- Email: [email protected]
In vivo depletion of T-cells and cytokines during primary exposure of sheep to parasites
Veterinary
Immunology
ELSEVIER
and Immunopathology
Veterinary immunology and immunopathology
54 (1996) 83-90
In vivo depletion of T-cells and cytokines during primary exposure of sheep to parasites S.J. McClure a,*, R.L. Davey a, D.L. Emery a, I.G. Colditz b, J.B. Lloyd ’ a
CSIRODivision
ofAnimu1
Health.
McMuster
Luborutory.
LockedBag No 1I Blucktown.NSW2148.
Australia b Pustorul ’ Depurtment
Research Laboratory.
r?f’Veterinury
Pathology.
Armidale. University
NSW 2350, Australia
oj’ Sydney, NSW 2006, Auutraliu
Abstract This study examined the role of CD8 + and WC1 + T-cells and of interferon (IFN)-y in the development of protective immunity against infection with the enteric nematode parasite Trichostrongylus colubri’urmis in sheep. Monoclonal antibodies (mAb) were administered during
induction of the immune response to deplete or neutralise these components. Protection against the primary and challenge infections was assessed by faecal egg count and total worm count. Prolonged administration of mAb recognising IFN-y and CD8 resulted in significantly increased protection during the 6 week primary infection and following challenge. CD8 + cells were depleted from blood but not from intestinal mucosa. After injection of mAb (CCl.5) recognising the surface antigen WC 1, WC 1 + and TcryG + cells were depleted from blood but not markedly from enteric mucosa, and protection against challenge, although variable, was increased by up to 88%. It appears that CD8 + and WC1 + /y6 + cells and EN--y all retard the potential development of naturally acquired immunity against the parasite. Keyword.~:
Sheep;
Parasite; Depletion; T-cells; Interferon
1. Introduction Trichostrongylus (T.) colubriformis is a browsing nematode parasite of sheep which spends its host-dwelling lifetime in the lumen and sub-epithelium of the proximal jejunum. It causes major production loss, and alternative control methods to anthelmintic
* Corresponding 0165.2427/96/$15.00 P/I
author. Tel: 02 8402964; Copyright
SOl65-2427(96)05694-2
0
fax: 02 8402939.
1996 Elsevier
Science B.V.
All rights reserved.
chemicals are being sought. One of these is vaccination. This requires the ability to predict the desirable responses to be induced by a vaccine. and therefore an understanding of which immune responses are protective in natural infection. The intestinal response to natural infection with T. cw/z~br+mnis is very complex and involves cellular. humoral and inflammatory components. These include increased numbers of T-cells (CD4 + , CD8 + . WC1 + and TcryG + ) in enteric lamina propria and epithelium and in efferent mesenteric lymph. and an increased number of interferon-y (IFN-y) + cells in lamina propria (Gorrell et al.. 198X: McClure et al.. 1992. McClure. unpublished). For the purposes of vaccination, it is important to know which of these responses is essential and/or sufficient for protection. In order to examine the role of CD8 + and WCI + T-cells and of IFN-y in the development of protective immunity. rnonoclonal antibodies (mAb) were administered during the induction of the immune response to deplete or neutralize these components. Protection against first and challenge infections with the parasite was assessed by faecal egg count and finally by total worm count.
2. Materials
and methods
2. I. Productiotl utld pur(ficutiotr mAbs were purified from cell culture aupernates on Protein G (Pharmacia) columns and filtered (0.2 pm filter). The concentration of mAb was estimated from the protein content (Bio-Rad Protein Assay) and the purity of the preparation as assessed from PAGE gradient gels (Plate 1).
) \hw.lng the mAh Cc‘/T xhnmistrrcd I nwlccuh wcl~ht marker\. Tracl, 7. culture
Plate I. PAGE gradient grl (J-7O“i bupematc on Protein G. Trad Track + unbound supmate.
to sheep, purliicd t’rom culture hupcmate; Tracl, 3- bound mAh;
SJ. McClure
Plate 2. Tcrgd+
Ed 01. /Veterinary
ceils in the jejunum
challenge with T. coluhrijwmrc
Immunology
owl Immunoputhology
of (a) a non-immune
(immunoperoxidase
WC I (Ccl 5, (a) and (b)), and Tcrgd (86-D.
8.5
sheep and (b) an immune sheep, 7 days after
stamin g with mAb 86-D).
(c) and cd)).
54 (1996) 83-90
Immunoperoxidase
slaining for
Table I Experimental Group
design ,I
mAb injected
Specificity
Isotype
Reference
Saline CC56 17D 197 cc15 7c2 IFN-9
Bovine B cells CD4 WCI WC1 CD8 IFN-y
lgG2a I@ I IgG2b IgG2a IgG2a IgG I
Howard, I99 I Mackay et al., 1988 McClure et al.. 1989 Howard et al., 1989 Mackay, CR. (unpublished) Rothel et al., 1990
I
6
II
6
111
2
IV v VI Vll
2 6 6 (1
2.2.
Esprrimental
design
Thirty-four
one-year-old merinos, raised worm-free, were infected with 30000 T. larvae (TcL3, McMaster strain). After 6 weeks they were drenched with Levamisole, and one week later challenged with 30000 TcL3. For the 6 weeks of primary infection. sheep were given 4 mg mAb in 10 ml saline three times weekly by i.v. injection (Table I). Faecal egg counts were performed weekly, and the sheep killed 6 weeks after challenge for total worm counts. Jejunal biopsies taken 3 weeks after first infection and at necropsy were cryosectioned and stained by immunoperoxidase for T-cell phenotype (Plate 2). The in vivo efficacy of IFN-9 wa:, confirmed in a parallel sheep study in which the prefemoral efferent lymphatic duct was cannulated. IFN-9 was injected i.v. during primary and/or secondary responses to locally administered ovalbumen in Quil A or dextran sulphate. adjuvants that normally elicit IFN-7. The efferent lymph was assayed for IFN-y by ELBA (Rothel et al., 1990) colubrijormis
4
800
3 25; ’
400 0” 1
Oa, L C 3*
0 800
*
u” 2:: .-
400
1 2 0 0
8
16
f E . : tT
0
SJ. McClure er al./Veterinary
Immunobgy
Ural Immunoparhology
$4 (1996) 83-90
87
3. Results When EN-9 was administered to cannulated sheep during immunisation, IFN-9 depleted the increased levels of IFN=y normally detected in efferent lymph following immunisation (Fig. 1). Addition of 17D to cultured lymphocytes inhibited both the proliferation and the production of EN-y induced by con A and antigen, the effect being greater on antigen-induced stimulation. However, administration of 17D during primary immunisation neither depleted CD4 + cells, nor affected protection. In contrast, 7C2 depleted CD8 + cells in blood from 15% to 7% but did not affect CD8 + cell numbers in jejunum. (Fig. 2). CC 15 depleted WC 1 + and TcryS + cells in blood from 8-27% to < 1% post-first injection and while injections continued. The number of CD8 + cells in these animals was also reduced from 17% to 9%. WC1 +
y
OI 1
2
3
4
Day
control anti-CD4 anti-CD8 anti-WC1 anti-IFNy
1
2
3
4
3
4
Day
1
2 Day
Fig. 2. Phenotype of blood mononuclear cells from sheep treated with mAb to CD4. CD8 or WCI lymphocyte surface antigens. (mean + SEM, n = 9). Reproduced by kind permission of Elsevier.
CD5 CD4 CD8 negative
relative
Fig. 3. Phenotype
sheep given mAb CCI 5 (WCI. analysed by FAG.
fluorescence
of blood mononuclear Reproduced
lntenslty
(
log
ceils from (a) a sheep given mAb
cytotoxic).
The samples were collected
by kind permission
of Blackwell
1
197 (WCI,
not cyloroxic)
and (b) a
2 weeks after the last injection
and
Science.
cells in the jejunum were not ablated. but were somewhat reduced in number in the crypt region and in the villi associated with the Peyer‘s patches. Ccl.5 also reduced the staining intensity of CD4 + and CD8 + cells in blood 2 weeks after the last injection, without affecting that of new WC1 or Tcry6 cells (Fig. 3). Sheep treated with CC15 (anti-WCl) showed a 20% reduction in mean cumulative FEC during the 6 weeks of first infection, 88%, reduction in mean FEC during the challenge infection. and 44% reduction in mean worm count at kill (Fig. 4). In vitro. addition of CC15 to lymphocytes cultured with Con A or parasite antigen inhibited both cell proliferation and IFN-y production. The effect was more marked in the case of mitogen, in contrast to the effect of 17D. The other mAb recognising WC 1, 197, did not deplete cells, affect staining intensity. or affect protection. In vitro, however, its effect was similar to that of CC15. as was that of the anti-Tcry6 mAb 86D. Sheep treated with IFN-9 (anti-IFNyl showed 42% reduction in FEC during first infection (P < 0.05). 83% reduction in FEC during challenge and 52% reduction in final worm count.
SJ. McClure
et al./
0
1
Veterinary
2
3
Immunology
and Immunopathology
4
7
5
6
8
9
10
11
54 (1996) H-90
89
12 WormCount
Weeks After Primary Infection Fig. 4. Faecal egg counts and worm counts after primary and challenge infection with T. colubriformis sheep treated with mAb during primary infection (mean + SEM, n = 6). Reproduced by kind permission Blackwell Science.
of of
Sheep treated with 7C-2 (anti-CD8) showed 42% reduction in FEC during first infection (P < 0.0.9, 62% reduction in FEC during challenge and 60% reduction in final worm count. In vitro, 7C2 inhibited cell proliferation and IFN-y production in cultures stimulated with Con A but not with antigen.
4. Discussion Depletion of the blood WC 1 + /TcryG + population may have affected protection by loss of cytokine production or via interference with the DTH reaction, leaving cytokines free to induce other mechanisms. The efficacy of blood depletion implies a role for immigrating as opposed to resident yS cells. The diminished staining intensity of lymphocytes from sheep treated with anti-WC1 indicates reduced levels of surface molecules on non-targeted lymphocytes and suggests an increased proportion of immature cells. This may have been a non-specific effect of cell-lysis, or possibly a direct effect of loss of WC1 + cells if, for example, Tcr-yS + cells play a role in the generation of other T-cells. Mast cell/IgE hypersensitivity responses are assoociated with rejection of T. colubriformis (McClure et al., 1992) and such responses are favoured, at least in mice, by IL-4 (Finkelman et al., 1991). There is evidence in mice that IL-4 synthesis requires repeated antigenic stimulation (Marshall et al., 1993) and this would be consistent with the prolonged exposure period required for immunity against this parasite. Neutralisation of IFN-y may thus encourage the IL-4 bias and induce a Th2 type of response more rapidly than is achieved normally.
As IFN-9 recognises a cytokine rather than cells, and does not cause cell lysis, the increased protection resulting from administration of IFN-9 cannot be due to a nonspecific effect of cell lysis. It is still possible that decreased worm establishment may occur as a result of depletion of any of the local responses at the time of initial exposure to T. colubriformis. Further experiments, with a wider range of mAbs, are currently under way to test this possibility. If it does prove to be specific for IFN-y/TcryG/IgE, it would suggest that the hypersensitivity response is critical for protection against this parasite, and is a desirable goal for an effective vaccine. In either case, the finding that the immune response can be manipulated to reduce the time required for the development of protective immunity. has major implications for the design of enteric parasite vaccines.
Acknowledgements This work was supported by Australian woolgrowers Secretariat, and by Biotechnology Australia.
through the International
Wool
References Finkelman. F.D., Gorrell, M.D..
Pearce, E.J., Urban, J.F. and Sher. A., 1991. Parasuol. Today. 7: A62
Willis, G., Brandon, M.R. and Lascelles. A.K..
1988. Clin. Exp. Immunol., 72: 274.
Howard. C.J., Sopp, P.. Parsons, K.R. amd Finch, J., 1989. Eur. J. Immunol., Howard. C.J.. 1991. Vet. Immunol. Immunopathol.. Mackay. CR.,
Hein, W.R.. Brown, M.H. and Matzinger.
McClure. S.J., Hein. W.R.. Yamaguchl, Biol.. 67: 215. McClure,
19: 757.
27: 77. P., 1988. Eur. J. lmmunol.,
K.. Dudler. l... Beya. M.-F. and Miyasaka,
S.J.. Emery, D.L.. Wagland, B.M. and Jones, W.0..
Marshall. J.D.. Wen, Y., Abrams. J.S. and Umetsu. D.T..
19: 1477. M.. 1989. Immunol. Cell
1992. lnt. J. Parasitol.. 22: 227.
1993. Cell. Immunol..
152: 18.
Rothel. J.S.. Jones. S.L.. Comer, L.A.. Cox. J.C. and Wood, P.R.. 1990. Aust. Vet. J., 67: 134.
Immunology
ELSEVIER
and Immunopathology
Veterinary immunology and immunopathology
54 (1996) 83-90
In vivo depletion of T-cells and cytokines during primary exposure of sheep to parasites S.J. McClure a,*, R.L. Davey a, D.L. Emery a, I.G. Colditz b, J.B. Lloyd ’ a
CSIRODivision
ofAnimu1
Health.
McMuster
Luborutory.
LockedBag No 1I Blucktown.NSW2148.
Australia b Pustorul ’ Depurtment
Research Laboratory.
r?f’Veterinury
Pathology.
Armidale. University
NSW 2350, Australia
oj’ Sydney, NSW 2006, Auutraliu
Abstract This study examined the role of CD8 + and WC1 + T-cells and of interferon (IFN)-y in the development of protective immunity against infection with the enteric nematode parasite Trichostrongylus colubri’urmis in sheep. Monoclonal antibodies (mAb) were administered during
induction of the immune response to deplete or neutralise these components. Protection against the primary and challenge infections was assessed by faecal egg count and total worm count. Prolonged administration of mAb recognising IFN-y and CD8 resulted in significantly increased protection during the 6 week primary infection and following challenge. CD8 + cells were depleted from blood but not from intestinal mucosa. After injection of mAb (CCl.5) recognising the surface antigen WC 1, WC 1 + and TcryG + cells were depleted from blood but not markedly from enteric mucosa, and protection against challenge, although variable, was increased by up to 88%. It appears that CD8 + and WC1 + /y6 + cells and EN--y all retard the potential development of naturally acquired immunity against the parasite. Keyword.~:
Sheep;
Parasite; Depletion; T-cells; Interferon
1. Introduction Trichostrongylus (T.) colubriformis is a browsing nematode parasite of sheep which spends its host-dwelling lifetime in the lumen and sub-epithelium of the proximal jejunum. It causes major production loss, and alternative control methods to anthelmintic
* Corresponding 0165.2427/96/$15.00 P/I
author. Tel: 02 8402964; Copyright
SOl65-2427(96)05694-2
0
fax: 02 8402939.
1996 Elsevier
Science B.V.
All rights reserved.
chemicals are being sought. One of these is vaccination. This requires the ability to predict the desirable responses to be induced by a vaccine. and therefore an understanding of which immune responses are protective in natural infection. The intestinal response to natural infection with T. cw/z~br+mnis is very complex and involves cellular. humoral and inflammatory components. These include increased numbers of T-cells (CD4 + , CD8 + . WC1 + and TcryG + ) in enteric lamina propria and epithelium and in efferent mesenteric lymph. and an increased number of interferon-y (IFN-y) + cells in lamina propria (Gorrell et al.. 198X: McClure et al.. 1992. McClure. unpublished). For the purposes of vaccination, it is important to know which of these responses is essential and/or sufficient for protection. In order to examine the role of CD8 + and WCI + T-cells and of IFN-y in the development of protective immunity. rnonoclonal antibodies (mAb) were administered during the induction of the immune response to deplete or neutralize these components. Protection against first and challenge infections with the parasite was assessed by faecal egg count and finally by total worm count.
2. Materials
and methods
2. I. Productiotl utld pur(ficutiotr mAbs were purified from cell culture aupernates on Protein G (Pharmacia) columns and filtered (0.2 pm filter). The concentration of mAb was estimated from the protein content (Bio-Rad Protein Assay) and the purity of the preparation as assessed from PAGE gradient gels (Plate 1).
) \hw.lng the mAh Cc‘/T xhnmistrrcd I nwlccuh wcl~ht marker\. Tracl, 7. culture
Plate I. PAGE gradient grl (J-7O“i bupematc on Protein G. Trad Track + unbound supmate.
to sheep, purliicd t’rom culture hupcmate; Tracl, 3- bound mAh;
SJ. McClure
Plate 2. Tcrgd+
Ed 01. /Veterinary
ceils in the jejunum
challenge with T. coluhrijwmrc
Immunology
owl Immunoputhology
of (a) a non-immune
(immunoperoxidase
WC I (Ccl 5, (a) and (b)), and Tcrgd (86-D.
8.5
sheep and (b) an immune sheep, 7 days after
stamin g with mAb 86-D).
(c) and cd)).
54 (1996) 83-90
Immunoperoxidase
slaining for
Table I Experimental Group
design ,I
mAb injected
Specificity
Isotype
Reference
Saline CC56 17D 197 cc15 7c2 IFN-9
Bovine B cells CD4 WCI WC1 CD8 IFN-y
lgG2a I@ I IgG2b IgG2a IgG2a IgG I
Howard, I99 I Mackay et al., 1988 McClure et al.. 1989 Howard et al., 1989 Mackay, CR. (unpublished) Rothel et al., 1990
I
6
II
6
111
2
IV v VI Vll
2 6 6 (1
2.2.
Esprrimental
design
Thirty-four
one-year-old merinos, raised worm-free, were infected with 30000 T. larvae (TcL3, McMaster strain). After 6 weeks they were drenched with Levamisole, and one week later challenged with 30000 TcL3. For the 6 weeks of primary infection. sheep were given 4 mg mAb in 10 ml saline three times weekly by i.v. injection (Table I). Faecal egg counts were performed weekly, and the sheep killed 6 weeks after challenge for total worm counts. Jejunal biopsies taken 3 weeks after first infection and at necropsy were cryosectioned and stained by immunoperoxidase for T-cell phenotype (Plate 2). The in vivo efficacy of IFN-9 wa:, confirmed in a parallel sheep study in which the prefemoral efferent lymphatic duct was cannulated. IFN-9 was injected i.v. during primary and/or secondary responses to locally administered ovalbumen in Quil A or dextran sulphate. adjuvants that normally elicit IFN-7. The efferent lymph was assayed for IFN-y by ELBA (Rothel et al., 1990) colubrijormis
4
800
3 25; ’
400 0” 1
Oa, L C 3*
0 800
*
u” 2:: .-
400
1 2 0 0
8
16
f E . : tT
0
SJ. McClure er al./Veterinary
Immunobgy
Ural Immunoparhology
$4 (1996) 83-90
87
3. Results When EN-9 was administered to cannulated sheep during immunisation, IFN-9 depleted the increased levels of IFN=y normally detected in efferent lymph following immunisation (Fig. 1). Addition of 17D to cultured lymphocytes inhibited both the proliferation and the production of EN-y induced by con A and antigen, the effect being greater on antigen-induced stimulation. However, administration of 17D during primary immunisation neither depleted CD4 + cells, nor affected protection. In contrast, 7C2 depleted CD8 + cells in blood from 15% to 7% but did not affect CD8 + cell numbers in jejunum. (Fig. 2). CC 15 depleted WC 1 + and TcryS + cells in blood from 8-27% to < 1% post-first injection and while injections continued. The number of CD8 + cells in these animals was also reduced from 17% to 9%. WC1 +
y
OI 1
2
3
4
Day
control anti-CD4 anti-CD8 anti-WC1 anti-IFNy
1
2
3
4
3
4
Day
1
2 Day
Fig. 2. Phenotype of blood mononuclear cells from sheep treated with mAb to CD4. CD8 or WCI lymphocyte surface antigens. (mean + SEM, n = 9). Reproduced by kind permission of Elsevier.
CD5 CD4 CD8 negative
relative
Fig. 3. Phenotype
sheep given mAb CCI 5 (WCI. analysed by FAG.
fluorescence
of blood mononuclear Reproduced
lntenslty
(
log
ceils from (a) a sheep given mAb
cytotoxic).
The samples were collected
by kind permission
of Blackwell
1
197 (WCI,
not cyloroxic)
and (b) a
2 weeks after the last injection
and
Science.
cells in the jejunum were not ablated. but were somewhat reduced in number in the crypt region and in the villi associated with the Peyer‘s patches. Ccl.5 also reduced the staining intensity of CD4 + and CD8 + cells in blood 2 weeks after the last injection, without affecting that of new WC1 or Tcry6 cells (Fig. 3). Sheep treated with CC15 (anti-WCl) showed a 20% reduction in mean cumulative FEC during the 6 weeks of first infection, 88%, reduction in mean FEC during the challenge infection. and 44% reduction in mean worm count at kill (Fig. 4). In vitro. addition of CC15 to lymphocytes cultured with Con A or parasite antigen inhibited both cell proliferation and IFN-y production. The effect was more marked in the case of mitogen, in contrast to the effect of 17D. The other mAb recognising WC 1, 197, did not deplete cells, affect staining intensity. or affect protection. In vitro, however, its effect was similar to that of CC15. as was that of the anti-Tcry6 mAb 86D. Sheep treated with IFN-9 (anti-IFNyl showed 42% reduction in FEC during first infection (P < 0.05). 83% reduction in FEC during challenge and 52% reduction in final worm count.
SJ. McClure
et al./
0
1
Veterinary
2
3
Immunology
and Immunopathology
4
7
5
6
8
9
10
11
54 (1996) H-90
89
12 WormCount
Weeks After Primary Infection Fig. 4. Faecal egg counts and worm counts after primary and challenge infection with T. colubriformis sheep treated with mAb during primary infection (mean + SEM, n = 6). Reproduced by kind permission Blackwell Science.
of of
Sheep treated with 7C-2 (anti-CD8) showed 42% reduction in FEC during first infection (P < 0.0.9, 62% reduction in FEC during challenge and 60% reduction in final worm count. In vitro, 7C2 inhibited cell proliferation and IFN-y production in cultures stimulated with Con A but not with antigen.
4. Discussion Depletion of the blood WC 1 + /TcryG + population may have affected protection by loss of cytokine production or via interference with the DTH reaction, leaving cytokines free to induce other mechanisms. The efficacy of blood depletion implies a role for immigrating as opposed to resident yS cells. The diminished staining intensity of lymphocytes from sheep treated with anti-WC1 indicates reduced levels of surface molecules on non-targeted lymphocytes and suggests an increased proportion of immature cells. This may have been a non-specific effect of cell-lysis, or possibly a direct effect of loss of WC1 + cells if, for example, Tcr-yS + cells play a role in the generation of other T-cells. Mast cell/IgE hypersensitivity responses are assoociated with rejection of T. colubriformis (McClure et al., 1992) and such responses are favoured, at least in mice, by IL-4 (Finkelman et al., 1991). There is evidence in mice that IL-4 synthesis requires repeated antigenic stimulation (Marshall et al., 1993) and this would be consistent with the prolonged exposure period required for immunity against this parasite. Neutralisation of IFN-y may thus encourage the IL-4 bias and induce a Th2 type of response more rapidly than is achieved normally.
As IFN-9 recognises a cytokine rather than cells, and does not cause cell lysis, the increased protection resulting from administration of IFN-9 cannot be due to a nonspecific effect of cell lysis. It is still possible that decreased worm establishment may occur as a result of depletion of any of the local responses at the time of initial exposure to T. colubriformis. Further experiments, with a wider range of mAbs, are currently under way to test this possibility. If it does prove to be specific for IFN-y/TcryG/IgE, it would suggest that the hypersensitivity response is critical for protection against this parasite, and is a desirable goal for an effective vaccine. In either case, the finding that the immune response can be manipulated to reduce the time required for the development of protective immunity. has major implications for the design of enteric parasite vaccines.
Acknowledgements This work was supported by Australian woolgrowers Secretariat, and by Biotechnology Australia.
through the International
Wool
References Finkelman. F.D., Gorrell, M.D..
Pearce, E.J., Urban, J.F. and Sher. A., 1991. Parasuol. Today. 7: A62
Willis, G., Brandon, M.R. and Lascelles. A.K..
1988. Clin. Exp. Immunol., 72: 274.
Howard. C.J., Sopp, P.. Parsons, K.R. amd Finch, J., 1989. Eur. J. Immunol., Howard. C.J.. 1991. Vet. Immunol. Immunopathol.. Mackay. CR.,
Hein, W.R.. Brown, M.H. and Matzinger.
McClure. S.J., Hein. W.R.. Yamaguchl, Biol.. 67: 215. McClure,
19: 757.
27: 77. P., 1988. Eur. J. lmmunol.,
K.. Dudler. l... Beya. M.-F. and Miyasaka,
S.J.. Emery, D.L.. Wagland, B.M. and Jones, W.0..
Marshall. J.D.. Wen, Y., Abrams. J.S. and Umetsu. D.T..
19: 1477. M.. 1989. Immunol. Cell
1992. lnt. J. Parasitol.. 22: 227.
1993. Cell. Immunol..
152: 18.
Rothel. J.S.. Jones. S.L.. Comer, L.A.. Cox. J.C. and Wood, P.R.. 1990. Aust. Vet. J., 67: 134.