ArticlePDF Available

Taxonomy and Pathogenicity of Erwinia cacticida sp. nov.

Authors:

Abstract and Figures

A total of 108 pectolytic, soft-rotting Erwinia strains were collected from 11 types of cacti growing in Arizona, Texas, northern Mexico, and Australia between 1958 and 1989. Four strains were collected from soils beneath or close to naturally rotting saguaro cacti. Collectively, these strains caused soft rots of saguaro, organ pipe, and senita cacti, Opuntia (cactus) fruits and pads, tomato fruits, and potato slices, but only occasionally caused soft rots of slices of carrot roots. A numerical cluster analysis showed that 98 of the 112 strains formed a uniform group (cluster 1A) that was distinguished from other pectolytic erwinias by an API 20E code of 1205131, by negative reactions in API 50CHE tests for L-arabinose, myo-inositol, D-cellobiose, melibiose, and D-raffinose, and, in supplemental tests, by positive reactions for malonate and growth at 43°C. The average levels of DNA relatedness of 22 cluster 1A strains to the proposed type strain (strain 1-12) as determined by the hydroxyapatite method were 88% in 60°C reactions (with 1% divergence within related sequences) and 87% in 75°C reactions. The levels of relatedness to the type strains of other Erwinia spp. were ≤38% in 75°C reactions. Cluster 1A strains also had a characteristic cellular fatty acid profile containing cyclo-(11,12)-nonadecanoic acid (C(19:0 Cyclo C11-12)) and missing tridecanoic acid (C(13:0)), heptadecanoic acid (C(17:0)), and cis-9-heptadecenoic acid (C(17:1 CIS 9)), which separated them from other pectolytic erwinias. Collectively, these data indicate that the members of cluster 1A are members of a new species that is closely related to E. cacticida; these strains are designated E. cacticida-like pending the availability of additional strains for testing. The remaining cactus strains (in cluster 4) have the physiological, DNA, and fatty acid profiles of Erwinia carotovora.
Content may be subject to copyright.
INTERNATIONAL
JOURNAL
OF
SYSTEMATIC BACTERIOLOGY,
Apr.
1991,
p.
197-212
0020-77 13/91/020197- 16$02.00/0
Copyright
0
1991,
International Union
of
Microbiological Societies
Vol.
41,
No.
2
Taxonomy and Pathogenicity
of
Erwinia cacticida
sp.
nov.
S.
M. ALCORN,l T.
V.
ORUM,’* ARNOLD G. STEIGERWALT,* JOAN
L.
M. FOSTER,3
JAMES
C.
FOGLEMAN,3
AND
DON J. BRENNER2
Department
of
Plant Pathology, University
of
Arizona,
Tucson,
Arizona’; Centers
for
Disease Control,
Atlanta, Georgia2; and Department
of
Biological Science, University
of
Denver, Denver,
Colorado3
A
total of 108 pectolytic, soft-rotting
Erwinia
strains were collected from
11
types of cacti growing in Arizona,
Texas, northern Mexico, and Australia between 1958 and 1989. Four strains were collected from soils beneath
or
close to naturally rotting saguaro cacti. Collectively, these strains caused soft rots of saguaro, organ pipe,
and senita cacti,
Opuntia
(cactus) fruits and pads, tomato fruits, and potato slices, but only occasionally caused
soft rots of slices of carrot roots.
A
numerical cluster analysis showed that 98 of the 112 strains formed a
uniform group (cluster
1A)
that was distinguished from other pectolytic erwinias by an
API
20E code of
1205131, by negative reactions in API 5OCHE tests for L-arabinose, myo-inositol, D-cellobiose, melibiose, and
D-raffinose, and, in supplemental tests, by positive reactions for malonate and growth at 43°C. The average
levels of DNA relatedness of 22 cluster
1A
strains to the proposed type strain (strain 1-12) as determined by the
hydroxyapatite method were
88%
in 60°C reactions (with 1% divergence within related sequences) and
87%
in 75°C reactions. The levels of relatedness to the type strains of other
Erwinia
spp. were 138% in 75°C
reactions. Cluster
1A
strains also had a characteristic cellular fatty acid profile containing cyclo-( 11,12)-
nonadecanoic acid
(C19:o
cycle
cll-lz)
and missing tridecanoic acid (CI3J, heptadecanoic acid
(C1,:o),
and
cis-9-heptadecenoic acid
(Cl,:l
g),
which separated them from other pectolytic erwinias. Collectively, these
data indicate that the members of cluster
1A
are members of a new species, which we name
Erwinia
cacticida.
Three cactus strains in cluster 1B appear to represent a second new species that is closely related to
E.
cacticida;
these strains are designated
E.
cacticida-like pending the availability of additional strains for testing. The
remaining cactus strains (in cluster 4) have the physiological, DNA, and fatty acid profiles of
Erwinia
carotovora.
To
our knowledge, Johnston and Hitchcock (22) were the
first workers to describe a bacterial soft-rot disease of cacti
in the United States. The cultures of these authors were
isolated from prickly pear cacti
(Opuntia tomentella
Berger
and
OpuntiaJicus-indica
(L.)
Mill.) that were originally from
Guatemala and Columbia but were growing in the
U.S.
Department of Agriculture plant introduction garden in
Florida. The bacterium was briefly characterized as “an
actively motile, gram-negative, aerobic, and facultative an-
aerobic bacillus” which produced an acid reaction when it
was grown in broth containing “glucose, saccharose, man-
nite, and salicin but none in maltose, lactose, dulcite, and
arabinose” (22). Subsequently, another new bacterial spe-
cies,
Erwinia carnegieana
Standring
1942
(23), was de-
scribed (3,
6,
23, 28) as being the causal agent of bacterial
necrosis (a soft-rot disease) of saguaro cacti
(Carnegiea
gigantea
Britt.
&
Rose). Among the soft-rot erwinias, this
species was unique in that it was gram positive, a character-
istic also noted by Boyle
(6),
and had a host range limited to
the saguaro cactus. Alcorn
(1)
isolated a number of soft-rot
erwinias from saguaro cacti and demonstrated that they had
a broad host range. Schuyler (26) compared pectolytic
strains from saguaro cactus postblooms (permanently closed
flowers) with selected Alcorn strains and concluded that all
of these organisms were gram-negative bacteria with char-
acteristics similar to those of members of the
Erwinia
carotovora
group. More recently, Fucikovsky and Jaimes
(17)
isolated three biochemically different cultures of soft-rot
erwinias from
Opuntia
spp.
in
Mexico; all were gram-
negative rods with peritrichous flagella.
*
Corresponding author.
1-
University
of
Arizona Agriculture Experiment Station publica-
tion no.
7244.
Because of the anomalies in the original description of
“Erwinia carnegieana,”
the disparity between that descrip-
tion and the very brief account by Johnston and Hitchcock
of their soft-rot pathogen of
Opuntia
spp., and the observa-
tions of Alcorn and Schuyler on their saguaro cactus strains,
taxonomic tests
of
bacteria isolated from various naturally
infected, soft-rotting cacti were initiated. During this work,
it was determined that apparently no culture with the char-
acteristics of
Erwinia carnegieana
as described previously
exists. (The type culture is a
Klebsiella pneumoniae
strain
which is not pathogenic in cacti.) Furthermore, repeated
attempts to isolate the previously described bacterium
failed. Therefore, a recommendation to reject the name
Erwinia carnegieana
was made (3). We report here other
results of
our
taxonomic studies, describe a new
Erwinia
species, and indicate how this species can be identified by
using fatty acid analyses and API 20E and API
SOCHE
test
strips (Analytab Products, Inc.
,
Plainview, N.Y.).
(A preliminary report of our early taxonomic studies has
been presented previously
[2].)
MATERIALS AND METHODS
Collection and isolation techniques.
The data reported
below are based on
88
cultures isolated from
11
varieties or
species of naturally infected cacti that generally were grow-
ing in native stands in Arizona
(76
strains), Texas
(5
strains),
Mexico
(1
strain), and Australia
(6
strains) (Table
1).
Each
strain listed in Table
1
was assigned a sequence number
(S/N)
and was placed in
a
cluster on the basis of the results
of the numerical analyses described below. The Arizona
strains included
1
strain recovered from
Acanthocereus
pentagonus
(L.)
Britt.
&
Rose growing in the cactus collec-
tion at the Boyce Thompson Arboretum, Globe, Ariz., and
two strains from
0.
ficus-indica
specimens growing as land-
197
198 ALCORN ET AL.
INT.
J.
SYST. BACTERIOL.
TABLE
1.
Erwinia
strains and other strains studied, listed in the sequence defined by a cluster analysis
of
114 phenotypic characteristics
-
Year
of
isolation Site
of
S,"
Cluster Taxon StrainU Synonym(s)" Sourcea Host
or
year isolation
received'
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
I.
8
11
9
20
Z!
1
22
:!3
24
25
26
27
2
8
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
1A
1A
1A
1A
1A
1A
1A
1A
1A
1A
1A
1A
1A
1A
1A
1A
1A
1A
1A
1A
1A
1A
1A
1A
1A
1A
1A
1A
1A
1A
1A
1A
1A
1A
1A
1A
1A
1A
1A
1A
1A
1A
1A
1A
1A
1A
Erwinia cacticida
E.
cacticida
E. cacticida
E. cacticida
E.
cacticida
E.
cacticidu
E.
cacticida
E.
cacticida
E. cacticida
E. cacticida
E.
cacticida
E.
cacticida
E. cacticida
E. cacticida
E.
cacticida
E.
cacticida
E. cacticida
E.
cacticida
E.
cacticida
E.
cacticida
E.
cacticida
E. cacticida
E. cacticida
E. cacticida
E.
cacticida
E. cacticida
E. cacticida
E.
cacticidu
E. cacticida
E. cacticida
E.
cacticidu
E. cacticidu
E.
cacticida
E.
cacticida
E. cacticida
E. cacticida
E.
cacticida
E.
cacticida
E.
cacticida
E. cacticida
E. cacticida
E.
cacticidu
E. cacticida
E.
cacticida
E.
cacticida
E.
cacticida
67-53
71-3-3
106
64-28
64-30
66-186
66-50
67-106-6
66-78
71-8-22
66-190
67-115-11
67-1 18-9
67-52
ICMP 7449-81
89-6-1
ICMP 7451-81'
83-41-2
84-19-el
64-20
66-42
583-10-10
65-145-2
78-25-2'
78-29b
67-1 16-4
78-28'
66-93-3
82-la'
DU 7
87-5a
83-38b
66-187'
66-188
Texas 28'
65-164a
Texas 27
66-31
62-69-1"
Texas 31
62-55
62-59-5
66-36
62-70-2'
1-12=
62-OP-41-2
ICPB EC189
ICPB EC290
ICMP 7453-81,
ICPB EC292
ICPB EC297, 71-
8-3 (lost)
ICPB 283, 71-4A
(lost)
ICMP 7448-81,
ICPB EC282
ATCC 49482,
ICPB EC283
ICPB EC296
ATCC 49484,
ICPB EC286
ICPB EC221
ATCC 49487
ATCC 49481T,
ICMP 1551-66T,
ICPB EC186T,
Dye EH-3''
ICMP
ICMP
Fogleman
Richard son
Richardson
Richardson
Carnegiea
gig
an
tea
C.
gigantea
C.
gigantea
C.
gigantea
C.
gigantea
C.
gigantea
postbloom
C.
gigantea
C.
gigantea
Stenocereus thurberi
C.
gigantea
C.
gigantea
postbloom
C.
gigantea
postbloom
C.
gigantea
postbloom
C.
gigantea
C.
gigantea
Soil beneath "leaking"
C.
gigantea
Acanthocereus
pentagonus
C.
gigantea
S.
thurberi
C.
gigantea
C.
gigantea
C.
gigantea
C.
gigantea
postbloom
C.
gigantea
S.
thurberi
C.
gigantea
postbloom
Stenocrreus
gummosus
C.
gigantea
Opuntia jicus-indica
S.
gummosus
C.
gigantea
C.
gigantea
C.
gigantea
postbloom
C.
gigantea
postbloom
Opuntia phaeacantha
C.
gigantea
postbloom
0.
phaeucantha
var.
discata
0.
phaeacantha
var.
discuta
Ferocactus wislezenii
(barrel cactus)
0.
phaeacantha
var.
discuta
0.
phaeacantha
var.
major
0.
phaeacantha
var.
major
0.
phaeacantha
var.
discata
Opuntia fulgida
(jump-
ing cholla cactus)
C.
gigantea
var.
discata
0.
phaeacantha
var
d is ca ta
1967(i)
1971(i)
1958(i)
1965(i)
1964(i)
1966(i)
1966(i)
1967(i)
1966(i)
Arizona
Arizona
Arizona
Arizona
Arizona
Arizona
Arizona
Arizona
Arizona
1971(i) Arizona
1966(i) Arizona
1967(i) Arizona
1967(i) Arizona
1967(i) Arizona
1988(r) Arizona
1989(i) Arizona
1988(r) Arizona
1983(i) Arizona
1984(i) Arizona
1964(i) Arizona
1966(i) Arizona
1959(i) Arizona
1965(i) Arizona
1978(i) Arizona
1978(i) Arizona
1967(i) Arizona
1978(i) Mexico
1967(i)
1982(i)
1986(r)
1983(i)
1987(i)
1966( i)
1966(i)
1971(r)
Arizona
Arizona
Mexico
Arizona
Arizona
Arizona
Arizona
Texas
1965(i) Arizona
1971(r) Texas
1966(i) Arizona
1962(i) Arizona
1971(r) Texas
1962(i) Arizona
1962(i) Arizona
1966(i) Arizona
1962(i) Arizona
1958(i) Arizona
1962(i) Arizona
Continued on following page
VOL.
41, 1991
TABLE
1-Continued
ERWZNZA
CACTZCZDA
SP. NOV. 199
~~
Year
of
isolation Site
of
or
year isolation
receivedb
S/N
Cluster Taxon Strain" Synonym(
s)O
Source"
Host
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
1A
1A
1A
1A
1A
1A
1A
1A
1A
1A
1A
1A
1A
1A
1A
1A
1A
1A
1A
1A
1A
1A
1A
1A
1A
1A
1A
1A
1A
1A
1A
1A
1A
1A
1A
1A
1A
1A
1A
1A
1A
1A
1A
1A
1A
1A
1A
1A
E.
cacticida
E.
cacticida
E.
cacticida
E.
cacticida
E.
cacticida
E.
cacticida
E.
cacticida
E.
cacticida
E.
cacticida
E.
cacticida
E.
cacticida
E.
cacticida
E.
cacticida
E.
cacticida
E.
cacticida
E.
cacticida
E.
cacticida
E.
cacticida
E.
cacticida
E.
cacticida
E.
cacticida
E.
cacticida
E.
cacticida
E.
cacticida
E.
cacticida
E.
cacticida
E.
cacticida
E.
cacticida
E.
cacticida
E.
cacticida
E.
cacticida
E.
cacticida
E.
cacticida
E.
cacticida
E.
cacticida
E.
cacticida
E.
cacticida
E.
cacticida
E.
cacticida
E.
cacticida
E.
cacticida
E.
cacticida
E.
cacticida
E.
cacticida
E.
cacticida
E.
cacticida
E.
cacticida
E.
cacticida
87-9-2~'
88-6
84-19-e2
Texas 29
82-lb
66-24
67-58
65-279-8
64-OP-4
64-OP-8
NCPPB 671
88-3a
83-42a
88-21-1
66-19-1
83-42b
DU 89-6.1'
DU 89-20.1
66-34
89-3-2'
DU 89-7.3
DU 89-8.1'
DU 89-5.1
72-1
67-3
66-12
64-OP-7
67-54
556-1-2'
87-10-1~
87-10-2~
87-9-1~
DU
89-3.3
62-47-2'
Texas 30
623-2'
89-5
NCPPB 672'
62-63'
85-13b'
71-8-23
84-19-e3
71-8-18
71-8-8'
75-14-2
65-95-5
78-29a'
67-1 11-7'
Dye EH-4, 622
(lost)
ATCC 49483
ICMP 7452-81,
ICPB EC188
Dye EH-1, ICMP
1381, ICPB
EC187
ICPB EC223
623-2
ATCC 49485,
ICPB EC293
ICPB EC294
ATCC 49486
Richardson
NCPPB
Fogleman
Fogleman
Fogleman
Fogleman
Fogleman
Fogleman
Richardson
NCPPB
C. gigantea
C. gigantea
S.
thurberi
0.
phaeacantha
var.
discata
0.
Jicus-indica
0.
phaeacantha
var.
discata
C. gigantea
C. gigantea
0.
phaeacantha
var.
major
0.
phaeacantha
var.
discata
C. gigantea
C. gigantea
C. gigantea
Soil beneath "leaking"
C. gigantea
0.
phaeacantha
var.
major
Soil beneath "leaking"
C. gigantea
Opuntia stricta
0.
stricta
0.
phaeucantha
var.
discata
C. gigantea
0.
stricta
0.
stricta
0.
stricta
0.
fulgida
0.
phaeacantha
var.
major
C.
gigantea
Opuntia violacea
var.
C.
gigantea
0.
phaeacantha
var.
major
C. gigantea
C. gigantea
C.
gigantea
0.
stricta
0.
phaeacantha
var.
discata
0.
phaeacantha
var.
discata
C. gigantea
C.
gigantea
C. gigantea
0.
phaeacantha
var.
major
C.
gigantea
C. gigantea
S.
thurberi
C. gigantea
C. gigantea
C.
gigantea
C. gigantea
postbloom
S.
thurberi
C. gigantea
postbloom
macrocentra
1987(i)
1988(i)
1984(i)
1971(r)
1982(i)
1966(i)
1967(i)
1965(i)
1964(i)
1964(i)
1980(r)
1988(i)
1988(i)
1983(i)
1966(i)
1983(i)
1989(r)
1989(r)
1966(i)
1989(i)
1989( r)
1989(r)
1989(r)
1958(i)
1958(i)
1966(i)
1964(i)
1967 (i)
1958(i)
1987(i)
1987(i)
1987(i)
1989(r)
1962(i)
1971(r)
1959(i)
1989(i)
1980(r)
1962(i)
1985(i)
1971(i)
1984(i)
197 1( i)
197
1
(i)
1975(i)
1965(i)
1978(i)
1967(i)
Arizona
Arizona
Arizona
Texas
Arizona
Arizona
Arizona
Arizona
Arizona
Arizona
Arizona
Arizona
Arizona
Arizona
Arizona
Arizona
Australia
Australia
Arizona
Arizona
Australia
Australia
Australia
Arizona
Arizona
Arizona
Arizona
Arizona
Arizona
Arizona
Arizona
Arizona
Australia
Arizona
Texas
Arizona
Arizona
Arizona
Arizona
Arizona
Arizona
Arizona
Arizona
Arizona
Arizona
Arizona
Arizona
Arizona
Continued
on
following page
200 ALCORN ET AL.
INT.
J.
SYST.
BACTERIOL.
TABLE 1-Continued
-
Year
of
isolation Site
of
receivedb
S/N
Cluster
Taxon Straina Synonym(s)a Source"
Host
or
year isolation
C. gigantea
C. gigantea postbloom
C. gigantea
C. gigantea
C. gigantea
C. gigantea
C. gigantea
ATCC Chrysanthemum sp.
1989(i) Arizona
1966(i) Arizona
1988(i) Arizona
1989(i) Arizona
1988(i) Arizona
1989(i) Arizona
1988(i) Arizona
1988(r)
95 1A
96 1A
917 1A
98
1A
99 1B
100
1B
101 1B
102
2
103 3
1.04
3
105
3
YO6 4
YO7 4
'LO8 4
109 4
110 4
111
4
112 4
113 4
114 4
115 4
116 4
117 4
118 4
119 4
120 4
121 4
122 4
123 4
124 4
125 4
126 4
127 4
128 4
129 4
130 4
131 4
132 4
133
4
134
4
135
4
136
4
137 4
138 4
139 4
140 4
141 4
142 4
143
4
E.
cacticida
E. cacticida
E.
cacticida
E.
cacticida
E. cacticida-like
E. cacticida-like
E. cacticida-like
Erwinia chrysan-
Enterobacter agglo-
Erwinia herbicola
Erwinia nigrifuens
themi
merans
89-9
66-185
88-10-b"
89-10"
88-2'
89-4-1
88-4'
ATCC 11663T'
ATCC 2715jT' 1988(r)
ATCC Human
ATCC 33243T'
ATCC 1302tJT"
78-31-1
ATCC 1988(r)
ATCC Juglans regia (walnut) 1988(r)
Beta vulgaris (sugar 1978(i) Arizona
beet)
ICMP
B.
vulgaris 1988(r) California
ICMP 4226-75T' Dye FD-lT, UR-7T,
NCPPB 2795T,
UCPPB 193T,
ATCC 43762T
Erwinia carotovora
subsp. betavas-
culorum
Stanghellini
B.
vulgaris 1978(i) Arizona
ICPB 1978(r)
78-30
E 173-78 Erwinia carotovora
E. carotovora
Erwinia atroseptica
subsp. atroseptica
subsp. atroseptica
ICPB 173
ATCC 33260T' ATCC Solanum tuberosum
CUCPB
(potato)
S.
tuberosum
S.
tuberosum
0.
phaeacantha var.
major
0.
violacea var.
macrocentra
ICPB
CUCPB
CUCPB
CUCPB
ATCC
S.
tuberosum
1988(r)
E
28-78
70-37-A1
70-37-B3
64-OP-5
CUCPB E28 1978(r)
1970(i) Arizona
1970(i) Arizona
1964(i) Arizona
64-0P-6" 1964(i) Arizona
Erwinia carotovora
E.
carotovora
E.
carotovora
Erwinia aroideae
Erwinia carotovora
subsp. carotovora
Erwinia atroseptica
EC 101
EC 6-78
EC 14-78
EA 14-78
ATCC 15713='
1970(r)
1978 (r
)
1978(r)
1978(r)
1989(r)
CUCPB EC6
CUCPB EC14
CUCPB EA14
ICPB EA179 ICPB 1978(r)
Stanghellini Brassica oleracea var. 1987(r) Arizona
Stanghellini B. vulgaris 1974(r) Arizona
botrytis (cauliflower)
E
179-78
87-2-2
Erwinia carotovora
subsp. atroseptica
KSB 10
80-91 Stanghellini
Stone
Stone
Stone
Stone
Stanghellini
ICPB
CUCPB
Stone
ICPB
Cucurbita foetidissima
(buffalo gourd)
Capsicum annuum
(chili fruit)
Vegetable
Vegetable
Vegetable
Vegetable
S.
tuberosum
1980(r)
1971(i)
1966(i)
1966(i)
1966(i)
1966(i)
1974(r)
1978(r)
1978(r)
1966(ij
1965(i)
1967(i)
1967(i)
1967(i)
1967(i)
1967(i)
1967 (i)
1965(i)
1965 (i)
1978(r)
Arizona
Arizona
Arizona
Arizona
Arizona
Arizona
Arizona
Arizona
Arizona
Arizona
Arizona
Arizona
Arizona
Arizona
Arizona
Arizona
Arizona
71-33~
Erwinia sp.
Erwinia sp.
Erwinia sp.
Erwinia sp.
S-29
s-10
s-11
S-8
74-4
E 153-78
EA
13-78
S-17
65-214'
67-1 11-3
67-114-1
67-110-8
67-112-11
67-117-14
67-113-7
65-82-6
Erw inia atrosep tica
Erwinia aroideae
Erwinia
sp.
ICPB EA153
CUCPB EA13
Vegetable
C. gigantea postbloom
C. gigantea postbloom
C. gigantea postbloom
C. gigantea postbloom
C. gigantea postbloom
C. gigantea postbloom
C. gigantea postbloom
Soil beneath decayed
C. gigantea
C. gigantea
65-46-
1"
EC 208-78 Erwinia carotovora ATCC 495, Dye
EG23, ICPB EC208
Continued on following page
VOL.
41, 1991
TABLE
1-Continued
ERWZNZA CACTZCZDA
SP. NOV.
201
Year of
isolation Site of
or year isolation
receivedb
S/N Cluster Taxon Strain" Synonym(s)" Source" Host
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
4
4
5
5
5
5
5
5
5
5
6
6
6
6
7
7
7
E. carotovora
EC
218-78
Erwinia atroseptica
E 70-78
Erwinia quercina
ATCC
29281TC
Erwinia salicis
ATCC
15712T'
Erwinia tracheiphila
ATCC
33245T'
Erwinia mallotivora
ATCC
29573='
Erwinia rubrifaciens
ATCC
2929lTC
Erwinia psidii
IBSBF
435'
Erwinia amylovora
ATCC
15580T'
Erwinia stewartii
ATCC
8199T'
Erwinia ananas
ATCC
33244T'
Erwinia uredovora
ATCC
19321T'
Erwinia
sp. ATCC
11773'
Erwinia rhapontici
ATCC
29283Tc
Klebsiella oxytoca
ICPB
3866
Klebsiella
ATCC
13883T'
Escherichia coli
ATCC
1177ST'
pneumoniae
CUCPB
E70
Dye FB-lT, ICPB
EQ1OlT,
NCPPB
1852T
Dye
EX2T,
ICMP
1587aT,
NCPPB
447T
PDDC
8426
ICPB
CUCPB
ATCC
ATCC
ATCC
ATCC
ATCC
IBSBF
ATCC
ATCC
ATCC
ATCC
ATCC
ATCC
ICPB
ATCC
ATCC
1978(r)
1978(
r)
Quercus
sp. (oak)
1988(r)
Salix alba
(willow)
1988(r)
Cucumis melo
Mallotus japonicus
Juglans regia
Pyrus
communis
(pear)
Ananas comosus
(pineapple)
Cereal
rust
uredia
C. gigantea
Rheum rhabarbarum
(rhubarb)
Cucumis sativus
(cucumber)
1988(r)
1988(r)
1988(r)
1988(r)
198 8
(r)
1988(r)
1988(r)
1988(r)
1988(r)
Arizona
1988(r)
1978(r)
1988(r)
Urine
1988(r)
a
Abbreviations: ATCC, American Type Culture Collection, Rockville, Md.
;
Dye, D.
W.
Dye, International Collection of Micro-Organisms from Plants, Plant
Disease Division, New Zealand Department of Scientific and Industrial Research, Auckland, New Zealand; CUCPB, Cornell University Collection of
Phytopathogenic Bacteria, Ithaca,
N.Y.
;
Fogleman, J. C. Fogleman, Department of Biological Sciences, University
of
Denver, Denver, Colo.; IBSBF, Instituto
Biol6gico de Sao Paulo, Sao Paulo, Brazil; ICMP and PDDC, International Collection of Micro-Organisms from Plants, Plant Disease Division, New Zealand
Department of Scientific and Industrial Research, Auckland, New Zealand; ICPB, International Collection of Phytopathogenic Bacteria, University of California,
Davis; NCPPB, National Collection of Plant Pathogenic Bacteria, Harpenden, United Kingdom; Richardson, R.
H.
Richardson, Department of Zoology,
University of Texas, Austin; Stone,
W.
T. Stone, formerly of the Department
of
Plant Pathology, University of Arizona, Mesa; Stanghellini, M.
E.
Stanghellini,
Department of Plant Pathology, University
of
Arizona, Tucson; UCPPB, Department of Plant Pathology, University
of
California, Berkeley.
i, Year isolated; r, year received.
Isolate used for DNA studies.
scape plants in Tucson, Ariz. In addition to the 88 strains
isolated from naturally infected cacti,
1
strain came from a
rot following artificial injury to an agria cactus
(Stenocereus
gummosus
(Engelm.) Gibson
&
Horak) in Mexico (16), 19
strains came from postbloom saguaro cactus flowers that did
not show obvious symptoms (26),
4
strains came from soils
beneath or close to rotting saguaro cacti,
1
strain came from
a saguaro cactus and was deposited with the American Type
Culture Collection by Bryan and May (strain ATCC 11773
[=
SIN 156]), and
14
strains came from cultivated crops in
Arizona. In addition, 20 type strains and 13 other reference
strains were included in the study. Strains
S/N
103 and S/N
104 were received as type strains
of
Enterobacter
agglom-
erans
and
Erwinia herbicola,
respectively, and are identified
as such throughout this paper even though both of these
organisms have recently been classified as
Puntoea
agglom-
erans
strains (18). The 14 strains isolated from cultivated
crops in Arizona included five subcultures
(S/N
126 through
S/N
129 and
S/N
133) of strains used
in
studies by Stone (31)
and one subculture
(S/N
123) tested by de MendonCa and
Stanghellini
(11).
Except as noted below, the data given are
from tests conducted in 1988 and 1989 on strains collected
between 1958 and 1989 (Table
1).
Bacteria that were pectolytic under aerobic conditions
when they were streaked onto pectate medium (13) (modified
by deleting yeast extract) were subcultured onto either
potato dextrose peptone agar (PDP) (40 g of potato dextrose
agar [Difco Laboratories, Detroit, Mich.], 5 g
of
agar, and 20
g of peptone in
1
liter of distilled water; sufficient
5
N
NaOH
was added to adjust the pH to ca.
7.2
after autoclaving) or
glucose yeast extract agar
(20
g of glucose, 10 g
of
yeast
extract, and
10
g of peptone in
1
liter of distilled water).
Single colonies were subjected to at least three cycles of
streaking and subculturing to confirm their purity and were
retested for pectolytic activity before being stored in nutrient
broth (Difco) and on glucose yeast extract carbonate agar
slants (glucose yeast extract agar containing 1% CaCO,) at
4
to 6°C. Stored cultures were transferred every 6 to 9 months.
Taxonomic studies have been conducted intermittently
over the last 30 years and have included not only many
strains maintained at the University of Arizona but also
subcultures from strains sent previously to other laborato-
ries. In 1988 and 1989 the largest number of strains (160
strains) and control cultures were tested concurrently (Table
1). Included in the
160
strains was one duplicate
(S/N
82 and
S/N
84), which varied consistently in fatty acid characteris-
tics. In addition, data from tests performed with
13
duplicate
cultures were used to estimate the variability of responses to
tests; these duplicate cultures were not included in further
analyses.
Morphological
and
physiological
tests.
KOH tests (32) and
Gram staining (9), in which we used alkaline gentian violet,
Kopeloff-Beerman iodine, and 0.25% aqueous safranin
0
and decolorized with
30%
acetone in 95% ethanol, were
202
ALCORN
ET
AL.
INT.
J.
SYST.
BACTERIOL.
conducted by using ca. 13- to 24-h cultures (grown on PDP or
glucose yeast extract carbonate agar) of all strains. Smears
of known gram-positive and gram-negative bacteria also
were included on each slide as controls. Motility was de-
tected by observing suspensions of cultures
of
similar ages in
nutrient broth under a compound microscope (magnification,
x
600 with phase-contrast or dark-field illumination). The
flagellation
of
representative strains was observed by using
an electron microscope or a compound microscope or both.
For compound microscopy the bacteria were first stained by
using the rapid technique of Mayfield and Inniss (24). Cata-
lase activity was determined by adding 3%
H,02
to bacteria
growing on PDP or in an API mannitol cupule. The Hugh-
Leifson
(21)
method was used for oxidation-fermentation
tests.
Most of the media used for physiological tests were
inoculated with sterile distilled water (SDW) suspensions of
20- to 28-h cultures grown on PDP. The exceptions were the
tests involving API 20E and API 5OCHE strips; these test
preparations were inoculated with bacteria grown on nutri-
ent agar (Difco). Inoculated substrates were incubated at
30"C, and tests were read by using the standard protocol (9)
or the manufacturer's instructions. The API procedures
which we used were essentially those of Mergaert et al. (25)
and Verdonck et al. (33). Over the years, we have found that
the API 20E Voges-Proskauer, citrate, and nitrate tests
occasionally give false-negative results. Therefore, every
isolate which gave a negative result for these characteristics
in 1988 and 1989 was retested and scored negative in our data
set only if the repeat test result was also negative.
In supplemental tests, acid production from carbohydrates
was detected in purple broth base (Difco) in 1978, in 1%
peptone (Difco) broth base containing bromthymol blue
indicator in 1971, and in phenol red broth base (Difco) in
1964 and 1962. The production of reducing substances from
sucrose was detected by using the procedures of Dye (13)
with (1989 tests) and without beef extract. Brown, orange, or
olive brown reactions after
24
h
were recorded as positive;
green, marine, and blue reactions were recorded as negative.
Malonate and tartrate media were made in 1989 as described
by Dye (13) or by substituting malonate and tartrate for
citrate in Simmons citrate agar (Difco). All other tests were
performed by using the methods of Edwards and Ewing
(15)
or the methods described in the
Manual
of
Microbiological
Methods
(9).
Growth and acid production at 30, 36, and 43°C also were
determined. For these tests, dilute suspensions (ca. no.
1
on
the McFarland scale [turbidity standards were obtained from
Remel Microbiology Products, Lenexa, Kans.]) of 18- to
24-h cultures were made in
SDW.
Single loopfuls (loop
inside diameter,
3
mm) of the suspensions were transferred
to 10-ml portions of Difco purple broth base (containing 1%
dextrose) and to Dye
YS
broth
(13).
Tests were carried out
in a circulating water bath at 43"C, in a circulating air
incubator at 36"C, and in a standard radiant heat incubator at
30°C. The tubes were preincubated for
5
days at the desired
temperatures to check for contamination, inoculated, and
then immediately returned to the appropriate incubation
units. Readings were made after
5
days. To prevent contam-
ination, the caps of all of the tubes in the water bath were
sealed with parafilm (American Can Co., Greenwich,
Conn.).
Soft-rot
tests.
The soft-rot capabilities
of
the strains tested
were determined in 1989 by using single, greenhouse-grown
seedlings of saguaro and organ pipe
(Stenocereus thurberi
(Engelm.)
S.
Buxb.) cacti that were ca.
3
to
12
cm tall. Each
cactus was injected with ca. 0.25 ml
of
a suspension of a ca.
24-h culture
in
SDW with
a
turbidity between ca. no.
1
and
4 on the McFarland scale (ca. 2
X
lo8
to 8
X
lo8
CFUlml).
Control plants were similarly treated with SDW alone. All of
the cacti were incubated in a greenhouse at ca. 34°C.
Because of quarantine restrictions, plants inoculated with
Erwinia ananas, Erwinia mallotivora, Erwinia psidii, Er-
winia sakis,
and strains isolated from Australian material
were incubated under 12 h of fluorescent and incandescent
light at 34°C
in
a locked growth chamber. The cacti used for
all inoculations were deprived
of
water for several weeks
prior to inoculation. Following injection, the plants were
watered daily and assessed after
4
days. To do this, each
plant was severed at the groundline, and the aerial portion
was weighed. Subsequently, any rotted material was exca-
vated and weighed. The test was repeated three times for
each strain with each of the two cactus species. Strains were
considered to be pathogenic if there was at least 8% rot in at
least two of the three plants inoculated. This criterion was
chosen
so
that only those strains that unequivocally caused
soft rot were scored as positive.
In various years, suspensions
of
selected strains were
injected into unripened fruits of the Indian fig cactus
(0.
$ficus-indica)
and ripened tomato fruits
(Lycopersicon escu-
lentum
(L.) Mill.) that had been surface sterilized with 2%
Amphyl (Lehn and Fink Industrial Products Div.
,
Sterling
Drugs, Inc., Montvale, N.J.). Suspensions (ca.
1
ml) of
selected strains also were injected into separate pads of
0.
ficus-indica
(infiltrated tissues temporarily appeared to be
water soaked) and into 2.5- to 18-cm-tall saguaro and senita
(Lophocereus schottii
(Engelm.) Britt.
&
Rose) cacti. At
least two separate inoculations were made per strain. Refer-
ence strains were inoculated similarly. The inoculated ma-
terial was incubated appropriately and observed for 7 to 10
days. A strain was considered to be pathogenic only when a
major portion of the seedling or fruit rotted or when disease
in the pads progressed beyond the area of initial water
soaking.
Carrot
(Daucus carota
L.)
roots and potato
(Solanum
tuberosum
L.)
tubers were first thoroughly scrubbed in
running warm tap water (detergent was added in 1989) and
then either surface sterilized with 2% amphyl or (in 1989)
soaked in a
0.5%
sodium hypochlorite solution containing
ca.
1
ml of Tween
80
per
4
liters for
5
min. The treated carrot
roots and potato tubers were sliced aseptically into ca.
2-mm-thick sections and either placed directly on two sheets
of new, 9.0-cm-diameter filter paper in sterile, plastic petri
plates (in 1989) or placed on sterile, glass V-shaped rods that
were resting on similar filter paper in petri dishes (previous
years). The filter papers were moistened to saturation with
SDW. Two drops of a bacterial suspension (turbidity in the
range of McFarland standards no.
1
to 4) were placed on the
center of each slice (three slices per strain in 1973 and one
slice per strain in 1978 and 1989), after which the plates were
sealed with either Parafilm or masking tape (Shurtape,
Hickory, N.C.) and incubated at 30°C. Reactions were
considered positive if tissue softening or maceration oc-
curred within
2
days as determined by gentle probing of the
slices with a sterile glass rod. In 1989 the controls consisted
of end and middle slices of potato tubers and carrot roots.
Each tuber or root slice was placed
in
a separate plate and
inoculated with SDW. If the controls rotted, all bacterial
inoculations of slices from the same potato tuber
or
carrot
root were repeated. For inoculations in previous years, the
control slices were placed on V rods in the same plates as
VOL.
41.
1991
ERWINIA
CACTICIDA
SP.
NOV.
203
inoculated slices. If the controls rotted, the test was re-
peated.
Fatty acid analyses. Quantitative analyses of cellular fatty
acid compositions were performed by using a gas-liquid
chromatographic procedure similar to that described by De
Boer and Sasser
(10).
Strains were grown on tryptic soy
broth agar (Difco) for 24 h at 28°C. Several loopfuls of cells
were added to
1.0
ml of 3.75 M NaOH in 50% aqueous
methanol and heated for 30 min in
a
sealed tube in
a
boiling
water bath. The saponified material was then acidified by
adding
1.08
ml of 6 M hydrochloric acid, the fatty acids were
methylated by adding 0.92 ml of methanol, and the resulting
mixture was heated in a sealed tube for
10
min at 80°C. After
cooling to room temperature, the fatty acid methyl esters
were extracted with 1.2 ml of hexane-diethyl ether
(1:l)
and
washed with 3.0 ml of
0.3
M NaOH.
Fatty acid determinations were made by using a microbial
identification system (MIS) (Hewlett-Packard, Palo Alto,
Calif.) which included a Hewlett-Packard model 5890A gas
chromatograph equipped with a flame ionization detector.
With this system, we used a phenylmethyl silicone fused
silica capillary column (25 m by 0.2 mm) and hydrogen as the
carrier gas. The gas chromatograph was temperature pro-
grammed from
170
to 270°C at a rate of 5"Urnin. Identifica-
tion of cellular fatty acids (via their methyl esters) was an
automated function of the MIS in which a Hewlett-Packard
model 9000 series
300
computer was used along with the MIS
software program (version
3.0).
The Hewlett-Packard model
3392A integrator component of the MIS provided the data
used by the system to estimate the relative percentages
of
the fatty acids that were present in each sample.
Numerical analyses. Test error and reproducibility were
analyzed by using data from
13
code-numbered, duplicate
strains and formulas recommended by Austin and Priest
(5).
The following three categories of duplicates were analyzed:
recent (1988) subcultures (five pairs), pre-1977 subcultures
and subcultures of our strains returned to
us
from other
collections (six pairs), and the same reference strain from
two different collections (two pairs). Individual test vari-
ances, probability of error for an individual test, pooled
variance, and average probability of an erroneous test result
were calculated.
Each of the
160
bacterial strains was evaluated for 130
characteristics, including soft-rot capability and the occur-
rence of 39 separate fatty acids (the total number of fatty
acids identified from all strains). Thirteen of these charac-
teristics were either positive or negative for all 160 strains
and were not used to generate dendrograms. Three (phos-
phatase and fatty acids and
C15:1
11)
of the 130
characteristics were not used because analysis of duplicate
strains revealed an error rate that was too high. Most of the
characteristics analyzed were coded
0
for negative or
1
for
positive; the exceptions were the API SOCHE tests, in which
positive reactions after
3,
6, 24, and 48 h
of
incubation at
30°C were coded 5,4,
3,
and 2, respectively, as described by
Mergaert et al.
(25).
A modified Gower similarity index
(S,)
(27) was calculated, with the API 5OCHE codes weighted by
dividing by 5 (the possible range of values) and with all
characteristics (negative, partial, and positive matches)
counted in the denominator of the index (the
S,
was
modified by including negative matches). A dendrogram was
generated by using the Cluster program of SPSS/PC+
(SPSS,
Inc., Chicago, Ill.) and the unweighted pair group
average method of clustering, Weighted coded data for each
strain were used by the SPSS program to calculate a distance
matrix, using the city block measure. The resulting dendro-
FIG.
1. Peritrichous flagella on a cell
of
Erwiniu
cctcticidu
1-12=
(A)
and fimbriae
on
a cell
of
the same strain
(B).
Bars
=
1.0
pm.
Cells were negatively stained in
1%
phosphotungstic acid
for
2
rnin
and photographed by using an Hitachi model
H500
electron micro-
scope.
gram and agglomeration schedule are the same as those
based on the modified Gower similarity matrix, after trans-
formation by the following formula:
SG
=
1
-
(Dcb/114),
where
D,,
is
the distance between strains or clusters in the
city block distance matrix,
S,
is the corresponding element
in the similarity matrix, and 114 is the number of character-
istics per strain.
DNA
hybridization. DNA hybridization studies were done
by using the hydroxyapatite method and previously de-
scribed techniques (7). Strains 1-12T
(=
S/N
4ST) (T
=
type
strain), 88-2
(=
S/N
99), and DU 89-8.1
(=
S/N 68) were
labeled enzymatically in vitro by using [32P]dCTP provided
in a nick translation reagent kit (Bethesda Research Labo-
ratories, Inc., Gaithersburg, Md.) as directed by the manu-
facturer. Labeled DNAs were reacted with DNAs from
selected cultures (Table
1)
that represented the range of
strains isolated from cacti, as identified by dendrographic
analyses, and with DNAs from the various type cultures. A
total of 49
of
the DNA source strains were tested as
numbered strains; the identities of these strains were not
revealed until the conclusion of the study. Two strains
(S/N
63 and
S/N
68)
were identified as cactus strains from
Australia and were tested separately from the others. Three
strains, strain ICPB EC186T,
Erwinia cypripedii
EC155T,
and
Erwinia milletiae
9572-82T, were tested as known
strains; these three strains were used only in DNA tests and
were not included in the tests used for numerical analysis.
RESULTS
The typical pectolytic cactus strain (for convenience, this
term includes the pectolytic strains isolated from soils be-
neath or close to rotting saguaro cacti but not the nonpec-
tolytic strain isolated from
a
saguaro cactus, strain ATCC
11773) was a capsule-producing, gram-negative, nonspore-
forming, rod-shaped organism
(0.5
to 1.2 by 0.5 to 1.7
pm)
that usually occurred singly (Fig.
l),
but occasionally oc-
curred in chains of several cells. The rapid,
3%
KOH Gram
test (32) for all cactus strains was negative (stringy); this
result is in agreement with traditional Gram stain results.
Certain strains (e.g., strains 64-30
[=
S/N
51,
65-164a
[=
S/N
204
ALCORN
ET
AL.
INT.
J.
SYST.
BACTERIOL.
O/o
SG
70
80
90
I00
I
1
I
I
I
1
1
I
I
I
I
I
I
I
I
I
70 80
90
I00
'10
SG
Cluster
ldentif icotion
IA
16
2
3
4
5
6
7
Type Strains in Cluster
Eiwinia
cacticida
chrysanthemi
Enterobacter agglomeruns,
Erwinia
herbicolu,
nigrifluens
E
caroto
vo
ra
a
fro sept icu,
E
carot
o
vo
fa
&
e to
vus
cu
lo
ru
m,
E.
corotoYoro
curofovoru
€.
quercina,
E
solicis,
tracheiphilu,
€.
ononus,
E.uredovora,
E
rhopontici
Klebsie
No
pneumoniae, Escherichio
GO/;
E
mallotivoro,
E
rubrifaciens,
E
psidih
E
amylo
voru,
stewartii
FIG.
2.
Simplified dendrogram showing major clusters and the type strains located in them.
SG
is the Gower similarity index modified to
count negative matches. The unweighted pair group average method
of
clustering was used. The length
of
the base
of
each triangle is
proportional to the number
of
strains in the cluster.
A
total
of
160
strains are represented in the dendrogram, which
is
based on 114
characteristics per strain.
361,
and 87-10-2x
[
=
S/N
771)
appeared to form more chains
than others. The cells were motile by means of peritrichous
flagella; fimbriae could occur (Fig. 1).
Colonies that were 18 to 24 h old on
PDP
were translucent
ivory and convex with entire margins. Such colonies fre-
quently had striations of dense material, giving
a
barred
appearance. Older colonies darkened slightly, becoming
opaque and losing the striations.
Numerical analyses.
The following characteristics were
positive or negative for all 160 strains tested and thus were
not included when we generated the dendrogram: oxidation
in the oxidation-fermentation test (positive), fermentation in
the oxidation-fermentation test (positive), catalase (posi-
tive), Gram staining (negative), oxidase (negative),
API
20E
arginine dihydrolase test (negative) and
H,S
test (negative),
API
SOCHE
L-xylose test (negative) and P-methylxyloside
test (negative), and fatty acids
C12:o
(positive),
C14:o
(posi-
tive),
c16:o
(positive), and
C16:1
(positive). The follow-
ing
API
SOCHE
characteristics were positive for all strains
but were included in cluster analysis (Fig.
2)
because strains
turned positive at different rates: D-glucose, D-fructose,
D-mannitol, and N-acetylglucosamine tests. The analysis of
duplicate strains indicated that the average probability
of
an
erroneous test result for the entire data set was less than
2%.
However, three characteristics (phosphatase and fatty acids
and
11)
showed individual error rates of more than
15%
and therefore were deleted from the database. This left
a total of 114 observations per strain that were used in
developing a dendrogram (Fig.
2).
The data utilized were
from the 1988 and 1989 physiological tests, from the 1989
soft-rot tests, and from motility, temperature, and fatty acid
c,,:,
VOL.
41, 1991
ERWZNZA CACTZCZDA
SP.
NOV.
205
TABLE
2.
Soft-rot capabilities
of
cactus and reference strains categorized into clusters
by
numerical analysis"
Host
Cluster 1A Cluster 1B
No.
of
%
of
strains No.
of
%
of strains
strains positive strains positive
Clusters
2
and
4
~
No.
of
%
of
strains
strains positive
~~
Clusters
3,
5,
and
6
No.
of
%
of
strains
No.
of
%
of strains
strains positive strains positive
Cluster
7
Cacti
Carnegiea gigantea
Lophocereus schottii
Opuntia ficus-indica
Pads
Fruits
Stenocereus thurberi
Daucus carota
root slices
Lycopersicon esculentum
fruits
Solanurn tuberosum
tuber slices
Noncac ti
98
14
9
68
98
98
58
98
83 3 100
71
78
91
93
3
100
9
3
0
97
3
100
100
41
44
6
100
1
100
41 66
41
37
41 76
23 70
41 100
15
0
3
8
8
15
15
0
3
15
0
3
5
15 6
3
0
0
0
0
0
0
0
0
~ ~~
a
Data from 1989 inoculations of saguaro and organ pipe cacti, carrots, and potatoes, 1978 inoculations of senita and Indian
fig
cacti, and 1973 inoculations of
tomatoes. The clusters correspond to the following taxa: cluster lA,
Erwinia cacticidcr;
cluster lB,
Erwinia cacticida-like;
cluster
2,
Erwinia chrysantherni;
cluster
4,
Erwinia carotovora;
clusters
3,
5,
and 6, other
Erwinia
spp.; cluster 7,
Eschericia coli
and
Klebsiella
spp.
studies (see below). The cluster analysis indicated that most
cactus strains fell into one major group, cluster
1A
(Fig.
2
and Table 1). This cluster contained 83 strains isolated from
soft rots of
11
types of cacti from Arizona, Texas, northern
Mexico, and Australia,
12
strains isolated from the corolla
areas of apparently normal saguaro cactus postblooms, and
3 strains from soils associated with rotting cacti. Three
cactus strains made up cluster
lB,
while
11
strains (mostly
isolated from postblooms) were members of cluster 4. Clus-
ter
1
was distinct at the 83% level from cluster
4,
which
contained not only several cactus strains but also a number
of agricultural strains and the type cultures for the subspe-
cies of
Erwinia carotovora.
In turn, the cluster 4 strains were
distinct from the remaining
Erwinia
and
non-Erwinia
refer-
ence strains. Strain ATCC 11773 of Bryan and May, which
was isolated from a saguaro cactus, was in cluster 6 along
with the
Erwinia
species that were least similar (71%) to the
majority of the cactus strains.
Soft-rot tests.
Cactus strains, as well as
Erwinia caroto-
vora
and
Erwinia chrysanthemi
reference strains, caused
soft decays of a number of the cacti tested and of tomato
fruits and potato slices (Table 2). However, only a few of the
cactus strains rotted carrot slices. (Inoculations of saguaro
cacti in 1962, 1964, 1969, and 1978 and of carrots and
potatoes in 1973 and 1978 gave similar results.) With one
exception,
Erwinia, Klebsiella,
and
Escherichia coli
refer-
ence strains did not cause tissue softening in any test; the
only exception was
Erwinia rhapontici,
which softened
potato slice tissue. Strain ATCC 11773 of Bryan and May,
which differed
so
greatly from other cactus strains in phys-
iological tests, also did not cause soft rot in any test.
Physiological studies.
The reactions of all of the strains
tested in API
20E
and API SOCHE tests and in supplemental
physiological tests are shown in Table 3. Strains were
grouped according to the clusters defined by the dendrogram
(Fig.
2).
A taxonomically and ecologically interesting obser-
vation is that only the cluster
1A
strains grew at 43°C.
It should be noted that the API 20E and API SOCHE strips
contained nine carbohydrates in common. However, the test
conditions differed (e.g., API SOCHE cupules were covered
with sterile mineral oil). Thus, the results from API SOCHE
cupules were often different, with the amygdalin and
myo-
inositol tests showing the most pronounced differences (Ta-
ble 3). As recommended by Verdonck et al. (33), data from
both types of strips were included in the numerical analysis
to generate Fig.
2.
When the test results for the nine API 20E
carbohydrates, as well as acid production from glucose at 36
and
43°C
(somewhat duplicated by growth in
YS
broth at 36
and 43"C), were omitted from the data set, the resulting
dendrogram (not shown) had the same major groupings
as
those shown in Fig.
2.
The only difference was that cluster
5
(in Fig.
2)
was inserted between clusters 3 and
4;
clusters
1A
and
1B
were unchanged. Also, in identifying cluster
1A
isolates, the API SOCHE L-arabinose test (no strain positive)
was somewhat more reliable than the API 20E arabinose test
(6% of strains positive) (Table 3).
To
determine the stability of test reactions, data from
previous determinations were reviewed. Nineteen physio-
logical tests have been repeated with available cactus strains
at intervals ranging from 10 to 26 years. Except for the
variability in the API
20E
test results noted below, results
for the following tests were consistent for tests conducted in
1962 to 1964, 1970 and 1971, 1977 and 1978, and 1988 and
1989: pectate (positive), oxidase (negative), nitrate (posi-
tive), Voges-Proskauer (positive), indole (negative), glucose
(positive), and sucrose (positive). Results for the following
tests were consistent for tests conducted
in
1962 to 1964,
1970 and 1971, and 1988 and 1989: salicin (positive), fructose
(positive), galactose (positive), dulcitol (negative), and mal-
onate (positive). Results for the following tests were consis-
tent for tests conducted in 1962 to 1964, 1977 and 1978, and
1988 and 1989: mannitol (positive) and rhamnose (positive).
Results for the following tests were consistent for tests
conducted in 1970 and 1971, 1977 and 1978, and 1988 and
1989: melibiose (negative) and a-methylglucoside (negative).
Results for the following test were consistent for tests
conducted in 1962 to 1964,1970 and 1971, and 1977 and 1978:
5%
NaCl (positive). Results for the following tests were
consistent for tests conducted in 1962 to 1964 and 1970 and
1971: formate (positive) and KCN (negative). With the
exception of the NaC1, formate, and KCN reactions, the
results of all of these tests were included
in
the dendro-
graphic analysis. The 1978 API 20E test results agreed with
the 1988 results with some variability in the citrate, Voges-
Proskauer, and nitrate tests (the API 20E media in 1978 were
inoculated with suspensions of bacteria grown on PDP [as
opposed to nutrient agar in 19881 and were more frequently
negative for citrate in 1978).
206
ALCORN ET
AL.
INT. J.
SYST.
BACTERIOL.
TABLE
3.
Percentage
of
positive reactions in physiological and other tests for strains in clusters determined by numerical analysis"
Test
%
of
positive
strains
Cluster
1A
Cluster
1B
Cluster
2
Cluster
3
Cluster
4
Cluster
5
Cluster
6
Cluster
7
(n
=
98)b
(n
=
3)
(n
=
1)
In
=
3)
(n
=
40)
(n
=
8)
(n
=
4)
(n
=
3)
API 20E tests
ONPG hydrolysis"
Lysine decarboxylase
Ornithine decarboxylase
Citrate
Urease
Tryptophan deaminase
Indole
Voges-Proskauer (acetoin)
Gelatin liquefaction
Glucose
Mannitol
Inositol
Sorbitol
Rhamnose
Saccharose
Melibiose
Am ygdalin
L-
Arabinose
Nitrate reduction
API 5OCHE tests
Glycerol
(
l)d
Erythritol (2)
D-Arabinose
(3)
L-Arabinose
(4)
D-Ribose
(5)
Adonitol
(8)
D-Galactose
(10)
D-Mannose
(13)
L-Sorbose
(14)
L-Rhamnose
(15)
Dulcitol
(16)
myo-Inositol
(17)
D-Sorbitol
(19)
a-Methyl-D-mannoside
(20)
a-Methyl-D-glucoside
(21)
Amygdalin
(23)
Arbutin
(24)
Esculin (25)
Salicin
(26)
D-Cellobiose
(27)
D-Maltose
(28)
Lactose
(29)
Melibiose
(30)
Saccharose
(31)
Trehalose
(32)
Inulin
(33)
Melezitose
(34)
D-Raffinose
(35)
Starch
(36)
Glycogen
(37)
Xylitol
(38)
P-Gentiobiose
(39)
D-Turanose
(40)
D-Tagatose
(42)
D-Fucose
(43)
L-Fucose
(44)
D-Arabitol
(45)
L-Arabitol (46)
Gluconate
(47)
2-Keto-gluconate
(48)
5-Keto-gluconate
(49)
D-Xylose
(6)
D-LyXOSe
(41)
Pectate
100
0
0
95
0
0
0
96
0
100
100
0
0
99
100
0
100
6
98
83
0
0
0
100
29
0
100
100
0
100
0
0
1
0
0
7
99
100
100
0
5
20
0
100
77
0
1
0
0
0
0
91
1
0
0
0
0
0
0
4
0
0
100
100
0
0
100
0
0
0
33
0
100
100
0
0
33
100
0
100
100
100
67
0
0
100
100
0
0
67
67
0
100
0
0
0
0
0
0
100
100
100
0
0
0
0
100
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
33
100
100
0
0
100
0
0
100
100
0
100
100
100
0
100
100
100
100
100
100
100
0
0
100
100
100
0
100
100
0
100
0
100
0
0
0
0
100
100
100
100
0
0
100
100
0
100
0
100
0
0
0
0
0
0
0
0
0
0
0
0
0
0
100
67
0
0
0
0
0
0
100
0
100
100
33
33
100
100
33
100
100
67
67
0
0
100
100
100
0
100
100
0
100
0
67
33
0
0
0
100
100
100
0
67
0
33
100
100
0
0
33
0
0
0
33
0
0
0
33
0
100
0
67
67
0
0
100
0
0
90
0
0
20
98
8
98
100
0
8
95
100
65
100
100
100
75
0
3
100
100
98
0
100
100
0
95
3
93
8
0
35
68
100
100
100
100
33
93
85
100
93
3
0
90
0
0
0
100
8
0
0
0
3
8
0
23
0
0
100
50
0
0
38
0
0
0
88
0
100
88
0
0
0
100
13
25
38
25
75
0
0
50
88
25
0
75
75
0
25
0
25
13
0
25
0
38
38
38
0
13
13
25
100
50
0
0
25
0
0
0
25
13
0
0
0
0
0
0
38
0
0
0
100
0
0
75
0
50
50
75
0
100
100
50
75
100
100
100
100
100
75
75
25
0
100
100
75
25
100
100
25
100
50
100
75
50
0
75
100
75
100
75
100
75
100
100
100
0
50
100
25
0
50
75
50
50
0
0
0
50
25
100
75
100
0
100
100
33
67
33
0
67
33
0
100
100
67
100
100
67
100
67
100
100
100
0
67
100
100
67
67
100
100
67
100
67
67
100
0
67
0
67
100
67
67
100
100
100
67
100
0
33
67
67
33
33
100
33
0
100
0
100
67
33
100
67
33
33
Continued
on
following
page
VOL.
41, 1991 ERWINIA CACTICIDA
SP.
NOV.
207
TABLE
3-Continued
Test
%
of positive strains
Cluster 1A Cluster
1B
Cluster
2
Cluster
3
Cluster
4
Cluster
5
Cluster
6
Cluster
7
(n
=
98)b
(n
=
3)
(n
=
1)
(n
=
3)
(n
=
40)
(n
=
8)
(n
=
4)
(n
=
3)
Phosphatase
Reducing substances
Motility
Erythromycin
Malonate
Tartrate
Rots saguaro cactus
Rots organ pipe cactus
Rots carrot slice
Rots potato slice
Acid produced from glucose
Growth at
36°C
in
YS
broth
Acid produced from glucose
Growth at
43°C
in
YS
broth
Yellow pigment
produced from sucrose
at
36°C
at
43°C
47
0
95
1
100
0
83
93
9
97
100
100
99
99
0
67
0
100
0
100
0
67
100
0
100
100
100
0
67
0
100
0
100
100
100
0
100
100
100
100
100
100
0
0
0
67
0
67
33
67
33
0
0
0
0
100
100
0
0
67
68
30
80
5
0
0
43
35
75
100
75
75
0
0
0
38
25
63
88
0
13
0
0
0
0
25
25
0
0
13
75
25
50
75
25
50
0
0
0
25
25
25
0
0
50
100
0
33
0
33
33
0
0
0
0
100
100
100
100
0
See text for the numerical analysis procedures which we used. For API
SOCHE
tests the results given are the percentages of strains that were positive within
n
is
the number of strains tested. Clusters were based
on
the dendrogram shown
in
Fig.
2.
ONPG,
o-Nitrophenyl-P-D-galactopyranoside.
The numbers in parentheses are the numbers of the cupules on API
SOCHE
strips; a missing number indicates that the test was positive or negative for all
48
h. See text for tests
in
which reactions were positive or negative for all strains.
strains.
Fatty acid analyses.
The following fatty acids were present
in all 160 strains tested: dodecanoic acid (C12J, tetrade-
canoic acid (C14:o), hexadecanoic acid (C16:J, and
cis-9-
hexadecenoic acid (C16:l
9).
In addition, 31 identified and
2 unidentified fatty acids and 2 long-chain alcohols were
recovered from the 160 bacterial strains tested (Table 4). The
chain lengths of the known acids and alcohols ranged from 10
to 21 carbon atoms; 26 (66.7%) of these compounds were
saturated (Table 4). The following fatty acids were particularly
useful for distinguishing between the cactus strains in clusters
1A and 1B and the soft-rot strains in cluster 4: hendecanoic
acid (Cl1:J, 3-hydroxydodecanoic acid (Clzz0 tridecanoic
acid (C13:J, 3-hydroxypentadecanoic acid
30H),
cis-7-
pentadecenoic acid (C15:l
,),
heptadecanoic acid (C1,:J,
cis-Pheptadecenoic acid (Cl,:l
J,
cyclo-(11,12)-nonade-
canoic acid (C19:O
cycle
cll-12),
and unknown acid 13.961.
Decanoic acid (Clo:o) was present in most of the cluster 4
strains and in many of the cluster
1
strains but was absent in all
other strains except the
Erwinia stewartii
strain
(S/N
153).
The standard MIS library (version 3
.O)
identified 37 cluster
1A strains
as
Enterobacter (Pantoea) agglomerans,
17
strains as
Erwinia sakis,
14 strains as
Escherichia vulneris,
1
strain as
Hafnia alvei,
and
1
strain as
Serratia plymuthica
at an acceptable confidence level. The remaining cluster 1A
strains were not identified to species. In dendrographic
analyses (data not shown) based only on fatty acid charac-
teristics, the strains in cluster
1
and
Erwinia chrysanthemi
segregated together and were separate from all other soft-rot
strains (cluster 4). However, also included with these organ-
isms were the following typically non-soft-rot species:
Er-
winia herbicola, Erwinia mallotivora, Erwinia nigrifuens,
Erwinia quercina, Erwinia rubrifaciens, Erwinia sakis,
and
Erwinia tracheiphila.
We also studied the ratios of fatty
acids as suggested previously (10) for subspecies of
Erwinia
carotovora,
but variability in the ratios precluded their use.
DNA
relatedness.
All
of
the strains tested were screened
for relatedness to strain 1-12T
(=
S/N
45T) in 75°C reactions,
in which only closely related sequences can reassociate. All
of the strains that exhibited more than 70% relatedness at
75°C were subsequently tested for relatedness and diver-
gence in 60°C reactions. Two subcultures of strain 1-12T
(strains ICPB EC186T and ICMP 1551-66T) were included in
the DNA tests and exhibited
100
and 93% relatedness to
labeled strain 1-12T at 60"C, respectively. A total of 22 other
cluster 1A strains were highly related to cluster 1A strain
1-12T. The average levels of relatedness were 88% (range, 82
to 99%) in 60°C reactions (with 1.0% divergence in related
sequences) and 87% (range, 76 to
100%)
in 75°C reactions
(Table
5).
Strains 88-2
(=
S/N
99) and 88-4
(=
S/N
lOl),
members of cluster
lB,
were 62 and 69% related to labeled
strain 1-12T in 60°C reactions. When 88-2 was labeled, 88-4
was closely related at 60°C
(85%),
whereas 1-12T was not
closely related (53%). When labeled Australian strain DU
89-8.1
(=
S/N
68) was compared with Australian strain DU
89-6.1
(=
S/N 63), cluster 1A strain 1-12T, and cluster
1B
strain 88-2, the levels of DNA relatedness at 60°C were 97,
84, and 60%, respectively. Conversely, cluster 4 cactus
strains 64-OP-6
(=
S/N
115), 65-215
(=
S/N
134), and 65-46-1
(=
S/N
142) exhibited 24, 31, and 44% relatedness, respec-
tively, to labeled 1-12T at 75°C. All type strains exhibited low
levels of relatedness to 1-12T; the type strains of the
Erwinia
carotovora
group (25 to 38% relatedness at 75°C) were the
closest. Strain ATCC 11773
(=
S/N
156)
(Erwinia
sp.), which
was isolated by Bryan and May from a saguaro cactus (4),
was only 11% related to strain 1-12T at 75°C. Strains ATCC
27155T
(=
S/N
103)
(Enterobacter agglomerans),
ATCC
33243T
(=
S/N
104)
(Erwinia herbicola),
and ATCC 15712T
(=
S/N
147)
(Erwinia sakis),
type strains of species that
were closely identified with cluster 1A strains in the fatty
acid analysis, exhibited only 9, 22, and 12% relatedness,
respectively, to 1-12~ at 75°C.
208
ALCORN
ET
AL.
INT.
J.
SYST.
BACTERIOL.
TABLE
4.
Percentages
of
strains in each cluster having detectable fatty acids and long-chain alcohols"
-
%
of
strains in:
Fatty acid methyl
ester structure Cluster
1A
Cluster
1B
Cluster
2
Cluster
3
Cluster
4
Cluster
5
Cluster
6
Cluster
7
(n
=
98)b
(n
=
3)
(n
=
1)
(n
=
3)
(n
=
40)
(n
=
8)
(n
=
4)
(n
=
3)
lo:o
10:0
3
OH
1l:O
12:O 2
OH
12:O 3
OH
13:O
13:O
IS0
14:O
IS0
3
14:O 2
OH
15:O
15:O
IS0
15:O
ANTEISO
15:O 3
OH
15:l
CIS
7
16:O
(alcohol)
16:O
IS0
16:O 3
OH
16:l
CIS
7
16:l
CIS
11
17:O
17:O ANTEISO
17:O 3
OH
17:O
Cyclo
17:l
CIS
9
17:l
CIS
10
17:l
CIS
11
17:l
ISO,
A5
18:O
18:l
TRANS
9
(alcohol)
18:l
CIS
9
19:O
IS0
21:l
TRANS
11
Unknownsc
13.961
14.503
19:O
CyClO
Cll-12
73
19
0
0
31
1
0
0
0
35
1
1
0
0
0
0
0
0
16
8
0
0
99
0
1
0
1
81
0
0
10
90
1
28
99
33
0
0
0
0
33
0
0
0
100
0
0
0
0
0
0
0
0
0
100
0
0
100
0
0
0
0
67
0
0
0
67
0
0
100
0
0
0
0
0
0
0
0
0
100
0
0
0
0
0
0
0
0
0
100
0
0
100
100
0
0
0
100
0
0
0
100
0
0
100
0
0
0
0
0
0
0
0
0
67
0
0
0
0
0
0
0
0
0
33
0
0
100
0
0
0
0
67
0
33
0
33
0
0
100
98
0
55
0
58
100
0
3
0
100
0
3
63
55
0
3
18
3
53
100
3
3
45
100
0
45
3
90
0
0
0
0
3
73
95
13
0
0
0
50
13
0
0
0
38
0
0
0
0
0
0
13
0
25
25
0
0
63
13
0
0
0
63
0
0
0
25
0
50
100
0
0
0
25
25
100
25
0
100
100
0
0
0
0
25
0
0
0
25
100
0
0
100
25
0
0
0
100
25
0
0
75
0
100
100
0
0
0
33
0
67
0
0
33
100
0
0
0
0
0
0
0
0
33
67
0
0
100
33
0
0
0
67
0
0
0
67
0
33
100
a
See text for fatty acids which were detected in
all
isolates.
n
is the number
of
strains tested.
Designations are the retention times relative to the retention times
of
known, straight-chain index peaks.
DISCUSSION
By using numerical taxonomy, tissue-macerating bacterial
strains isolated from cacti and soils beneath or adjacent to
rotting cacti (i.e., cactus strains) were compared with type
strains and reference strains belonging to the genus
Erwinia
and related genera. Analyses of 114 characteristics (includ-
ing API, fatty acid, soft-rot, and temperature sensitivity
tests) indicated that 98 (87.5%) of the cactus strains fell into
cluster 1A (Fig. 2), which had an internal similarity level
of
93%. This cluster joined cluster 1B (three cactus strains) at
the 89% level. No reference strains segregated in either
cluster 1A or cluster 1B; the closest reference organisms
were
Erwinia chrysanthemi
(cluster 2) (similarity level, 83%)
and
Enterobacter agglomerans, Erwinia herbicola,
and
Er-
winia nigrijluens
(cluster 3) (similarity level, 82%). The
members of all other clusters exhibited less than 80% simi-
larity
to
clusters 1A and 1B.
DNA studies confirmed the uniqueness
of
cluster 1A; the
levels of relatedness to strain 1-12T ranged from 82 to 99% at
60°C for
22
selected cactus strains representing the spectrum
of cluster 1A. Conversely, DNAs from 22 reference type
cultures, including
20
cultures of organisms belonging
to