FEMS MicrobiologyLetters 6 (1979) 95-98
© Copyright Federation of European MicrobiologicalSocieties
Published by Elsevier/North-HollandBiomedicalPress
95
A FIFTH CLASS OF AMINOGLYCOSIDE 3'-PHOSPHOTRANSFERASE FROM
ANTIBIOTIC-PRODUCING STRAINS OF STREPTOMYCES
J.E. DOWDING
26 HaslingfleldRoad, Barton, CambridgeshireCB3 7AG, England
Received 10 April 1979
1. Introduction
Resistance to the arninoglycoside antibiotic neomycin is commonly associated with the presence of
phosphotransferase activity. Four classes of aminoglycoside-3'-phosphotransferase (APT I-IV) have so
far been identified in bacteria; three of these occur in
a wide variety of antibiotic-resistant clinical isolates
[1,2] and the fourth is synthesised by Bacillus circu.
lans, the butirosin-producing organism [3]. These
enzymes all phosphorylate neomycin in vitro but can
be differentiated by their abilities to modify a range
of structurally related aminoglyeosides. For a recent
review on aminoglycoside-modifyingenzymes see
Davies and Smith [4]. It has been suggested that antibiotic-producing organisms may represent a source of
the resistance determinants found on R-factors in
clinical bacterial isolates [5] and a number of aminoglycoside-modifying enzymes have already been
described in antibiotic-producing strains [6]. In this
paper it is shown that four strains of Streptornyces
which produce neomycin or a closely related antibiotic all synthesise a very similar aminoglycoside
phosphotransferase. This enzyme modifies a considerably restricted range of compounds and is thus
easily distinguished from those of the other four
classes. It is not clear whether this enzyme is
responsible for the resistance of the producing strains
to their own antibiotics.
2. Materials and Methods
2.1. Bacterial strains
APT I was extracted from Escherichia cell W677
pJR35, APT II from E. coil W677 pJR67, APT III
from Staphylococcus aureus 20240 and APT IV from
Bacillus circulans NRRL B-3312. The strains of
Streptomyces were S. fradiae UC 2046 and S. albogriseolus NRRL B-1305 which both produce neomycin, S. rimosus forma paromomyeinus NRRL
2455 which produces paromomycin and S. ribosidificus ATCC 21294 which produces ribostamycin.
2. 2. Preparation of cell-free extracts
Cultures of E. cell, S. aureus and B. circulans were
grown and extracts prepared as described by Haas and
Dowding [7]. Streptomyces strains were grown from
a spore inoculum in 500 ml of Difco Bacto-Nutdent
Broth supplemented with 0.5% (w/v) glucose. Growth
was in an orbital incubator for 24-48 h at 30°C. The
mycelium was harvested, washed in buffer A (10 mM
Tris-HCl, pH 7.8, 10 mM MgC12, 25 mM NI-I4C1, 2
mM dithiothreitol) and resuspended in 5 ml of buffer
A. The resulting paste was sonicated briefly to reduce
viscosity and passed twice through a FCench pressure
cell at 10-12 000 lb/sq, in. The lysate ~,as first centrifuged at low speed to remove unbroken mycelium
and large debris and then at 105 000 X g for 2 h. The
supernatant was dialysed against 2 1 of buffer A and
stored in aliquots at -20°C.
2.3. Enzyme assays
Aminoglycoside-modifyingenzymes were assayed
as described by Haas and Dowding [7], Each assay
contained 10 ~l of phosphotransferase buffer, l0 ~l
of [a2P]ATP solution, 10 IA of extract and 2 IA of
antibiotics solution (1 mg/ml free base in water).
Assays were incubated for 20 min at 35°C and 25 ~l
96
samples were removed to phosphocellulose paper
(Whatman P-81) for counting in a scintillant containing 0.3% PPO and 0.01% POPOP (w/v) in toluene.
i
A
jii iii ii'ii:::
__=___~_ i
__~_~_~
E. coli WS?? pJR35
E
S. fradiae UC 2046
Staph. aureus 20240
~1
S. rim0sus forma
150
3. Results
m
Sections A - D of Fig. 1 show the substrate profiles
of typical examples of class I - I V aminoglycoside-3'phosphotransferases respectively. The various activities can be differentiated as follows: class I enzymes
do not phosphorylate the butirosins, class II enzymes
do not phosphorylate the lividomycins and class III
enzymes phosphorylate both these compounds. The
class IV enzyme is similar to class II but has a narrower substrate range. Sections E - H of Fig. I show
the substrate profiles (determined under the same
conditions) of the phosphotransferase extracted from
the four strains of Streptomyces. It is clear that t h e
IO0
'~ 50
e
z
o 150
.~ 100
ig
~
50
m
C
Paromom~cinus
100
5O
L
D
Fig. 1. Substrate specificities of aminoglycoside 3'-phosphotransferase types I - I V (Sections A - D ) and type V (Sections
E-H). Results are expressed relative to neomycin B (100%)
and are averages of two or three determinations.
L
Bacillus circulans
H
s. ribosidificus
100
50
~ T I ~ \
R2
R
R3
~O
2
~
'"'2 I
NH2~
'~a 2J"
Fig. 2. Structures of the neomyeins and related compounds. Ribostamyein is rings I, II and 11I of neomycin B. Butirosins A and
B are isomers of ribostamycin with a -CO-Ctt(OH)CIt2-CH2NH 2 substituent at the 1-NH2 of deoxystreptamine (ring II). The
arrow indicates the probable site of phosphorylation by APT V.
RI
R2
R3
R4
Rs
Neomycin B
Neomycin C
NH2
NH 2
OH
OH
H
CH2NH 2
CH 2 NH 2
H
H
H
Paromomycin
OH
OH
OH
OH
H
H
CH2 NH2 )
H
CH2 NH2
CH 2 NH 2
H
Lividomycin A
Lividomyein B
{H
CH2NH 2
H
H
mannose
H
97
same (or a very similar) enzyme is present in all four
strains and that it can modify comparatively few of
the aminoglycosides of the neomycin group (the
structures of these compounds are shown in Fig. 2).
Experiments with S. fradiae show that, in this
organism, APT V is not released by osmotic shock or
polymyxin B treatment [7] and that there is little
or none of the phosphotransferase in the growth
medium (no enzyme activity could be detected in
medium concentrated 50-100-fold by Diaflo ultrafiltration). The S. fradiae enzyme has a broad pH
optimum (pH 5.7-8.5) for the phosphorylation of
neomycin B, kanamycin B and ribostamycin. The
phosphorylation of neomycin B is not affected by
equimolar concentrations of the 3'-deoxyaminoglycosides tobramycin, gentamicin Cla or lividomycin
A; modification of the less efficient substrate kanamycin B is inhibited 30-50% under the same conditions. Cultures of the four strains, when grown under
conditions where there was no antibiotic synthesis,
all produced the phosphotransferase, suggesting that
the enzymes are synthesised constitutively. In common with many antibiotic-producing strains, however
[8], S. fradiae and S. albogriseolus are not resistant to
their own antibiotic at all stages of growth. Although
the strain of S. fradiae studied was found to produce
600 ~ugneomycin/ml (and is reported to produce
2 mg/ml under optimal conditions), spore germination was inhibited >99.9% on plates containing 30 #g
neomycin/ml.
4. Discussion
A fifth class of aminoglycoside phosphotransferase
with a characteristic narrow substrate specificity has
been detected in extracts of four aminoglycosideproducing strains of Streptomyces. Enzymes of types
I - I V are known to phosphorylate neomycin B at the
3'-OH of ring I (4) (see Fig. 2) and it appears likely
that APT V also modifies at this position; paromomycin is a substrate for the enzyme whereas 3'-deoxyparomomycin (lividomycin B) is not and, likewise,
kanamycin B is a substrate whereas 3'-deoxykanamycin B (tobramycin) is not.
There is so far no satisfactory explanation of the
role of modifying enzymes such as APT IV and V in
antibiotic-producing bacteria; the observation that
spores of S. fradiae and S. albogriseolus are sensitive
to neomycin suggests that the antibiotic can penetrate
spores, bind to ribosomes and inhibit growth, presumably before any resistance mechanism can be
expressed. This argues that resistance is not conferred by a permanent structural alteration to the
ribosome (although resistance due to a transient
ribosome modification remains a possibility). Many
strains, such as B. circulans, the four described here
and other [6], synthesise enzymes which will, in
vitro, inactivate the compounds they produce. It
seems likely, a priori, that such enzymes confer
resistance on these organisms in vivo and, in the case
of APT IV, it has indeed been shown that this
enzyme will mediate aminoglycoside resistance when
transferred to E. coli [3]. A number of cases are now
known, however, in which a strain produces an antibiotic-modifying enzyme but either a different antibiotic from those modified or none. Examples include
the two spectinomycin producers S. spectabilis and
S. flavopersicus which both contain a streptomycin
phosphotransferase ([6]; Dowding, unpublished
results) and several Streptomyces spp. which do not
produce chloramphenicol but do contain chloramphenicol acetyltransferase [9]. Presumably such
enzymes are not unselected relics from a time when
these strains did produce antibiotics; their functions
are not yet understood and they may not necessarily
be involved in antibiotic resistance directly. In addition to APT V the ribostamicin-producer, S. ribosidificus, synthesises an aminoglycoside acetyltransferase
(Dowding, unpublished results); it can thus phosphorylate or acetylate its own product when either
activity would, presumably, be sufficient to confer
resistance. A very similar case of such enzyme duplication has also been described for a gentamicin-resistant
clinical isolate of Staphylococcus [10]. In these cases,
perhaps because of the permeability properties of the
organisms involved, it may be that two modifying
enzymes are required for effective resistance to
aminoglycosides.
Acknowledgements
I am grateful to Julian Davies of the Department
of Biochemistry, University of Wisconsin-Madison, in
whose laboratory this work was carried out. It was
98
supported by grants to him from the National Science
F o u n d a t i o n and the National Institutes of Health.
References
[ 1] Brzezinska, M. and Davies, J. (1973) Antimicrob. Ag.
Chemother. 3,266-269.
[2] Courvalin, P. and Davies, J. (1977) Antimicrob. Ag.
Chemother. 11,619-624.
[3] Courvalin, P., Weisblum, B. and Davies, J. (1977) Proc.
Natl. Acad. Sei. USA 74,999-1003.
[4] Davies, J. and Smith, D.I. (1978) Annu. Rev. Mierobiol.
32,469-518.
[5] Walker, M.S. and Walker, J.B. (1970) J. Biol. Chem.
245, 6683-6689.
[6] Benveniste, R. and Davies, J. 91973) Proc. Natl. Acad.
Sci. USA 70, 2276-2280.
[7] Haas, M.J. and Dowding, J.E. (1975) Methods in Enzymol. 43,611-628.
[8] Demain, A.L. (1974) Ann. NY Acad. Sci. 235,601 612.
[9] Shaw, W.V. and Hopwood, D.A. (1976) J. Gen. Microbiol. 94, 159-166.
[10] Dowding, J.E. (1977) Antimicrob. Ag. Chemother. 1I,
47-50.