VETERINARSKI ARHIV 68 (4), 135-142, 1998

ISSN 0372-5480
Printed in Croatia

Levamisole-caused association between neutrophil and eosinophil granulocytes in dairy cows after parturition

Igor Vojtic*

Biotechnology Research Centre, Maribor Veterinary Hospital, Maribor, Slovenia

* Contact address:
Doc. Dr. Igor Vojtic,
Biotechnology Research Centre, Maribor Veterinary Hospital, SL-2000 Maribor, Šentiljska c. 109, Slovenia,
Phone: 386 62 224 346, Fax: 386 62 224 347

VOJTIC, I.: Levamisole-caused association between neutrophil and eosinophil granulocytes in dairy cows after parturition. Vet. arhiv 68, 135-142, 1998.


Sixteen Holstein-Friesian cows with retained foetal membranes or purulent uterine discharge were used to determine relationships among different leukocyte types from peripheral blood. Eight animals received 10 mg of oestradiol-dipropionate and antibiotics, while eight animals were additionally challenged with 750 mg of levamisole by intra-muscular route on the second, fourth and sixth day after calving. During levamisole challenge, on day 12 after parturition a significant negative correlation between neutrophil and eosinophil leukocytes appeared (r=-0.712, P<0.01), similar to this was noted on neutrophil and lymphocyte cells (r=-0.926, P<0.001) in the same group of cows, and non-challenged cows, respectively (r=-0.964, P<0.001). Neither a positive nor a negative association between neutrophil and eosinophil granulocytes was detected in non-challenged cows (r=0.295, P>0.05). The different expressions of such a relationship suggest that a levamisole enhanced cell immune reaction by mediator(s) through the same cell-surface receptors appeared in lymphocyte and eosinophil granulocytes.

Key words: levamisole, cow, parturition, leukocytes


In recent years a number of substances have been identified that offer a potential as immunomodulatory drugs. One of these, tetrahydro-6-phenylimidazo-thiazole (levamisole), was initially a synthetic broad-spectrum antihelmintic introduced in veterinary practice. Since 1971, tetramisole (a steric-isomer of levamisole-hydrochloride) has been recognised as an immunostimulating drug, which augmented the protective effect of a Brucella vaccine in mice (Symoens and Rosenthal, 1977).

Several studies have provided evidence supporting the hypothesis of the immunotherapeutic potential of levamisole (Hank and Schiller, 1990; Kimball et al., 1991; Zomborszky et al., 1993; Nicak et al., 1994; Stogaus and King, 1995). In general, it is accepted that levamisole increased phagocytosis by polymorphonuclear leukocytes and macrophages in cases of a hypofunctional state of these cells. Lymphokynes, complement, acute phase proteins and leukocyte interferon production could also be enhanced by levamisole. In contrast, levamisole had little or no effect on serum globulin levels and absolute lymphocyte counts. It would seem that responsiveness to levamisole depends on the immune status of the animal, concentration of antigen, dosage of levamisole and on the timing and duration of challenge (Symoens and Rosenthal, 1977).

One difficulty with previous studies that attempted to establish the effect of levamisole on leukocyte count was the lack of evaluating the interrelationship among particularly leukocyte cell types. Whether such effects of levamisole can be observed in postpartum cows is not documented. This, combined with the complex nature of the immune response, can contribute to a better understanding of how levamisole is involved in cell mediated immune reaction. The present study was designed to determine statistical relationship between different leukocyte cells from venous blood in levamisole-challenged cows during the pathological postpartum period.

Materials and methods

Sixteen Holstein-Friesian cows served as experimental animals. These cows were selected from a commercial dairy herd on the basis of their retained foetal membranes (not expelled within 24 hours after parturition) and/or purulent uterine discharge observed up to the fourth day after calving. Cows were placed in individual calving pens ten days before expected delivery (day 0 of experiment). Initially, they were offered a basal diet (2:1 ratio of mixed corn and grass silage) ad libitum (estimated consumption up to 30 kg) supplemented with soya expeller (1 kg) and a vitamin-mineral mix. After parturition, concentrates containing 18% of crude proteins were given up to an amount of 7.5 kg twice daily.

Animals were randomly assigned to one of the treatments: Eight animals received 10 mg of oestradiol-dipropionate in corn oil via intramuscular injection, and intrauterine antibiotic tablets (500 mg of gentamicine) on the second, third and fourth day after parturition. Details of this scheme were referred to previously (Vojtic, 1985). Eight cows subjected to the same treatment were additionally challenged with 750 mg of levamisole (Nilverm, Pliva, Zagreb, Croatia) regardless of their body mass by intramuscular route on the second, fourth and sixth day of experiment. On day 12 from the beginning of experiment, postprandial blood samples were taken from the jugular vein into the Vacuette system (1.8 mg K2-EDTA/ml of blood and gel-separation tubes, Greiner, Germany). Blood smears were prepared on slides and were stained using the Pappenheim method (May-Gruenwald-Giemsa solution in phosphate buffer pH=6.8). The blood smear from each animal was evaluated according to the different types of leukocytes, using a light microscope (Laboval 3, Zeiss, Germany).

The relation between two variates was evaluated using sample correlation coefficient (r). To predict the values of a dependent variate a simple linear regression procedure (y=a+bx) and appropriate coefficients were used. In this calculation, all leukocyte types were treated as fixed items. For some variates a multiple regression equations (y=a+b1x1+b2x2+b3x3+b4x4) were calculated. In this case, a special technique of forward stepwise regression was used (Glantz and Slinker, 1990). Briefly, the fixed variates are entered into the equation simultaneously and the F-value of each is checked with a computing algorithm (Fox et al., 1995). All variates with small F-values below a specified F are removed from a model. This process is repeated until the addition of variables does not significantly improve the prediction of the dependent variate. The goodness-of-fit for testing regression is based on analysis of variance. The null hypothesis was that both slopes (b1-4) and intercept (a) of the population were zero.

Results and discussion

The composition of white blood cells in cattle depends on several factors. The results of correlation analysis described in Table 1 showed some well known and, conversely, unexpected, relationships concerning total leukocyte count, and particularly leukocyte type, in peripheral blood of dairy cows with retained foetal membranes. This agrees with findings provided by Graen (1985), who reported similarly related variates. A very strong correlation was found between lymphocyte and neutrophil granulocytes in both levamisole-challenged and non-challenged animals. Therefore, the mechanisms co-ordinating these two major leukocyte sub-populations in peripheral blood may be independent of the immunomodulating action of levamisole. In contrast, this effect was more marked for neutrophil and eosinophil leukocytes, resulting in a difference between these two coefficients, with regard to challenged and non-challenged animals, as shown in Table 1 (-0.712** vs. 0.295). It is also clearly apparent that the population correlation coefficient to be expected in levamisole treated animals is significantly different from zero (P<0.05).

Table 1. Correlation matrix for total leukocyte count and single leukocyte type for levamisol challenged group (below diagonal) and control group of cows (above diagonal). Signifficant negative correlation between eosinophil and neutrophil leukocytes appeared after challenge with 750 mg of levamisole.


Total leukocytes





Total leukocytes






























**P<0.01; ***P<0.0001

Additionally, this observation is quantitatively confirmed in Table 2. It is interesting to note that such a large residual, with six degrees of freedom in non-challenged animals, leads to a poorly fitting (P=0.478) sample regression line. Neither positive nor negative significant correlations were detected among other leukocyte types.

The question that remains is how levamisole treatment influences peripheral neutrophil-eosinophil interaction. It has been recognised that neutrophil leukocytes comprise the major cellular element in lochia of cows with impaired puerperium (Mendonca et al., 1989; Kaneko et al., 1997). Nevertheless, lochial phagocytic activity of neutrophil granulocytes and expression of MHC-class-I molecules in these cells was down-modulated, compared with the autologous blood sub-population (Zerbe et al., 1996). Additionally, a recent study by Cobb and Watson (1995) characterises the distribution of macrophages and lymphocytes throughout normal endometrium. The findings described in this paper suggest that after an antigen exposure, B lymphocytes appeared in endometrial sections. They, as well as Boos et al. (1996) and Subandrio and Noakes (1997) noted that this feature is widely influenced by the stage of the oestrous cycle. These, however, were estimates made from an immunohistological aspect. Whether such interrelationships can be observed in peripheral blood of cows with impaired puerperium is not documented. Also unknown is the importance of the eosinophilic sub-population in both intra-epithelial and peripheral events after foetal membranes were retained. Concerning this, Zerbe et al. (1996) speculated that some mediators are involved (possibly cytokines).

Table 2. Algebraic form of analysis of variance for testing simple regression goodnes-of-fit between neutrophil (y) and eosinophil (x) leukocytes in levamisol challenged (experimental, r2=0.506) and intact cows (control, r2=0.087). Significant slope (-2.026, P=0.048) was detected if the source of variation was levamisole challenge.

Source of variation


Sum of squares

Mean square

F value

P value

Levamisole challenge


















No challenge


















DF=degrees of freedom

The number of studies on levamisole and cytokine family mediators has been reviewed in the past few years. The results of peroral levamisole treatment in mice show an enhanced production of interleukin-1 (IL-1) in isolated peritoneal macrophages (Kimball et al., 1991). Also, levamisole has little effect on neutrophil recovery in fluorouracil-caused neutropenia in murine species (Kimball, 1993). Furthermore, besides augmented IL-1 production, levamisole decreases TNF and IL-6 biosynthesis during lipopolysaccharide challenge in vitro (Kimball et al., 1992). Collectively, the various studies reveal, first, that the immunostimulating action of levamisole is IL-1 mediated and, secondly, the ability of levamisole to normalise is either a depressed or exuberant immune response.

Concerning results shown in Table 1, during levamisole challenge a significant negative correlation between neutrophil and eosinophil leukocytes appeared (-0.712, P<0.01), similar to that noted on lymphocyte cells (-0.926, P<0.001). The rise in the neutrophil count leads to a drop in the eosinophil number, due to their migration to the previous site of placenta attachment. A similar effect of eosinophil tissue accumulation has previously been seen in IL-5-stimulated mice (Mould et al., 1997). This suggests that eosinophil leukocytes may be involved in cell mediated immune reaction, although their phagocytic activity is not extensive (Elsner et al., 1996; Subandrio and Noakes, 1997). Until now, one chemokine (eotaxin, Ponath et al. 1996) and two cytokines (IL-5 and IL-8, Baggiolini et Dahinden, 1994) have been recognised as regulators of blood eosinophilia during allergic inflammation. Recently, it was elucidated that human eotaxin receptors with adjacent cysteines (CCR3, HaElens et al., 1996) are also expressed by T-helper lymphocytes (Sallusto et al., 1997). This fact supports our findings, which exhibit a parallel between eosinophil and lymphocyte cells. This action was obviously levamisole enhanced. Concerning this, further research is needed to establish more definitively the exact patterns of levamisole-eosinophil interaction.

Table 3. Prediction of total leukocyte count using multiple regression model for single leukocyte type as a fixed items. Analysis of variance shows a stright line as a poor fit for both, levamisole challenged group (F=0.223, R=0.479, P=0.909) and no-challenged group of cows (F=0.588, R=0.663, P=0.696).




T value

P value

Levamisole challenge

Intercept (a)






b1 neutrophils





b2 lymphocytes





b3 monocytes





b4 eosinophils





No challenge

Intercept (a)






b1 neutrophils





b2 lymphocytes





b3 monocytes





b4 eosinophils





SE=standard error

Neither in the experimental nor in the control group of cows is multiple linear regression able to provide an acceptable prediction of the total leukocyte count (Table 3). These findings can be explained by the fact that sample-based multi-co-linearity (not seen in Table 1) was present in both groups of animals. The number of studies on leukocyte sub-populations in cattle have shown that the composition of these depends on several factors (Ayoub and Yang, 1997; Kruger, 1996), but our observation (Table 3) confirms equilibrium among particular leukocyte types as a basic homeostatic mechanism during pathological and immunomodulated puerperium, respectively.

The results of forward stepwise regression analysis are in quantitative agreement with our data referred to above. Adding the lymphocyte, monocyte and eosinophil variates to the regression equation for both groups of animals, F-to-enter values (0.0278, 0.0072 and 0.0528 for experimental and 0.0555, 1.051 and 0.184 for control group of cows) were below the F-cutoff specified with algorithm-proposed step options (4.0, P=0.086 and 3.9, P=0.089 for the experimental and control groups of cows, respectively). This suggests that these variables did not significantly improve the ability of the equation to predict the total leukocyte count in the final equation.


In summary, using levamisole as an immunomodulatory drug during pathological puerperium, significant interactions in peripheral blood on day 12 postpartum were detected between neutrophil vs. lymphocyte and eosinophil leukocytes. Neither a positive nor a negative association between eosinophil and neutrophil granulocytes was detected in non-challenged cows. The different expression of such a relationship suggests that a levamisole enhanced cell immune reaction by mediator(s) through the same cell-surface receptors appeared in lymphocyte and eosinophil granulocytes.

This study was supported by "Pliva" Pharmaceutical Industries Inc., Zagreb, Croatia, Representative Office Ljubljana, Slovenia.


Ayoub, I. A., T. J. Yang (1997): Age-dependent changes in peripheral blood lymphocyte subpopulations in cattle: a longitudinal study. Devl. Comp. Immunol. 20, 353-363.

Baggiolini, M., C. A. Dahinden (1994): CC chemokines in allergic imflamation. Immunol. Today 15, 127-133.

Boos, A., W. Meyer, R. Schwarz, E. Grunert (1996): Immunohistochemical assessment of oestrogen receptor and progesterone receptor distribution in biopsy samples of the bovine endometrium collected throughout the oestrous cycle. Anim. Reprod. Sci. 44, 11-21.

Cobb, S. P., E. D. Watson (1995): Immunohistochemical study of immune cells in the bovine endometrium at different stages of the oestrous cycle. Res. Vet. Sci. 59, 238-241.

Elsner, J., R. Hoechstetter, D. Kimming, A. Kapp (1996): Human eotaxin represents a potent activator of the respiratory burst of human eosinophils. Eur. J. Immnol. 26, 1919-1925.

Fox, E., K. Shotton, C. Ulrich (1995): SigmaStat User's Guide. Jandel. San Rafael CA.

Glantz, S. A., B. K. Slinker (1990): Primer of applied regression and analysis of variance. McGraw-Hill. New York.

Graen, R. (1985): Untersuchung über der Einfluss von Leukozytenzahl und Leukozytenaktivität im Blut auf den Nachgeburtsabgang beim Rind. Vet. med. Diss. Tierartzliche Hohschule. Hannover.

Haelens, A., A. Wuyts, P. Proost, S. Struyf, G. Opdenakker, J. Vandamme (1996): Leukocyte migration and activation by murine chemokines. Immunobiol. 195, 499-521.

Hank, J. A., J. H. Schiller (1990): Immunomodulatory effects of levamisole. Vet. Cancer Soc. Newslett. 14, 12-13.

Kaneko, K., S. Kawakami, M. Miyoshi, T. Abukawa, S. Yamanaka, M. Mochizuki, S. Yoshihara (1997): Effects of retained placenta on subsequent bacteriological and cytological intrauterine environment and reproduction in Holstein dairy cows. Theriogen. 48, 617-624.

Kimball, E. S., M. C. Clark, C. R. Schneider, F. J. Perisco (1991): Enhancement of in vitro lipopolysaccharide-stimulated interleukin-1 production by levamisole. Clin. Immunol. Immunopathol. 58, 385-398.

Kimball, E. S., C. R. Schneider, M. C. Fisher, M. C. Clark (1992): Levamisole causes cytokine expression by elicited mouse peritoneal macrophages. J. Leukoc. Biol. 52, 349-356.

Kimball, E. S. (1993): Experimental modulation of IL-1 production and cell surface molecule expression by levamisole. Ann. N.Y. Acad. Sci. 685, 259-268.

Kruger, T. E. (1996): Immunomodulation of peripheral lymphocytes by hormones of the hypothalamus-thyroid axis (review). Advanc. Neuroimmnol. 6, 387-395.

Mendonca, M., J. Schilz, K. Eulenberger, K. Beck (1989): Phagozytose und pH-Wert in den Lochien von Rindern mit gestörtem Puerperverlauf. Mh. Vet. Med. 44, 346-349.

Mould, A. W., K. I. Matthaei, I. G. Young, P. S. Foster (1997): Relationship between interleukin-5 and eotaxin in regulating blood and tissue eosinophilia in mice. J. Clin. Invest. 99, 1064-1071.

Nicak, A., L. Mirossay, J. MojZiš, A. Kohut (1994): Influence of long term administration of levamisole and muramyl-dipeptid on fagocyting activity of polymorphonuclear leukocytes in dependence on ageing in rats. Bratisl. Lek Listy 95, 21-23.

Ponath, P. D., S. Qin, D. J. Ringler, I. Clark-Lewis, J. Wang, N. Kassam, H. Smith, X. Shi, J. A. Gonzalo, W. Newman, J. C. Gutierrez-Ramos, C. R. Mackay (1996): Cloning of the human eosinophil chemoattractan, eotaxin. J. Clin. Invest. 97, 604-612.

Sallusto, F., C. R. Mackay, A. Lanzavecchia (1997): Selective expression of the eotaxin receptor CCR3 by human T helper 2 cells. Science 277, 2005-2007.

Stogaus, R., M. G. King (1995): Is oral levamisole immunostimulation in rats mediated by reduced levels free plasma corticosterone? Int. J. Immunopharmacol. 17, 635-640.

Subandrio, A. L., D. E. Noakes (1997): Neutrophil migration into the uterine lumen of the cow - the influence of the endogenous and exogenous sex steroid hormones using two intrauterine chemoattractants. Theriogen. 47, 825-835.

Symoens, J., M. Rosenthal (1977): Levamisole in the modulation of the immune response: the current experimental and clinical state. J. Reticuloendothel. Soc. 21, 175-221.

Vojtic, I. (1985): Optimiranje servis perioda sustavnom kontrolom krava u puerperiju. Stocarstvo 39, 343-348.

Zerbe, H., H. J. Schuberth, M. Hoedemaker, E. Grunert, W. Leibold (1996): A new model system for endometritis: basic concepts and characterisation of phenotypic and functional properties of bovine uterine neutrophils. Theriogen. 46, 1339-1356.

Zomborszky, Z., E. Horn, S. Tuboly, Z. Megyeri, P. Tilly, Cs. Szabo (1993): T-cell deficiency in suffolk lambs. Acta Vet. Hung. 41, 427-436.

Recceived: 9 January 1998
Accepted: 23 June 1998

VOJTIC, I.: Odnos neutrofilnih i eozinofilnih granulocita pod utjecajem levamisola u mlijecnih krava u puerperiju. Vet. arhiv 68, 135-142, 1998.


Istrazili smo povezanost izmedu pojedinih vrsta leukocita u perifernoj krvi tijekom patoloskog puerperija uzrokovanog zaostajanjem posteljice ili gnojnim maternicnim iscjetkom u sesnaest krava muzara crno-sare pasmine. Osam krava smo lijecili injekcijama 10 mg estradiol dipropionata i antibioticima dok smo preostalih osam lijecili usto sa 750 mg levamisola drugi, cetvrti i sesti dan po teljenju. Na dvanaesti dan po porodu smo utvrdili znakovitu i negativnu korelaciju izmedu neutrofilnih i eozinofilnih leukocita (r=-0,712, P<0,01) i to slicnu onoj izmedu neutrofilnih leukocita i limfocita (r=-0,926, P<0,001) u istih zivotinja i zivotinja koje nismo lijecili levamisolom (r=-0,964, P<0,001). Istovremeno nije bilo povezanosti izmedu neutrofilnih i eozinofilnih leukocita u krava, koje nisu bile lijecene levamisolom (r=0,295, P<0,05). Rezultati istrazivanja pokazuju, da su razlike u povezanosti izmedu pojedinih vrsta leukocita uvjetovane djelovanjem levamisola. Pretpostavljamo, da je djelovanje levamisola vezano na zajednicke stanicne receptore eozinofilnih leukocita i limfocita.

Kljucne rijeci: levamisol, krava, puerperij, leukociti