Isolation of a T-lymphotropic retrovirus from a patient at risk for acquired immune deficiency syndrome (AIDS)

Abstract

A retrovirus belonging to the family of recently discovered human T-cell leukemia viruses (HTLV), but clearly distinct from each previous isolate, has been isolated from a Caucasian patient with signs and symptoms that often precede the acquired immune deficiency syndrome (AIDS). This virus is a typical type-C RNA tumor virus, buds from the cell membrane, prefers magnesium for reverse transcriptase activity, and has an internal antigen (p25) similar to HTLV p24. Antibodies from serum of this patient react with proteins from viruses of the HTLV-I subgroup, but type-specific antisera to HTLV-I do not precipitate proteins of the new isolate. The virus from this patient has been transmitted into cord blood lymphocytes, and the virus produced by these cells is similar to the original isolate. From these studies it is concluded that this virus as well as the previous HTLV isolates belong to a general family of T-lymphotropic retroviruses that are horizontally transmitted in humans and may be involved in several pathological syndromes, including AIDS.

Full text

Isolation of a T-Lymphotropic Retrovirus from a Patient at Risk for Acquired
Immune Deficiency Syndrome (AIDS)
F. Barré-Sinoussi; J. C. Chermann; F. Rey; M. T. Nugeyre; S. Chamaret; J. Gruest; C. Dauguet;
C. Axler-Blin; F. Vézinet-Brun; C. Rouzioux; W. Rozenbaum; L. Montagnier
Science, New Series, Vol. 220, No. 4599. (May 20, 1983), pp. 868-871.
Stable URL:
http://links.jstor.org/sici?sici=0036-8075%2819830520%293%3A220%3A4599%3C868%3AIOATRF%3E2.0.CO%3B2-D
Science is currently published by American Association for the Advancement of Science.
Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at
http://www.jstor.org/about/terms.html. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained
prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in
the JSTOR archive only for your personal, non-commercial use.
Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at
http://www.jstor.org/journals/aaas.html.
Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.
JSTOR is an independent not-for-profit organization dedicated to and preserving a digital archive of scholarly journals. For
more information regarding JSTOR, please contact support@jstor.org.
http://www.jstor.org
Wed Jun 6 12:49:30 2007

Isolation of
a
T-Lymphotropic Retrovirus from a Patient
at Risk for Acquired Immune Deficiency Syndrome (AIDS)
Abstract.
A retrovirus belonging to the fhmily cfrc.cently discovcrc.d human T-cdl
leukemia viruses (117'LVj, hut
clearly
distinct from each previoins isolatc, has hecn
isolatc~cifrom a Caricasian patic.nt with signs and symptorns that cgten prc.c,ecic the
acquired immrinc c1yfic.ic.nc.y syndrome (AIDS). This i'irus is a typical type-C' KNA
tumor virus, buds from the c.cll membrane, prcfi.rs rnagnc~.siinrn for revclrsc transcrip-
tax uc.tivity, und has an internal antigen
(~25)
sirnilur to IITI,V
p24.
Antibodies
jkom
serum oj'this patient react with protc,ins Jrorn viru.sc.s ofthe 117'LV-I suhgrorip, hint
type-specific antisc>ra to HTLV-I do not prc>c.ipitatc> proteins of the new isolatc.
The.
virris ,from this
patient
has hccn transmitted into cord hlooci lymphocytc~s, ancl the
virins produced by these cc.lls is similar to the original isolatr. From thcsc. stinciic~s it is
conclrirlcd that this virus as ~x.11 as the prrvioins HT1.V isolates hclong to a
general
family
cg
7lymphotropic retrovirmsc.s that arc horizontally transmitt~d in hmnlans
and may bc involvc.ci in sc.vc.ral pathologicx~l syncirornc.~, inclrrding AIDS.
The acquired immune deficiency syn-
drome (AIDS) has recently been recog-
nized in several countries
(I).
The dis-
ease has been reported mainly in homo-
sexual males with multiple partners, and
epidemiological studies suggest horizon-
tal transmission by sexual routes (2) as
well as by intravenous drug administra-
tion
(3),
and blood transf~~sion (4). The
pronounced depression of cellular immu-
nity that occurs in patients with AIDS
and the quantitative modifications of
subpopulations of their
7'
lymphocytes
(5)
suggest that T cells or a subset of
1'
cells might be
a
preferential target fhr the
putative infectious agent. Alternative1 y,
these modifications may result from sub-
sequent infections.
The depressed cellu-
lar immunity may result in serious op-
portunistic infections in AIDS patients,
many of whom develop Kaposi's sarco-
ma
(I).
However, a picture of persistent
multiple lymphadenopathies has also
been described in homosexual males
(6)
and inhnts
(7)
who may or may not
develop AIDS
(8).
The histological as-
pect of such lymph nodes is that of
reactive hyperplasia. Such cases may
correspond to an early or
a
milder form
of the disease. We report here the isola-
tion of
a
novel retrovirus from a lymph
node of
a
homosexual patient with multi-
ple lymphadenopathies. The virus ap-
pears to be a member of the human
T-cell leukemia virus (HI'L,V) Ibmily
(9).
The retrovirus was propagated in cul-
tures of 'S lymphocytes from a healthy
adult donor and from umbilical cord
blood of newborn humans. Viral core
proteins were not immunologically relat-
ed to the
p24 and p19 proteins of sub-
group
I
of HTLV
(9).
However, serum of
the patient reacted strongly with surface
antigen (or antigens) present on HTLV-
I-infected cells. Moreover, the ionic re-
quirements
ofthe viral reverse transcrip-
tase were close to that of HTLV. lie-
cently, a type-C retrovirus was also iden-
tified in T cells from
a
patient with hairy
cell leukemia. Analysis of the proteins of
this virus showed they were related to,
but clearly
diKerent fi-om, proteins of
previous H'SLV isolates (10). Moreover,
recent studies
of
the nucleic acid se-
quences of this new virus show it is less
than 10 percent homologous to the earli-
er HTLV isolates (11). This virus was
called
H'SLV-I1 to distinguish it from all
the earlier, highly related viruses termed
0
5
10
Fractions
Fig.
1.
Analysis of virus from patient
I
on
sucrose gradients. Cord blood
T
lymphocytes
infected with virus from patient
I
were la-
beled for 18 hours with 13Fl]uridine
(28
Cil
mmole, Amersham; 20 I~Ci/ml). Cell-free su-
pernatant was ultracentrifuged for
1
hour at
50,000 revlmin. The pellet was resuspended in
200 p1 of
NTE
buffer (10 mM tris,
pH
7.4, 100
mM NaCI, and
1
mM
EDTA)
and was centri-
fuged over a 3-ml linear sucrose gradient (10
to 60 percent) at 55,000 revlmin for 90 minutes
in an JEC type
SB
498 rotor. Fractions (200
p1) were collected, and 30 p1 samples of each
fraction were assayed for DNA polymerase
activity with 5 mM Mg2' and poly(A)
.
oligo-
(dr),?
IK
as template primer; a 20-p1 portion of
each fraction was precipitated with 10 percent
trichloroacetic acid and then filtered on a
0.45-pm Millipore filter. The 'H-labeled acid
precipitable material was measured in a Pack-
ard
p
counter.
HTLV-I. The new retrovirus reported
here appears to also differ from HI'LV-
11. We tentatively conclude that this vi-
rus, as well as all previous H'SLV iso-
lates, belong to
a
family of l'-lympho-
tropic retroviruses that are horizontally
transmitted in humans and may be in-
volved in several pathological
5yn-
dromes, including AIDS.
The patient wa5
a
33-year-old homo-
sexual male who sought medical consul-
tation in December 1982 for cervical
lymphadenopathy and asthenia (patient
1).
Examination showed axillary and in-
guinal lymphadenopathies. Neither fever
nor recent loss of weight were noted.
The patient had
a
history of several
episodes of gonorrhea and had been
treated for syphilis in September 1982.
During interviews he indicated that he
had had more than 50 sexual partners per
year and had traveled to many countries,
including North Africa, Greece, and
In-
dia. His last trip to New York was in
1979.
Laboratory tests indicated positive se-
rology (immunoglobulin G) for
cytomeg-
alovirus (CMV) and Epstein-Rarr virus.
Herpes simplex virus was detected in
cells from his throat that were cultured
on human and monkey cells. A biopsy of
a cervical lymph node was performed.
One sample served for histological ex-
amination, which revealed
follicular hy-
perplasia without change of the general
architecture of the lymph node. Immu-
nohistological ctudies revealed, in para-
cortical areas, numerous T lymphocytes
(OKT3+). Typing of the whole cellular
suspension indicated that 62 percent of
the cells were
7'
lymphocytes (OK7'3
'),
44 percent were l'-helper cells (OKT4
'),
and
16
percent were suppressor cells
(OKT8
'
).
Cells of the same biopsied lymph node
were put in culture medium with phyto-
hemagglutinin (PHA), T-cell growth fac-
tor (TCGF), and antiserum to human
a
inte~feron (12). 'She reason for using this
antiserum was to neutralize endoge-
nous interferon which is secreted by
cells chronically infected by viruses, in-
cluding retroviruses. In the mouse sys-
tem. we had previously shown that anti-
serum to interferon could increase retro-
virus production by
a
lbctor of 10 to
50
(13).
After 3 days, the culture was con-
tinued in the same medium without
PHA. Samples were regularly taken for
assay of reverse transcriptase and for
examination in the electron microscope.
After 15 days of culture, a reverse
transcriptase activity was detected in the
culture supernatant by using the ionic
conditions described
fix HTLV-I (14).
Virus production continued for 15 days
SCIENCE,
VOL.
220
868

and decreased thereafter, in parallel with
the decline of lymphocyte proliferation.
Peripheral blood lymphocytes cultured
in the same way were consistently nega-
tive for reverse transcriptase activity,
even after 6 weeks. Cytomegalovirus
could be detected, upon prolonged co-
cultivation with MRCS cells, in the origi-
nal biopsy tissue, but not in the cultured
T lymphocytes at any time of the culture.
Virus transmission was attempted
with the use of a culture of T lympho-
cytes established from an adult healthy
donor of the Blood Transfusion Center at
the Pasteur Institute. On day 3, half of
the culture was cocultivated with lym-
phocytes from the biopsy after centrifu-
gation of the mixed cell suspensions.
Reverse transcriptase activity could be
detected in the supernatant on day I5 of
the coculture but was not detectable on
days 5 and 10. The reverse transcriptase
had the same characteristics as that re-
leased by the patient's cells and the
amount released remained stable for 15
to 20 days. Cells of the uninfected cul-
ture of the donor lymphocytes did not
release reverse transcriptase activity
during this period or up to 6 weeks when
the culture was discontinued.
The cell-free supernatant of the infect-
ed coculture was used to infect 3-day-old
cultures of
T
lymphocytes from two um-
bilical cords, LC1 and LC5, in the pres-
ence of Polybrene (2 pg/ml). After a lag
period of
7
days, a relatively high titer of
reverse transcriptase activity was detect-
ed in both of the cord lymphocyte cul-
tures. Identical cultures, which had not
been infected, remained negative. These
two successive infections clearly show
that the virus could be propagated on
normal lymphocytes from either new-
borns or adults.
That this new isolate was a retrovirus
was further indicated by its density in a
sucrose gradient, which was 1.16, and by
its labeling with [3H]uridine (Fig. 1).
Electron microscopy of the infected um-
bilical cord lymphocytes showed charac-
teristic immature particles with dense
crescent (C-type) budding at the plasma
membrane (Fig.
2).
Virus-infected cells from the original
biopsy as well as infected lymphocytes
from the first atld second viral passages
were used to determine the optimal re-
quirements for reverse transcriptase ac-
tivity and the template specificity of the
enzyme. The results were the same in
all instances. The reverse transcriptase
activity displayed a strong affinity
for poly(adeny1ate
.
oligodeoxythymid-
ylate) [poly(A)
.
oligo(dT)I, and required
~g~+ with an optimal concentration (5
mM) slightly lower than that for HTLV
20
MAY
1983
(14) and an optimal pH of
7.8.
The reac-
tion was not inhibited by actinomycin D.
This character, as well as the preferential
specificity for riboseadenylate
.
deoxy-
thymidylate over deoxyadenylate
.
de-
oxythymidylate, distinguish the viral en-
zyme from DNA-dependent polymer-
ases.
We then determined whether or not
this isolate was indistinguishable from
HTLV-I isolates. Human T-cell leuke-
mia virus has been isolated from cultured
T lymphocytes of patients with T lym-
phomas and T leukemias [for a review,
see (9)l. The antibodies used were spe-
cific for the p19 and p24 core proteins of
HTLV-I. A monoclonal antibody to p19
(15) and a polyclonal goat antibody to
p24
(16)
were used in an indirect fluores-
cence assay against infected cells from
the biopsy of patient 1 and lymphocytes
obtained from a healthy donor and in-
fected with the same virus. As shown in
Table 1, the virus-producing cells did not
react with either type of antibody,
whereas two lines of cord lymphocytes
chronically infected with HTLV (17) and
used as controls showed strong surface
fluorescence.
When serum from patient
1
was tested
against infected lymphocytes from the
biopsy the surface fluorescence was as
Table
1.
Indirect immunofluorescence assay. Cells were washed with phosphate-buffered saline
(PBS) and resuspended in the same buffer. Portions (5
x
lo4
cells) were spotted on slides, air-
dried and fixed for 10 minutes at room temperature in acetone. Slides were stored at -80°C until
use. Twenty microliters of either monoclonal antibody to HTLV p19 (diluted 11400 in PBS) or
goat antibody to HTLV p24 (diluted 11400
in
PBS) or serum from patient
1
diluted 1/10
in
PBS
was applied to cells and incubated for 45 minutes at 37OC. The appropriate fluorescein-
conjugated antiserum (antiserum to mouse, goat, or human immunoglobulin
G)
was diluted and
applied to the fixed cells for 30 minutes at room temperature. Slides were then washed three
times
in
PBS. Cells were stained
with
Evans blue solution for I5 minutes and then washed
extensively with water before microscopic examination.
Immunofluorescence (percent positive)
Cell type
Antibody to Antibody to Serum from
~19 ~24 patient
I
Normal blood lymphocytes
N
10916
-
-
-
LC
I
-
-
-
HTLV-producing cells
C9,/PL
+
(90 to
100)
+
(90
to 100)
+
(90 to 100)
CtdMJz
+
(90 to 100)
+
(90 to
100)
+
(90
to
100)
Virus-producing cells from
Patient
1
-
-
+
(90
to
100)
LCllpatient
1
- -
-+
(0.5 to 2)
Patient 2
-
-
+
(90
to
100)
Fig.
2.
Electron microscopy of thin sections of virus-producing cord lymphocytes.
The
inset
shows various stages of particle
budding
at the cell surface.

intense as that of the control HTLV-
producing lines. This suggests that se-
rum of the patient contains antibodies
that recognize a common antigen present
on HTLV-I-producing cells and on the
patient's lymphocytes. Similarly, cord
lymphocytes infected with the virus from
patient
1
did not react with antibodies to
p19 or p24. Only a minor proportion of
the cells (about
1
percent) reacted with
the patient's serum. This may indicate
that only this fraction of the cells was
infected and produced virus. Alterna-
tively, the antigen recognized by the
patient's serum may contain cellular de-
terminants that show less expression in
T lymphocytes of newborns.
We also cultured T lymphocytes from
a lymph node of another patient (patient
2) who presented with multiple adenopa-
thies and had been in close contact with
an AIDS case. These lymphocytes did
not produce viral reverse transcriptase;
however, they reacted in the immunoflu-
orescence assay with serum from patient
1. Moreover, serum from patient 2 react-
ed strongly with control HTLV-produc-
ing lines (not shown). In order to deter-
mine which viral antigen was recognized
by antibodies present in the two patients'
sera, several immunoprecipitation ex-
periments were carried out. Cord lym-
phocytes infected with virus from patient
1
and uninfected controls were labeled
with [3SS]methionine for 20 hours. Cells
were lysed with detergents, and a cyto-
plasmic S10 extract was made. Labeled
virus released in the supernatant was
banded in a sucrose gradient. Both mate-
rials were immunoprecipitated by antise-
rum to HTLV-1 p24, by serum from
patients
1
and 2, and by serum samples
from healthy donors. Immunocomplexes
were analyzed by polyacrylamide gel
electrophoresis under denaturing condi-
tions. Figure 3 shows that a p25 protein
present in the virus-infected cells from
patient
1
and in LC1 cells infected with
this virus, was specifically recognized by
serum from patients
1
and 2 but not by
antiserum to HTLV-1 p24 or serum of
normal donors. Conversely, the p24
Fig. 3. Immunoprecipitation of 3SS-labeled viral proteins. Cord blood T-lymphocytes infected
with virus from patient
I
were incubated overnight in culture medium containing one-fifth of the
normal concentrations of methionine in minimum essential medium, [35S]methionine (1500 Cil
mmole, Amersham; 50 pCiIml), and 10 percent dialyzed fetal calf serum. The virus was purified
by banding on a sucrose gradient as described in Fig.
I.
Labeled cells were resuspended in 10 pI
of saline and then lysed with
90
pl of RIPA buffer (18) containing aprotinin (500 Ulml;
Zymofren, Specia) at 4°C for 15 minutes. The supernatant of a 10,000g centrifugation of the cell
extract was used for immunoprecipitation. A similar extract was made from HTLV-producing
C9,1PL cells (17). (A) Portions (20 pl) of cell extracts were mixed with
6
p1 of serum, incubated
for 2 hours at 37OC and overnight at +4"C. Then,
60
p1 of a suspension of Protein ASepharose
(10 mglml in RIPA buffer) were added. After 45 minutes of incubation at 4"C, immunocom-
plexes bound to Protein A-Sepharose were washed five times with RIPA buffer by centrifuga-
tion, heated for
3
minutes at 100°C in denaturing buffer and electrophoresed on 12.5 percent
polyacrylamide-SDS slab gel (19). Lanes 1 to 5: Extract of LC, cells infected with virus from
patient
I
and tested against 1, serum from patient 1; 2, serum from patient 2; 3, serum of a
healthy donor; 4, goat antiserum to HTLV-Ip24; 5, normal goat serum. Lanes
6
to 10: C91lPL
(HTLV-producing) cell extract tested with:
6,
serum from patient 1; 7, serum from patient 2;
serum of a healthy donor; 4, goat antiserum to HTLV-Ip24; 5, normal goat serum.
(B)
Portions
(20 p1) of the band containing virus from patient 1 were treated with various antisera and
processed as described for cell extracts. Lane I, serum from patient 1; 2, serum from patient 2;
3,
serum of a healthy donor; 4, serum of another healthy donor; 5, goat antiserum to HTLV-
Ip24. Arrows indicate the p24-p25 protein. Molecular weights (in thousands) are indicated on
the left.
present in control HTLV-infected cell
extracts was recognized by antibodies to
HTLV but not by serum from patient
1.
A weak band (lane 2, Fig. 3B) could
hardly be seen with serum from patient
2, suggesting some similarities of the p25
protein from this patient's cells with
HTLV-1 p24. When purified, labeled vi-
rus from patient
1
was analyzed under
similar conditions, three major proteins
could be seen: the p25 protein and pro-
teins with molecular weights of 80,000
and 45,000. The 45K protein may be due
to contamination of the virus by cellular
actin which was present in immunopre-
cipitates of all the cell extracts (Fig.
3).
These results, together with the immu-
nofluorescence data, indicate that the
retrovirus from patient
1
contains a ma-
jor p25 protein, similar in size to that of
HTLV-I but different immunologically.
The DNA sequences of these and other
members of the HTLV family are being
compared. All attempts to infect other
cells such as a B-lymphoblastoid cell line
(Raji), immature or pre-T cell lines
(CEM, HSB2), and normal fibroblasts
(feline and mink lung cell lines) were
unsuccessful.
The role of this virus in the etiology of
AIDS remains to be determined. Patient
1
had circulating antibodies against the
virus, and some of the latter persisted in
lymphocytes of his lymph node (or
nodes). The virus-producing lympho-
cytes seemed to have no increased
growth potential in vitro compared to the
uninfected cells. Therefore, the multiple
lymphadenopathies may represent a host
reaction against the persistent viral in-
fection rather than hyperproliferation of
virus-infected lymphocytes. Other fac-
tors, such as repeated infection by the
same virus or other bacterial and viral
agents may, in some patients, overload
this early defense mechanism and bring
about an irreversible depletion of T cells
involved in cellular immunity.
F. BARRC-SINOUSSI,
J.
C. CHERMANN
F. REY, M. T. NUGEYRE
S.
CHAMARET,
J.
GRUEST
C. DAUGUET, C. AXLER-BLIN
Institut Pasteur, DPpartement de
Virologie, 75724 Paris CPdex I5
F.
V~ZINET-BRUN, C. ROUZIOUX
Hbpital Claude Bernard, Laboratoire
Central-Virologie, 10 avenue de la
Porte d'Aubervilliers, 75019 Paris
W. ROZENBAUM
Hbpital La PitiP-SqlpPtriPre,
DPpartement de SantP Publique et
MPdecine Tropicale,
97 Boirlevard de I'Hbpital, 75013 Paris
L. MONTAGNIER
Institut Pasteur, Departement de
Virologie, 75724 Paris CPdex 15
SCIENCE.
VOL.
220

References and Notes
1.
Centers for Disease Control, Task Force on
Kaposi's Sarcoma and O~portunistic Infections,
N. Engl.
J.
Med.
306,
248 (1982).
2.
M. Marmor
et al., Lancet
1982-11, 1083 (1982);
S. Fannin
et al., Morbid. Mortal. Weekly Rep.
31, 305 (1982).
3.
Centers for Disease Control, Task Force on
Kaposi's Sarcoma and Opportunistic Infections.
Morbid. Mortal. Weekly ~eb.
31, 507 (1982).
4.
M. C. Poon
et al., !bid.,
p. 644.
5.
R.
E.
Stahl, A. Fnedman-Kien, R. Dubin, M.
Marmor, S. Zolla Parner,
Am.
J.
Med.
73, 171
(1982).
6.
D. Mildvan
et al., Morbid. Mortal. Weekly Rep.
31, 249 (1982).
7.
R. O'Reilly
et al., ibid.,
p.
665.
8.
W.
ROzenbaum
et
Lancet
1982-11,
572
(1982).
9.
R. C. Gallo and M. S. Reitz, Jr.,
J.
Natl. Cancer
Inst.
69
(No.
6), 1209 (1982).
10.
V. S. Kalyanaraman
er a/., Science
218, 571
(1982).
11.
E.
Gelmann, F. Wong-Staal, R. Gallo, personal
communication.
12.
The cells were grown in RPMI-1640 medium
supplemented with antibiotics,
10-'M
p-mercap-
toethanol,
10
percent fetal calf serum,
0.1
per-
cent sheep antibody to human
a
interferon (neu-
tralizing titer,
6
IU at
lo-'),
and
5
percent
TCGF,
free
ofp~~,
13.
F. Ban&-Sinoussi
el Ann.
Microbial.
(Inst.
Pasreur)
l30B 349 (1979).
14.
B.
J.
Poiesz
'et a/., Proc. Natl. Acad. Sci.
U.S.A.
77, 7415 (1980).
15.
M. Robert-Guroff
et al.,
J.
Exp. Med.
154, 1957
(1981).
16.
V,
S,
~~l~~~~~~~~~
et
J,
viral,
38,
906
(1981).
17.
M. Popovic, P. S. Sarin, M. Robeft-Guroff, V.
S. Kalyanaraman, D. Mann, J. Minowada, R. C.
~~ll~,
science
219, 856 (1983).
18.
R.
E.
Karess
et al., Proc. Natl. Acad. Sci.
U.S,A,
76, 3154 (1979).
19.
U. K. Laemmli,
Nature (London)
227. 608
(1970).
20.
We thank Dr. Fradellizi for gifts of T-cell growth
factor, R. C. Gallo for providing ant~bod~es to
gaining weight (3 to 5 days after capture),
they were divided into experimental
groups of similar mean body weight (460
g) and welght distribution. Blood
(1
ml)
was taken from each bird, and experi-
ments were begun immediately. In ex-
periment
1,
the birds were given either
10 or 20 ml of PBCO
(8)
per kilogram of
body weight per day in gelatin capsules;
controls received empty capsules. Dos-
ing continued daily. Blood samples were
drawn into heparinized tubes on day 5,
and routine hematological measurements
HTLV
and
for
HTLV-~~~~~~~~~
cells,
M~~.
L~
Franqois for preparation of the cord lympho-
were made (9). Values for all groups
cytes,
M.
Lavergne (Institut Pasteur, Produc-
tion) for gifts of fluorescein-conjugated antisera,
F,
Huraud for performing some of the HTLV
tests, and members of the French Working
G~~~~
on
AIDS
for
helpful discussion,
I9
April
1983
prior to dosing and for controls on day
5
of dosing did not differ significantly. On
day 5, blood taken from oil-dosed birds
was dark brown and failed to redden on
mixing with air; this result suggested a
marked reduction in oxygen-carrying he-
moglobin. The birds receiving oil were
severely anemic, with packed cell vol-
umes (PCV) reduced by 43 and 50 per-
cent (Table 1). The plasma from these
birds was rusty red, an indication of
hemolysis either intravascularly or dur-
ing sample handling. A strong regenera-
tive response to the anemia was evident
in the high reticulocyte count (Table
1).
After the birds were killed, there was no
evidence of trauma, enteric bleeding, or
other hemorrhage. These data are suffi-
cient to permit the classification of this
anemia arising from oil ingestion as a
hemolytic anemia (10).
Heinz bodies were abundant in eryth-
rocytes from oil-dosed birds (Table 1).
These were identified by two different
staining techniques applied to fresh sam-
ples and in sections studied by light and
transmission electron microscopy (Fig.
1)
(9). Heinz bodies are dense granular
masses in red cells thought to consist of
precipitates of hemoglobin oxidized in
the protein moiety. They are a classical
feature of toxic hemolytic anemias pro-
duced by a variety of dissimilar chemi-
cals linked mechanistically by the ability
to cause destructive oxidative reactions
in red cells (11). The presence of Heinz
bodies is good evidence of a primary
toxicosis of red cells and of an oxidative
biochemical mechanism of toxicity. To
probe other possible sites of destructive
oxidative processes in these red cells, we
measured methemoglobin, sulfhemoglo-
bin, and reduced glutathione (GSH) in
whole blood and extractable fluores-
cence (EF) in red cell membranes as an
indicator of membrane lipid peroxidation
(12).
No sulfhemoglobin was detected.
The range of methemoglobin values was
wide in all groups, and experimental
animals did not differ significantly from
controls. Means and standard deviations
for the percentages of methemoglobin
Heinz-Body Hemolytic Anemia from the Ingestion of Crude Oil:
A Primary Toxic Effect in Marine Birds
Abstract. Hemolytic anemia developed in young herring gulls and Atlantic puflns
given daily oral doses of a Prudhoe Bay crude oil. Anemia developed 4 to
5
days after
the initiation of oil ingestion and was accompanied by Heinz-body formation and a
strong regenerative response. The data evince a toxic effect on circulating red blood
cells involving an oxidative biochemi~al mechanism and the first clear evidence of a
primary mechanism of toxicity from the ingestion of crude oil by birds.
Petroleum oils regularly enter the ma-
rine environment through spills, runoff,
and seepage (I). Large numbers of birds
have died in association with marine oil
spills
(2),
and the effects of oil on birds
have been studied experimentally. Birds
that become oiled ingest oil while preen-
ing
(3),
and oral doses of several petro-
leum oils have produced a wide range of
sublethal toxic changes affecting growth,
reproduction, osmoregulation, steroid
metabolism, and hepatic function
(4-6).
Thus there is a firm basis for concern
that oil pollution may produce subtle,
sublethal effects in wild birds that impair
reproduction or survival. This concern is
heightened by the current emphasis on
offshore oil development.
We report here that young herring
gulls (Larus argentatus) and Atlantic
puffins (Fratercula arctica) developed a
severe hemolytic anemia after several
days of oral dosing with a Prudhoe Bay
crude oil (PBCO). Our data indicate that
this was a primary toxic effect in which
oxidative chemical processes damaged
red blood cells in the peripheral circula-
tion, and they constitute the first clear
evidence of a primary toxic mechanism
in experimental studies of the toxicity of
ingested crude oil in birds.
In our initial experiments we used
herring gull nestlings 2 to
3
weeks old,
20
MAY
1983
taken from a coastal colony. These
young birds adjust well to captivity and
tolerate the manipulation required in lab-
oratory work. Gulls were collected on
Great Island, 50 km south of St. John's,
Newfoundland, and held in pens at Me-
morial University of Newfoundland, St.
John's
(7).
Pens were partially bedded
with hay, and the birds were fed unlimit-
ed amounts of capelin (Mallotus villosus)
and seawater. When all the birds were
Fig. 1. Transmission electron micrograph
of
a
Heinz body attached to the plasma membrane
of a red blood
cell
from a herring gull which
ingested 20 ml of Prudhoe Bay crude oil per
kilogram per day for
4
days (experiment
1).
This
cell
is a ghost erythrocyte, with most of
the free hemoglobin lost from its cytoplasm;
bar.
200
x
lo-' mm.