofolate reductase is enclosed.
tide sequence is the structural gene for the enzyme. This
sequence is
enclosed
in Fig. 3.
Comparison of the amino acid sequence of the R388 and R67
dihydrofolate reductases. The complete amino acid sequence deduced from the nucleotide sequence of the R388 dihydrofolate
705
Nucleic Acids Research
reductase gene is shown in Fig. 4. Extensive homologies can be
found to the amino acid sequence of the dihydrofolate reductase
isolated from the trimethoprim-resistance plasmid R67 (26). Both
enzymes have exactly the same number of amino acid residues and
61 of the 78 amino acids are identical. The remaining 17 amino
acids, for the most part, are conservative replacements, eight
of which can be explained by transitions of single bases (amino
acid residues 3,6,8,9,17,18, and 21; Fig. 4) and two by transversions of single bases (residues 49 and 61) . At least two substitutions are needed to explain the remaining 7 exchanges
(residues 2,10,15,20,26,77, and 78).
Organization of the dihydrofolate reductase gene. The dihydrofolate reductase gene (cf. Fig. 3) is only 234 bp long. Its
promoter contains the sequence CATAAT (pos. 744-749) and
GTTCGATTT (pos. 726-734) which differ only minimally from the
-10 and -3 5 consensus sequences described by Rosenberg and Court
(27). In analogy to other mRNAs by which the dihydrofolate'
reductase mRNA is expected to start at the A at position 7 56,
seven nucleotides downstream of the ubiquitous T of the -10 hexamer (pos. 749). Between the start of the mRNA and the AUG initia-
>
S
10
>S
20
ATG GTT CAA AGT AGC GAT GAA GCC AAC GCT CCC GTT GCA GGG CAG TTT GCG CTT CCC CTG
Gly
Gin
Asp
Olu
Aim
Asn
Aim
Pro
Vml
Aim
O.V
Olu
Arg
Asn
Glu
Vml
Smr
Am
Pro
Vml
Aim
Gly
25
30
35
<0
AGT GCC ACC TTT GGC TAA GGG GAT CGC GTA CGC AAG AAA TCT GGG GCC GCT TGG CAG GGT
Aim
Thr
Phm
Oly
Oly
Asp
A,3
Vml
Arg
Lys
Lyi
Smr
Oly
Aim
Aim
Tip
Oln
Gly
Aim
Thr
Phm
Oly
Oly
Asp
Arg
Vml
Arg
Lys
Lym
Smr
Oly
Aim
Aim
Tip
Oln
Oly
45
50
55
SO
CAA GTC GTC GGT TGG TAT TGC ACA AAA CTC ACT CCT GAA GGC TAT GCG GTC GAG TCC GAA
ft 388
ft 67
Gin
Vml
oty
Trp
Tyr
Cys
Thr
Lmu
Thr
Pro
Glu
Oly
Tyr
Aim
Vml
Olu
Smr
Olu
Gin
Vml
Oly
Trp
Tyr
Cyt
Thr
Lmu
Thr
Pro
Olu
Oly
Tyr
Aim
Vml
Clu
Smr
Glu
05
70
75
79
TCC CAC CCA GGC TCA GTG CAA ATT TAT CCT GTG GCT GCA CTT GAA CGT GTG GCC TAA
R 388
ft67
Smi
His
Pro
Oly
Smr
Vml
Gin
Urn
T,r
Pro
Vml
Aim
Aim
Lmu
Olu
Arg
Aim
His
Pro
Oly
Smr
Vml
Gin
Urn
Tyr
Pro
Vml
Aim
Aim
Lmu
Glu
Aig
Figure 4. Amino acid sequence of the R388 dihydrofolate reductase. The sequence is compared to that of the R67 enzyme (26).
Identical amino acids are enclosed. The number above the nucleotide triplets refer to the amino acid residues.
706
Nucleic Acids Research
tion codon there is an untranslated region of 16 nucleotides
which contains the sequence AAGGA (pos. 761-765) that could
serve as a ribosome binding site (28) . The mRNA can form a hairpin structure in this region by pairing GGAUUC (763-768) to
GAAGCC (pos. 790-795). The initiation codon thus would be exposed in a non-paired loop.
Two GC-rich sequences with dyad symmetries are found at the
31 side of the gene: the sequence GCGTCA (pos. 1O54-1O59) and
TGACGC (pos. 1O6 5-1O7O) and the sequence AGTCG (pos. 1094-1099)
and CGACT (pos. 11O8-1112). Both inverted repeat structures
resemble the terminator signals (27) and thus might be involved
in termination of mRNA transcription. There are two stop codons
in phase between the end of the coding sequence and those structures (TAA 1OO6-1OO8 and TAG 1O51-1O53) .
DISCUSSION
The amino acid sequence of the R388 dihydrofolate reductase
shown in Fig. 4 bears no apparent homology to the sequences of
other prokaryotic or eukaryotic dihydrofolate reductases (for a
recent compilation see Ref. 29). This lack of homology raises
the intriguing question of whether the R388 and the similarly organized R67 dihydrofolate reductase represent a novel type of
dihydrofolate reductases or if these enzymes are non-reductase
proteins with fortuitous reductase activities. The latter possibility is suggested by the low turnover number which can be deduced from the specific activity of the purified R388 enzyme.
Its value (50-100 moles/min/mole of enzyme) is one to two orders
of magnitude lower than those of other dihydrofolate reductases
(3) . However, Smith and Burchall (30), who have asked the same
question for the R67 enzyme, could detect no other activity
besides that of the dihydrofolate reducing one.
A comparison of the amino acid sequences of the R388 and
the R67 dihydrofolate reductase gives instructive insights into
the organization of the two enzymes. Of the 17 amino acids which
are different, all but four map at the amino- and carboxy termini
of the enzymes (Fig. 4 ) . The sequences between amino acid
residues 22 and 76 are virtually identical. This suggests that
only the central part of the molecule is essential for the acti-
707
Nucleic Acids Research
vity. This is supported by the conspicuous accumulation of
functional groups therein (e.g. Arg-Val-Arg-Lys-Lys-Ser (29-34),
Trp-Tyr-Cys-Thr (45-48), Thr-Pro-Glu-Gly-Tyr (51-55), Glu-SerGlu (58-60), or Ile-Tyr-Pro-Val (68-71)). In this respect it
would be interesting to know if there are kinetic differences
between the R67 and the R388 dihydrofolate reductase since glutamine residue 49 in the R67 dihydrofolate reductase is replaced
by a lysine residue in the R388 enzyme.
The amount of R388 dihydrofolate reductase activity is
rather low in E.coli (4,7). This might be due to a low level of
gene expression. However, the organization of the promoter
region is as in other prokaryotic genes. The nucleotide sequences
of the "Pribnow box" (31), of the RNA polymerase recognition
site (32,33), and of the ribosome binding site (28) have no more
deviations from the prototype sequences than other genes (27).
This suggests that the low amount of enzyme in E.coli is mostly,
if not exclusively due to the low specific activity/of the
enzyme. However, it should be recalled that we never have obtained a trimethoprim-resistant recombinant plasmid that is
smaller than about 12OO bp. This might well indicate that far
more sequence information is needed to express the gene than
the information stored in the putative promoter region and that
the auxilary sequences modulate the rate of gene expression.
The nucleotide sequence at the end of the gene contains two
termination codons in phase. It is conceivable that the first
stop codon is translated in suppressor positive strains. This
would add the sequence X-Gln-Leu-Ala-Gln-Glu-Gly-Ser-Pro-Pro-AlaGlu-Leu-Cys-Arg to the carboxy end of the protein and increase
the molecular weight by about 1600. The recent finding of Fling
and Elwell that the R388 dihydrofolate reductase is somewhat
larger than the R67 enzyme in minicell producing strains (9)
might be explained by this mechanism.
The R388 dihydrofolate reductase gene is unusually small in
size. With a length of less than 24O bp it should be possible
to insert the gene into viral genomes without grossly affecting
the size of the viral DNA. Since the R388 dihydrofolate reductase gene also confers resistance to methotrexate, which is
toxic for fungi (34,35) , plants (36) , and animal cells (2) ,
708
Nucleic Acids Research
recombinants between the dihydrofolate reductase gene and the
viruses could possibly be used as selectable DNA vectors to
transfer foreign genes into the cells of fungi, plants, and
animals.
ACKNOWLEDGEMENTS
We would like to thank Dr. H.G. Zachau for helpful suggestions and discussions during this work. Dr. E. Wachter for performing the amino acid analysis, and Ms. Ch. Graf, Ms. I. Wagner,
and Mr. Neumaier for their excellent technical assistance. This
work was supported in parts by the Deutsche Forschungsgemeinschaft (Forschergruppe Genomorganisation) and the Fonds der Chemischen Industrie. J.W. Zolg was supported by Cusanuswerk,
Bischofliche Studienforderung.
REFERENCES
1. Present address: Institute of Animal Genetics, University of
Edinburgh, West Mains Road, GB Edinburgh EH9 3JN
2. Burchall, J.J. and Hitchings, G.H. (1965) Mol. Pharm. 1,
126-136
3. Blakley, R.L. (1969) The biochemistry of folic acid and
related pteridines, pp. 139-187, North-Hoiland, Amsterdam
4. Amyes, S.G.D. and Smith, J.T. (1974) Biochem. Biophys. Res.
Comm. 58, 412-418
5. Skold, 0. and Widh, A. (1974) J. Biol. Chem. 249, 4324-4325
6. Pattishall, K.H., Acar, J., Burchall, J.J., Goldstein, F.W.,
and Harvey, R.J. (1977) j . Biol. Chem. 2 52, 2319-2323
7. Zolg, J.W., Hanggi, U.J., and Zachau, H.G. (1978) Molec.
Gen. Genet. 164, 15-29
8. Fling, M.E. and Elwell, L.P. (1978) in Genetic Engineering,
Boyer, H.W. and Nicosia, S.» Eds., pp. 173-180, Elsevier/
North-Ho Hand, Amsterdam
9. Fling, M.E. and Elwell, L.P. (198O) J.Bacteriol. 141, 779-785
10. Barth, P.T. and Grinter, N.J. (1977) J.Mol.Biol. 113, 455-474
11. Ward, J.M. and Grinsted, J. (1978) Gene 3, 87-95
12. Blakley, R.L. (I960) Nature 188, 231-232
13. Peter, G., Hanggi, U.J., and Zachau, H.G. (1979) Molec.
Gen. Genet. 175, 333-341
14. Laemmli, U.K. (197O) Nature 227, 68O-685
15. Hanggi, U.J. and Littlefield, J.W. (1974) J. Biol. Chem.
249, 1390-1397
16. Vogelstein, B. and Gillespie, D. (1979) Proc. Natl. Acad.
Sci. USA 76, 615-619
17. Green, P.J., Heyneker, H.L., Bolivar, F., Rodriguez, R.L.,
Betlach, M.C., Covarrubias, A.A., Backman, K., Russel, D.J.,
Tait, R., and Boyer, H.W. (1978) Nucl. Acids Res. 5,
2373-2380
18. Maxam, A.M. and Gilbert, W. (1980) Methods in Enzymol. 65,
499-56O
709
Nucleic Acids Research
19. Pech, M. . Streeck, R.E., and Zachau, H.G. (1979) Cell 18,
883-893
20. Gray, C.P., Sommer, R., Polke, C , Beck, E., and Schaller, H.
(1978) Proc. Natl. Acad. Sci. USA 75, 5O-53
21. Ohmori, H., Tomizawa, J., and Maxam, A.M. (1978) Nucl. Acids
Res. 5, 1479-1485
22. Watson, D.H., Harvey, M.J., and Dean, P.D.G. (1978)
Biochem. J. 173, 591-596
23. Smith, S.L., Stone, D., Novak, P., Baccanari, D.P., and
Burchall, J.J. (1979) J. Biol. Chem. 254, 6222-6225
24. Wachter, E. and Werhahn, R. (1979) Anal. Biochem. 97, 56-64
25. Beaven, G.H. and Holiday, E.R. (1952) Adv. Protein chem. 7,
319-386
26. Stone, D. and Smith, S.L. (1979) J. Biol. Chem. 254,
10857-10861
27. Rosenberg, M. and Court, D. (1979) Ann. Rev. Genet. 13,
319-353
28. Shine, J. and Dalgarno, L. (1974) Proc. Natl. Acad. Sci.
USA 71, 1342-1346
29. Kumar, A.A., Blankenship, D.T., Kaufman, B.T., and Freisheim, J.H. (198O) Biochem. 19, 667-678
30. Smith, S.L. and Burchall. J.J. (198O) Fed. Proc. 39, 1771
31. Pribnow, D. (1979) Proc. Natl. Acad. Sci. USA 72, 784-788
32. Sugimoto, K., Sugisaki, H-, Okamoto, T., and Takanami, M.
(197 5) Nucl. Acids Res. 2, 2O91-21OO
33. Seeburg, P.H., Nusslein, C , and Schaller, H. (1977)
Eur. J. Biochem. 74, 1O7-113
34. Scholz, K. and Jaenicke, L.(1968) Eur. J. Biochem. 4,448-457
35. Reddy, V.A., Kalghatgi, K.K., Rao, P.V.S., and Rao, N.A.
(1976) Indian J. Biochem. Biophys. 13, 1O1-1O5
36. Reddy, V.A. and Rao, N.A. (1975) Indian J. Biochem. Biophys.
12, 7 5-8O
710