From: TSS ()
Subject: Amyloidogenic potential of foie gras
Date: June 22, 2007 at 2:23 pm PST
Amyloidogenic potential of foie gras
Alan Solomon*†, Tina Richey*, Charles L. Murphy*, Deborah T. Weiss*,
Jonathan S. Wall*, Gunilla T. Westermark‡,
and Per Westermark§
*Human Immunology and Cancer Program, Department of Medicine, University of
Tennessee Graduate School of Medicine, Knoxville, TN 37920;
‡Division of Cell Biology, Linko¨ ping University, SE-58185 Linko¨ ping,
Sweden; and §Department of Genetics and Pathology, Uppsala University,
SE-75185 Uppsala, Sweden
Communicated by D. Carleton Gajdusek, Institut de Neurobiologie Alfred
Fessard, Gif-sur-Yvette, France, January 30, 2007 (received for review
October 10, 2006)
The human cerebral and systemic amyloidoses and prionassociated
spongiform encephalopathies are acquired or inherited
protein folding disorders in which normally soluble proteins
or peptides are converted into fibrillar aggregates. This is a
nucleation-dependent process that can be initiated or accelerated
by fibril seeds formed from homologous or heterologous amyloidogenic
precursors that serve as an amyloid enhancing factor (AEF)
and has pathogenic significance in that disease may be transmitted
by oral ingestion or parenteral administration of these conformationally
altered components. Except for infected brain tissue,
specific dietary sources of AEF have not been identified. Here we
report that commercially available duck- or goose-derived foie gras
contains birefringent congophilic fibrillar material composed of
serum amyloid A-related protein that acted as a potent AEF in a
transgenic murine model of secondary (amyloid A protein) amyloidosis.
When such mice were injected with or fed amyloid
extracted from foie gras, the animals developed extensive systemic
pathological deposits. These experimental data provide evidence
that an amyloid-containing food product hastened the development
of amyloid protein A amyloidosis in a susceptible population.
On this basis, we posit that this and perhaps other forms of
amyloidosis may be transmissible, akin to the infectious nature of
amyloid protein A amyloidosis amyloid-enhancing factor protein
aggregation rheumatoid arthritis transmissibility
Amyloid protein A amyloidosis (AA) occurs in patients with
rheumatoid arthritis and other chronic inflammatory diseases
and results from a sustained elevation of the apolipoprotein
serum amyloid A (SAA) protein produced by hepatocytes
under regulation by interleukin (IL)-1, IL-6, and tumor necrosis
factor (1). This acute-phase reactant is cleaved into an 76-
residue N-terminal fragment deposited as amyloid predominately
in the kidneys, liver, and spleen. The disorder also can be
induced experimentally in susceptible strains of mice by inflammatory
stimuli that result in an 1,000-fold increase in SAA
concentration (2). Further, the lag phase of this process is greatly
decreased by injecting or feeding animals extracts of amyloidladen
spleens of affected mice (2–5).
To determine whether amyloid-containing food products exhibit
amyloid enhancing factor (AEF) activity, we used a more
robust in vivo murine model of AA amyloidosis involving mice
carrying the human IL-6 (hIL-6) gene under control of either the
murine metallothionein-1 (MT-1) (MT-1/hIL-6) or histocompatibility
H2-Ld (H2/hIL-6) promoter (6). Typically, AA amyloid
develops in these animals at 5 mo of age and is initially located
predominately in the perifollicular regions of the spleen. Over
the next 2–3 mo, the deposits spread rapidly into the liver
parenchyma, renal glomerular and intertubular regions, cardiac
muscle, tongue, and gastrointestinal tract and lead to death at
8–9 mo. However, by injection into 8-wk-old transgenic mice
of a single 100-g i.v. dose of an exogenous source of AA fibrils,
amyloid deposits are formed within 3 wk, and severe systemic
disease (akin to that found in 8-mo-old animals) occurs within
2 mo, at which time the resultant pathology is lethal (7).
AA amyloid deposits are commonly found in waterfowl,
particularly Pekin ducks, in which the liver is predominately
involved (8–10). This pathological alteration is noticeably increased
in birds subjected to stressful environmental conditions
as well as to the forced feeding that is used to produce foie gras
(8). This culinary product, derived from massively enlarged fatty
livers results from gorging young ducks or geese up to three times
daily over a 4-wk period with corn-based feed.
We now report the results of our studies that have shown that
AA-containing fibrils extracted from duck or goose foie gras
have potent AEF activity when administered by i.v. injection or
gavage into our IL-6 transgenic mice.
Results and Discussion
We analyzed several commercial sources of foie gras histochemically
and found amyloid to be present. Microscopic examination
of hematoxylin/eosin- and Congo red-stained sections cut from
formalin-fixed, paraffin-embedded specimens revealed virtual
replacement of the normal hepatic parenchyma by fat; additionally,
green birefringent congophilic areas in residual vasculature
were noted by polarizing microscopy (Fig. 1 a and b). Further,
these deposits were immunostained by a specific anti-AA antiserum
(Fig. 1c). Similar material was found in marketed paˆte´s
prepared from duck or goose liver (Fig. 2).
The AA composition of the hepatic amyloid deposits was
confirmed chemically through analysis of material derived from
acetone-defatted specimens extracted first with 0.15MNaCl and
then distilled water. The isolates were strongly congophilic, and,
when examined by transmission electron microscopy, contained
fibrils with the typical ultrastructural features of amyloid;
namely, 10-m-thick unbranched structures (Fig. 3a). Electrophoresis
of the water-suspended product on a SDS/
polyacrylamide gel in the presence of 0.1 M DTT and 8 M urea
revealed, after Coomassie blue staining, a protein band with aMr
of 6,000, comparable to that of amyloid extracted from the
spleen of a mouse with AA amyloidosis (Fig. 3b). After transfer
to a PVDF membrane, this component was subjected to automated
Edman degradation with which 14 residues identical in
amino acid sequence to that of the N-terminal portion of duck
SAA were detected. In a similar study of tryptic digests obtained
from cleavage of this molecule after reduction and alkylation, six
peptides that included 45 of the first 60 residues of duck SAA
were identified by MS/MS (Fig. 3c) (9).
Author contributions: A.S., J.S.W., and P.W. designed research; T.R. and
research; J.S.W., G.T.W., and P.W. analyzed data; A.S. and D.T.W. wrote the
The authors declare no conflict of interest.
Freely available online through the PNAS open access option.
Abbreviations: AA, amyloid protein A amyloidosis; AEF, amyloid enhancing
interleukin; SAA, serum amyloid A protein.
†To whom correspondence should be addressed at: University of Tennessee
School of Medicine, 1924 Alcoa Highway, Knoxville, TN 37920. E-mail:
This article contains supporting information online at
© 2007 by The National Academy of Sciences of the USA
10998–11001 PNAS June 26, 2007 vol. 104 no. 26
To determine whether amyloid-containing duck- or goosederived
foie gras had AEF activity, groups of up to nine
MT-1/hIL-6 or H2/hIL-6 mice received tail vein injections of
either 100 g of extract suspended in 0.1 ml of PBS or the
equivalent volume of PBS alone. Both sets were euthanized 8 wk
later and multiple organs (liver, spleen, kidney, pancreas, heart,
lung, tongue, and intestines) were obtained at necropsy for
histochemical analysis. Examination by polarizing microscopy of
Congo red-stained sections revealed the presence of varying
amounts of amyloid deposits in one or more tissues of virtually
all of the treated mice; most affected were the liver, spleen, and
to a lesser extent, the kidneys and pancreas (Fig. 4a). In contrast,
control animals that received PBS had no detectable amyloid.
Similar results were obtained in the conventional murine
model of AA amyloidosis in which SAA overexpression was
induced by an inflammatory stimulus (4). Two groups of wildtype
BALB/c mice were given two 0.5-ml s.c. injections of
aqueous 1% AgNO3 (days 1 and 10), and one set also were
injected i.v. with 100 g of the foie gras extract; the others
(controls) received PBS only. At the time of euthanizing (day
21), 8 of 10 mice from the first group had detectable amyloid in
the liver and spleen. In contrast, no amyloid was found in the
The amyloid induced by administration of fibril-containing
foie gras into both the wild-type and transgenic mice was
immunostained by an anti-AA antibody. Further, the deposits
were AA in nature as confirmed by MS of protein extracted from
the spleen of a recipient animal. MS/MS analyses of tryptic
peptides generated from an HPLC-purified reduced and alkylated
water pellet identified residues 19–56 of murine SAA.
AA-containing foie gras extracts also had AEF activity when
administered orally to the hIL-6 transgenic animals. Five of eight
mice that were gavaged for 5 consecutive days with 100 g of
material suspended in 50 l of PBS were found 8 wk later to have
amyloid deposits in virtually all organs examined, and, as in the
case of animals injected i.v. with this material, this effect was
most pronounced in the liver and spleen (Fig. 4b).
The AEF activity of foie gras was reduced, but not abolished,
by cooking, as specified by the supplier. Intravenous injections
into nine hIL-6 transgenic mice of 100-g doses of extracts
prepared from liver that had been heated to 95°C for 20 min
in an oven resulted in 4, 2, and 1 hepatic and/or splenic
amyloid deposits in two, one, and two animals, respectively (in
four cases, no amyloid was found). In contrast, when this
material was dissolved in 6 M guanidine HCl, incubated at 37°C
for 24 h, dialyzed against PBS, and injected into six transgenic
Fig. 1. AA deposition in foie gras. (a) Large venule surrounded by residual,
extensively vacuolated fatty hepatic tissue (hematoxylin/eosin stain). (b)
birefringent amyloid deposits in the blood vessel wall (Congo red stain).
Immunohistochemical identification of vascular AA amyloid. (Scale bar, 62
Fig. 2. Tissue fragment with amyloid in duck paˆ te´ . Congo red stain.
<----> < ----------
--> <-> <-
Fig. 3. Ultrastructural and chemical characterization of amyloid extracted
from foie gras. (a) Fibrillar nature of proteins contained in the pellet
micrograph, negative uranyl acetate stain). (Scale bar, 200 nm.) (b)
of congophilic components extracted from duck foie gras and the spleen of a
mouse with AA amyloidosis (Coomassie blue stain). The Mr values of the
standard proteins are given; arrows show the location of AA-containing
protein bands. (c) Comparison of the amino acid sequence of duck foie grasAA
amyloid with that of duck SAA (9). Homologous residues are indicated by
Solomon et al. PNAS June 26, 2007 vol. 104 no. 26 10999
mice, only traces of splenic amyloid were found in two mice. A
summary of the results of all of the above studies is presented in
Table 1 and supporting information (SI) Fig. 5.
The prevalence of AA amyloidosis in the human population is
unknown. In most developed countries, sustained inflammatory
processes, particularly rheumatoid and juvenile chronic arthritis
(as opposed to infectious diseases), now account for the majority
of such cases (11). Notably, there is a marked geographic
variation in incidence among such patients, e.g., the number is
relatively high in Europe compared with the United States but
particularly high in parts of Papua New Guinea (12). This
difference has been attributed, in part, to genetic factors,
including expression of a more amyloidogenic SAA allotype (13)
or other genes encoding inflammatory molecules (11). Notably,
increased blood levels of SAA do not necessarily result in
amyloidosis (11, 14).
Given our experimental findings, exposure to exogenous
substances with AEF activity also may be an important epigenetic
or environmental factor in the development of AA amyloidosis
in a susceptible population. In this regard, it would seem
prudent for children and adults with rheumatoid arthritis or
other diseases who are at risk for this disorder to avoid foods that
may be contaminated with AA fibrils (15). In addition to foie
gras, meat derived from sheep (16) and seemingly healthy cattle
(17) may represent other dietary sources of this material.
Further, the fact that chemically heterologous fibrils can serve as
AEF, as demonstrated in experimental models of AA (18–20)
and AApoAII amyloidosis (21–23), suggests that it may be
hazardous for individuals who are prone to develop other types
of amyloid-associated disorders, e.g., Alzheimer’s disease or type
II diabetes, to consume such products.
Materials and Methods
Materials. Whole fresh duck and goose liver (foie gras) was
purchased from three commercial venders located in the United
States and France.
Mice. Mice carrying the hIL-6 gene under control of the mouse
MT-1 or histocompatibility H2-Ld promoter were obtained from
Gennaro Ciliberto and Michael Potter, respectively, and generated
as described previously (6). At 4 wk the mice were weaned,
and the presence of the transgene was confirmed through
analysis of genomic DNA derived from tail snips. Wild-type
BALB/c mice were purchased from Charles River Laboratories
(Boston, MA). Animals were housed in groups of eight in a
positive pressure environment with a 12-h light/dark cycle and
provided filtered tap water and a standard laboratory rodent
chow (Harlan Teklad, Madison, WI) ab libitum. The animals
were treated in accordance with National Institutes of Health
regulations under the aegis of a protocol approved by the
University of Tennessee’s Animal Care and Use Committee.
Histochemical and Immunohistochemical Analyses. Samples of foie
gras and mouse organs obtained at necropsy were placed in 10%
buffered formalin (Fisher Scientific International, Inc., Hampton,
NH) and embedded in paraffin. To detect amyloid, 6-mthick
deparaffinized sections were stained with a freshly prepared
solution of alkaline Congo red and examined under
polarizing microscopy. A qualitative assessment of amyloid
deposition was made by an experienced microscopist (T.R.)
based on the relative extent of green birefringence seen in at
least 10 fields at 20 magnification; a value of 1, 2, 3, or
4 was assigned if such material occupied, respectively, trace,
minimal, moderate, or extensive portions of the sections studied,
and these values were corroborated by quantitative image analyses.
Electron microscopy on Epon-embedded sections stained
n e d r u B d i o l y m A e v i t a l e R
1 2 3 4 5 6 7 8
1 2 3 4 5 6 7
n e d r u B d i o l y m A e v i t a l e R
Fig. 4. AEF activity of foie gras. Hepatic and splenic amyloid deposits
hIL-6 transgenic mice 8 wk after they were injected (a) or gavaged (b) with
fibrils extracted from foie gras. The extent of amyloid burden in the liver
spleen (o), kidney (), and pancreas (u) is as indicated. (Scale bars, 100
Table 1. Summary of the amyloidogenic potential of foie gras preparations
Group Mice (n) Treatment Route Positive (%) Mean score*
1 H-2/hIL-6 (8) Extract 1 i.v. 3 (37.5) 3
2 H-2/hIL-6 (6) Extract 2 i.v. 2 (33.3) 4
3 H-2/hIL-6 (7) Extract 3 i.v. 6 (85.7) 1.9
4 H-2/hIL-6 (5) PBS i.v. 0 (0) 0
5 H-2/hIL-6 (7) Extract 4 i.v. 7 (100) 3.1
6 MT-1/hIL-6 (8) Extract 4 Gavage 5 (62.5) 4
7 MT-1/hIL-6 (7) Extract 5 i.v. 5 (71.4) 4
8 H-2/hIL-6 (9) Extract 5/cooked i.v. 5 (55.6) 1.75
9 H-2/hIL-6 (6) Extract 5/guanidine HCl i.v. 2 (33.3) 0.5
10 BALB/c (10) Extract 5 i.v. 8 (80) 1.6
11 BALB/c (5) PBS i.v. 0 (0) 0
*Mean score of amyloid-positive spleens in each group.
11000 www.pnas.orgcgidoi10.1073pnas.0700848104 Solomon et al.
negatively with 1% uranyl acetate on Formvar-coated copper
grids was performed with a Hitachi H-800 electron microscope
(Hitachi High Technology, Pleasanton, CA). Specimens were
immunostained using the avidin-biotin complex technique (Vector
Laboratories, Burlingame, CA). An anti-marmoset AA
murine mAb (24) and a biotinylated sheep anti-mouse Ig antiserum
(BioGenex, San Ramon, CA) were used as the primary
and secondary reagents, respectively, and the reactions were visualized
with the Super Sensitive Link-labeled HRP Detection
System (BioGenex), under conditions specified by the
Amyloid Extraction and Characterization. Thirty- to 80-gram portions
of foie gras were cut into 0.5-cm3 pieces and, after defatting
by a series of four acetone washes, were homogenized in 0.15 M
NaCl using an Omni-Mixer blender (Omni International, Marietta,
GA) and centrifuged (15,000 g) for 30 min at 4°C. This
step was repeated until the OD280 of the supernatant was 0.10.
The saline-extracted sediment was similarly homogenized in cold
distilled water, and the resultant pellet was lyophilized. After
reextraction with chloroform and ether, the protein was dissolved
in 0.25 M TrisHCl buffer, pH 8.0, containing 6 M
guanidine HCl, reduced and alkylated, and purified by reversephase
HPLC (25). The peak UV-absorbing fractions were dried
in a vacuum centrifuge, reconstituted in sample loading buffer,
and, after electrophoresis on 10% NuPage SDS/PAGE gels
(Invitrogen, Carlsbad, CA), transferred to a PVDF membrane.
Coomassie blue-stained bands were excised and subjected to
automated sequence analysis by Edman degradation using an
ABI model 494 pulsed liquid sequencer (Applied BioSystems,
Foster City, CA). The amino acid sequences of tryptic peptides
generated from HPLC-purified fibrillar tissue extracts were
determined by MS/MS using an ion-trap instrument (Thermo
Finnigan, Waltham, MA), as described previously (25).
We thank T. K. Williams, S. D. Macy, C. Wooliver, S. Wang, and J.
Dunlap for technical assistance; A. Lehberger for manuscript preparation;
and R. Wetzel for critical input. This work was supported in part
by Public Health Research Grant CA 10056 from the National Cancer
Institute, by the Aslan Foundation, and by the Swedish Research Council
and Torsten and Ragnar So¨derberg’s Foundations. A.S. is an American
Cancer Society Clinical Research Professor.
1. Benson MD (2001) in The Metabolic and Molecular Bases of Inherited
eds. Scriver CR, Beaudet AL, Sly WS, Valle D (McGraw–Hill, New York), pp
2. Kisilevsky R, Boudreau L (1983) Lab Invest 48:53–59.
3. Hardt F, Ranløv PJ (1976) Int Rev Exp Pathol 16:273–334.
4. Johan K, Westermark GT, Engstro¨m U, Gustavsson Å, Hultman P, Westermark
P (1998) Proc Natl Acad Sci USA 95:2558–2563.
5. Lundmark K, Westermark GT, Nystro¨m S, Murphy CL, Solomon A, Westermark
P (2002) Proc Natl Acad Sci USA 99:6979–6984.
6. Solomon A, Weiss DT, Schell M, Hrncic R, Murphy CL, Wall J, McGavin MD,
Pan HJ, Kabalka GW, Paulus MJ (1999) Am J Pathol 154:1267–1272.
7. Wall JS, Kennel SJ, Paulus MJ, Gleason S, Gregor J, Baba J, Schell M,
T, O’Nuallain B, Donnell R, et al. (2005) Amyloid 12:149–156.
8. Cowan DF, Johnson WC (1970) Lab Invest 23:551–555.
9. Guo JT, Aldrich CE, Mason WS, Pugh JC (1996) Proc Natl Acad Sci USA
10. Landman WJM, Gruys E, Gielkens ALJ (1998) Avian Pathol 27:437–449.
11. Hazenberg BP, van Rijswijk MH (2000) Ann Rheum Dis 59:577–579.
12. McAdam, KPWJ, Raynes JG, Alpers MP, Westermark GT, Westermark P
(1996) Papua New Guinea Med J 39:284–296.
13. Booth DR, Booth SE, Gillmore JD, Hawkins PN, Pepys MB (1998) Amyloid
14. Scheinberg MA, Benson MD (1980) J Rheumatol 7:724–726.
15. Cathcart ES, Elliott-Bryant R (1999) Amyloid 6:107–113.
16. Mensua C, Carrasco L, Bautista MJ, Biescas E, Fernandez A, Murphy CL,
Weiss DT, Solomon A, Lujan L (2003) Vet Pathol 40:71–80.
17. Tojo K, Tokuda T, Hoshii Y, Fu X, Higuchi K, Matsui T, Kametani F, Ikeda
S-I (2005) Amyloid 12:103–108.
18. Ganowiak K, Hultman P, Engstro¨m U, Gustavsson Å, Westermark P (1994)
Biochem Biophys Res Commun 199:306–312.
19. Cui D, Kawano H, Takahashi M, Hoshii Y, Setoguchi M, Gondo T, Ishihara
T (2002) Pathol Int 52:40–45.
20. Lundmark K, Westermark GT, Olse´n A, Westermark P (2005) Proc Natl Acad
Sci USA 102:6098–6102.
21. Fu X, Korenaga T, Fu L, Xing Y, Guo Z, Matsush!ta T, Hosokawa M, Naiki
H, Baba S, Kawata Y, et al. (2004) FEBS Lett 563:179–184.
22. Korenaga T, Fu X, Xing Y, Matsusita T, Kuramoto K, Syumiya S, Hasegawa
K, Niaiki H, Ueno M, Ishihara T, et al. (2004) Am J Pathol 164:1597–1606.
23. Korenaga T, Yan J, Sawash!ta J, Matsush!ta T, Naiki H, Hosokawa M, Mori
M, Higuchi K, Fu X (2006) Am J Pathol 168:898–906.
24. Ludlage E, Murphy CL, Davern SM, Solomon A, Weiss DT, Glenn-Smith D,
Dworkin S, Mansfield KG (2005) Vet Pathol 42:117–124.
25. Murphy CL, Eulitz M, Hrncic R, Sletten K, Westermark P, Williams T, Macy
SD, Wooliver C, Wall J, Weiss DT, et al. (2001) Am J Clin Path 116:135–142.
Solomon et al. PNAS June 26, 2007 vol. 104 no. 26 11001
Post a Followup