The comparative toxicity of a reduced, crude comfrey (Symphytum officinale) alkaloid extract and the pure, comfrey-derived pyrrolizidine alkaloids, lycopsamine and intermedine in chicks (Gallus gallus domesticus)


Brown, A.W.; Stegelmeier, B.L.; Colegate, S.M.; Gardner, D.R.; Panter, K.E.; Knoppel, E.L.; Hall, J.O.

Journal of Applied Toxicology 36(5): 716-725

2016


Comfrey (Symphytum officinale), a commonly used herb, contains dehydropyrrolizidine alkaloids that, as a group of bioactive metabolites, are potentially hepatotoxic, pneumotoxic, genotoxic and carcinogenic. Consequently, regulatory agencies and international health organizations have recommended comfrey be used for external use only. However, in many locations comfrey continues to be ingested as a tisane or as a leafy vegetable. The objective of this work was to compare the toxicity of a crude, reduced comfrey alkaloid extract to purified lycopsamine and intermedine that are major constituents of S. officinale. Male, California White chicks were orally exposed to daily doses of 0.04, 0.13, 0.26, 0.52 and 1.04 mmol lycopsamine, intermedine or reduced comfrey extract per kg bodyweight (BW) for 10 days. After another 7 days chicks were euthanized. Based on clinical signs of poisoning, serum biochemistry, and histopathological analysis the reduced comfrey extract was more toxic than lycopsamine and intermedine. This work suggests a greater than additive effect of the individual alkaloids and/or a more potent toxicity of the acetylated derivatives in the reduced comfrey extract. It also suggests that safety recommendations based on purified compounds may underestimate the potential toxicity of comfrey.

Journal
of
Research
article
Received:
9
April
2015,
Revised:
1
June
2015,
Accepted:
2
June
2015
Published
online
in
Wiley
Online
Library.
14
July
2015
(wileyonlinelibrary.com
)
DOI
10.1002/jat.3205
The
comparative
toxicity
of
a
reduced,
crude
comfrey
(Symphytum
officinale)
alkaloid
extract
and
the
pure,
comfrey-derived
pyrrolizidine
alkaloids,
lycopsamine
and
intermedine
in
chicks
(Gallus
gallus
domesticus)
Ammon
W.
Brown
a
*,
Bryan
L.
Stegelmeier
a
,
Steven
M.
Colegate",
Dale
R.
Gardner',
Kip
E.
Panter
a
,
Edward
L.
Knoppel
a
and
Jeffery
0.
Hall`
ABSTRACT:
Comfrey
(Symphytum
officinale),
a
commonly
used
herb,
contains
dehydropyrrolizidine
alkaloids
that,
as
a
group
of
bioactive
metabolites,
are
potentially
hepatotoxic,
pneumotoxic,
genotoxic
and
carcinogenic.
Consequently,
regulatory
agencies
and
international
health
organizations
have
recommended
comfrey
be
used
for
external
use
only.
However,
in
many
locations
comfrey
continues
to
be
ingested
as
a
tisane
or
as
a
leafy
vegetable.
The
objective
of
this
work
was
to
compare
the
toxicity
of
a
crude,
reduced
comfrey
alkaloid
extract
to
purified
lycopsamine
and
intermedine
that
are
major
constituents
of
S.
officinale.
Male,
California
White
chicks
were
orally
exposed
to
daily
doses
of
0.04,
0.13,
0.26,
0.52
and
1.04
mmol
lycopsamine,
intermedine
or
reduced
comfrey
extract
per
kg
bodyweight
(BW)
for
10
days.
After
another
7
days
chicks
were
euthanized.
Based
on
clinical
signs
of
poisoning,
serum
biochemistry,
and
histopathological
analysis
the
reduced
comfrey
extract
was
more
toxic
than
lycopsamine
and
intermedine.
This
work
suggests
a
greater
than
additive
effect
of
the
individual
alkaloids
and/or
a
more
potent
toxicity
of
the
acetylated
derivatives
in
the
reduced
comfrey
extract.
It
also
suggests
that
safety
recommendations
based
on
purified
compounds
may
underestimate
the
potential
toxicity
of
comfrey.
Published
2015.
This
article
has
been
contributed
to
by
US
Government
employees
and
their
work
is
in
the
public
domain
in
the
USA.
Keywords:
comfrey;
lycopsamine;
intermedine;
pyrrolizidine
alkaloid;
N-oxide;
tisane
Cn
Introduction
Comfrey
has
been
used
for
a
wide
variety
of
medicinal
purposes
for
over
2000
years
(Rode,
2002),
and
it
continues
to
be
consumed
as
an
herbal
tea
or
a
vegetable
in
many
countries
(Mei
et
al.,
2010).
More
recently,
controlled
studies
have
found
multiple
comfrey-
based
topical
treatments
to
be
beneficial
for
treatment
of
a
variety
of
muscle
and
joint
pains
(Koll
et
al.,
2004;
Staiger
2012).
The
anti-
inflammatory
and
analgesic
effects
of
comfrey
are
thought
to
be
as
a
result
of
the
imidazolidinylurea
allantoin
and
the
phenylpropanoid
rosmarinic
acid
(Staiger,
2012).
Comfrey
also
contains
dehydropyrrolizidine
alkaloids
(DHPAs)
which
are
pro-toxins
that
are
hepatotoxic,
and
potentially
pneumotoxic,
genotoxic
and
carcinogenic
(Mei
et
al.,
2010).
As
a
result
of
the
health
hazards
posed
by
DHPAs,
health
organizations
and
food
and
drug
safety
agencies
in
several
countries
have
developed
regulations
and
recommendations
regarding
the
sale
and
use
of
comfrey.
In
2001,
the
US
Food
and
Drug
Administration
sent
an
ad-
visory
letter
to
manufacturers
of
dietary
supplements
requesting
that
they
remove
all
comfrey
products
intended
for
consumption
from
the
market
(FDA,
2001).
In
Germany,
the
Federal
Institute
for
Risk
Assessment
(BfR)
conducted
a
risk
assessment
for
DHPAs
and
concluded
that
exposure
should
be
kept
as
low
as
possible
limiting
tolerable
daily
intake
to
0.007
jig
of
unsaturated
pyrrolizi-
dine
alkaloids
per
kg
bodyweight
(BW)
(BfR,
2011).
In
2008,
the
UK
Committee
on
Toxicity
of
Chemicals
in
Food,
Consumer
Products
and
the
Environment
released
a
statement
on
DHPAs
in
food,
which
supported
that
of
the
BfR,
limiting
daily
oral
exposures
to
less
than
0.007
jig
DHPAs
kg
-1
BW
(Committee
on
Toxicity
of
Chemicals
in
Food,
2008).
The
World
Health
Organization,
Dutch
National
Institute
for
Public
Health
and
the
Environment,
and
European
Food
Safety
Authority
have
all
conducted
similar
reviews
with
similar
concerns
and
recommendations
(RIVM,
2005;
EFSA,
2011;
WHO,
2011).
The
Food
Standards
Australia
New
Zealand
Authority
recommends
a
somewhat
higher
tolerable
exposure
of
1
jig
DHPAs
kg
-1
BW
per
day
based
exclusively
on
hepatotoxicity
as
opposed
to
potential
carcinogenicity
(FSANZ,
2001).
*Correspondence
to:
Ammon
W.
Brown,
USDA/ARS
Poisonous
Plant
Research
Laboratory,
Logan,
UT
84341
USA.
E-mail:
°
USDA/ARS
Poisonous
Plant
Research
Laboratory,
Logan,
UT
84341,
USA
b
Animat
Dairy
and
Veterinary
Sciences,
Utah
State
University,
Logan,
UT
84341,
USA
`Utah
Veterinary
Diagnostic
Laboratory,
Utah
State
University,
Logan,
UT
84341,
USA
J.
Appl.
Toxicot
2016;
36:
716-725
comfrey
root
extract
C
intermedine
20
80
lycopsamine
80
40
20
6
Time
(min)
da
t
ive
Abu
n
da
nce
(
%
100
Bo
as
20
138
A
or
B:
intermedine/lycopsamine
94
120
t t
220 240
600
0i0
100
D:
7-acetylintermedine/lycopsamine
180
so
20
10
1.
80
180
200
.0
240 280
280
300 320
m/z
0
80
100
100
sa
70
eo
20
10
0
80
120
100
160
Comparative
toxicity
of
comfrey,
lycopsamine
and
intermedine
DHPA-related
toxicity
has
been
associated
with
its
in
vivo
oxida-
tion
to
a
didehydropyrrolizidine
alkaloid
metabolite,
referred
to
as
the
'pyrrolic'
or
dihydropyrrolizine
(DHP)
metabolite,
that
is
a
bi-
functional
alkylating
agent
of
macrobiomolecules
forming
adducts
with
proteins
and
DNA
(Edgar
et
al.,
2014).
The
indirect
detection
of
these
'pyrrolic'
adducts
by
their
oxidative
cleavage
from
the
tis-
sue
adducts
has
been
used
as
an
indicator
of
exposure
to
the
DHPAs
(Mattocks
and
Jukes,
1990,
1992a,
1992b;
Winter
et
al.,
1990;
Stegelmeier,
et
al.
1996;
Winter,
et
al.
1990;
Lin
et
al.,
2011).
Therefore,
the
observation
of
adverse
clinical
signs
or
pathological
changes
in
this
study
will
be
correlated
with
the
indirect
detection
of
'pyrrolic'
adducts
in
an
attempt
to
quantify
exposure.
A
California
White
male
chick
model
has
recently
been
investi-
gated
to
assess
the
comparative
toxicity
of
pure
DHPAs
in
a
biologically-relevant
manner
and
to
use
less
alkaloid
than
some
other
models
(Stegelmeier
et
al.,
unpublished
data).
Therefore,
the
objective
of
this
study
was
to
use
this
model
to
compare
the
toxicity
of
a
reduced,
crude
comfrey
alkaloid
extract
with
that
of
pure
lycopsamine
and
intermedine,
the
two
major
alkaloids
isolated
from
the
reduced
comfrey
extract.
Materials
and
methods
Animals
Three-day-old
male,
California
White
chicks
(Gallus
gallus
domesticus)
were
purchased
from
Privett
Hatchery
(Portales,
New
Mexico)
through
the
Intermountain
Farmers
Association
(Hyde
JH
0
yw
i
0
OH
R
2
Journal
of
d
Toxi
co
logy
Park,
Utah).
California
White
chickens
are
a
commercial
hybrid
resulting
from
the
cross
of
a
White
Leghorn
hen,
and
a
California
Grey
rooster
touted
as
a
hardy
breed
that
is
somewhat
easier
to
handle
than
the
White
Leghorn.
During
a
3-day
acclimation
period
before
initiation
of
treatment,
the
chicks
were
weighed,
and
out-
liers
culled
such
that
test
birds
were
of
a
uniform
size
and
body
condition.
Chicks
were
housed
in
heated
brooder
cages
that
pro-
vided
a
thermal
gradient
within
their
microenvironment
ranging
from
roughly
16
to
25
°C.
All
animals
had
free
access
to
fresh
water,
and
a
commercial
20%
protein
poultry
starter
purchased
from
the
Intermountain
Farmers
Association
(Salt
Lake
City,
UT,
USA).
The
brooder
cages
were
housed
in
a
windowed
room
and
thus
exposed
to
ambient
sunlight.
Room
lights
were
used
such
that
there
was
a
minimum
of
12
h
of
light
each
day.
Humidity
in
the
brooder
cages
was
essentially
room-ambient
humidity
that
ranged
from
approximately
30%
to
60%
during
this
study.
This
research
was
conducted
with
the
approval
of
the
Utah
State
University
Animal
Care
and
Use
Committee
(IACUC
Protocol
#2055).
Preparation
of
dehydropyrrolizidine
alkaloid
test
samples
Dry,
powdered
root
of
common
comfrey
(Symphytum
officinale)
was
purchased
from
Starwest
Botanicals
(Cordova,
CA,
USA)
or
Take
Herb
(Alhambra,
CA,
USA).
After
mixing
the
powdered
com-
frey
root
from
both
sources,
a
crude
extract
of
the
comfrey
root
al-
kaloids
(including
reduction
of
N-oxides
to
their
free
base
forms)
and
the
monoester
DHPAs
lycopsamine
and
intermedine
(Fig.
1)
were
isolated
from
a
reduced,
crude
alkaloidal
extract
of
the
R
3
0
R
1
=
R
3
=
H;
R
2
=
OH
:
intermedine
R
2
=
R
3
=
H;
R
1
=
OH
:
lycopsamine
R
1
=
H;
R
2
=
OH;
R
3
=
COCH
3
:
7-acetylintermedine
R
1
=
OH;
R
2
=
H;
R
3
=
COCH
3
:
7-acetyllycopsamine
Figure
1.
The
structures
of
lycopsamine,
intermedine
and
their
7-acetyl
derivatives.
The
HPLC-esi(+)MS
ion
chromatograms
show
a
high
purity
for
the
isolated
lycopsamine
(B)
and
intermedine
(A)
and
a
mixture
of
lycopsamine,
intermedine
and
their
7-acetyl
derivatives
(D)
in
the
comfrey
reduced
alkaloid
sample.
Also
present
in
the
latter
are
minor,
residual
concentrations
of
intermedine
and
lycopsamine
N-oxides
(C).
The
MS/MS
profiles
confirm
the
identity
of
lycopsamine
and intermedine,
and
C7
as
the
site
of
acetylation
rather
than
C13.
J.
Appl.
Toxicol.
2016;
36:
716
-
725
wileyonlinelibrary.com/joumal/jat
of
'A u
OredToxicology
powdered
comfrey
as
previously
described
(Colegate
et
al.,
2014).
The
extract
and
purified
samples
were
analyzed
using
high-pressure
liquid
chromatography
coupled
to
an
electrospray
ionization
mass
spectrometer
operated
in
the
positive
ion
full
scan
and
tandem
mass
spectrometry
mode
[HPLC-esi(+)MS
and
MS/MS].
Experimental
design
Eighty
chicks
were
randomly
divided
into
16
groups
of
five
and
assigned
to
a
dose
group
for
comfrey
extract,
lycopsamine
or
intermedine.
Each
test
compound(s)
was
administered
at
five
doses
i.e.
0.04, 0.13,
0.26,
0.52
and
1.04
mmols
DHPA
free
base
kg
-1
BW
per
day.
Five
animals
were
assigned
to
the
control
group
that
received
a
volume
of
absolute
ethanol
equal
to
the
largest
volume
received
by
any
of
the
other
groups.
Dosing
was
based
on
purified
alkaloid
toxicity
studies
using
the
same
model
(Stegelmeier
et
al.,
unpublished
data).
In
the
purified
alkaloid
stud-
ies,
doses
were
0.01,
0.04,
0.13
and
0.26
mmols
DHPA
free
base
kg
-1
BW
per
day.
Because
lycopsamine
was
found
to
be
less
toxic
than
the
other
purified
alkaloids
tested,
the
two
higher
doses
were
added.
Daily
doses
were
divided
into
two
equal
portions
half
of
which
was
given
in
the
morning,
and
the
other
half
was
given
in
the
afternoon
via
gavage.
Water
was
added
to
all
doses
such
that
the
final
volume
was
0.5
ml.
All
animals
were
weighed
three
times
per
week,
and
doses
were
recalculated
prior
to
the
next
treatment.
The
chicks
were
dosed
for
10
days
and
maintained
for
another
7
days
before
they
were
euthanized
with
carbon
dioxide
for
necropsy
examination.
Chicks
were
monitored
at
least
twice
daily
for
the
duration
of
the
study.
Any
animals
that
became
moribund
were
euthanized
with
carbon
dioxide
and
necropsied.
A.
W.
Brown
et
al.
Tissue
preparation
At
post-mortem
examination,
gross
abnormalities
were
recorded
and
the
entire
brain,
heart,
spleen
and
right
liver
lobe,
and
repre-
sentative
samples
of
lung,
kidney,
testicle,
crop,
proventriculus,
ventriculus,
small
intestine,
colon,
cecum
and
bursa
of
Fabricius
were
collected,
preserved
in
neutral-buffered
formalin,
embedded
in
paraffin
and
subsequent
sections
were
stained
with
hematoxy-
lin
and
eosin
via
standard
procedures
for
histopathologic
analysis.
A
random
number
chart
was
used
to
determine
the
start
location
for
sectioning
the
right
liver
lobe.
The
entire
right
liver
lobe
was
sectioned
every
3
mm
transversely
along
the
long
axis
prior
to
embedding
and
staining.
Blood
samples
were
taken
immediately
after
euthanasia
via
cardiac
puncture,
divided
into
serum
and
packed
cells,
and
frozen.
The
left
liver
lobe
was
collected
and
stored
at
-80
°C
for
'pyrrole'
adduct
analysis.
Histologic
grading
Liver
size
was
quite
variable
between
animals,
and
consequently
all
sections
of
the
right
liver
lobe
could
occupy
from
one
to
four
slides.
Each
slide
was
graded
by
examining
all
sections
on
that
slide
and
scored
for
necrosis
as
follows:
0
-
no
observed
hepatocel-
lular
necrosis;
1
-
scattered
individual
cell
death;
2
-
multifocal
areas
of
necrosis
which
encompassed
up
to
25%
of
the
sections
on
that
slide;
3
-
larger
areas
of
necrosis
were
present
accounting
for
between
25
and
50%
of
the
sections;
and
4
-
greater
than
50%
of
the
observed
sections
were
necrotic
(Fig.
2).
The
hepatic
necrosis
score
for
each
animal
was
computed
by
taking
the
average
of
the
necrosis
scores
for
each
slide
of
liver
tissue
from
that
animal.
'
4,
Figure
2.
Hematoxylin
and
eosin
(H&E).
(A)
200x.
Necrosis
score
0.
Normal
liver
from
a
control
chick.
(B)
400x.
Necrosis
score
1.
Liver
from
a
chick
exposed
to
0.52
mmol
kg
-1
body
weight
(BW)
per
day
comfrey
extract
with
a
necrosis
score
of
one.
Note
the
individual
cell
death
(arrow).
(C)
200x.
Necrosis
score
2.
Liver
from
a
chick
exposed
to
1.04
mmol
kg
-1
BW
per
day
comfrey
extract
with
a
necrosis
score
of
two.
Note
the
individualized
hepatocytes
that
are
surrounded
by
hemorrhage
(arrows)
and
the
lake
of
hemorrhage
(*)
where
hepatic
cords
should
be.
(D)
100x.
Necrosis
score
3.
Liver
from
a
chick
exposed
to
1.04
mmol
kg
-1
BW
per
daywith
a
necrosis
score
of
three.
Note
the
large
lakes
of
hemorrhage
(*),
with
occasional
rafts
of
hepatocytes.
wileyonlinelibrary.com/journal/jat
J.
Appl.
Toxic&
2016;
36:
716-725
Metabolic
Oxidation
Comfrey
Alkaloids
Didehydro-Alkaloid
A
N
Pyrrole
Adduct
AgNO
3
/Et0H
—,..
o
O
Protein
Adduction
o
-
100-
100-
4.55
Chick
Liver
Extract
5.99
6
46
TIC
4.3
50:
3.12
3.
6
Pyrrole-DMABA
Chick
Liver
50
4.18
RIC
m/z
296
6
7
8
9
10
11
12
13
14
Time
(min)
Re
la
tive
Abun
dan
C
j
A
u
Ol
of
edToxi co
I
ogy
Comparative
toxicity
of
comfrey,
lycopsamine
and
intermedine
'Pyrrole'
adduct
detection
The
indirect
detection
of
hepatic
'pyrrolic'
adducts
was
an
adaptation
of
methods
originally
reported
by
Mattocks
and
Jukes
(1992a,
1992b)
and
more
recently
described
by
Lin
et
al.
(2011)
(Fig.
3).
Monocrotaline
(crotaline),
tetrachloro-o-benzoquinone
(o-chloranil,
97%),
boron
trifluoride
diethyletherate
[BF
3
.O(C
2
H
5
)
2
]
and
silver
nitrate
were
purchased
from
Sigma-Aldrich
(St.
Louis,
MI,
USA).
p-Dimethylaminobenzaldehyde
(DMABA)
was
from
ICN
Biochemical
(Cleveland,
OH,
USA).
Acetonitrile
was
HPLC
grade
(Sigma-Aldrich),
and
absolute
ethanol
was
(reagent
grade).
Formic
acid
(98%;
Sigma-Aldrich)
used
for
HPLC
was
diluted
with
purified
water
(18.2
MO
cm
-1
)
(WaterPro
PS
Station,
Labconco,
Kansas
City,
MO,
USA)
to
a
concentration
of
0.1%
(V/V).
Ehrlich's
solution
was
prepared
by
dissolving
DMABA
(0.2
g)
in
ethanol
(10
ml)
with
purified,
redistilled
BF
3
.O(C
2
H
5
)
2
(2
ml).
0
RO
0
OH
OH
RO
Dehydromonocrotaline
(DHMC),
used
for
the
positive
control
standard,
was
prepared
by
oxidizing
monocrotaline
using
a
modi-
fication
of
procedures
previously
reported
(Culvenor
et
al.,
1970).
o-Chloranil
(30
mg)
was
added
to
a
solution
of
monocrotaline
(30
mg)
in
chloroform
(4
ml)
and
shaken
for
30
to
60
s.
A
saturated
solution
of
potassium
hydroxide
(KOH)
containing
2%
(w/v)
so-
dium
borohydride
(NaBH
4
)
(4
ml),
purchased
from
Mallinckrodt
(Paris,
Kentucky),
was
added
to
the
oxidation
solution
and
shaken
until
the
color
faded.
The
solvent
layers
were
separated
by
centri-
fugation;
the
organic
layer
was
filtered
through
anhydrous
sodium
sulfate;
and
the
solvent
was
removed
by
roto-evaporation
at
40
°C
and
reduced
pressure
to
afford
the
oxidation
product
(30
mg).
Integration
and
comparison
of
the
H2, H3,
H7
and
H9
resonance
signals
for
DHMC
and
monocrotaline
in
the
1
H-NMR
spectrum
(Culvenor
et
al.,
1970)
(Fig.
4)
of
the
oxidation
product
indicated
an
87:
13
mixture
of
DHMC
to
residual
monocrotaline.
A
measured
,
Protein/DNA
RO
0
OH
OH
LC-MS/MS
N
Ehrlich's
/
Reagent
100:
TIC
Pyrrole-DMABA
Standard
B
50:
o
100--
252.17
80
70
60
MS/MS
Spectrum
50
40-
5
.
30-
20-
10-
267.15
.
I
100
120 140
160 180
200
220
240
260 280
300
320 340
Mk
Figure
3.
Schematic
outline
of
the
indirect
detection
of
pyrrole
liver
adducts
starting
with
the
metabolic
oxidation
of
comfrey
alkaloids,
hydrolysis
of
pyrrolic
adducts
and
formation
of
pyrrole-DMABA
complex
(A).The
HPLC-esi(+)MS/MS
total
ion
chromatograms
(TIC)
and
reconstructed
ion
chromatograms
(RIC)
from
pyrrole-DMABA
standard
(top)
and
then
that
obtained
from
chick
liver
(B:
middle
and
bottom).
MS/MS
spectrum
of
the
pyrrole-DMABA
standard
(C).
J.
Appl.
Toxicol.
2016;
36:
716-725
wileyonlinelibrary.com/joumal/jat
Re
la
tive
Abu
n
da
nce
Pyrrole
-
DMABA
Standard
29
.
19
34
.23
monocrotaline
5.0
I.5
1.0
11
SOON
dehydromonocrotaline
H2
I
H7
H9
of
'A u
OredToxicology
A.
W.
Brown
et
al.
HO
CHa
0
Hi"
H9
O
OH
°
O
H7
I
H2
H3
monocrotaline
dehydromonocrotaline
Figure
4.
H-NMR
Spectrum
of
dehydromonocrotaline.
The
ratio
of
dehydromonocrotaline
to
that
of
the
residual
monocrotaline
was
measured
from
inte-
gration
of
the
H2
and
H3
protons
from
dehydromonocrotaline
versus
the
H7
and
H9
protons
from
the
residual
monocrotaline.
?0.
CIS
946
9.76
9.70
9.45
CIO
9
34
CAA
9.25
20.
9.15
919
*I
66
-0.05
quantity
of
oxidation
product
was
dissolved
in
absolute
ethanol
to
give
a
2.69
µmol
m1
-1
pyrrole
equivalent
standard
stock
solution.
Ethanolic
silver
nitrate
was
prepared
by
dissolving
silver
nitrate
(625
mg)
in
deionized
water
(0.5
ml)
with
sonication
for
1-2
min.
Absolute
ethanol
(25
ml)
was
added
and
the
mixture
and
soni-
cated
until
all
of
the
silver
nitrate
had
redissolved
(approximately
5
min).
'Pyrrole'
calibration
standards
were
prepared
by
diluting
an
ali-
quot
(50
µI)
of
the
pyrrole
standard
stock
solution
with
ethanol
(900
µI)
and
Ehrlich's
solution
(50
µI)
to
afford
a
'pyrrole'
concentra-
tion
of
135
nmol
m1
-1
.
An
aliquot
(50
µI)
was
further
diluted
with
ethanol
(950
µI)
i.e.,
6.8
nmol
'pyrrole'
m1
-1
which
was
further
serially
diluted
(1/5)
to
give
standards
at
1.35,
0.27,
0.054,
0.11
and
0.002
nmol
'pyrrole'
m1
-1
.
Entire
left
liver
lobes
frozen
from
individual
chicks
were
freeze-
dried
using
a
Labconco®
freeze
dryer.
Portions
of
the
lyophilized
liver
lobes
were
placed
in
conical
tubes
with
small
copper-coated
steel
spheres.
The
conical
tubes
were
then
placed
in
a
Retech®
MM301
shaker
at
20
revolutions
s
-1
for
10
min.
An
accurately
weighed
portion
(c.
25
mg)
of
the
crushed,
lyophilized
chicken
liver
was
then
placed
into
a
plastic
snap-cap
conical
tube.
Ethanolic
sil-
ver
nitrate
(1.0
ml)
was
added
to
the
sample
and
the
sample
mixed
for
30
min
on
an
auto
rotator.
Trifluoracetic
acid
(10
µI)
was
added
to
the
mixture
after
30
min,
and
samples
further
mixed
on
the
auto
rotator
overnight
(-16
h).
The
samples
were
then
centrifuged
at
13000
rpm
(16000
G)
for
10
min.
An
aliquot
(20
µI)
of
the
superna-
tant
was
added
to
absolute
ethanol
(170
µI)
containing
Ehrlich's
reagent
(10
µI)
in
an
HPLC
autosampler
vial.
The
HPLC-esi(+)
MS/MS
system
consisted
of
a
Agilent
1260
Infin-
ity
HPLC
System
(Agilent
Technologies,
Santa
Clara,
CA,
USA).
Samples
(5
µI)
were
injected
onto
a
Synergi
Hydro
RP
column
(75
x
2
mm,
4
µ)
fitted
with
a
guard
column
of
a
similar
adsor-
bent.
A
gradient
flow
(300
µI
min
-1
)
of
0.1%
formic
acid
in
water
(A)
and
acetonitrile
(B)
was
used
to
elute
the
sample
components
from
the
column
using
the
following
linear
gradient:
20%
B
(0-1
min);
20%-45%
B
(1-2
min);
45%-75%
B
(2-11
min);
75%
B
(11-15
min);
75%-20%
B
(15-16
min);
and
20%
B
(16-21
min)
for
re-equilibration
of
the
column.
Flow
from
the
HPLC
column
was
connected
to
a
Velos
Pro
LTQ
mass
spectrometer
(Thermo
Scientific,
Waltham,
MA,
USA)
equipped
with
a
heated
electrospray
ionization
(HESI)
source.
The
capillary
temperature
was
set
at
275
°Q
the
ionization
spray
voltage
at
3.45
kV,
the
HESI
source
heater
temperature
at
305
°C
and
the
sheath
gas
flow
was
40
units
with
an
auxiliary
flow
of
5
units.
The
mass
spectrometer
was
monitored
in
a
positive
ion
MS/MS
mode,
scanning
product
ions
from
a
mass
range
of
m/z
90-350
after
fragmentation
of
the
parent
ion
(MH
+
=
341)
using
a
relative
collision
energy
setting
of
45%
with
high-energy
collision-
induced
dissociation
(HCD).
Under
these
conditions
the
'pyrrole'-
DMABA
compound
eluted
with
a
retention
time
of
4.4
min
and
the
resulting
MS/MS
spectrum
contained
major
fragment
ions
at
m/z
252
and
296.
The
detected
'pyrrole'
peak
area
was
measured
from
the
reconstructed
ion
chromatogram
displaying
m/z
296
(Fig.
3)
and
quantitated
based
on
an
external
calibration
curve
established
from
the
standards.
The
resulting
'pyrrole'
concen-
tration
(nmol
m1
-1
)
of
the
injected
sample
was
converted
to
nmol
g
-1
liver.
Serum
biochemistry
analysis
Frozen
serum
samples
were
outsourced
(Antech
Laboratories,
Indianapolis,
IN,
USA)
for
analysis
of
total
protein,
glucose,
triglycerides,
cholesterol,
gamma
glutamyl
transferase
(GGT),
aspartate
aminotransferase
(AST),
lactate
dehydrogenase
(LDH),
sorbitol
dehydrogenase
(SDH),
bile
acid
and
creatine
phosphoki-
nase
(CPK).
wileyonlinelibrary.com/journal/jat
J.
Appl.
Toxic&
2016;
36:
716-725
9/
a
(4.5)
117
a
(86)
9A
a
(7.6)
56"
(55)
2.5"
(1.0)
15"
(3)
3.3"
(32)
24"
(17)
2.2"
(0.5)
35"
(11)
0.7"
(0.4)
13"
(10)
1.4"
(0.9)
12"
(8)
0.6"
(0.5)
15"
(8)
0.9"
(1.0)
28"
(12)
0.5"
(0.6)
23"
(10)
1.7"
(1.9)
14"
(7)
0.2"
(02)
15"
(7)
2.3
1
'(12)
13
b
(4)
0.6"
(0.6)
18"
(6)
1.8"
(0.9)
18"
(7)
1.8"
(0.9)
21"
(6)
187
a
(101)
349
b,c,d,e
(67)
400
c
'
d
'
e
(57)
383
b,c,d,e
(45)
452
e
(68)
273
a,b,c
(33)
355
b,c,d,e
(88)
313
a,b,c,d
(68)
398
c,d,e
(56)
403
c
'
d
'
a
(52)
263
x
'
1
'(29)
344
b,c,d
(47)
357
b,c,d,e
(56)
400
c
'
d
'
e
(50)
385
b,c,d,e
(40)
431
d
'
e
(40)
Journal
of
iedToxicology
Comparative
toxicity
of
comfrey,
lycopsamine
and
intermedine
Statistical
Analysis
Statistical
analyses
were
performed
using
Proc
GLM
and
Proc
Freq
in
SAS
(SAS
Institute,
Cary,
NC,
USA)
9.3.
A
two-way
ANOVA
with
Tukey-Kramer
adjustment
for
multiple
comparisons
was
used
to
compare
average
daily
gain,
liver
weight/body
weight,
necrosis
scores,
'pyrrole'
concentration
and
serum
biochemistries.
A
Pearson's
correlation
was
used
to
determine
the
relationship
be-
tween
pyrrole
and
dose
for
each
alkaloid.
Proc
Freq
was
used
to
compare
the
frequency
of
ascites
present
at
necropsy
between
the
different
groups.
A
P-value
of
<
0.05
was
considered
significant.
Results
Test
samples
HPLC-esi(+)MS
and
MS/MS
analysis
showed
that
the
reduced
crude
extract
comprised
mainly
of
the
monoester
DHPA
lycopsa-
mine,
its
C13
epimer
intermedine
and
their
C7-acetylated
derivatives
(Fig.
1).
Minor
concentrations
of
intermedine
and
lycopsamine
N-oxides
observed
were
residual
owing
to
an
incom-
plete
reduction
step.
Traces
(<1%
based
on
integrated
areas
of
HPLC
ion
chromatogram
peaks)
of
the
open
chain
diester
DHPAs
symlandine
and
symphytine,
were
also
observed.
No
other
DHPAs
reported
to
be
found
in
comfrey
(e.g.
echimidine
or
its
N-oxide)
were
observed
in
the
reduced
comfrey
extract.
Based
on
the
rela-
tive
integrated
areas
of
the
HPLC-MS
base
ion
chromatogram
the
ratio
of
lycopsamine,
intermedine
and
the
combined
acetylated
derivatives
was
1.3:
1.5:
1.
The
test
sample
solutions
were
prepared
by
accurately
weighing
sub-samples
of
the
crude
alkaloidal
extract
or
the
purified
lycopsa-
mine
and
intermedine
and
dissolving
them
in
ethanol
to
concen-
trations
of
54.5
mg
total
alkaloid
m1
-1
;
59.8
mg
lycopsamine
m1
-1
;
and
60.4
mg
intermedine
m1
-1
.
The
calculated
concentrations
were
confirmed
by
quantitative
HPLC-esi(+)MS
analysis
against
a
lycopsamine
standard
calibration
curve.
Furthermore,
pre-
and
post-dosing
HPLC-esi(+)MS
analysis
of
the
samples
did
not
reveal
any
degradation
in
ethanol
solution
over
the
10-day
dosing
period.
Clinical
At
the
doses
used
in
this
study,
the
chicks
in
the
lycopsamine
or
intermedine-dosed
groups
did
not
show
any
difference
in
weight
gain
over
the
course
of
the
study
compared
with
the
controls.
The
reduced
comfrey
extract-dosed
animals
showed
a
dose-
dependent
decrease
in
average
daily
gain
(Table
1).
The
chicks
in
the
1.04
and
0.52
mmol
kg
-1
groups
gained
an
average
of
only
1.6
and
4.2
g
day
-1
respectively,
whereas
the
controls
gained
an
average
of
7.4
g
day
-1
.
All
five
of
the
chicks
in
the
1.04
and
one
of
the
chicks
in
the
0.52
mmol
kg
-I
I
-
educed
comfrey
extract
groups
developed
ascites
over
the
course
of
the
study
(Table
1).
Two
high
dose
(1.04
mmol
kg
-1
)
reduced
comfrey
extract-
exposed
chicks
had
to
be
euthanized
at
7
and
10
days
into
the
study.
Serum
biochemistry
Serum
activity
of
the
cytosolic
hepatocellular
enzyme
SDH
was
in-
creased
in
the
two
highest
doses
of
comfrey
(Table
1)
when
com-
pared
with
the
controls.
Bile
acid
was
also
increased
in
the
chicks
from
the
highest
dose
reduced
comfrey
extract
group
when
com-
pared
with
all
the
other
groups.
Serum
glucose
concentrations
were
lower
in
chicks
from
all
of
the
1.04
mmol
kg
-1
groups
com-
pared
with
the
controls
but,
only
the
chicks
from
the
1.04
mmol
kg
-1
reduced
comfrey
extract
group
were
significantly
different
compared
with
the
lower
dose
cohorts.
No
differences
were
Table
1.
Group
averages
and
(SD)
of
chicks
dosed
with
comfrey,
lycopsamine,
intermedine
and
the
control
at
0.4,
0.13,
0.26,
0.52
and
1.04
mmol
kg
-1
body
weight
(BW)
per
day.
Comparisons
were
made
with
average
daily
gain
(ADG),
percent
of
liver
weight
to
total
body
weight,
number
of
animals
with
ascites
at
necropsy,
the
average
necrosis
score
for
the
group,
the
average
'pyrrole'
concentration
in
the
liver,
and
serum
sorbitol
dehydrogenase
(SDH),
bile
acid,
and
glucose
at
death.
Different
superscript
letters
denote
differences
at
P
<
0.05.
All
comparisons
were
made
using
a
two-way
ANOVA
with
the
exception
of
number
of
animals
with
ascites
where
a
Fisher's
exact
test
was
used.
Numbers
in
parenthesis
are
one
standard
deviation
(SD)
Dose
ADG
Liver
wt.
%
Ascites
Necrosis
"pyrrole"
Conc.
SDH
(1U/I)
Bile
Acid
Glucose
(mmol/kg)
(
g/day)
Body
wt.
Score
(nmol/g)
(jimoVI)
(mg/dl)
5/5
a
1/5"
0/5"
0/5"
0/5"
0/5"
0/5"
0/5"
0/5"
0/5"
0/5"
0/5"
0/5"
0/5"
0/5"
0/5"
Comfrey
1.04
0.52
0.26
0.13
0.04
Lycopsamine
1.04
0.52
0.26
0.13
0.04
I
ntermed
ine
1.04
0.52
0.26
0.13
0.04
Control
0
1.6
a
(1.0)
4.2"
(1.8)
6.7
1',c
(1.5)
7.8`
(0.9)
7.4`
(0.8)
7.2`
(1.5)
7.5`
(0.8)
8.2`
(0.6)
7.9`
(0.4)
7.9`
(0.7)
6.3
1
'
4
(2.1)
8.4`
(1.0)
7.3`
(0.5)
7.8`
(0.9)
7.5`
(1.2)
7.4`
(1.0)
2.8
a
(0.8)
3.6
a
'"
(0.6)
3.5
a
'"
(0.4)
3.7
a,b
(0.5)
3.6
a
'
b
(0.2)
4.0"
(0.5)
4.2"
(0.5)
4.1"
(0.3)
3.9
a,b
(0.3)
4.1"
(0.4)
3.8
a
'
b
(0.6)
3.9"
(0.6)
4.0"
(0.5)
4.1"
(0.4)
4.0
1
'(0.4)
4.0
1
'(0.4)
2.3
a
(0.4)
1.1"
(1.2)
0.6
1
'
4
(1.0)
0.4"
4
(0.8)
0.0`
(0)
0.0`
(0)
0.3
1
'
4
(0.7)
0.0`
(0)
0.0`
(0)
0.0`
(0)
0.1
1
'
4
(0.2)
0.1
1
'
4
(0.2)
0.0`
(0)
0.0`
(0)
0.0`
(0)
0.0`
(0)
0.59
a
(0.12)
0.29
a
'"'`
(0.03)
0.20
4
(0.04)
0.13
1
'
4
(0.07)
0.00
c
(0)
0.38
a
'
1
'(0.41)
0.29
a
'"'`
(0.3)
0.04`
(0.06)
0.01`
(0.02)
0.00
c
(0.01)
0.07
1
'
4
(0.06)
0.01`
(0.02)
0.00
c
(0)
0.00
c
(0)
0.01`
(0.03)
0.01`
(0.03)
J.
Appl.
Toxicol.
2016;
36:
716-725
wileyonlinelibrary.com/joumal/jat
A
Z.•
'
.
. .
-
'
4
4
of
'A u
OredToxicology
detected
between
any
groups
at
any
concentration
for
total
protein,
triglycerides,
cholesterol,
GGT,
AST,
LDH
and
CPK.
Gross
observations
and
histopathology
At
necropsy,
all
livers
were
weighed
after
excising
and
emptying
the
gall
bladders.
The
percent
of
the
liver
weight
to
body
weight
was
low
in
the
chicks
from
the
1.04
mmol
kg
-1
reduced
comfrey
extract
group,
which
was
the
only
group
that
differed
significantly
from
the
controls.
However,
they
were
not
significantly
different
from
the
chicks
in
other
reduced
comfrey
extract
groups
or
to
the
chicks
in
the
1.04
mmol
kg
-1
intermedine
group.
The
most
serious
histologic
change
was
centrilobular,
or
zone
three,
hepatic
necrosis
that
extended
through
the
entire
lobule
in
the
most
severe
cases.
In
less
severely
affected
individuals
there
were
occasional
small
foci
or
individual
hepatocytes
that
were
surrounded
by
hemorrhage
and
had
pyknotic
nuclei.
Histologic
hepatic
necrosis
scores
tended
to
be
higher
in
the
three
highest
doses
of
the
reduced
comfrey
extract-treated
animals.
The
average
Figure
5.
Hematoxylin
and
eosin
(H&E)
200x.
Angiectasis
(peliosis
hepatis)
in
a
chick
liver
from
the
1.04
mMol
kg
-1
group.
A.
W.
Brown
et
al.
hepatic
necrosis
scores
were
significantly
higher
than
controls
in
both
the
1.04
and
0.52
mmol
kg
-1
-reduced
comfrey
extract
groups,
but
only
the
1.04
mmol
kg
-1
group
was
significantly
higher
than
the
corresponding
lycopsamine
and
intermedine
groups.
Angiectasis
(Fig.
5)
was
present
in
three,
four
and
three
of
the
0.26,
0.52
and
1.04
mmol
kg
-1
lycopsamine-treated
animals,
respectively.
One
chick
each
of
the
0.13
and
0.26
mmol
kg
-1
,
and
two
and
three
of
the
0.52
and
1.04
mmol
kg
-1
-reduced
com-
frey
extract
groups,
respectively,
also
had
angiectasis.
Angiectasis
was
not
observed
in
any
of
the
animals
in
the
intermedine-dosed
groups.
In
this
study,
regeneration
was
an
uncommon
finding,
only
ob-
served
in
two
reduced
comfrey
extract-treated
animals
(one
each
in
0.26
and
1.04
mmol
kg
-1
dose
groups).
It
presented
primarily
in
portal
areas
and
was
composed
with
increased
numbers
of
bile
ducts
with
fewer,
more
basophilic-staining
hepatocytes,
indicative
of
immature
hepatocytes
and
occasional
mitotic
figures.
Capsular
serositis
(Fig.
6)
was
only
present
in
chicks
from
the
two
higher
dose
reduced
comfrey
extract
groups
and
ranged
from
mild
to
severe.
The
capsules
were
diffusely
and
variably
thickened
and
composed
of
spindle
cells
(likely
fibroblasts),
a
variably
dense
eosinophilic
matrix,
and
mononuclear
inflammatory
cells.
Occa-
sionally
vascularization
of
the
capsule
was
present.
In
the
1.04
mmol
kg
-1
dose
group
four
of
five
animals
had
a
capsular
serositis.
The
one
animal
that
did
not
was
euthanized
on
day
7
of
the
study.
One
of
the
five
animals
in
the
0.52
mmol
kg
-1
group
also
had
capsular
serositis.
Virtually
every
section
of
liver
from
all
animals
in
all
groups,
in-
cluding
controls,
showed
mild
to
moderate
periportal
inflamma-
tion.
There
appeared
to
be
no
difference
in
frequency,
intensity,
type
or
distribution
of
inflammation
between
the
groups.
Tissue
'pyrrole'
adduct
detection
'Pyrrole'
was
detected
(Fig.
3)
in
chicks
from
the
0.13, 0.26,
0.52
and
1.04
mmol
kg
-1
BW
reduced
comfrey
extract
and
the
lycopsamine-dosed
groups.
Similarly,
'pyrrole'
was
detected,
albeit
at
lower
concentrations
than
for
the
chicks
in
the
reduced
comfrey
extract
and
lycopsamine
groups,
in
chicks
from
the
0.04,
0.52
and
V
Figure
6.
Hematoxylin
and
eosin
(H&E)
(A)
100x
Liver
from
a
comfrey-treated
chick
(1.04
mMol
kg
-1
group)
with
a
proliferative
capsule
(double
headed
arrow).
(B)
200x
is
an
enlargement
of
the
box
in
A
The
capsule
is
composed
of
spindle
cells
and
a
variably
dense
fibrillar
matrix
and
occasional
small
caliber
vessels.
(C)
200x
is
a
normal
liver
from
a
control
chick.
Note
the
very
thin
capsule
(arrow).
wileyonlinelibrary.com/journal/jat
J.
Appl.
Toxic&
2016;
36:
716-725
Journal
of
iedToxicology
Comparative
toxicity
of
comfrey,
lycopsamine
and
intermedine
1.04
mmol
kg
-1
BW
intermedine-dosed
groups.
The
'pyrrole'
concentration
tended
to
increase
with
dose
for
all
alkaloids.
Correlation
coefficients
for
dose
and
reduced
comfrey
extract,
lycopsamine
and
intermedine
were
0.93,
0.58
and
0.58
respec-
tively.
The
'Pyrrole'
concentration
tended
to
be
higher
in
the
comfrey
extract
and
lycopsamine-exposed
chicks,
however,
the
variation
between
animals
was
too
great,
with
the
number
of
animals
used,
to
determine
significance
(Table
1).
Indeed,
among
the
equivalent-dose
groups,
the
only
statistically
significant
difference
that
was
observed
was
between
the
1.04
mmol
kg
-1
groups
dosed
with
the
reduced
comfrey
extract
or
intermedine.
Discussion
California
White
chicks
are
highly
sensitive
to
DHPA
toxicity
and
this
model
allowed
direct
comparison
of
various
clinical
(
general
health
and
weight
gains)
and
pathologic
(gross
pathology,
serum
biochemistry,
histopathology
and
tissue
adducts)
indicators
of
toxicity.
The
pure
lycopsamine
and
intermedine,
and
the
reduced,
crude
comfrey
root
alkaloid
extract
all
resulted
in
some
evidence
of
toxicity.
However,
the
total
alkaloid
extract
of
comfrey
root
was
more
toxic
than
an
approximately
molar
equivalent
dose
of
either
pure
lycopsamine
or
pure
intermedine
with
respect
to
clinical,
serum
biochemical,
tissue
'pyrrole'
accumulation
and
histopathological
comparisons.
This
is
unexpected
as
lycopsamine
and
intermedine
are
the
major
DHPAs
in
this
comfrey
sample,
and
they
are
enatimers
as
they
differ
only
in
rotation
of
the
chiral
carbon
on
the
ester
side
chain.
Previously
reported
toxicities
of
lycopsamine
and
intermedine
are
both
1500
mg
kg
-1
[LD
50
for
single
intraperitoneal
(i.p.)
injection
in
rats]
(Cheeke
and
Shull,
1985).
No
data
are
available
for
i.p.
exposure
of
comfrey
in
rats,
but
a
diet
containing
30%
comfrey
or
about
1000
mg
kg
-1
day
-1
for
21
days
resulted
in
a
significantly
reduced
weight
gain
(Garrett
et
al.,
1982).
It
is
difficult
to
compare
these
estimates
because
both
DHPA
content
and
quantity
are
highly
variable
in
comfrey
(Tice,
1997).
It
is
uncertain
why
the
comfrey
extract
was
more
toxic
than
lycopsamine
and
intermedine
in
this
study.
However,
the
signifi-
cant
presence
of
the
acetylated
derivatives
in
the
comfrey
extract
may
be
the
key
factor.
This
is
especially
so
since
equimolar
exposure
of
the
chicks
to
either
lycopsamine
or
intermedine
in
the
comfrey
extract
was
far
less
than
when
these
DHPAs
were
dosed
in
pure
form.
The
decreased
polarity
of
the
acetylated
derivatives
may
affect
initial
absorption
of
the
pro-toxins
resulting
in
an
overall
increase
in
the
DHPA
load
that
reaches
the
liver.
However,
if
absorption
rates
are
approximately
the
same,
then
the
potential
for
the
acetylated
derivatives
to
be
more
toxic
per
se
is
a
consideration.
In
this
regard,
it
has
been
shown
that
the
acet-
ylated
derivatives
were
more
genotoxic
in
a
wing
spot
test
of
Drosophila
melanogaster
after
oral
administration
of
the
alkaloids
(Frei
et
al.,
1992).
Other
possible
factors
that
may
have
affected
toxicity
are:
a
greater
than
additive
combined
effect
of
lycopsamine
and
intermedine,
the
toxicity
of
the
trace
amounts
of
other
minor
alkaloids
such
as
symphytine
or
symlandine,
or
competition
for
hepatic
metabolizing
enzymes
from
some
other
compound
in
the
comfrey
extract.
Data
comparing
pure
intermedine,
pure
lycopsamine
and
a
combination
of
the
two
are
not
available,
thus
determining
the
combined
effect
is
not
possible
without
further
research.
Symphytine
and
symlandine
are
potentially
about
10
times
more
toxic
than
either
lycopsamine
or
interme-
dine
(LD
50
for
symphytine
is
130
mg
-1
ki
1
i.p.,
male
rat)
(Hirono
et
al.,
1979).
It
would
be
unlikely
for
the
proportion
of
symphytine/symlandine
present
in
the
comfrey
extract
to
be
overtly
toxic
alone,
but
it
is
conceivable
that
in
conjunction
with
the
other
DHPAs
it
may
have
increased
the
toxicity.
The
last
consideration
is
competition
for
hepatic
metabolizing
enzymes
from
other
compounds
in
the
crude
comfrey
extract.
If
these
other
compounds,
undetected
using
the
HPLC-esi(+)MS
analytical
method,
have
a
greater
affinity
for
DHPA-detoxifying
enzymes,
then
conceivably,
if
upregulation
of
these
enzymes
is
slow,
there
will
be
more
DHPA
to
be
activated
and
lead
to
'pyrrole'-tissue
adducts
and
consequent
damage.
Histologically
the
lesions
tended
to
be
most
severe
in
the
comfrey
extract
groups,
followed
by
the
lycopsamine
groups,
and
least
severe
in
the
intermedine
exposed
animals.
In
the
comfrey
extract-exposed
animals,
the
highest
dose
group
had
significantly
more
widespread
hepatocellular
necrosis.
Addition-
ally,
only
animals
exposed
to
the
two
highest
doses
of
comfrey
extract
developed
capsular
serositis.
Regeneration
characterized
by
oval
cell
hyperplasia
and
bile
duct
reduplication,
which
is
thought
to
be
an
attempt
to
regenerate
hepatic
paranchyma
when
paranchymal
cells
have
lost
this
capacity
(Stalker
and
Hayes,
2007),
was
only
observed
in
two
comfrey-treated
animals.
Angiectasis,
which
is
thought
to
be
a
result
of
loss
or
weakening
of
the
sinusoidal
walls
and/or
supporting
tissue
(Thoolen
et
al.,
2010)
may
be
an
earlier
and
perhaps
more
sensitive
indicator
of
hepatic
insult
in
this
study.
Only
animals
exposed
to
pure
lycopsamine
or
the
comfrey
extract
developed
angiectasis.
Angiectasis
is
a
feature
that
has
been
reported
in
mice
exposed
to
DHPAs
(
Ruebner
et
al.
1970;
Brown
et
al.,
2015;)
and
may
be
important
because
it
has
been
suggested
that
angiectasis
may
be
a
preneoplastic
lesion
in
some
DHPA-induced
neoplasms.
The
concentration
of
'pyrole'
adduct
tended
to
be
higher
in
comfrey
extract
and
lycopsamine
exposed
chicks
compared
with
those
exposed
to
intermedine
that
also
suggests
that
lycopsamine
may
be
slightly
more
toxic
than
intermedine.
If
this
is
the
case,
it
suggests
that
the
stereochemistry
around
the
13th
carbon
is
important
to
toxicity.
The
toxicity
of
the
DHPAs
is
entirely
related
to
metabolic
oxida-
tion
to
form
the
electrophilic
'pyrrolic'
mono-and
di-alkylating
agents
that
react
with
nucleophiles
on
biomacromolecules
such
as
proteins
and
DNA
(Edgar
et
al.,
2014).
Therefore,
it
was
recog-
nized
that
extraction
and
chemical
release
of
these
tissue
conju-
gates
could
be
used
diagnostically
to
confirm
a
pyrrolizidine
alkaloidosis
(Mattocks
and
Jukes
1990,
1992a,
1992b;
Winter
et
al.,
1990).
Thus,
'pyrrole'
tissue
adducts
were
originally
indirectly
detected
as
the
diethylethers
of
1-hydroxymethyl-7-hydroxy-6,7-
dihydropyrrolizine
(the
'pyrrole')
after
oxidative
cleavage
from
tis-
sue
adducts
with
silver
using
GC-MS.
The
procedure
was
adapted
more
recently
to
use
HPLC-MS
in
which
the
oxidatively-cleaved
'pyrrole'-diethylether
was
reacted
with
Ehrlich's
reagent
to
indi-
rectly
detect
tissue
'pyrrole'
adducts
in
a
human
patient
suffering
from
DHPA-induced
hepatic
sinusoidal
obstruction
syndrome
(Lin
et
al.,
2011).
In
the
present
study,
the
generation
of
a
'pyrrole'
calibration
curve
from
dehydromonocrotaline
enabled
a
semi-quantitative
comparison
of
in
vivo
'pyrrole'
formation
for
each
of
the
treatments
groups.
The
HPLC-MS
method
has
advantages
over
the
previous
GC-MS
method,
specifically
it
is
simpler
to
perform
(it
has
fewer
steps)
and
it
is
more
reliable
and
more
quantitative.
Ideally
detection
of
pyrrole
tissue
adducts
in
tissue
can
be
more
widely
and
effectively
used
in
the
diagnostic
arena.
This
method
needs
to
be
validated
in
multiple
species,
and
a
greater
understanding
of
the
kinetics
of
J.
Appl.
Toxicol.
2016;
36:
716-725
wileyonlinelibrary.com/joumal/jat
of
'A u
OredToxicology
tissue
pyrrole
adducts
is
necessary
prior
to
full
implementation
as
a
diagnostic
test.
In
the
present
study,
there
appears
to
be
a
dose-dependent
in-
crease
in
'pyrrole'
concentration
particularly
in
the
comfrey
extract
and
lycopsamine-exposed
animals.
Similar
to
the
histologic
analy-
sis,
the
comfrey
extract
exposed
animals
tend
to
be
most
severely
affected
followed
by
the
lycopsamine,
and
the
intermedine
ex-
posed
animals
appear
least
affected
although
there
is
no
statistical
difference.
Additional
research
may
be
helpful
to
determine
the
relevance
of
this
trend.
The
comfrey
extract
used
in
this
study
is
somewhat
different
than
what
would
be
expected
from
natural
exposure.
Comfrey
(Symphytum
officinale
and
S.
uplandicum)
has
been
reported
to
produce
several
monoester
and
open
chain
diester
DHPAs
and
their
N-oxides.
In
this
present
study,
the
crude
alkaloid
extract
of
the
S.
officinale
powdered
root
was
treated
in
the
usual
way
with
zinc
and
sulphuric
acid
to
reduce
most
of
the
N-oxides
to
the
free
base
DHPA
forms.
The
resultant
crude,
reduced
comfrey
extract
contained
lycopsamine,
intermedine
and
their
acetylated
derivatives
as
the
predominant
alkaloids
(Fig.
1).
When
ingested
DHPA-N-oxides
are
reduced
in
the
gastro-intestinal
tract
to
their
parent-free
base,
and
although
it
has
been
estimated
that
N-oxides
are
similar
in
toxicity
to
their
free
base
form
(Frei
et
al.,
1992),
in
the
present
study,
the
toxicity
of
N-oxides
was
not
tested.
Different
comfrey-derived
(e.g.
leaf,
roots
and
flowers)
products,
comfrey-based
products
or
products
that
contain
comfrey
as
a
mi-
nor
additive
will
contain
different
concentrations
of
the
alkaloids
and
their
N-oxides.
Therefore,
any
products
that
include
comfrey
could
be
analyzed
and
the
concentrations
of
the
alkaloids
deter-
mined
and
assessed
against
the
chick
toxicity
data
presented
herein
to
determine
the
level
of
risk.
In
this
study,
comfrey
root
was
used
as
the
source
of
relatively
high
concentrations
of
the
al-
kaloids
to
facilitate
the
dosing.
In
the
assessment
of
risks
presented
to
human
health
by
dietary
supplements,
medicinal
herbs, nutraceuticals
or
new
functional
foods,
the
application
of
data
acquired
using
various
in
vitro
or
in
vivo
models
is
a
major
challenge
with
respect
to
extrapolation
to
the
human
situation.
Therefore,
models
are
continually
changed
or
modified
in
an
attempt
to
establish
meaningful
indications
of
toxicity.
In
vivo
toxicity
assessments
of
the
pro-toxic
DHPAs
are
complicated
by
considerable
variations
in
species-,
gender-
and
age-related
susceptibilities.
Additionally,
assessment
of
entire
products,
crude
extracts
or
purified
components
presents
poten-
tial
complications
associated
with
any
intrinsic
additive
or
protec-
tive
effects
of
the
whole
plant
or
crude
extract
relative
to
the
purified
components.
In
this
present
study,
male
California
White
chicks,
that
have
been
shown
to
be
sensitive
to
the
DHPAs
in
terms
of
dose,
duration
of
exposure
and
time
for
clinical
and
pathological
signs
to
develop,
were
used
to
compare
the
hepatotoxicity
of
an
orally-administered
reduced
comfrey
extract
to
that
of
the
purified
monoester
DHPAs
lycopsamine
and
intermedine
(Fig.
1).
Conclusions
The
pure
lycopsamine
and
intermedine,
and
the
reduced,
crude
comfrey
root
alkaloid
extract
all
resulted
in
some
evidence
of
toxicity.
In
this
California
White
male
chick
model,
the
total
alkaloid
extract
of
comfrey
root
was
more
toxic
than
an
approximately
equivalent
(to
the
total
alkaloid
content
of
the
reduced
comfrey
extract)
dose
of
either
pure
lycopsamine
or
pure
intermedine
with
respect
to
clinical,
serum
biochemical,
tissue
'pyrrole'
accumula-
tion
and
histopathological
comparisons.
This
work
suggests
a
A.
W.
Brown
et
al.
greater
than
additive
effect
of
the
individual
major
components
of
the
reduced
comfrey
extract
and/or
a
more
potent
toxicity
of
the
acetylated
derivatives
in
the
reduced
comfrey
extract.
This
may
result
in
safety
recommendations,
based
on
selected
purified
compounds,
underestimating
the
potential
toxicity
of
comfrey-
derived
products.
Acknowledgments
The
authors
would
like
to
thank
Joseph
Jacobsen
for
histology
support
This
work
was
funded
through
the
USDA
Agriculture
Research
Service
NP215-2080-32630-012-00.
Conflict
of
interest
The
Authors
did
not
report
any
conflict
of
interest.
References
BfR.
2011.
Chemical
analysis
and
toxicity
of
pyrrolizidine
alkaloids
and
as-
sessment
of
the
health
risk
posed
by
their
occurrence
in
honey.
BfR
Opinion
no.
038/2011
11.
Brown
AW,
Stegelmeier
BL
Colegate
SM,
Panter
KE,
Knoppel
EL
Hall
JO.
2015.
Heterozygous
p53
knockout
mouse
model
for
dehydropyrrolizidine
alkaloid-induced
carcinogenesis.
J.
Appl.
Toxicol.
DOI:
10.1002/jat.3120.
PubMed
PMID:
25690669.
Cheeke
P,
Shull
L
1985.
Natural
toxicants
in
feeds
and
livestock
AVI
Publishing
Inc.:
West
Port,
CT.
Colegate
SM,
Gardner
DR,
Betz
JM,
Panter
KE.
2014.
Semi-automated
Separation
of
the
Epimeric
Dehydropyrrolizidine
Alkaloids
Lycopsamine
and
Intermedine:
Preparation
of
their
N-oxides
and
NMR
Comparison
with
Diastereoisomeric
Rinderine
and
Echinatine.
Phytochem.
Anal.
25:
429-438.
Committee
on
Toxicity
of
Chemicals
in
Food,
Consumer
Products
and
the
Environment
(COT).
2008.
COT
statement
on
pyrrolizidine
alkaloids
in
food.
Available:
http://cot.food.gov.uk/pdfs/cotstatementpa200806.pdf
[Accessed
6
June
2015].
Culvenor
C,
Edgar
J,
Smith
L,
Tweeddale
H.
1970.
Dihydropyrrolizines.
III.
Preparation
and
reactions
of
derivatives
related
to
pyrrolizidine
alkaloids.
Aust.
J.
Chem.
23:
1853-1867.
Edgar
JA,
Molyneux
RJ,
Colegate
SM.
2014.
Pyrrolizidine
Alkaloids:
Potential
Role
in
the
Etiology
of
Cancers,
Pulmonary
Hypertension,
Congenital
Anomalies
and
Liver
Disease.
Chem.
Res.
Toxicol
28:
4-20.
EFSA
2011.
EFSA
Panel
on
Contaminants
in
the
Food
Chain
(CONTAM).
Scientific
Opinion
on
Pyrrolizidine
alkaloids
in
food
and
feed.
EFSA
J.
9:
1-134.
FDA.
2001.
FDA
advises
dietary
supplement
manufacturers
to
remove
com-
frey
products
from
the
market.
U.S.
Food
and
Drug
Administration,
Center
for
Food
Safety
and
Applied
Nutrition.
Available:
http://www.
cfsan.fda.gov/-dms/dspltr06.html
[Accessed
6
July
2001].
Frei
H,
Luthy
J,
Brauchli
J,
Zweifel
U,
Wurg
ler
FE,
Schlatter
C.
1992.
Structure/
activity
relationships
of
the
genotoxic
potencies
of
sixteen
pyrrolizidine
alkaloids
assayed
for
the
induction
of
somatic
mutation
and
recombina-
tion
in
wing
cells
of
Drosophila
melanogaster.
Chem.
Biol.
Interact.
83:
1-22.
FSANZ.
2001.
Pyrrolizidine
alkaloids
in
food.
A
toxicological
review
and
risk
assessment.
In
Food
Standards
Australia
New
Zealand
Tech.
Report
se-
ries
No.
2.
Canberra,
Australia;
Garrett
BJ,
Cheeke
PR,
Miranda
CL
Goeger
DE,
Buhler
DR.
1982.
Consump-
tion
of
poisonous
plants
(Senecio
jacobaea,
Symphytum
officinale,
Pteridium
aquilinum,
Hypericum
perforatum)
by
rats:
chronic
toxicity,
mineral
metabolism,
and
hepatic
drug-metabolizing
enzymes.
Toxicol.
Lett.
10:
183-188.
Hirono
I,
Haga
M,
Fujii
M,
Matsuura
5,
Matsubara
N,
Nakayama
M,
Furuya
T,
Hikichi
M,
Takanashi
H,
Uchida
E.
1979.
Induction
of
hepatic
tumors
in
rats
by
senkirkine
and
symphytine.
J.
Natl.
Cancer
Inst.
63:
469-472.
Koll
R,
Buhr
M,
Dieter
R,
Pabst
H,
Predel
H-G,
Petrowicz
0,
Giannetti
B,
Klingenburg
5,
Staiger
C.
2004.
Efficacy
and
tolerance
of
a
comfrey
root
extract
(Extr.
Rad.
Symphyti)
in
the
treatment
of
ankle
distorsions:
results
of
a
multicenter,
randomized,
placebo-controlled,
double-blind
study.
Phytomedicine
11:
470-477.
wileyonlinelibrary.com/journal/jat
J.
Appl.
Toxicol.
2016;
36:
716-725
Comparative
toxicity
of
comfrey,
lycopsamine
and
intermedine
Lin
G,
Wang
JY,
Li
N,
Li
M,
Gao
H,
Ji
Y,
Zhang
F,
Wang
H,
Zhou
Y,
Ye
Y,
Xu
HX,
Zheng
J.
2011.
Hepatic
sinusoidal
obstruction
syndrome
associated
with
consumption
of
Gynura
segetum.
J.
Hepatol.
54:
666-673.
Mattocks
AR,
Jukes
R.
1990.
Recovery
of
the
pyrrolic
nucleus
of
pyrrolizidine
alkaloid
metabolites
from
sulphur
conjugates
in
tissues
and
body
fluids.
Chem.
Biol.
Interact.
75:
225-239.
Mattocks
AR,
Jukes
R.
1992a.
Chemistry
of
sulphur-bound
pyrrolic
metabo-
lites
in
the
blood
of
rats
given
different
types
of
pyrrolizidine
alkaloid.
Nat.
Toxins
1:
89-95.
Mattocks
AR,
Jukes
R.
1992b.
Detection
of
sulphur-conjugated
pyrrolic
me-
tabolites
in
blood
and
fresh
or
fixed
liver
tissue
from
rats
given
a
variety
of
toxic
pyrrolizidine
alkaloids.
Toxicol.
Lett.
63:
47-55.
Mei
N,
Guo
L,
Fu
PP,
Fuscoe
JC,
Luan
Y,
Chen
T.
2010.
Metabolism,
genotoxicity,
and
carcinogenicity
of
comfrey,
J.
Toxicol.
Environ.
Health
8
Crit.
Rev.
13:
509-526.
RIVM.
2005.
Advisory
Report
on
Pyrrolizidine
Alkaloids
in
Herb
Preparations.
Available:
http://www.rivm.nVen/Documents_and_publications/Scien-
tffic/Reports/2015/aprilifyrrolizidine_alkaloids_in_herbal_preparations
[Accessed
8
April
2015].
Rode
D.
2002.
Comfrey
toxicity
revisited.
Trends
Pharmacol.
Sd.
23:
497-499.
Ruebner
BH,
Watanabe
K,
Wand
J.
1970.
Lytic
necrosis
resembling
peliosis
hepatis
produced
by
lasiocarpine
in
the
mouse
liver.
A
light
and
electron
microscopic
study.
Am.
J.
PathoL
60:
247.
Journal
of
d
Toxi
co
logy
Staiger
C.
2012.
Comfrey:
a
clinical
overview.
Phytother.
Res.
26:
1441-1448.
Stalker
M,
Hayes
M.
2007.
Liver
and
binary
system.
Jubb,
Kennedy,
Palmer's
PathoL
Domestic
Animals
2:
297-388.
Stegelmeier
BL,
Gardner
DR
James
LF,
Molyneux
RJ.
1996.
Pyrrole
detection
and
the
pathologic
progression
of
Cynoglossum
officinale
(houndstongue)
poisoning
in
horses.
J.
Vet.
Diagn.
Investig.:
Official
publication
of
the
American
Association
of
Veterinary
Laboratory
Diagnosticians,
Inc
8:
81-90.
Thoolen
B,
Maronpot
RR,
Harada
T,
Nyska
A,
Rousseaux
C,
Nolte
T,
Malarkey
DE,
Kaufmann
W,
Kuttler
K,
Desch!
U,
Nakae
D,
Gregson
R,
Vinlove
MP,
Brix
AE,
Singh
B,
Belpoggi
F,
Ward
JM.
2010.
Proliferative
and
nonproliferative
lesions
of
the
rat
and
mouse
hepatobiliary
system.
Toxicol.
PathoL
38:
5s-81s.
Tice
R.
1997.
Comfrey
and
One
of
Its
Constituent
Alkaloids,
Symphytine:
Re-
view
of
the
Toxicological
Literature.
National
Institute
of
Environmental
Health
Sciences:
Research
Triangle
Park,
North
Carlina.
WHO.
2011.
Discussion
Paper
on
Pyrrolizidine
Alkaloids,
Joint
FAO/WHO
Food
standards
Programme,
Codex
Committee
on
Contaminants
in
Foods.
The
Hague,
The
Netherlands.
Winter
H,
Seawright
AA,
Mattocks
AR,
Jukes
R,
Tshewang
U,
Gurung
BJ.
1990.
Pyrrolizidine
alkaloid
poisoning
in
yaks.
First
report
and
confirmation
by
identification
of
sulphurbound
pyrrolic
metabolites
of
the
alkaloids
in
preserved
liver
tissue.
Aust
Vet.
J.
67:
411-412.
J.
Appl.
Toxicol
2016;
36:
716-725
wileyonlinelibrary.com/joumal/jat