462

VOl81 2 MAY 1992

SAMJ

------------------------------------------

Encephalitis and chorioretinitis associated with

neurotropic African horsesickness virus infection
in laboratory w-orkers
Part IV. Experimental infection of the vervet monkey (Cercopithecus
pygerythrus)
M. B. TAYLOR, C. H. VAN DERMEYDEN, B.J. ERASMUS, R. REID,
L. DREYER, o. W. PROZESKY

J. H. LABUSCAGNE,
Summary

Neurotropic vaccine strains of African horsesickness (AHS) virus types I and 6 were iInplicated as
the possible aetiological agents in 4 cases of
encephalitis and uveochorioretinitis in laboratory
workers accidentally exposed to the freeze-dried
vaccine preparations of the virus. To date, AHS
virus has not been known to infect man. To ascertain whether or not primates were susceptible to
infection with AHS virus, vervet monkeys (Cercopithecus pygerythrus) were inoculated, either
transnasally or intracon;unctivally, with vaccine
strains of AHS virus types I and 6. The course of
infection was monitored using parameters such as
behavioural changes, febrile reaction, cerebrospinal fluid pleocytosis, serology, magnetic
resonance imaging and autopsy. Encephalitis,
manifested by varying deg-rees of fever,
behavioural changes and pleocytosis, but no
chorioretinitis was detected in all 6 transnasally
infected monkeys. This was confirmed by autopsy, where a meningo-encephalitis affecting the
medial temporal lobe but no lesions in the eyes
was demonstrated. Neither virus appeared to
infect the aniInals after intracon;unctival inoculation. These findings support the theory that the
patients were infected by the inhalation of freezedried vaccine preparations. The pathogenesis of
the eye lesions, however, remains uncertain.
S Air Med J 1992; 81: 462467.

RICA..N" horsesickness (AHS), a highly infectious

and often fatal disease affecting equines,' has, to
date, not been known to affect man.' arural

Departments ofMedical Virology and Anatomical Pathology,

Institute of Pathology, University of Pretoria
M. B. TAYLOR, D.Se.
L. DREYER, M.D.
Departments of Neurology and Ophthalmology, University of
Pretoria and H. F. Verwoerd Hospital, Pretoria
C. H. VAN DER MEYDEN, M.B. CH.B., Fe.P. (SA) (Present
address; Department of Neurology, Universiry of the
Orange Free State, Bloernfontein)
R. REID, M.B. CH.B., M.MED. (OPHTH.), F.e.S. (SA.) (OPHTH.),
Fe.O. (Present address; 98 Beethoven St, Vanderbijlpark,
Tvl)
Veterinary Research Institute, Onderstepoort, Tvl
B. J. ERASMUS, B.v.se.
470 Myburgh Street, Capital Park, Pretoria
J. H. LABUSCAGNE, M."'16. (RAD. D.)
South African Medical Research Council, Parowvallei, CP
O. W. PROZESKY, M.B. CH.B., M.D.
Accepted 22 i\1ar 1991.

infection, predominantly via biological transmission by

Culicoides midges,'-' is usually restricted to equines,'
although outbreaks in dogs have been reponed. 6 The
experimental infection of a number of unrelated species,
including ferrets, mice, rats and guinea-pigs, has been
documented. I - 7
The aetiological agent, AHS virus, is a viscerotropic
arbovirus' of the genus Orbivirus. The AHSvirus serogroup is composed of 9 serotypes, comprising a number
of antigenically related strains, which show a variation in
virulence. The clinical features of the disease, especially
in horses, are dependent on the type of virus and the
immune status of the infected animal. Control of the
disease is predominantly by vaccination. of susceptible
animals and stabling in insect-proof enclosures. I
One of the vaccines currently in use contains 8 serologically distinct attenuated virus serotypes, administered as two quadrivalent doses. Of these vaccine
strains of AHS virus, 3 have been plaque attenuated by
passage in cell culture while 5 strains, A501 (type 1),
OD (type 2), L (type 3), VH (type 5) and 114 (type 6),
were attenuated by serial intracerebral passage in adult
mice. Studies on AHS virus attenuated in adult mouse
brain have indicated that the virus strains differ
considerably in their neurotropic properties for horses
and show a predilection for different pans of the horse's
brain. 1O One of the strains, Karen (type 7), caused a fatal
encephalitis while another strain, 7/60 (type 9), affected
the optic tracts and optic centres with resultant blindness when administered to horses intranasally.'o The
other neurotropic attenuated strains of AHS virus also
have the potential to cause encephalitis and blindness.
Recently AHS virus, types 1 and 6, have been implicated as the possible aetiological agents in 4 cases of
encephalitis and chorioretinitis in workers exposed to
the freeze-dried vaccine preparations."- l3 Cenain viruses
previously not known to infect man naturally have been
found to be infectious for man in the laboratory situation via an unnatural route of infection. I < Arboviruses,
especially via aerosol transmission, are documented as
being responsible for a large percentage of laboratoryassociated infections.'5-17 As this, to our knowledge, is
the first known repon of AHS in man it was essential to
try to establish the nature and pathogenesis of this infection in man to confirm the aetiological role of AHS virus
in the 4 clinical cases.
Experimental infection of non-human primates has
played an imponant role in elucidating the pathogenesis
of a number of human viral infections. I 8-'O Consequently
the vervet monkey (Cercopithecus pygerythms) was chosen
as a putative non-human primate model for initial investigations. These studies were instigated to ascenain the
possible route of infection of the patients and whether
the clinical picture was due to simultaneous infection
with both types of AHS virus or superinfection with one
type in patients previously sensitised by another type.

VOL 81 2 MAY 1992

463

SAMJ
.--------------------------------------------------

Wild-caught monkeys were therefore infected, either

transnasally or intraconjunctivally, with suspensions of
attenuated vaccine strains of types 1 and 6 of AHS
virus.

Materials and methods

AHS virus inocula
Freeze-dried mouse-brain artenuated AHS virus type 1
(strain A501) and type 6 (strain 114), reconstituted in
distilled water, with titres of 10,,9 suckling mouse
LD'o/rnl, were used for the initial infection procedures.
Freshly cell-cultured mouse-brain artenuated AHS
virus type 1 (Strain A50l) and type 6 (strain 114), resuspended in distilled water, with titres 1080 and 10 7,9 suckling mouse LD,oIml, respectively, were used for the
superinfection studies.

C. pygerythrus monkeys
Five newly wild-caught male monkeys, all lacking antibodies to AHS virus, were used for experimentation
after a quarantine period of 6 weeks. In addition,S wildcaught females, monkeys which had been in captivity for
up to 5 years, all lacking detectable AHS virus antibodies, were infected. The animals were housed in 0,6 x
0,6 x 0,8 m stainless steel cages suspended from a
wall. To monitor any cross- or airborne infection, an
uninfected control animal was housed in the same
room, approximately 2,0 m apart from the experimental
animals. One week before infection, the animals' temperatures were taken, lumbar punctures were performed
and blood samples collected. Serum was tested for the
presence of antibodies to AHS virus and the cerebrospinal fluid (CSF) for cell count, and protein, glucose
and chloride values. All sampling procedures were
carried out under ketamine hydrochloride (10 mg/kg)
sedation.
After infection, the animals were monitored daily for
any behavioural changes. Every 2 - 3 days post-infection, for a period of 28 days, temperatures were taken
and blood samples collected for antibody assays and
virus isolation. Thereafter, blood specimens and temperatures were taken weekly. Lumbar punctures were
performed when indicated by the occurrences of symptoms, and magnetic resonance imaging (MRI) was done
on selected animals. The animals' eyes were all examined 2 - 4' weeks post-infection. If marked clinical symptoms of encephalitis were evident, the animal was
humanely sacrificed using pentobarbitone (200 mg/kg)
and at autopsy selected samples of tissues were removed
aseptically for virus isolation and the remainder preserved in 10% phosphate-buffered formalin for
histopathological examination. The remaining animals
were all sacrificed, autopsies performed and specimens
processed (as above) 25 days after superinfection.

Infection programme
In two consecutive experiments, monkeys of different
sex and mass, were infected either intranasally or intraconjunctivally with the attenuated vaccine strains of
either AHS virus type 1 or AHS virus type 6 or a mixture of both viruses:
1. Initial infection: 3 monkeys were inoculated
intraconjunctivally with 0,2 rnl of either type 1 (CZl!84,
female, 3,6 kg), type 6 (C2l!85, female, 3,0 kg) or both
viruses (C39/89, male, 5,5 kg). Six monkeys were
instilled intranasally with 1 rnl of either type 1 (C52/85,
female, 4,2 kg; C31/89, male, 5,9 kg), type 6 (C53/85,
female, 4,3 kg; Cll/89, male, 5,9 kg) or both viruses
(C20/85, female, 4,6 kg; C44/89, male, 6,2 kg).

2. Superinfection: The monkeys used for the initial

infection procedure were re-infected 74 days post-infection, with the alternative virus strain to that used initially. The 3 monkeys that were infected intraconjunctivally were re-infected by the same route with 0,2 ml of
either type 1 (C21/85), type 6 (C21/84) or both types
(C39/89). The 6 monkeys infected intranasally were
also re-infected by the same route ""rith 1 ml of either
type 1 (C53/85; Cll/89), type 6 (C52/85; C31/89) or
both types (C20/85; C44189).
Serological methods. Antibody titres to AHS virus
types 1 and 6 in both serum and CSF specimens were
determined using the plaque reduction neutralisation
(PRN) assay as described previously.'
Viral isolation. Blood, CSF and autopsy tissue'
specimens were treated as described previously" and the
preparations used to infect BHK-21 cell cultures as well
as to inoculate suckling mice intracerebrally.
Eye examinations. After pupil dilation with tropicamide (1%) and phenylephrine HCl (10%), the animals' eyes were examined by indirect opthalmoscopy
with a hand-held 20 and 14 dioptre condensing lens.
MRI. The animals were sedated with pentobarbitone
(1,5 mg/kg) and studies done in a 1,5T magnet system
using a circular polarised head coil. The following
parameters were used to obtain T2 weighted images:
TR == 2,800 ms; TE == 90 ms; FOV == 14 cm; ~TEX == 2;
slice thickness 3 mm; interslice gap 0,6 mm.
Histopathology. Formalin-fixed eyes were cut sagittally and selected blocks of brain tissue including
frontal, temporal, parietal, occipital lobes, basal ganglia,
midbrain, cerebellum, olfactory tract and spinal cord
were examined for evidence of infection. The tissue was
routinely processed and sections were stained with
haematoxylin and eosin. Sections from the temporal
lobes and olfactory tracts were also stained with luxol
fast blue stain for myelin." Immunocytochemistry, using
the standard peroxidase-antiperoxidase method, was
used for the demonstration of glial fibrillary acidic protein (GFAP) and AHS virus.

Results
Clinical features and CSF changes
No clinical features associated with encephalitis, i.e.
behavioural changes, febrile reaction or lymphocytic
pleocytosis or signs indicative of chorioretinitis, were
noted in the uninfected control animal or any of the
intraconjunctivally infected animals after either initial or
superinfection.
Behavioural changes, varying from mild subtle
changes, such as passiveness, lethargy, anorexia and
sleepiness (predominantly in the female monkeys) to
marked changes, including nervous hyperactivity and
startle reactions (C31/89) and extreme aggression
(C44189), were noted in all animals infected intranasally
6 - 12 days after initial infection (Table I). Only 2 animals, C31/89 and to a lesser extent C44189, showed any
behavioural changes after superinfection. Animal
C3l!89 became nervously agitated and hyperactive with
a marked startle reaction 3 days after superinfection.
Febrile reactions (temperatures above 40C) were
detected in all intranasally infected animals 8 - 14 days
after initial infection (Table I) and only in animal
C3l!89 (day 10) after superinfection. PleoCytosis (lymphoCytes > 5 mm' and neutrophilis > mm') and also
an increase in CSF protein levels were detected in the 6
intranasally infected animals 10 - 20 days after the initial
infection (Table I) and in animal C3l!89 (day 10) after
superinfection. TO evidence of chorioretinitis was
detected in any of the intranasally infected animals' eyes
either after initial or superinfection.

464

VOL 81 2 MAY 1992

SAMJ

TABLE I.

Response of C. pygerythrus to initial infection with AHS virus

Monkey

Sex
Mass (kg)
Infection
AHS virus type
Route
Clinical (d*)
Behaviour changes
Febrile reaction
Pleocytosis
Chorioretinitis
MRI
CSF
Red blood cells (l1l1)
Neutrophils (/1l1)
Lymphocytes (l1l1)
Protein (mgll)
Glucose (mg/l)
Bacteria - propagation
AHSV antibody titre
Type 1
Type 6
Virus isolation
Blood
CSF
Seroconversion (d*)
Type 1
Type 6
'Duration in days post-infection
IN ~ intranasally; - = negative; NO

C52185

C31/89

C53/85

C11/89

C20/85

CMI89

F
4,2

M
5,9

F
4,3

M
5,9

F
4,6

M
6,2

1
IN

1
IN

6
IN

6
IN

1 &6
IN

1 &6
IN

6-14
8-10
20;32

14-23
14-18
14-20

10-16
10
10;14
20;32

9-19
10
10;14
20;32

8-14
8
10;14
20;32

12-18
14-16
14;16
20;32

NO
Day 32
0
0
5
150
2,9

Day 16
0
0
15
300
3,0

NO
Day 20
0
0
10
150
2,3

NO
Day 32
0
0
20
250
2,3

NO
Day 20
0
0
5
390
2,5

Day 20
0
0
24
340
2,3

Day 32
<10
80

Day 32
<10
40

Day 32
30
< 10

Day 48
<10
<10

48
28

48

28
48

Day 32
20
<10

NO

28
= nol done.

TABLE 11.

serum antibody titres to AHS virus in AHS virus-infected monkeys

Days post-infection
Monkey
C25/89
C39/89
C20/85
CMI89
C21/84
C52189
C31/89
C21/85
C53/85
C11/89

AHS virus
antibodies
Type 1
Type 6
Type 1
Type 6
Type 1
Type 6
Type 1
Type 6
Type 1
Type 6
Type 1
Type 6
Type 1
Type 6
Type 1
Type 6
Type 1
Type 6
Type 1
Type 6

-3

0-23

28

48

70
-4

74
0

78
4

81
7

84
10

86
12

<10
<10
<10
<10
<10
<10
<10
<10
<10
<10
<10
<10
<10
<10
<10
<10
<10
<10
<10
<10

<10
<10
<10
<10
<10
<10
<10
<10
<10
<10
.<10
<10
<10
<10
<10
<10
<10
<10
<10
<10

<10
<10
<10
<10
640
<10
<10
<10
<10
<10
40
<10
<10
<10
<10
<10
<10
20
<10
<10

<10
<10
<10
<10
640
20
<10
<10
<10
<10
80
<10
<10
<10
<10
<10
20
20
<10
20

<10
<10
<10
<10
1280
40
<10
<10
<10
<10
80
<10
<10
<10
<10
<10
20
20
<10
80

<10
<10
<10
<10
NT
NT
<10
<10
<10
<10
NT
<10
<10
<10
<10
<10
NT
NT
<10
NT

<10
<10
<10
<10
NT
NT
<10
<10
<10
<10
NT
<10
<10
<10
<10
<10
NT
NT
<10
NT

<10
<10
<10
<10
2560
80
<10
<10
<10
<10
80
<10
<10
<10
<10
<10
20
40
<10
40

<10
<10
<10
<10
10240
80
<10
<10
<10
<10
160
<10
<10
<10
<10
<10
40
40
<10
40

<10
<10
<10
<10
NT
NT
<10
<10
<10
<10
NT
<10
<10
<10
<10
<10
NT
NT
NT
NT

88
14

92
18

99'
25t

<10
<10
<10
<10
<10
<10
<10
<10
<10
<10
<10
<10
10240 10240 10240
160
160
160
<10
<10
<10
<10
<10 - <10
<10
<10
<10
<10
<10
<10
320
80
80
<10
<10
<10
<10
<10
<10
<10
<10
<10
<10
<10
80
40
80
80
40
80
<10
<10
<10
80
40
40

Innial infection.
Superinfection.
NT = nol tested.

Viral isolation and serological results

No viruses were isolated from any of the blood, CSF or
autopsy tissue specimens.
Serum antibodies to AHS virus were not detected
in the uninfected control animal or in 2 of the intraconjunctivally infected an4nals (C21/85 and C39/89) or
2 of the intranasally infected animals (C31/89 and

C44189) after the initial or the superinfection (up to 25

days post-infection) (Table II). Serum antibodies to
AHS virus types 1 and 6 were detected in the remaining
5 animals at varying times after initial infection and
superinfection (Tables I and
Antibodies to rype 1, a more virulent virus, were
detectable 28 days post-infection, while those to type 6
were detected only 48 days after the initial infection.

m.

VOl81 2 MAY 1992

465

SAMJ

---------------------------------------------------_........::.-MRI
Abnormalities were visualised by MRI in 1 of me 2
infected animals examined. Animal C3l/89 was first
examined 36 days after the initial infection and reexamined 13 days after superinfection. At the first
examination, two small poorly circumscribed areas of
increased signal intensity were seen in the medial
aspects of me left temporal lobe in me parahippocampal
gyrus on me T2 weighted images. An even weaker signal
increase was visible in me right parahippocampal gyrus.
When compared wim equivalent sections of me control
animal (Fig. 1) this was clearly abnormal.

The abovementioned changes in me temporal lobe

were thought re be due re areas of inflammation. This process then probably changed re necrotic lesions giving rise
re more clear demarcation and elevated signal intensity.
No changes were detected in animal C44/89 41 days
after me initial infection.

Histological findings
No histological evidence of infection was shown in me
eyes and optic nerves of me control animal or in any of
the infected monkeys. In animal CZ1/84, infected intraconjunctivally, hydrocephalus wim dilatation of bom
lateral ventricles was evident on gross examination.
Parasitic cYStS with the features of cysticercus were
present in me left lateral ventricle and me left parietal
conex. The control animal, C25/89, and me omer animals infected by me intraconjunctival route, did not
show any abnormalities in me brain sections.
The 6 animals infected via me intranasal route had a
marked meningo-encephalitis wim a mononuclear infiltrate in me subarachnoidal space, extending into me
Virchow-Robin space (Fig. 3). This was associated wim
a localised encephalitis affecting the olfactory tract
bilaterally.

FIG. 1.

MRI of the parahippocampal gyral region of an uninfected

monkey.

The second examination on the same monkey

revealed more clearly demarcated areas of increased
signal intensity in me left parahippocampal gyral region
on me T2 weighted images. These corresponded wim
me areas of abnormality depicted at me first examination. The signal change in me right parahippocampal
gyrus was also more intense, approaching the signal
intensity of CSF (Fig. 2). These changes also correlated
wim me changes displayed during me first examination.

FIG. 3.

section of temporal lobe (parahippocampal area) of

monkey C31/89, showing lymphocytic infiltrate in
Virchow-Robin space and showing brain substance
with focal necrosis (H and E x 180).

The parahippocampal gyrus and orbitofrontal lobes

were marked by focal necrosis of neurons and astrocyres
and an infiltrate of Iymphocyres and histiocyres (Fig. 3).
An increase in glial tissue in me area of encephalitis
was demonstrable on stains for GFAP (Fig. 4), but

AG.2..

FIG. 4.

MRI of C31/89 after superinfection showing an intense

signal change in the right parahippocampal gyrus.

section of temporal lobe (parahippocampal area) of

monkey C31/89, showing gliosis (GFAP x 180).

466

VOl81 2 MAY 1992

SAMJ

._.:.....-_--------------------------------demyelinisation was not evident. AHS virus 'was not

clearly demonstrilble by peroxidase-antiperoxidase staining and no inclusion bodies were seen. No virus particles were revealed by electron microscopy.
The anatomical distribution of the encephalitis was
very constant with the degree of inflammation varying
slightly from animal to animal. The location of the
lesions was suggestive of intranasal infection which had
spread directly through the olfactory nerves to the
parahippocampal gyrus and the orbitofrontallobes.
The histological features of the lesions were not
influenced by the sequence of infection with the different types or combinations of types of AHS virus.

Discussion
The successful transnasal infection of the vervet
monkey, albeit without the ocular lesions, with neurotropic vaccine strains of AHS virus types 1 and 6, indicates that these attenuated viral strains can, in fact,
infect primates and could therefore have been responsible for the human infections.
Infections of the central nervous system (CNS) are
usually a rare complication of an infection established
elsewhere in the body and, for reasonsthat are unclear,
only a small proportion of infected persons will develop
neural complications. 22 Viruses may penetrate the CNS
via neural, haematogenous or olfactory routesn and it
appears that cettain viruses have a predilection for particular parts of the CNS, which are often reflected in the
clinical signs of the disease. 22 During natural infections,
the direct exposure of a nerve to a virus is an unusual
event. Replication of the virus at peripheral sites may
possibly facilitate neural spread by amplifying the original inoculum. 24 In the olfactory mucosa, however, nerve
cells (receptor cells with olfactory rods) are exposed to
the exterior through the cribriform platen ,2. and experimental evidence has shown biologically inert particles to
move directly to the olfactory lobes via the olfactory
nerves. n The olfactory spread of aerosols of arboviruses
and rabies has been implicated in several laboratoryacquired CNS infections. I ',2;
The highly conserved, localised nature of the lesions
in the monkeys' brains corresponded with the predominantly temporal and orbitofrontallobe involvement seen
in patient 2" and supports the view that the patients
were infected transnasally, via the olfactory route, by
aerosol infection with pulverised freeze-dried AHS virus
vaccine from broken vials. The lack of any clinical features or histological evidence of infection in the control
and intraconjunctivally infected monkeys, fails to support the possible role of intraconjunctival infection or
casual contact or exposure in the human infections.
In the infected monkeys both AHS virus types 1 and
6, singly or in combination; were found to cause a
meningo-encephalitis with or without subsequent seroconversion. The PRN titres in the monkeys' sera varied
(Table IT) but in most cases were higher than those
demonstrated in the patients' sera." Monkey C20/85's
titre of 10 240 to AHS virus type 1 was also higher than
commonly observed in horses following natural infection with wild type virus 2 or following vaccination.
Either AHS virus type 1 or type 6 could therefore have
been responsible for the human infections. However, in
addition to the 4 patients who developed encephalitis
with subsequent seroconversion to AHS virus types 1
and 6, 10 other workers from the same vaccine-packaging facility, with no histories of clinical infection, were
shown to have antibodies to vaccine strains of AHS
virus types other than 1 and 6."'" The possibility of
another AHS virus type playing a role in the human
infections can therefore not be excluded, considering
that the monkeys with the most severe encephalitis
(C31189 and C44189) failed to seroconvert or to develop

ocular lesions, and serological profiles of workers in the

same facility demonstrated exposure to other AHS virus
types.
The lack of a demonstrable viraemia in the infectf j
monkeys is consistent with the findings in the humcn
infections" and correlates with results from equivale.:t
experiments in guinea-pigs, where a viraemia was on '{
detected in the terminal stages of the infection" TI ~
infection therefore appears to be contained within tl :
CNS and our inability to propagate virus or detect vir i
antigens in autopsy CNS specimens can be ascribed )
the late stage in the infection cycle at which the. animc ,
were sacrificed. Acute phase biopsies could possib .
have been more productive.
A true understanding of the pathogeriesis of an infe tion requires the tracking of the spread of the virus du ing both the prodromal as well as clinical phases of t1 :
infection. n The results of these investigations indica :
that these neurotropic vaccine strains of AHS vin 3
invaded the CNS directly from the olfactory mucos
The histological features suggest that the viral spre~ I
was directly from the receptor-cell processes along t1 ~
olfactory tracts to the parahippocamp'al gyrus an .
orbitofrontal lobes. The mechanism of viral tissl.' .
damage, however, is unclear. The infection then appea: ,
to have remained localised, as evidenced by the lack of
viraemia, and to have been short-lived, as no virus coul
be detected in or propagated from CNS specimens frO!
monkey C31/89 13 days after superinfection. Th
pathogenesis of the eye lesions observed in the huma
patients is, however, still an enigma.
The experimental findings, together with the result,
from the 4 patients"'" clearly indicate that neurotropi .
vaccine strains of AHS virus were most probably th
aetiological agents in the human infections. The result,
again higWight the need for precautionary measure,
when working with all supposedly 'safe' or laboratoI)adapted viral strains. Whether the wild-type AH..
viruses could infect man via the transnasal route i i
unknown and caution must be exercised before extra
polating results from accidental or experimental infections using neuro-adapted vaccine strains to naturai
arthropod-borne infections with the wild-type AH~;
virus.
We wish to thank Dr G. Beverley, Dr J. Picard, Mr J
Makwena and staff of the H. A. Grove Research Centre fo'
their interest and for providing facilities and assistance i;
these studies; Dr M. M. Lippert for performing the lumba .
punctures; the Departments of Microbiology and Chemicc
Pathology of the University of Pretoria for the CS]
analyses; Dr D. H. du Plessis for the rabbit anti-AHS"
sera; and Mr L. Pieterse for expert technical assistance
This work was supported by a grant from the South AfriC3J
Poliomyelitis Research Foundation.
REFERENCES
L Blood DC, Radostits OM, Henderson JA. Vererinary Medicine: ,
Texlbook of the Diseases of Cacc/e, Sheep, Pigs, GoalS and Horses. 60
ed. London: BaIliere Tindall, 1983: 719-722.
2. Gonoan BM, Taylor J. Orbiviruses. In: Fields B_ T, Knipe D..
Chanoe!< RM, el al., eds. Virology. New York: Raven Press, 198:
907-925.
3. Mellor PS, Boonoan J, Jennings M. The multiplication of Africa!
horsesickness virus in two species of Culicoides (Diptera
Ceratopogonidae). Arch Virol1975; 47: 351-356.
4. EraSIOUS BJ- Bluetongue and African horsesickness viruses. Pape
presented at the International Symposium on Reoviridae, Guelpb
16 - 17 May 1977.
5. Sellers RP, Pedgley DE, Tucker MR. Possible spread of Africa'
horsesickness on the wind. J Hyg (Lond) 1977; 79: 279-298.
6. Van Rensburg rn, De Klerk J, Groenewald HB, Botha WS. An out
break of African horsesickness in dogs. J S Afr Vel Assoc 1981; 52
323-325.
7. Alexander RA. PreliminaIy note on the infection of white mice an!
guinea-pigs with the virus of horsesickness. J S Afr Vel Med Asso
1933; 4: 1-9.
8. Howell PG. The isolation and identification of funher antigeni
types of African horsesickness virus. Onderscepoorc J Vel Res 1962
29: 139-149.

....VOl81 2 MAY 1992

467

SAMJ

----------------------------------------------------------9. Erasmus BJ. Preliminary observations on me value of me guineapig in derermining me innocuiry and antigeniciry of neurotropic
anenuated horsesickness strains. Ondersrepoorr] Vee Res 1963; 30:
11-22.
10. Erasmus
The anenuation of horsesickness virus: problems and
advantages associated with the use of different host systems. In:
Bryans IT, Gerber H, eds. Proceedings of ehe Ise Internaeional
Conference on Equine Infeccious Diseases (Stresa, Italy, 11-13 Ju!
1966). Lexington, Kentucky: Grayson Foundation, 1966: 208-213.
11. Van der Meyden CH, Erasmus BJ, Swanepoel R, Prozesky OW.
Encephalitis and chorioretinitis associated wim neurotropic African
horsesickness virus infection in laboratory workers: Pan 1. Clinical
and neurological observations. S Afr Med] 1992; 81: 451-454 (this
issue).
12. Reid R, Van der Meyden CH, Erasmus BJ, Meyer H, Hamilton
AMP. Encephalitis and chorioretinitis associated wim neurotropic
African horsesickness virus infection in laboratory workers: Pan IT.
Ophmalmological findings. S Afr Med] 1992; 81: 454-458 (this
issue):
13. Swanepoel R, Erasmus BJ, Williams R, Taylor ME. Encephalitis and
chorioretinitis associated wim neurotropic African horsesickness
virus infection in laboratory workers: Part ill. Virological and serological investigations. S Afr Med] 1992; 81: 458-461 (this issue).
14. Hanson RP, Sulkin SE, Buescher EL, Hammon WMcD,
McKinney RW, Work TH. Arbovirus infections in laboratory
workers. Science 1967; 158: 1283-1286.
15. Flewen TH. Safery in me virology laboratory. 1n: Waterson AP, ed.
Recene Advances in Clinical Virology. New York: Churchill
Livingstone, 1980: 169-187.

Br

16. Pike RM. Laboratory-associated infections: summary and analysis

00921 cases. Healeh Lab Sci 1976; 13: 105-114.
17. Pike RM, Richardson JR. Prevention of laboratory infections. 1n:
Lennene EH, Schmidt NJ, eds. Diagnoseic Procedures for Viral,
RikeccsiaJ and Chlamydial Infecriuns. 5m ed. Washington: American
Public Health Association, 1979: 49-63.
18. Deinhardt F. Hepatitis in primates. Adv Virus Res 1976; 20:
113-157.
19. Tabor E, Purcell RH, Gerery RJ. Primate animal models and
titered inocula for me study of human hepatitis A, hepatitis B, and
non-A, non-B hepatitis.] Med Primaco11983; 12: 305-318.
20. Soike KF, Rangan SRS, Gerone PJ. Viral disease models in primates. Adv Vee Sci Comp Med 1984; 28: 151-199.
21. Kluver H, Banera E. A method for the combined staining of cells
and fibres in the nervous system. ] Neuropachol Exp Neurol 1953;
12: 400-403.
22. White DO, Fenner FJ. Medical Virology. 3rd ed. New York:
Academic Press, 1986: 614-619.
23. Johnson RT. Viral Illfeccions of ehe Nervous Syseem. New York:
Raven Press, 1982: 37-60.
24. Shope RE. Viral diseases of the central nervous system. In:
Romschild H, Cohen JC, eds. Virology ill Medicine. New York:
Oxford Universiry Press, 1986: 133-151.
25. Winkler WG, Fasmnell TR, Leffingwell 1.., Howard P, Conomy JP.
Airborne rabies transmission in a laboratory worker.]AMA 1973;
226: 1219-1221.
26. Blackburn NK, Swanepoel R Observations on antibody levels associated with active and passive irnmuniry to African horsesickness.
TropAnim Healeh Prod 1988; 20: 203-210.

Effects of ganuna-linolenic acid on mitosis and nuclear

morphology in osteogenic sarcoma cells
M. DE KOCK,

;;ummary

J. C. SEEGERS,

H. J. ELS

In this study it is shown that gaIIlIIla-linolenic acid

(GLA) at concentrations of 10, 20 and 50 J.J.g/ml has
a dose-responsive inhibitory effect on mitosis in
osteogenic sarCOIIla cells after exposure for 24, 48
and 72 hours, respectively. GLA also has marked
effects on the morphology of the nucleus and
nucleolus of these cells. Decreased silver staining
of nucleolar phosphoproteins was also evident in
GLA-supplemented cells.
S Atr Med J1992; 81: 467-472.

OrrObin ' postulated that a deficiency of gammalinolenic acid (GLA) in malignant cells and that
consequently inadequate amounts of prostaglandin El (pGE,) and/or thromboxane A., (TXA.,) may
be pivotal to the metabolic abnormalities of cancer cells.
This author suggests that levels 'Of TXA., could be elevated by directing arachidonate metabolism away from
other eicosanoids of the 2-series and that PGE, could be
increased by supplementation by its immediate precursors.
The abnormality in question, common to all malignant cells according to Horrobin, I lies in the desaruration of linoleic acid. This author suggested that a key
event in malignant disease might be the loss of delta-6desarurase activity by malignant cells. He pointed out
that loss of this enzyme would lead to reduced levels of

)epartInent of Physiology and Electron Microscopy Unit, University

If South Africa, Pretoria
M. DE KOCK, M.Se.
H. J. ELS, M.Se.
)epartInent of Physiology, University of Pretoria
J. C. SEEGERS, DSC.
Accepted 26 lune 1991.

dihomo-GLA (DGLA), PGE" and cyclic adenosine

monophosphate (cAMP) and could be a key determinant of malignant change. If so, provision of GLA
to malignant cells in order to bypass the blocked delta6-desaturase enzyme should raise levels of DGLA,
PGE, and cAMP in such cells and should theoretically
induce reverse transformation.
Research into the possible anti-cancer effects of GLA
has recently been conducted by adding GLA at low
levels (10 and 20 ~ml) to normal cell lines, mouse
melanoma cells, human hepatoma cells and human
oesophageal carcinoma cells. 2-4 Dippenaar et at. 2 demonstrated that the fany acids (FAs) affected benign and
malignant cells differently. Mouse melanoma cells were
exposed to a range of doses of GLA. This resulted in a
statistically highly significant reduction of up to 70% in
the growth rate of these malignant cells within 7 days.2
Human osteogenic sarcoma (MG63) cells subjected to
the same treaunent resulted in a reduction of up to 76%
in their growth rate within 7 days.' Equivalent doses of
GLA in both studies had no effect on benign cells. The
addition of GLA to continuous cultures of human oesophageal carcinoma cells caused marked morphological
changes, culminating in cell death at high dosage levels. J
These changes appeared to be time- and dose-dependent, being more pronounced at GLA concentrations of
40 ~ml and 60 ~ml than at 10 ~ml or 20 ~ml.
This effect was influenced to some extent by the cell line
used.
The purpose of this study was to investigate this
effect on dividing cells at microscopic level. MG63 cells
in culture were exposed to various concentrations of
GLA for 24-, 48- and 72-hour periods. Cell proliferation was monitored by the determination of mitotic
indices in haematoxylin and eosin (H and E)-stained
cells.
Morphological effects seen in the nucleoli of H and
E-stained MG63 cells prompted us to investigate the
possible effects of GLA on the localisation of phospho-