ISSN 1580-4003 THE SCIENTIFIC JOURNAL OF THE VETERINARY FACULTY UNIVERSITY OF LJUBLJANA SLOVENIAN VETERINARY RESEARCH SLOVENSKI VETERINARSKI ZBORNIK Volume A 41 1 Slov Vet Res • Ljubljana • 2004 • Volume 41 • Number 1 • 1-60 THE SCIENTIFIC JOURNAL OF THE VETERINARY FACULTY UNIVERSITY OF LJUBLJANA SLOVENIAN VETERINARY RESEARCH SLOVENSKI VETERINARSKI ZBORNIK volume j 41 1 Slov Vet Res • Ljubljana • 2004 • Volume 41 • Number 1 • 1-60 The Scientific Journal of the Veterinary Faculty University of Ljubljana SLOVENIAN VETERINARY RESEARCH SLOVENSKI VETERINARSKI ZBORNIK Previously: RESEARCH REPORTS OF THE VETERINARY FACULTY UNIVERSITY OF LJUBLJANA Prej: ZBORNIK VETERINARSKE FAKULTETE UNIVERZA V LJUBLJANI 4 issues per year / izhaja štirikrat letno Editorial board / Uredniški odbor: Emer. Prof. Dr. Srdan V. Bavdek, Prof. Dr. Darinka Z. Doganoc, M.Sc. Gita Grecs Smole, Prof. Dr. Jože Jurca, Zdenka Karlin, Prof. Dr. Zlatko Pavlica, Prof. Dr. Uroš Pestevšek, Prof. Dr. Milan Pogačnik, Dr. Malan Štrbenc, Dr. Matjaž Uršič, Prof. Dr. Janko Žust, Veterinary Faculty University of Ljubljana / Veterinarska fakulteta Univerze v Ljubljani Prof. Dr. Ivor D. Bowen, Cardiff School of Biosciences, Cardiff, Wales, UK, Prof. Dr. Rudolf Cabadaj, University of Veterinary Medicine, Košice, Slovakia, Dr. Dr.h.c. Gerry M. Dorrestein, Faculteit der Diergeneeskunde Utrecht, The Netherlands, Assist. Prof. Dr. Wolfgang Henninger, Veterinärmedizinische Universität Wien, Austria, Prof. Dr. Dr.h.c. mult. Josef Leibetseder, Veterinärmedizinische Universität Wien, Austria, Dr. Louis Lefaucheur, INRA, Saint-Gilles, France, Prof. Dr. Bela Nagy, Veterinary Medical Research Institute Budapest, Hungary, Prof. Dr. Detlef Rath, Institut für Tierzucht, Forschungsbereicht Biotechnologie, Bundesforschungsanstalt für Landwirtschaft (FAL), Neustadt, Germany, Prof. Dr. Dr.h.c. Hans-Peter Sallmann, Tierärtzliche Hochschule Hannover, Germany, Prof. Dr. Duro Sulimanovic, Veterinary Faculty Zagreb, Croatia Editor in chief / glavni in odgovorni urednik: Emer. Prof. Dr. Srdan V. Bavdek Technical editor / tehnični urednik: Dr. Matjaž Uršič Language revision / lektorja: Prof. Dr. Viktor Majdič (Slovenian / slovensko), Petra Kaloh (English / angleško) Address: Veterinary Faculty, Gerbičeva 60, 1000 Ljubljana, Slovenia Naslov: Veterinarska fakulteta, Gerbičeva 60, 1000 Ljubljana, Slovenija Tel.: +386 (0)1 47 79 100, 47 79 129, Fax: +386 (0)1 28 32 243 E-mail: slovetres@vf.uni-lj.si Sponsored by the Ministry of Education, Science and Sport of the Republic of Slovenia Sofinancira: Ministrstvo za šolstvo, znanost in šport Republike Slovenije ISSN 1580-4003 Printed by / tisk: Tiskarna Pleško, d.o.o., Ljubljana Indexed in / indeksirano v: Agris, Biomedicina Slovenica, CAB Abstracts, Urlich's International Periodicals Directory http://www.vf.uni-lj.si SLOVENIAN VETERINARY RESEARCH SLOVENSKI VETERINARSKI ZBORNIK Slov Vet Res 2004; 41 (1) Review paper Butinar J, Mujagic E, Galac S. The oestrus cycle in the bitch: a review article ..............................5 Original research papers Juntes P, Demšar M. Morphometrical analysis of gastrin cells in the gastric mucosa of three-week-old pigs (Sus scrofa domesticus) and a comparison to with other gastric enteroendocrine cells...............13 Zdovc I, Švara T, Juntes P, Kotnik T, Ocepek M, Černe M, Pogačnik M. The role of Aeromonas hydrophila bacterium as a causative agent of septicaemia in dogs ..........................................23 Podpečan O, Pengov A, Hrastnik U. Treatment of subclinical staphylococcal mastitis........................31 Gregorc A, Planinc I. Using oxalic acid for varroa mite control in honeybee colonies during the beekeeping season ................................................................. 35 Piccione G, Arcigli A, Costa A, Fazio F, Caola G. Changes in the clotting times and fibrinogen concentrations in horses during a showjump.................................................. 41 Case reports Kadunc Kos V. Treatment of scrotal hernia in foals .................................................47 Kramarič P, Nemec A, Fležar M, Kern I, Pavlica Z. Clinical analysis of horses with the recurrent airway obstruction .....................................................................53 News ..................................................................................59 Slov Vet Res 2004; 41 (1): 5-11 UDC 619:636.7.082.45 Review paper THE OESTRUS CYCLE IN THE BITCH: A REVIEW ARTICLE Janos Butinar 1*, Emir Mujagic 1, Sara Galac 2 Addresses of authors: 1 Clinic for Small Animal Medicine and Surgery, Veterinary Faculty, Cesta v Mestni log 47, 1000 Ljubljana, Slovenia; 2 Department of Clinical Sciences of Companion Animals, Faculty of Veterinary Medicine, Utrecht University, The Netherlands *Corresponding author. E-mail: janos.butinar@vf.uni-lj.si Summary: The regulation of the oestrous cycle in the bitch is quite unique among the various animal species commonly encountered in veterinary medicine. The cycle, which has an average length of about 3 months, consists of the pro-oestrus, oestrus and metoestrus phases and is followed by the anoestrus phase that varies in duration. The duration and regulation of the luteal phase is the same in both cyclic and pregnant bitches. In contrast to some other species, the uterus is not involved in the regulation of the cyclic corpus luteum. While the first stage of the luteal phase is completely autonomous, the second depends on pituitary factors, mainly prolactin. It is still unclear whether LH has luteotrophic properties in the bitch. Recent studies centred on the role that hormones play in the oestrous cycle have led to developments such as the FSH threshold concept, which could lead to the induction and manipulation of the oestrous cycle. They have also led to the use of progesterone-receptor antagonists to control some of the physiological consequences of the luteal phase, such as pseudopregnan-cy. This article presents a review of the new scientific insights concerning the oestrous cycle of the bitch, with the emphasis on the regulation and complications of the luteal phase, such as the cystic endometrial hyperplasia-pyometra syndrome, acromegaly, insulin resistance, diabetes mellitus and the incidence of mammary tumours. Key words: oestrous cycle; luteal phase; bitch Introduction The oestrous cycle of the bitch consists of the pro-oestrus, oestrus, and metoestrus phases. After each oestrous cycle, which has a length of about 3 months, there is an anoestrus phase, which varies in duration. The mean interval from the onset of one oestrous cycle to the next is about 7 months, within a range of between 4 and 12 months. The inter-oestrous interval of individual bitches may be either regular or variable (1). Pro-oestrus is defined as the period when the bitch has become sexually attractive but is rejecting the male's advances until the first signs of its willingness to accept the male. As early behavioural signs may be indistinct, the onset of a serosanguineous vaginal discharge and swelling of the vulva are used to mark the first day of pro- Received: October, 2003 Accepted for publication: May, 2004 oestrus. The duration of pro-oestrus ranges from 3 to 17 days with an average of 9 days. Oestrus is the period when the female allows breeding and has an average duration of 9 days, within a range of 3 to 21 days. During oestrus the vulva begins to shrink and soften and the vaginal discharge generally diminishes. Metoestrus begins when the bitch will no longer accept the male and usually lasts about 70 days. The end of metoestrus can be defined in a number of ways such as when the progesterone secretions of the luteal phase subside, mammary development declines, progesterone secretions no longer effect the endometrium or when the plasma-progesterone concentration initially declines to a level of 1 ^g/L or less (1, 2). In addition to this behaviour-oriented classification of the oestrous cycle, a new and far more appropriate classification system has been introduced. It is based on the ovarian function and divides the oestrous cycle into four phases: the 6 J. Butinar, E. Mujagic, S. Galac follicular, the pre-ovulatory luteinization and ovulation, the luteal and the anoestrus phases (3, 4). Follicular phase Tertiary follicles developing in the ovaries produce oestradiol. The increased concentration of oestradiol is responsible for external signs of pro-oestrus, such as hyperaemia and oedema of the vulva, the bloody vaginal discharge and for behavioural changes. It also causes a lengthening and hyperaemia of the uterine horns, an enlargement of the cervix and a thickening of the vaginal wall (1, 5). Plasma-oestradiol levels increase constantly throughout the follicular phase and reach peak plasma values 1 to 2 days before the preovulatory LH surge (6, 7, 8). The oestrogen levels decline rapidly thereafter, while the level of plasma progesterone starts to increase as a result of the partial luteinization of the follicles (7). Both LH and FSH plasma concentrations are relatively low during the follicular phase (9, 10). Pre-ovulatory luteinization and ovulation The pre-ovulatory LH surge starts 1 to 2 days after the oestradiol peak and coincides with the declining oestradiol and rising plasma-progesterone concentrations (6, 7). It has been suggested that the pre-ovulatory LH surge is triggered by a decline in the plasma oestrogen: progesterone ratio in the latter stages of the follicular phase (7, 11). The pre-ovulatory LH surge, which lasts from 24 to 72 hours and produces a rapid and final enlargement and luteinization of the mature follicles, causes ovulation and in the process transforms oestrogen-secreting follicles into progesterone-secreting corpora lutea (7). Thus the LH surge represents the transition from the follicular phase to the luteal phase. Ovulation appears to occur synchronously about 36 - 48 hours after the LH peak (7, 12). Most ova in the bitch are ovulated in an immature state as primary oocytes (13) and cannot be fertilized until they undergo the first meiotic division to become secondary oocytes, which usually occurs about 60 hours after ovulation (12, 14). By this time the ova have descended through two thirds of the oviduct. Plasma-progesterone concentrations are between 2 and 4 pg/L at the LH peak and by the time ovulation occurs, usually 36 to 48 hours later, they rise from 5 to 8 pg/L (15). Concurrent with the LH peak, there is also a pre-ovulatory surge of FSH that reaches its peak concentration 1 to 2 days later (9). Luteal phase The concentration of progesterone, which originates from the corpora lutea, increases in the peripheral blood during the latter stage of oestrus and the onset of metoestrus and reaches its maximum level 10 to 30 days after the LH peak. Thereafter, in non-pregnant bitches, the progesterone secretion slowly declines and reaches a basal level of 1 pg/L for the first time about 75 days after the start of the luteal phase (15). The transition from oestrus to metoestrus takes place during the initial stage of the luteal phase. Regulation of the luteal phase In many species, the regulation of the cyclic corpus luteum is influenced by both luteotrophic and luteolytic factors. Prostaglandin F2? originating from the endometrium, which is the causative factor for luteolysis in the cow and sheep, is not present during the luteal phase of the cyclic dog (15). This is demonstrated by the fact that a hysterectomy does not influence the length of the luteal phase. Therefore the uterus is not involved in the regulation of the cyclic corpus luteum (15). Moreover, in the initial stage of the luteal phase the canine corpus luteum functions completely autonomously. Studies of dogs that had undergone hypophysectomy demonstrated that the canine corpus luteum functions independently of pituitary support for 24 to 28 days from the onset of the luteal phase (16). Administering aglepristone in the early part of the luteal phase does not effect its duration (17). During the second half of the luteal phase, pituitary luteotrophic factors - prolactin and possibly LH - are necessary to sustain the luteal function (16, 18, 19, 20). Whether LH has luteotrophic properties in the bitch is still unclear. Concannon et al. (1987) reported that passive immunization against LH caused a decline in the progesterone concentration. However, the luteotrophic role of LH has been brought into question by studies in which LH-inhi-bition had no effect on the plasma-progesterone concentration, whereas prolactin-inhibition caused it to fall abruptly, indicating that only pro- The oestrus cycle in the bitch: a review article 7 160 t 11 16 21 26 31 36 41 46 51 56 61 66 71 day of the luteal phase progesterone......prolactin Figure 1: Mean plasma concentrations of progesterone and prolactin in 3 healthy beagle bitches, starting from the day of ovulation (Day 1) to the end of the luteal phase. (From Galac S. The effect of aglepristone, the progesterone receptor antagonist, on the hypothalamic-pituitary-ovarian axis, pregnancy and luteal phase in bitches. In: doctoral thesis. Ljubljana, 2001. Reproduced with the author's permission) lactin is luteotrophic in cyclic dogs (19). It is possible that LH has an indirect luteotrophic role that is mediated by the secretion of prolactin (21). Anoestrus The transition from the luteal phase to anoestrus is gradual and varies considerably among bitches. The onset of anoestrus depends on which criteria are being used to define the end of the luteal phase. It can be defined as the period when mammary development subsides, which is usually after 2 to 3 months, or when the plasma-progesterone concentration reaches a level below 1 ^g/L for the first time or as the moment that the influence of progesterone on the endo-metrium is no longer evident (1, 22). In anoestrus, the normal bitch is neither attractive nor receptive to the male, the mucoid vaginal discharge is minimal and the vulva is small (5). Although anoestrus seems an inactive part of the oestrous cycle in the bitch, neither the ovaries nor pituitary are quiescent (9). From early to late anoestrus, the hypothalamus releases increasing amounts of GnRH (23), which the pituitary becomes increasingly sensitive to (24). Additionally, increases in ovarian responsiveness to gonado-trophins (25) and the level of basal LH-concentra-tions towards the end of the anoestrus (9), as well as a brief period of increased LH pulsatility (26) have been reported as important determinants of the initiation of a new follicular phase. It has even been suggested that changes in the LH secretion may be more important than changes in the FSH secretion in the initiation of the follicular phase leading to ovulation (11). In line with this, the administration of pharmacological doses of LH can terminate anoestrus in bitches by inducing the follicular phase (27). However, the progression from early to late anoestrus is associated with an increase in the basal plasma-FSH concentration, suggesting that in the bitch an increase in circulating FSH levels is a critical event in the initiation of ovarian folliculogenesis (10). In this respect, there are similarities with the situation in primates. Observations during gonadotrophin-indu- 8 J. Butinar, E. Mujagic, S. Galac 50 40 3 30 | 20 CL 10 0 Before During After Figure 2: The average (± SEM) of the mean plasma prolactin concentrations in 6 beagle bitches before, during and after treatment with aglepristone. Asterisks indicate significant difference. (From Galac S. The effect of aglepristone, the progesterone receptor antagonist, on the hypothalamic-pitui-tary-ovarian axis, pregnancy and luteal phase in bitches. In: doctoral thesis. Ljubljana, 2001. Reproduced with the author's permission) ced ovulation in women have emphasized that plasma FSH must exceed a certain concentration before preantral follicles reaching the FSH-de-pendent stage can progress to maturation (28). This has been labelled the FSH threshold concept. An increase of only 10 to 30 % above the threshold concentration of plasma FSH is sufficient to stimulate normal follicular development in women (29). The threshold for FSH may vary among individuals, and each follicle also has its own sensitivity to FSH (28). The study of Kooistra et al. (1999) indicate that the FSH threshold concept, as anticipated for women, could hold true for the dog as well. The oestrous cycle can begin at any time throughout the year and there appears to be little, if any, seasonal influence. Breed differences and strains within breeds can form the basis of variation in mean inter-oestrous intervals. Environmental factors can also affect the inter-oestrous interval: the onset of pro-oestrus in an anoestrus bitch can be brought forward by several weeks by placing her in close proximity to a bitch in oestrus. Furthermore, bitches housed together often have synchronous oestrous cycles (11, 30). Consequences of the oestrus cycle The relatively long exposure to high levels of progesterone circulating during each oestrous cycle may result in disorders such as cystic endome-trial hyperplasia-pyometra syndrome (31), acro-megaly, insulin resistance, diabetes mellitus, (32) and an increased incidence of mammary tumours. Cystic endometrial hyperplasia involves proliferation of the glandular epithelium and cystic dilatation of the endometrial glands with endome-trial fluid accumulated in their lumen. These conditions provide an excellent environment for bacterial growth, which can lead to the development of pyometra. Pyometra is often caused by entering the progesterone phase of the sexual cycle with an abnormal endometrium, which can result in an overgrowth of bacteria that is normally isolated from this area of the anatomy. Surgery is the preferred treatment for pyometra unless the owner adamantly wants to breed with the bitch. The medical treatment consists of administering antibiotics and prostaglandins. If the bitch is still in the luteal phase, progesterone-receptor blockers may also be administered to diminish the influence of progesterone (33). The oestrus cycle in the bitch: a review article 9 Acromegaly occurs as a consequence of excess secretion of the growth hormone (GH). Progesterone-induced GH secretions originate from the foci of hyperplastic ductular epithelium of the mammary gland (34, 35). In contrast to the GH from the pituitary gland, GH from the mammary gland is not pulsatile and cannot be stimulated by the GH-releasing hormone (GHRH) and nor can it be inhibited by somatostatin (34, 36). The progesterone-induced GH excess may lead to insulin resistance, exhaustion of the pancreatic ?-cells and consequently diabetes mellitus (32). If diabetes mellitus is diagnosed while there is a high level of progesterone secretion, it might have a reversible nature. However, the source of progesterone must be removed as early as possible. Therefore, an ovariectomy is advised if diabetes mellitus occurs during the luteal phase, although it is difficult to predict whether the pancreatic insulin production will completely recover. In any case, supportive therapy with insulin is recommended after the surgery. In order to prevent hypoglycaemia and to achieve the right dosage of insulin, daily blood glucose measurements are needed and the insulin dose adjusted accordingly (37). Pseudopregnancy is a syndrome that accompanies the extended luteal phase of all the non-pregnant ovarian cycles in the bitch (38). An important precipitating factor for pseudopregnan-cy appears to be a rapid decline in the plasma progesterone concentration, which is assumed to be the trigger for the release of prolactin, which in turn would give rise to pseudopregnancy (39). Correspondingly, an ovariectomy performed in the luteal phase often induces an overt pseudopreg-nancy. Studies using the progesterone-receptor antagonist aglepristone, have suggested that a sudden decline in the plasma progesterone concentration induces an increase in the concentration of prolactin (40, 41). The development of mammary gland tumours in the bitch is clearly hormone dependent. The role of progestins in the pathology of the mammary gland was revealed in 1969, when Schneider et al. published a study about the protective effect of an ovariohysterectomy on mammary tumour development. They estimated that in comparison with intact dogs, bitches that had been spayed prior to their first oestrus had a 0.05 % risk of developing malignant tumours. This increased to 8 % if spayed following their first oestrus and rose to 26 % if spayed after their second oestrus. The spaying of older dogs does not reduce their risk of developing malignant tumours, although an ovariectomy does appear to reduce their risk of developing benign tumours (42). The protective effect of an early pregnancy, which is well known in the human, has not been demonstrated in the dog. As in the normal mammary gland, GH receptors have been demonstrated in neoplastic tissue (43). It has been speculated that the maximal effect of progestins on the mammary gland might be facilitated by the additional local action of GH. Yet, it must still be proven whether progestin-induced GH acts as an intermediate in the prog-estin-stimulated development of canine mammary tumours. Conclusion Being familiar with the endocrinological events associated with the oestrous cycle in the bitch could help the clinician to understand any complications that may eventuate during the luteal phase and to provide the best possible treatment for them. Applying this knowledge to breeding management, which is based on the hormonal changes in the oestrous cycle, will provide better results than those produced by using the empirical, behaviour-oriented approach. Canine female reproduction is a rapidly developing field in veterinary medicine and the pharmaceutical industry has provided us with several new possibilities to improve breeding programmes or to treat maladies associated with the oestrous cycle. Therefore it is of great importance that the small animal clinician keeps up to date with newly emerging information and developments regarding the endocrinology of the oestrous cycle in the bitch. References 1. Concannon PW, McCann JP, Temple M. Biology and endocrinology of ovulation, pregnancy and parturition in the dog. J Reprod Fertil 1989; 39 (Suppl): 3-25. 2. Holst PA, Phemister RD. Onset of diestrus in the Beagle bitch: definition and significance. Am J Vet Res 1974; 35: 401-6. 3. Schaefers-Okkens AC. Ovaries. In: Rijnberk A, ed. Clinical endocrinology of dogs and cats. Dordrecht: Kluwer Academic Publishers, 1996: 131-56. 4. Schaefers-Okkens AC. Estrous cycle and breeding management of the healthy bitch. In: Ettinger SJ, Feldman EC, eds. Textbook of veterinary internal medicine. 5th ed. Philadelphia: WB Saunders, 2000: 1510-9. 5. Wildt DE, Seager SWJ, Chacraborty PK. Beha- 10 J. Butinar, E. Mujagic, S. Galac vioral, ovarian and endocrine relationship in the pubertal bitch. J Anim Sci 1981; 53 (1): 182-1. 6. Concannon PW, Hansel W, Visek VJ. The ovarian cycle of the bitch: Plasma estrogen, LH and progesterone. Biol Reprod 1975; 13: 112-21. 7. Concannon PW, Hansel W, McEntee K. Changes in LH, progesterone and sexual behavior associated with preovulatory luteinization in the bitch. Biol Reprod 1977; 17: 604-13. 8. Jones GE, Boyns AR, Cameron EHD, Bell ET, Christie DW, Parkes MF. Plasma estradiol. Luteinizing hormone and progesterone during pregnancy in the Beagle bitch. J Reprod Fertil 1973; 35: 187-9. 9. Olson PN, Bowen RA, Behrendt MD, Olson JD, Nett TM. Concentrations of reproductive hormones in canine serum throughout late anestrus, proestrus and estrus. Biol Reprod 1982; 27: 1196-206. 10. Kooistra HS, Okkens AC, Bevers MM, et al. Concurrent pulsatile secretion of luteinizing hormone and follicle-stimulating hormone during different phases of the estrus cycle and anestrus in beagle bitches. Biol Reprod 1999; 60: 65-71. 11. Concannon PW. Biology of gonadotrophin secretion in adult and prepubertal female dogs. J Reprod Fertil 1993; 47 (Suppl): 3-27. 12. Phemister RD, Holst PA, Spano JS, Hopwood ML. Time of ovulation on the beagle bitch. Biol Reprod 1973; 8: 74-82. 13. Wildt DE, Chacraborty PK, Panko WB, Seager SWJ. Relationship of reproductive behavior, serum lu-teinizing hormone and time of ovulation in the bitch. Biol Reprod 1978; 18: 561-70. 14. Holst PA, Phemister RD. The prenatal development of the dog: preimplantation events. Biol of Reprod 1971; 5: 194-206. 15. Okkens AC, Dieleman SJ, Bevers MM, Willemse AH. Evidence of non-involvement of the uterus in the lifespan of the corpus luteum in the cyclic dog. Vet Q 1985; 7: 169-73. 16. Okkens AC, Dieleman SJ, Bevers MM, Lubbe-rink AAME, Willemse AH. Influence of hypophysectomy on the lifespan of the corpus luteum in the cyclic dog. J Reprod Fertil 1986; 77: 187-92. 17. Galac S, Kooistra HS, Dieleman SJ, Cestnik V, Okkens AC. Effects of aglepristone, a progesterone receptor antagonist, administered during the early luteal phase in the non-pregnant bitch. Theriogenology 2003, Submitted. 18. Concannon PW, Weinstein R, Whaeley S, Frank D. Suppression of luteal function in dogs by luteinizing hormone antiserum and bromocriptine. J Reprod Fertil 1987; 81: 175-80. 19. Okkens AC, Bevers MM, Dieleman SJ, Willemse AH. Evidence for prolactin as the main luteotrophic factor in the cyclic dog. Vet Q 1990; 12: 193-201. 20. Onclin K, Verstegen JP. In vivo investigation of luteal function in dogs: effects of cabergoline, a dopa-mine agonist, and prolactin on progesterone secretion during mid-pregnancy and -diestrus. Dom Anim Endocrinol 1997; 14: 25-38. 21. Onclin K, Verstegen JP, Concannon PW. Time-related changes in canine luteal regulation: in vivo ef- fects of LH on progesterone and prolactin during pregnancy. J Reprod Fertil 2000; 118: 417-24. 22. Holst PA, Phemister RD. Temporal sequence of events in the estrous cycle of the bitch. Am J Vet Res 1975; 36: 705-6. 23. Tani H, Inaba T, Tamada H, Sawada T, Mori J, Torii R. Increasing gonadotropin-releasing hormone release by perifused hypothalamus from early to late ane-strus in the beagle bitch. Neurosci Lett 1996; 207: 1-4. 24. Van Haaften B, Bevers MM, Van Den Brom WE, et al. Increasing sensitivity of the pituitary to GnRH from early to late anoestrus in the beagle bitch. J Reprod Fertil 1994; 101: 221-5. 25. Jeffcoate IA. Endocrinology of anestrous bitches. J Reprod Fertil 1993; 47 (Suppl): 69-76. 26. Concannon PW, Whaeley S, Anderson SP. Increased LH pulse frequency associated with termination of anestrus during the ovarian cycle in the dog. Biol Reprod 1986; 34: 119. 27. Verstegen J, Onclin K, Silva L, Concannon P. Termination of obligate anestrus and induction of fertile ovarian cycles in dogs by administration of purified pig LH. J Reprod Fertil 1997; 111: 35-40. 28. Schoemaker J, Van Weissenbruch MM, Scheele T, Van Der Meer M. The FSH threshold concept in clinical ovulation induction. Baillere's Clin Obstet Gynaecol 1993; 7: 297-308. 29. Brown JB. Pituitary control of ovarian function: concepts derived from gonadotrophin therapy. Aust NZ J Obstet Gynaecol 1978; 18: 47-54. 30. Bouchard GF, Youngquist RS, Vaillancourt D, Krause GF, Guay P, Paradis M. Seasonality and variability of the interoestrus interval in the bitch. Theriogenology 1991; 36: 41-50. 31. Dow C. The cystic hyperplasia-pyometra complex in the bitch. Vet Rec 1958; 70: 1102-10. 32. Eigenmann JE, Eigenmann RY, Rijnberk A, Van Der Gaag I, Zapf J, Froesch ER: Progesterone controlled growth hormone overproduction and naturally occurring canine diabetes and acromegaly. Acta Endocrinol 1983; 104: 167-76. 33. Hoffmann B, Schuler G. Receptor blockers-gen-eral aspects with respect to their use in domestic animal reproduction. Anim Reprod Sci 2000; 60-61: 295312. 34. Selman PJ, Mol JA, Rutteman GR, Van Garderen E, Rijnberk A. Progestin-induced growth hormone excess in the dog originates in the mammary gland. Endocrinology 1994; 134: 287-92. 35. Van Garderen E, De Wit M, Voorhout WF, Rutteman GR, Mol JA, Nederbragt H, Misdorp W. Expression of growth hormone in canine mammary tissue and mammary tumours: evidence for a potential autocrine / paracrine stimulatory loop. Am J Pathol 1997; 150: 1037-47. 36. Watson ADJ, Rutteman GR, Rijberk A, Mol JA. Effect of somatostatin analogue SMS 201-995 and an-tiprogestin agent RU 486 in canine acromegaly. Front Horm Res 1987; 17: 193-8. 37. Rijnberk A. Endocrine pancreas. In: Rijnberk A, ed. Clinical endocrinology of dogs and cats. Dordrecht: Kluwer Academic Publishers, 1996: 95-118. The oestrus cycle in the bitch: a review article 11 38. Jöchle W. Prolactin in canine and feline reproduction. Reprod Domest Anim 1997; 32: 183-93. 39. Gerres S, Höveler R, Evers P, Hoffman B. Investigation of the role of progesterone (P4) in the endocrine control of overt pseudopregnancy in the bitch, application of an antigestagen. In: Proceedings of the 11th International Congress of Animal Reproduction and Artificial insemination. Dublin, 1988: 90-1. 40. Galac S, Kooistra HS, Butinar J, Bevers MM, Dieleman SJ, Voorhout G, Okkens AC. Termination of mid-gestation pregnancy in bitches with aglepristone, a progesterone receptor antagonist. Theriogenology 2000; 53: 941-50. 41. Fieni F, Martal J, Marnet PG, et al. Hormonal variation in bitches after early or mid-gestational pregnancy termination with aglepristone (RU534). J Reprod Fertil 2001; 57 (Suppl): 243-8. 42. Misdorp W. Progestogens and mammary tumors in dogs and cats. Acta Endocrinol (Kopenhagen) 1991; 125: 27-31. 43. Van Garderen E, Van Der Poel HJA, Swen-nehius JF, et al. Expression and molecular characterisation of the growth hormone receptor in canine mammary tissue and mammary tumours. Endocrinology 1999; 140: 5907-14. POJATVENI CIKLUS PRI PSICI: PREGLEDNI ČLANEK J. Butinar, E. Mujagic, S. Galac Povzetek: Regualacija pojatvenega ciklusa pri psici, ki ga sestavljajo proestrus, estrus in metestrus in traja približno 3 mesece, sledi pa mu različno dolg anestrus, je med živalskimi vrstami v veterinarski medicini edinstvena. Trajanje in regulacija lutealne faze se ne razlikujeta pri ciklični in breji psici. V nasprotju z nekaterimi drugimi živalskimi vrstmi maternica ni vpletena v regulacijo cikličnega rumenega telesa. Prav tako je prvi del lutealne faze popolnoma avtonomen, v nasprotju z drugim, ki je odvisen od hipofiznih dejavnikov, predvsem prolaktina. Še vedno ni pojasnjeno, ali ima LH pri psici luteotropno vlogo ali ne. V zadnjem času so prišli do nekaterih novih spoznanj o hormonskih dogajanjih, na primer uveljavljanje koncepta praga FSH, ki lahko sprožijo pojatveni ciklus. Prav tako so bile nedavno v raziskavah z antagonisti prog-esteronskih receptorjev osvetljene nekatere fiziološke posledice lutealne faze, na primer navidezna brejost (psevdogravid-nost). Dajanje aglépristona, antagonista progesteronskih receptorjev, v zgodnji lutealni fazi ni vplivalo na njeno trajanje. Članek predstavlja pregled novih znanstvenih spoznanj o pojatvenem ciklusu pri psici, s poudarkom na urejanju in zapletih lutealne faze, kot so cistična hiperplazija, sindrom endometrija - piometra, akromegalija, inzulinska odpornost in sladkorna bolezen ter povečano pojavljanje tumorjev mlečne žleze. Ključne besede: estrus; lutealna faza; psica Slov Vet Res 2004; 41 (1): 13-22 UDC 619:612.014.2:612.325:636.4 Original research paper MORPHOMETRICAL ANALYSIS OF GASTRIN CELLS IN THE GASTRIC MUCOSA OF THREE-WEEK-OLD PIGS (Sus scrofa domesticus) AND A COMPARISON WITH OTHER GASTRIC ENTEROENDOCRINE CELLS Polona Juntes 1*, Matej Demšar 2 Addresses of authors: 11nstitute of Pathology, Forensic and Administrative Veterinary Medicine, Veterinary Faculty, University of Ljubljana, Gerbičeva 60, 1000 Ljubljana, Slovenia; 2 Visoko pri Poljanah 2, 4220 Škofja Loka, Slovenia ""Corresponding author. E-mail: polona.juntes@vf.uni-lj.si Summary: The number and localisation of enteroendocrine cells from different parts of the gastric mucosa of three-week-old pigs were analysed. The basic method for the visualisation of enteroendocrine cells was immunohistochemistry and the results were compared to those derived through both the toluidine-blue staining for enteroendocrine cells and Grimelius silver staining methods. The greatest number of gastrin cells was in the pyloric mucosa with 124.66 ± 4.54 gastrin cells per mm2 in the antrum and 140.32 ± 5.30 per mm2 in the pyloric canal. The highest density of gastrin cells in relation to the thickness of the gastric mucosa was in the middle third of the mucosa with approximately 50 % of all gastrin cells. The remaining gastrin cells were evenly distributed in the upper and lower thirds of the mucosa. Cutaneous mucosa was negative for gastrin and other enteroendocrine cells. We found that the area of the small curvature, which has cutaneous and pyloric types of mucosa, had an average of 29.21 ± 2.97 gastrin cells per mm2 and that the areas with cardial and fundic mucosa had lower averages of 0.61 ± 0.14 cells per mm2 and 1.14 ± 0.11 gastrin cells per mm2, respectively. Morphometric analyses of the serotonin and somatostatin cells from the pyloric area, where there was the highest density of gastric cells, were performed. The numbers of both types of cells were much lower compared to the gastrin cells, with 17.7 ± 2.30 serotonin cells and 50.7 ± 2.76 somatostatin cells per mm2 of mucosa. The results of the silver staining and the toluidine-blue staining only partially correlated with the results of the immunohistochemical reaction. While many more cells reacted to the Grimelius silver staining (153.5 ± 0.32 cells per mm2) than to the toluidine-blue staining (51.15 ± 2.20 cells per mm2), there were still considerably less of them in comparison with the number of cells that reacted immunohistochemically. Key words: veterinary medicine; stomach; enteroendocrine cells; gastrin; somatostatin; serotonin; immunohistochemistry; Grimelius silver staining; toluidin blue; pig Introduction There are four distinct regions of gastric mucosa in pigs: the oesophageal, or proventricular part (pars oesophagealis s. proventricularis), which is located at the gastric entrance and has cutaneous mucosa, and three parts that have glandular mucosa (parsglandularis). The anterior third of glandular mucosa, which is located in the anterior part of ventricular body between the oesopha- Received: May, 2003 Accepted for publication: December, 2003 geal part and the genu ventriculi and includes the gastric diverticulum, has cardial mucosa. The second part is behind the genu and it contains fundic mucosa with fundic glands, and the third area is the caudal part of the stomach, which has pyloric mucosa and begins at the larger antrum and continues into the pyloric canal that contains numerous sero-mucosal glands in the propria (1). Enteroendocrine cells can be found scattered among the epithelial mucosal cells in all parts of the gastric mucosa (2). They belong to the diffuse endocrine system or the APUD system (amino precursor uptake and decarboxylation). In this sys- 14 P. Juntes, M. Demšar tem byogene amines and peptides are synthesised - neurohormones and neurotransmitters, suggesting their relation to neurones (2, 3). There are 6 types of enteroendocrine cells in the stomach: A, G, D, EC, P and ECL cells. The A cells synthesise gastro-glucagon, the G cells (argyrophil cells) produce gastrin, D cells produce somatostatin and the EC cells (enterochromaffin or argentaffin cells) produce 5-hydroxytryptamine (5-HT). The P cells produce both prostaglandin and motilin and the ECL cells (enterochromaffin-like cells), which can be visualised with the Grimelius silver staining, produce histamine (4, 5, 6). Gastrin is a heptadecapeptide with 17 amino acids produced by gastrin cells (G cells). Gastrin cells are located within the gastric epithelium and in humans, where they have a characteristically conical or oval form with microvilli on their surface (7), the highest concentrations are found in the antrum. Their ultrastructure depends on their activity status (8) and they contain three types of granules (6, 9, 10). Gastrin molecules are synthesised from precursor molecules (10). There are two chemical forms of gastrin, gastrin I and gastrin II (7), and several biologically active forms: large gastrin G34, small gastrin G17, mini gastrin G14 and component I (11). All forms have at C-end of the molecule tetragastrin, a prolongation molecule, which is also biologically active (2). Gastrin has an exocrine function as it secrets into the gastric lumen, it has also an endocrine function and influences distant tissues by secreting into the blood, and as it regulates the activities of neighbouring cells, it also has a paracrine effect (2). Gastrin is not only important for the normal functioning of the stomach, but also for the normal functioning of other parts of the digestive system - the duodenum, pancreas, liver. In humans it has been shown to stimulate the excretion of gastric acid and pepsin, the growth of gastric mucosa, the secretion of pancreatic enzymes and the secretion of water and electrolytes in the stomach, pancreas, liver and Brunner glands. It increases the absorption of water and electrolytes from the small intestine, stimulates contractions of the stomach and gall bladder and relaxes the muscles of the pyloric and the ileocaecal sphincter. It also stimulates the secretion of insulin, acetyl choline, somatostatin, pancreatic polypep-tide and calcitonin as well as affecting the proliferation of ECL cells (2, 12, 13, 14, 15). There is already a lot of published data regarding gastrin cells of different animal species. Three types of gastric cells have been described in newborn rats, according to their level of differentiation - primitive, transitional and typical gastrin cells (9). However, very little data is available that describes at what age they appear and if there are any age-dependent variations in their numbers and localisation within the gastric mucosa and physiological characteristics of the animal species at certain ages. We found very few references relating to the gastrin cells of the pig, and even fewer pertaining to the different pre- and postnatal developmental stages and to the adaptation to the changes in the feeding practices after birth and weaning. The majority of gastrin in the stomach of pigs is small gastrin 17 (16) and there are equal quantities of gastrin 17 and its precursor, gastrin 34 (17). The secretion of gastrin depends of several factors: the composition of food, the quantities of releasing and inhibiting peptides (GRP, GIP) in the circulation, nervous factors (vagus), and paracrine and endocrine factors (prostaglandin, somatostatin, histamine etc.) (18, 19). In this study we analysed the gastric mucosa and gastrin cells of three-week-old pigs. At this stage the pigs are weaned and their gastric mucosa has to quickly adapt to new types of food. It is also a period when piglets experience frequent gastrointestinal disorders. For comparison we also analysed the localisation and distribution of the somatostatin and serotonin cells, both of which are involved in controlling gastrin secretion. Material and methods Materials Ten weaned, three-week-old, farm pigs were killed and their stomachs, which were immediately removed from their abdominal cavity, were opened along the major curvature. The stomach content was removed and samples were taken from six parts of the gastric wall: from the cardia, fornix and gastric body, and from the antrum, pyloric canal and the minor curvature. The tissue was then fixed in buffered formalin for approximately 24 hours before being embedded in paraffin following standard laboratory procedures. The paraffin blocks were then cut into 7^m thick sections. Morphometriacal analysis of gastrin cells in the gastric mucosa of three-week-old pigs (Sus scrofa domesticus) 15 Histology One tissue section from each of the gastric regions was stained with haematoxylin and eosin for the histological examinations. Immunohistochemistry The immunohistochemistry was performed using the PAP method (peroxidase - anti-peroxi-dase) described by Sternberger (20). We applied a 1:1500 dilution of polyclonal rabbit antihuman gastrin-17 as primary antibodies (DAKO, cat. no. A568) to slides that were then incubated in a humidified chamber overnight at 4 oC. These antibodies also react with gastrin-34. The slides were then treated with a 1:50 dilution of goat anti-rabbit immunoglobulins (Amersham) prior to a final treatment with a 1:50 dilution of PAP complex (Sigma, cat. no P1291). All the dilutions were made with a 2.5 % solution of bovine serum albumin (BSA) in a buffered phosphate saline. DAB (3'-diaminobenzidin tetrahydrochloride, Sigma, cat. no. D5905) was used as a substrate and the tissue sections were counterstained with either haematoxylin or 1 % methyl green, dehydrated and mounted with synthetic resin. After we had completed the morphometric analyses of the PAP tissue sections the DAKO LSAB®2 System, HRP product (Dako, cat. no. K0675) became available on the market. By using it in accordance with the manufacturer's instructions we were able to assess the quality of the product by making a comparison of the immunohistochemical reactions of the PAP method and this new system. Grimelius silver staining (21) We used the Grimelius silver stain to determine the numbers of gastrin cells and compared these results with those of the PAP method. This histochemical method was also used to determine the population of all the enteroendocrine cells in the stomach that react with this type of silver staining. Toluidine-blue staining for enteroendocrine cells (22) The toluidine-blue staining method, which is adjusted for the detection of enteroendocrine cells, supposedly enables a partial differentiation of these cells without the use of immunohisto-chemistry. Cells containing gastrin should stain in a metachromatic manner and other ente-rochromaffin cells in an orthochromatic manner, e.g. cells containing serotonin, after a pre-treat-ment in hot hydrochloric acid. Treatment of serial tissue sections using all the methods and a double-PAP procedure for the localisation of gastrin and other enteroen- docrine cells Serial tissue sections were prepared using only the samples from the pyloric part of the stomach. One tissue section was treated with toluidine blue, on another tissue section we demonstrated gastrin cells using the PAP method and a third one was silver stained following the Grimelius silver-staining procedure. A fourth tissue section was immunohistochemically treated for a reaction to serotonin using the PAP method and a 1:1000 dilution of polyclonal rabbit anti-serotonin antibodies (Inc Chemical Credential). A fifth tissue section was used for a double-immunohistochem-ical reaction - the first layer for gastrin and the second layer for somatostatin (Dako, cat. no. A 566, diluted 1:1500). A sixth tissue section was used for a double-immunostaining procedure to determine serotonin (first layer) using a nickel-ammonium-sulphate-enhanced DAB substrate (23) that gives either a dark brown or a black insoluble-reaction product; and somatostatin (second layer) using a DAB substrate that gives an insoluble brown-reaction product. We also localised somatostatin-immunoreactive cells on another tissue section. As a comparative measure, the LSAB technique was also used later to detect serotonin, somatostatin and gastrin. The double-immunohistochemical reactions were achieved using the steps of the PAP method with the application of the first primary antibody and the incubation in the first substrate. However, instead of being dehydrated, the tissue sections were incubated with a second primary antibody in a humidified chamber overnight at 4 oC. The next day the procedure was concluded with repeating of all the consecutive steps of the PAP method that are usual after incubation with primary antibodies. The only change was the use of the different substrate. During the optimisation of the procedure we tested different combinations of antibodies, as this was found to be an impor- 16 P. Juntes, M. Demšar Table 1: Mean values (± SE ) of the thickness of gastric mucosa in different parts of the stomach of a three weeks old pig and mean number of gastin cells per mm2 (± SE) of mucosa Area Mucosal thickness in |jm ± SE Mean number (± SE) of gastrin cells per mm2 of mucosa 1 243,50 ± 0,81 0,00 ± 0,00 2 280,86 ± 1,04 0,61 ± 0,14 3 501,19 ± 1,33 1,14 ± 0,11 4 329,96 ± 1,38 124,66 ± 4,54 5 410,05 ± 0,94 140,32 ± 5,30 6 313,11 ± 1,12 29,21 ± 2,97 Legend: 1 - area with cutaneous mucosa (oesophageal part, cardia) 2 - area with cardial mucosa (fornix) 3 - area with fundic mucosa (gastric body) 4 - area with pyloric mucosa (antrum) 5 - area with pyloric mucosa (pyloric canal) 6 - area with cutaneous and pyloric mucosa (small curvature) Figure 1: Cardial mucosa of a three-week-old pig, a few gastric cells. PAP, DAB, counter-stained with haematoxylin, x 20 tant factor in the quality of the reaction. We achieved the best results using the aforementioned combination. Determination of the thickness of the gastric mucosa The average thickness of the gastric mucosa samples from the five pigs was measured using an ocular micrometer at a magnification of 100x. Ten measurements, from five locations on two consecutive tissue sections, were taken and the distances between the units on the ocular micrometer were calculated with an object micrometer. Determination of the number of gastrin cells The number of gastric cells was determined with a light microscope as were all the gastrinpositive cells on two tissue sections from each of the selected gastric regions. We counted the positive cells across the ten visual fields of each tissue section, i.e. 20 visual fields from each gastric Morphometriacal analysis of gastrin cells in the gastric mucosa of three-week-old pigs (Sus scrofa domesticus) 17 Figure 2: Cardial mucosa of a three-week-old pig, a few gastrin cells but more numerous than in figure 1 - comparative immunostaining to the PAP method (figure 1); DAKO LSAB®2 System, HRP; DAB, counterstained with haema-toxylin, x 20 region. A visual field was determined as the height of gastric mucosa by the width of the field of view at a magnification of 200x. At that magnification the visual field was 530 pm wide. The area of a visual field (VFA) was calculated with the formula: VFA = height of the mucosa x 530 (24) The number of cells was then calculated per mm2 of mucosa. Statistics The results were statistically evaluated using the Batch System software programme and evaluated with an analysis of variance. The calculations were made with a probability of 95 % (P = 0.05). Results and discussion A few references have been published in which the authors discuss the distribution of gastrin cells in the stomach of the pig (25) and their physiological significance at different ages (26, 27, 28). At three weeks the growth rate of a pig's stomach decreases in comparison with that of its body, which coincides with the adaptation of the gastric mucosa to the effects of active peptides, as was reported in a study on the effects of pentagastrin (29). It has also been reported that at the same age the speed of gastrin degradation in the gastric and intestinal lumen is higher in comparison with newborn and adult pigs, although it is not quite clear why (26). In the samples of the gastric mucosa of three-week-old pigs used in our study, the thickness of the mucosa from the different regions of the stomach varied significantly, except between the antral mucosa and the mucosa from the small curvature (average thickness 329.96 ± 1.38 pm and 313.11 ± 1.12 pm, respectively). The average thickness of cutaneous mucosa was 243.50 ± 0.81 pm, of the cardial mucosa 280.86 ± 1.04 pm and the fundic mucosa 501.19 ± 1.33 pm. The average thickness of the mucosa in the pyloric canal was 410.05 ± 0.94 pm, which was significantly more than it was in the antrum, which was 313.11 ± 1.12 pm thick (Table 1). The Grimelius silver staining revealed enteroendocrine cells in all parts of gastric mucosa, except in the areas with cutaneous mucosa. That correlated with both the immuno-histochemical results and those of the toluidine-blue staining. Based on their staining properties, two types of cells were found, which could be representative of two types of enteroendocrine cells. One type of cells gave a light-brown reaction product with the silver staining while the others were dark brown. The highest concentrations of positive cells revealed by the Grimelius silver staining method were in the fundic mucosa, the antral mucosa and in the mucosa of the pyloric canal. 18 P. Juntes, M. Demšar Figure 3: Fundic mucosa of a three-week-old pig, only one cell positive for gastrin; PAP, DAB, counterstained with hae-matoxylin, x 20 The cells were distributed throughout the mucosa with the highest concentrations being at the base of the glands with an intra-epithelial localisation. The toluidine-blue stain reacted with cells in all areas of the gastric mucosa, with the exception of the cutaneous mucosa. The cells stained either orthochromatically or metachromatically. While the orthochromatic cells were localised among epithelial cells (intra-epithelial localisation), similar to the silver-stained cells, the metachromatic cells were extra-epithelial. As with the silver-stained cells, the greatest number of orthochro-matic cells was in the mucosa of the fundus and the pyloric canal, and the least in the cardial mucosa. In all the regions, they were evenly distributed from the base to the luminal part of mucosa. Metachromatic cells were also found in all parts of gastric mucosa, with the exception of the cutaneous mucosa. Based on the localisation of the orthochromatic cells we believe that they are not gastrin cells as there was a very low number of immunohistochemically-positive gastrin cells in the fundic mucosa. Gastrin cells revealed immunohistochemically were found in all parts of the gastric mucosa, except in the oesophageal part with cutaneous mucosa. However, there was great variation between the quantities determined in the different parts of the stomach: from just a few positive cells in the cardial and fundic mucosa to the large number of positive cells found in the pyloric canal. The highest concentration of gastrin cells was in the mid-third of the mucosa and only partly correlated with the localisation and number of cells revealed with the non-specific methods, Grimelius silver staining and the toluidine-blue staining. The intensity of the positive immunohis-tochemical reactions varied from light to dark brown. Although similar distributions of gastrin and other enteroendocrine cells were detected by both immunohistochemical methods, the PAP and the DAKO LSAB®2 System, it was subjectively assessed that the latter revealed more positive cells than the former (Figures 1 and 2). The distribution of gastrin cells in the gastric mucosa of the three-week-old pigs was similar to that determined by Bussolati (25) using immunofluores-cence. However, as we do not know the ages of the pigs used in his study we cannot say that these results can be applied to pigs of all ages. The distribution of the other types of enteroen-docrine cells in parts of the stomach was distinctly different from the distribution of gastrin cells, with the exception of the cutaneous mucosa, which was completely negative for all enteroen-docrine cells. In the mucosa of the pyloric canal we found a small number of serotonin-positive cells and unlike the gastrin cells, which were concentrated in the middle third of the mucosa, the serotonin-positive cells were evenly dispersed throughout the mucosa. In the tissue sections with double-immunohistochemical reactions, the Morphometriacal analysis of gastrin cells in the gastric mucosa of three-week-old pigs (Sus scrofa domesticus) 19 Figure 4: Pyloric mucosa of a three-week-old pig near the transition to duodenal mucosa. Numerous gastrin cells (DAKO LSAB®2 System, HRP; DAB, counterstained with haema-toxylin, x 40) Figure 5: Pyloric mucosa of a three-week-old pig near the transition to duodenal mucosa - cells positive for somatostatin. DAKO LSAB®2 System, HRP; DAB, counterstained with haematoxylin, x 40 combinations of somatostatin with either gastrin or serotonin, we found both light and dark brown cells. This was similar to the variability in intensity of the gastrin-cell reaction product and could, in our opinion, be representative of different forms of the gastrin molecule. The number of gastrin cells in the cardiac and fundic mucosa were statistically insignificant, averaging 0.61 ± 0.14 cells per mm2 and 1.14 ± 0.11 cells per mm2, respectively (Figures 1, 2 and 3). A subjective assessment of the numbers of gastrin cells revealed by both the PAP and DAKO LSAB®2 System immunohistochemical methods and a comparison of them indicated that the number of positive cells revealed by the latter was somewhat higher. This is entirely understandable, as it is much more sensitive of the two systems (Figures 1 and 2). The number of gastrin cells was significantly higher in the pyloric mucosa, both in the antrum and the canal, and 20 P. Juntes, M. Demšar Figure 6: Pyloric mucosa of a three-week-old pig near the transition to duodenal mucosa - cells positive for serotonin. DAKO LSAB®2 System, HRP; DAB, counterstained with haematoxylin, x 40 Table 2: Mean values (± SE) for the number of cells, positive for gastrin, serotonin or somatostatin per unit of gastric mucosa (visual field) in pyloric canal of three weeks old pig, counted on consecutive serial sections, stained with toluidin blue (1), in Grimelius silver staining (3) and immunohis-tochemical reaction (IHC) for gastrin (2), serotonin (4) and somatostatin (5) Number of serial section Method / reaction for Number of cells per mm2 of mucosa ± SE 1 toluidin blue 51,15 ± 2,30 2 gastrin (IHC) 141,94 ± 5,53 3 Grimelius silver staining 153,5 ± 0,32 4 serotonin (IHC) 17,7 ± 2,30 5 somatostatin (IHC) 50,7 ± 2,76 in the area of the small curvature where cutaneous mucosa merges with the pyloric types of mucosa. The average values for the gastrin cells in these areas differed significantly. In the antral mucosa the average number of cells was 124.66 ± 4.54 per mm2, whereas the average number of gastrin cells in the pyloric canal was 140.32 ± 5.30 cells per mm2 and in the mucosa of the small curvature the average was 29.21 ± 2.97 cells per mm2 (Table 1). The highest concentration of gastrin cells was in the middle third of the pyloric mucosa, where the necks of the glands and the deep parts of gastric pits are located. The number of gastric cells in this part of the antral mucosa averaged 89.31 ± 7±.58 cells per mm2 and 214.30 ± 8.61 cells per mm2 in the mucosa of the canal. In relative terms that meant that 28.08 % of the gastrin cells in the antrum and 26.21 % of the gastrin cells in the canal were located in the basal part of the pyloric mucosa. The middle third of mucosa contained 50.62 % of all the gastrin cells in the antrum and 50.67 % of the gastrin cells in the canal; and the remaining 21.30 % of cells in the antral mucosa and 23.15 % of gastrin cells in the canal where located in the luminal third. This distribution pattern dissipated towards the duodenum where the cells were evenly distributed throughout all parts of the mucosa (Figure 4). In this area we also found cells that were immunoreactive to serotonin and somatostatin, which were evenly distributed among the gastrin cells (Figures 5 and 6). With a few exceptions the numbers of enteroendocrine cells revealed by the different staining methods used on the serial tissue sections of pyloric mucosa differed significantly. The exceptions were the numbers of serotonin-immunoreactive cells and the number of Morphometriacal analysis of gastrin cells in the gastric mucosa of three-week-old pigs (Sus scrofa domesticus) 21 orthochromatic cells revealed by the toluidine-blue staining, and between the numbers of metachromatic cells in the toluidine-blue staining and in most, but not all, of the number of dark-brown cells revealed by the silver staining (Table 2). With a comparison of the pictures taken from the serial sections it became obvious that the dark cells revealed by the silver staining were the same as the serotonin-immuno-positive cells. We were unable to confirm the same for the light-brown silver-stained cells and the gastrin cells. Orthochromatic cells were the same in both the toluidine-blue and the silver-stained cells. The results of the morphometric analysis of the distribution of gastrin cells in the gastric mucosa of the three-week-old pigs and the comparison with the localisation and distribution of other enteroendocrine cells of the gastric mucosa added new information to the existing body of data (25). The results could also be compared with data from some publications discussing the physiological role of gastrin in the postnatal development of gastric mucosa in pigs (25, 26, 27, 28, 29). They also provide an insight into the compar-itive sensitivities of some of the older non-specific methods for demonstrating enteroendocrine cells (Grimelius silver staining, toluidine blue for ente-rochromaffin cells) and the more specific immunohistochemical methods. The results of this study could provide a strong basis for further studies of the physiological and pathological processes of the digestive tract of the pig, given by the intensive postnatal morphological and functional changes in the digestive tract of this animal species. These are controlled and influenced by many factors (30) that could negatively affect the development of some or most of the segments of the digestive tract and enable conditions that are suitable for the development of pathological lesions in the stomach and intestines. References 1. Rigler L. Anatomija domačih živali. Splanhno-logija. Ljubljana: Univerza v Ljubljani, 1970: 138-46. 2. Durič DS. Osnovi neuroendokrinologije. Beograd: Zavod za udžbenike i nastavna sredstva 1985: 539 pp. 3. Pearse AGE. Histochemistry. Vol 2. 3 ed. Edinburgh: Churchill Livingstone, 1972: 1111-27. 4. Grimelius L, Wilander E. Silver stains in the study of endocrine cells of the gut and pancreas. Invest Cell Pathol 1980; 3: 3-12. 5. Parsons ME. Histamine and the pathogenesis of duodenal ulcer disease. Gut 1985; 26: 1159-64. 6. Rubin W. Endocrine cells in the normal human stomach. Gastroenterology 1972; 63 (5), 784-800. 7. Rooney PJ, Grennan D, Miller J. Gastrin: a review. Current Med Res Opin 1974; 2(5): 295-304. 8. Elouauer-Blanc L, Sobhani I, Ruszniewski P, et al.. Gastrin secretion by gastrinoma cells in long term culture. Am J Physiol 1988; 18: G569-G602. 9. Tanaka H, Ono K. Ultrastructure and immuno-histochemistry of antral gastrin cells in rats at birth. Z Mikrosk-Anat Forch 1987; 101(4): 609-18. 10. Hakanson R, Alumets J, Rehfeld JF, Ekelund M, Sundler F. The life cycle of the gastrin granule. Cell Tissue Res 1982; 222: 479-91. 11. Azuma T, Abrahm D, Walsh H. Immunocyto-chemical evidence for differential distribution of gastrin forms using regional-specific monoclonal antibodies. Gastroenterol Japon 1986; 21: 4: 319-24. 12. Feldman M, Walsh JH, Wong HC. Role of gas-trin heptadecapeptide in the acid secretory response to amino acids in man. J Clin Invest 1978; 61: 308-13. 13. Solcia E, Capella C, Vassalo G. Endocrine cells of gastric mucosa. Int Rew Cytol 1975; 42: 223-86. 14. Stumpf WE, Sar M, O'Brien LP, Morin J. Pyloric gastrin-producing cells and pyloric sphincter muscle cells are nuclear targets for 3H 1,25(OH)2 vitamin D3. Histochemistry 1988; 89: 447-50. 15. Tielemans Y, hakanson R, Sundler F, Willems G. Proliferation of enterochromaffine like cells in ome-prazole-treated hypergastrinemic rats. Gastroenterology 1989; 96: 723-9. 16. Christiansen LA, jensen SL, Rehfeld JF, Stadil F, Christensen KC. Antral content and secretion of gastrins in pigs. Scand J Gastroenterol 1978; 13(6): 719-25. 17. Jonsson A-C, Dockray GJ. Immunohistochemi-cal localization to pyloric antral G cells of peptides derived from porcine preprogastrin. Regulat Peptid 1984; 8: 283-90. 18. Holst JJ, Knuhtsen S, Orskov C, Skak-Nielsen T, Poulsen SS, Nielsen OV. GRP producing nerves control antral and somatostatin and gastrin secretion in pigs. Am J Physiol 1987; 16(6): G767-74. 19. Knuhtsen S, Holst JJ, Knigge U, Olesen M, Nielsen OV. Radioimmunoassay, pharmacokinetics and neuronal release of gastrin releasing peptide in anesthetized pigs. Gastroenterology 1984; 87: 372-8. 20. Bullock GR, Petrusz P. Techniques in immuno-histochemistry. Vol 1. London: Academic Press, 1982: 306 p. 21. Grimelius L, Wilander E. Silver stains in the study of endocrine cells of the gut and pancreas. Invest Cell Pathol 1980; 3: 3-12. 22. Gabe M. Histological techniques. Paris: Masson Springer Verlag, 1976: 763. 23. Shu S, Ju G, Fan L. The glucose oxydase-DAB-nickel metzhod in peroxidase histochemistry of the nervous system. Neurosc Lett 1988; 85: 169-71. 24. Bavdek S. Mikroskop in mikroskopiranje. Ljubljana: Univerza v Ljubljani, 1972: 62-6. 25. Bussolati G. Immunofluorescent localization of the gastrin-secreting G cells in the pyloric antrum of the pig. Histochemie 1970; 21: 1-4. 26. Xu RJ, Yuan-Li M, Man-Yin WT. Stability of ga-strin in the gastrointestinal lumen of suckling, weanling and adult pigs. Biol Neonate 1996; 70: 60-8. 27. Xu RJ, Cranwell PD. Gastrin metabolism in 22 P. Juntes, M. Demšar neonatal pigs and grower-pigs. Comp Biochem Physiol 1992; 101A: 177-82. 28. Xu RJ, Thngthanathanich P, Birtles MJ, Mellor DJ, Reynolds GW, Simpson HV. Growth and morphological changes in the stomach of newborn pigs during the first three days after birth. J Develop Physiol 1992; 17: 7-14. 29. Xu RJ, Cranwell PD. Development of gastric acid secretion in pigs from birth to thirty six days of age: the response to pentagastrin. J Dev Physiol 1990; 13(6): 315-26. 30. Xu RJ. Developmnet of newborn GI tract and its relation to colostrum/milk intake: a review. Reprod Fertil Dev 1996; 8(1): 35-48. MORFOMETRIČNA ANALIZA GASTRINSKIH CELIC V ŽELODČNI SLUZNICI TRI TEDNE STARIH PRAŠIČEV (Sus scrofa domesticus) TER PRIMERJAVA Z NEKATERIMI DRUGIMI ENTEROENDOKRINIMI CELICAMI V ŽELODCU P. Juntes, M. Demšar Povzetek: V sluznici želodcev tri tedne starih prašičev smo določali razporeditev in število enteroendokrinih celic v posameznih delih želodca. Izhodiščna za prikaz enteroendokrinih celic je bila imunohistokemična metoda, primerjalno pa smo enteroendokrine celice barvali s toluidinskim modrilom za tovrstne celice in jih srebrili po Grimeliusu. Največ gastrin-skih celic smo ugotovili v vratarjevi sluznici, v antrumu 124,66 ± 4,54 gastrinske celice na mm2 in v kanalu 140,32 ± 5,30 na mm2. Največja gostota gastrinskih celic, glede na višino sluznice, je bila v srednji tretjini, kjer je bilo približno 50 % vseh tovrstnih celic, preostale pa so bile dokaj enakomerno razporejene v zgornji in spodnji tretjini sluznice. Na področjih s kutano sluznico ni bilo niti gastrinskih niti drugih enteroendokrinih celic. Na mali krivini želodca, kjer se prekrivajo področja brezžlezne in vratarjeve sluznice, je bilo v povprečju 29,21 ± 2,97 gastrinske celice na mm2, na področjih s kardialno sluznico in sluznico pravih želodčnih ali fundusnih žlez pa je bilo njihovo število še manjše: 0,61 ± 0,14 celice na mm2 kar-dialne sluznice ter povprečno 1,14 ± 0,11 gastrinske celice na mm2 fundusne sluznice. Na področju vratarja, kjer je največ gastrinskih celic, smo morfometrično določili tudi število serotoninskih in somatostatinskih celic. Obeh vrst celic je bilo v sluznici tega področja značilno manj kot gastrinskih celic, in sicer serotoninskih 17,7 ± 2,30 na mm2 sluznice ter somatostatinskih 50,7 ± 2,76. Rezultati srebrenja in barvanja s toluidinskim modrilom so se le deloma ujemali z rezultati imuno-histokemične reakcije. Celic, ki so se v vratarjevi sluznici pobarvale s toluidinskim modrilom, je bilo bistveno manj kot smo jih ugotovili z imunohistokemično metodo (51,15 ± 2,20 celice na mm2), medtem ko je v postopku srebrenja reagiralo bistveno več celic, čeprav še vedno manj kot pri imunohistokemični reakciji (153,5 ± 0,32 celice na mm2). Ključne besede: veterinarska medicina; želodec; enteroendokrine celice; gastrin; somatostatin, serotonin; imunohis-tokemija; srebrenje po Grimeliusu; toluidinsko modrilo; prašič Slov Vet Res 2004; 41 (1): 23-9 UDC 619:616.94:582.231:637 Original research paper THE ROLE OF AEROMONAS HYDROPHILA BACTERIUM AS A CAUSATIVE AGENT OF SEPTICAEMIA IN DOGS Irena Zdovc 1, Tanja Švara 2, Polona Juntes 2, Tina Kotnik 3, Matjaž Ocepek 1, Manica Černe 2, Milan Pogačnik 2 Addresses of authors: 11nsitute for Microbiology and Parasitology, 2 Institute for Patology, Forensic Administrative Veterinary Medicine, 3 Clinic for Surgery and Small Animals, Veterinary Faculty, Gerbičeva 60, 1000 Ljubljana, Slovenia Corresponding author. E-mail: irena.zdovc@vf.uni-lj.si Summary: Aeromonas hydrophila is an organism commonly found in water, particularly during the warm season, though less commonly isolated as a pathogen in domestic animals and humans. The course of septicaemia caused by Aeromonas hydrophila in puppies is described to illustrate its bacterial pathogenic activity and the diagnostic procedures used. Seven of eight puppies, all clinically healthy at birth, died within ten days of birth, after receiving the same clinical course. Beside the gross pathology and histopathology, the internal organs were also bacteriologically examined. The puppies had diffuse acute fibrinous and necrotic bronchopneumonia. Infection with A. hydrophila, which was isolated from the liver, spleen, lungs and intestines, was determined as the cause of the sepsis and the consequent death. To establish the origin of the infection, cultures of milk, vaginal and rectal swabs taken from the dam were made and they were all negative for this bacterium. While the origin of the infection remains unknown, the underdevelopment of the puppies is thought to be a predisposing factor. Key words: dog; septicaemia; bacterial pneumonia; Aeromonas hydrophila Introduction Aeromonas hydrophila is a non-fastidious, gramnegative rod-shaped, motile bacterium. The 1974 edition of Bergey's Manual includes it in the family Vibrionaceae (1). Its morphological and cultural characteristics are the same as is indicated for Aeromonas caviae, therefore the differentiation is difficult. As the species' name hydrophila ("water lover") indicates, the natural habitat of the microorganism is both fresh and sea water. It has been observed in numerous species of freshwater fish, the occasional salt-water fish, and in amphibians, reptiles, cattle and humans all over the world. Aeromonas hydrophi-la is recognized as an opportunistic pathogen or a secondary invader (2, 3, 4). There are some reports that describe the role of Aeromonas hydrophila in the pathology of mammals, however, most of them relate to humans (5, 6, 7) and there only a few describing its role in dogs (8, 9). Received: April, 2004 Accepted for publication: June, 2004 The aim of this work is to show that infection with A. hydrophila can be severe enough to be considered a cause of death in mammals with weak immune responses. Material and methods Eight puppies were examined at birth and, at first, all appeared to be normal, healthy and strong. About 30 hours later, the first puppy showed signs of illness: it stopped sucking, became hypothermic and debilitated and died within 24 hours. All the puppies, except for one, developed the same symptoms and despite intensive care - warming, oxygen supplementation, supplemental feeding and glucose infusions - successively died within ten days. The bitch, a five-year-old Doberman pinscher, in good general condition was admitted to the veterinary clinic due to difficulties in parturition. The act began on the 60th day of pregnancy and was run with caesarean section because of dysto-cia. During the operation the bitch was treated 24 I. Zdovc, T. Švara, P. Juntes, T. Kotnik, M. Ocepek, M. Černe, M. Pogačnik Figure 1: Lung. Acute diffuse fibrinous bronchopneumonia with multifocal areas of necrosis and haemorrhages with antibiotics amoxycillin+clavulanic acid (Amoksiklav, Lek) and gentamicin (Gentamicin, Lek). Two puppies were submitted for post-mortem examinations immediately after death: a female (body weight 400 g) and a male (body weight 500 g). A necropsy was performed and several tissue specimens were taken for further laboratory examinations. For the histological examinations, tissue samples from the brain, kidneys, heart, liver, lungs, spleen and intestines were fixed in a 10 % buffered formalin, routinely processed in paraffin and then stained with haematoxylin and eosin (HE). The lung and spleen samples were also treated with Grocott's methenamine silver staining and periodic acid Schiff reaction (to exclude mycotic infections), as well as Goodpasture's stain method (gram staining for tissues). Cryostatic tissue sections of the lungs, liver and kidneys were stained with Sudan III for fat. Imprints of the pleural surface were prepared for cytology, air dried, fixed in methanol and stained with Giemsa. Samples for bacteriological examinations were taken from the liver, spleen, lungs and intestines. The material was inoculated on nutrient agar (Oxoid) supplemented with 5 % of ovine blood (BA) and Drigalski agar (DA) and incubated at 37 °C for 24 h, and on Sabouraud Dextrose Agar (SDA, bioMerieux, France) at 37 °C for five days. Subcultures for A. hydrophla identification were made on the BA. Simultaneously, the milk of the dam was examined and a few days later rectal and vaginal swabs were also taken from the dam. Bacteria that grew on the culture media were Gram stained (Difco-BBL) and tested for catalase and oxidase activity (Difco-BBL). The biochemical characteristics of the isolated bacteria were tested using classical biochemical tests and the Api 20NE, Api 20E and API Staph commercial systems (bioMerieux, France) in accordance with the manufacturer's instructions. Results The post-mortem findings of both animals were similar. The thoracic cavity was filled with a small amount of a serofibrinous, opaque inflammatory exudate. There were pulmonary lesions characteristic of acute fibrinous and necrotic bronchop-neumonia with acute fibrinous pleuritis (Fig. 1). The pleura was covered with a thick layer of fibrin, the lung texture was firm, and the majority of the pulmonary tissue was heavily congested and oedematous with multifocal grey areas of necrosis on the pleural and cut surfaces. Both the liver and the kidneys were enlarged and congested, the spleen was enlarged and the intestines displayed acute catarrhal enterocolitis. Microscopic lesions of the lung were consistent with the gross pathology findings - acute fib- The role of Aeromonas hydrophila bacterium as a causative agent of septicaemia in dogs 25 Figure 2: Liver. Many extrame-dullary haematopoietic foci and few macrophages located throughout the liver tissue and in portal areas. Haematoxylin & eosin, x 10 Figure 3: Culture of Aeromonas hydrophila with a large zone of haemolysis on 5 % ovine blood agar after 48 h incubation at 37 °C rinous bronchopneumonia with multifocal areas of coagulative necrosis and acute fibrinous pleuritis. Several bronchi and bronchiole had necrotic walls and were filled with desquamated epithelial cells, numerous macrophages and neutrophils. The alveoli were distended and filled with large quantities of fibrin and inflammatory cells; mostly macrophages and some neutrophils. There were also large areas of multifocal coagulative necrosis. In many parts of the lung, we found numerous small rod-shaped gram-negative bac- teria. Bacterial colonies of gram-negative rods were especially numerous and large in necrotic areas and were growing towards the periphery of such areas. In one puppy there were also large areas of multifocal haemorrhages. The intersti-tium around blood vessels was distended by oedema and contained a few macrophages, mono-cytes and neutrophils. Grocott's and PAS staining of the lungs and spleen established that they were negative for mycotic infectious agents. Goodpasture staining also established that the 26 I. Zdovc, T. Švara, P. Juntes, T. Kotnik, M. Ocepek, M. Černe, M. Pogačnik Figure 4: Gram stain of Aeromonas hydrophila ilu-strating gram-negative rods spleen was negative. The portal areas of the liver and interstitium around hepatic veins were oede-matous, lymph vessels in the portal areas were distended, and portal tissues were infiltrated with numerous macrophages and a few eosinop-hils. Macrophages were also found scattered within the sinusoids. There were swollen hepato-cytes as well as some disseminated necrotic he-patocytes, and karyolysis and cholestasis were also present. The Kupffer cells were swollen and many of them were in mitosis. There were many extramedullary haematopoietic foci with erythro-poietic cells and megakaryocytes located throughout the liver tissue (Fig. 2). The kidneys were underdeveloped for the age of animals. They had foetal cortices with numerous mitoses of epithelial cells in the nephrons, and small, hy-perchromatic glomerular cells forming palisades. The tubular cells were vacuolated and many revealed small accumulations of calcium salts within their cytoplasm due to tubular degeneration. The interstitium was distended by oedema. Erythrocytes in small vessels formed hyalinized cylindrical structures without obvious boundaries between cells, which was diagnosed as disseminated intravascular coagulation. The white and the red pulp of the spleen were not clearly separated. Lymphopenia was evident and a few necrotic lymphocytes were found in some follicles. The red pulp was highly cellular with diffuse extramedullary haematopoiesis, and the red pulp was congested. The epithelial lining of the small and large intestine was largely desquamated and the ratio between the intestinal villa and crypts in jejunum was approximately 4:1. Between the epithelial cells and the lamina propria in non-desquamated areas of mucosa there was a vac-uolated layer of severe oedema; the propria was also oedematous and infiltrated with a small number of eosinophils. The intestinal lumen contained desquamated cells, an amorphous mass with several haemosiderin granules, and many bacteria forming colonies (mostly short rods and small cocci). In one puppy, the intestinal mucosa was strongly congested with evidence of haemor-rhaging into the intestinal lumen. Lesions in other tissues were mostly congestion. Imprints of the pulmonary surface contained numerous short rod-shaped bacteria, degenerate macrophages - some of them containing bacteria in their cytoplasm, and a few neutrophilic granulocytes, erythrocytes and mesothelial cells (acute septic pleuritis). Bacteriological cultures of the organs yielded abundant growth of large colonies (2-3 mm) that were flat, greyish, circular and convex with an entire margin and surrounded by a large zone of beta-haemolysis (Fig. 3). Gram-stained cultures demonstrated gram-negative rod-shaped bacteria (Fig. 4). A presumptive diagnosis of an Aeromonas The role of Aeromonas hydrophila bacterium as a causative agent of septicaemia in dogs 27 species was based initially on a positive oxidase reaction and additionally on the fermentation of carbohydrates. Our isolate produced both acid and gas from glucose and acid from arabinose, manitol, sucrose and maltose, but not from inosi-tol or lactose. The numerical profile of the biochemical reactions in the Api 20NE system was 7577754. Aeromonas hydrophila was grown in an entirely pure culture, except with some rare colonies of non-haemolytic Escherichia coli in the intestines. Bacteriological examinations of all the samples taken from the dam were negative for A. hydrophi-la. The vaginal swab yielded only a few colonies of Staphylococcus haemolyticus, which belongs to the group of coagulase-negative staphylococci. The bacteriological culture of the dam's rectal swab yielded non-haemolytic Escherichia coli and alpha-haemolytic streptococci. Discussion The sudden death of seven, out of eight, newborn puppies in such a short period after parturition can have many causes but an intensive infection with Aeromonas hydrophila was probably the fatal one. The gross pathology, the histological determination of numerous short rod-shaped bacteria in many tissues and the isolation of this bacterium in a pure culture indicated an acute septic condition. Aeromonas hydrophila is a part of the normal flora of freshwater fish and is commonly present in fish ponds and tanks (10). Occasionally it can cause infections in humans, which range from soft-tissue infections, pneumonia, endocarditis and gastroenteritis to septicaemia (4, 11, 12). Cases of hospital infections with A. hydrophila were reported in humans as well (13). In dogs, A. hy-drophila was demonstrated in a few cases as an aetiologic agent of disease, usually in young adults where it was considered an opportunistic pathogen (8, 9). That means that some other stressful factor should have been present at first. Animals can be faecal carriers of Aeromonas spp. (14). Ghenghes and others presented an interesting study on the presence of the Aeromonas species in domestic dogs and cats. They found that Aeromonas are not uncommon in healthy dogs and cats. Furthermore, they found this organism occurred in the Doberman breed to a higher de- gree than in other breeds. They emphasized that haemolysin-producing Aeromonas species in the faeces of healthy domestic dogs and cats may present a public health problem for humans who came into contact with them. Numerous opportunistic bacteria can cause septicaemia in susceptible neonates. There are a few reports of neonatal A. hydrophila septicaemia in children that are comparable to our case. In these reported cases, the children were born at term and no signs of immaturity were present (15, 16). The aetiopathogenesis of infection in our case allows some speculation. Which factor facilitated the development of the extensive pneumonia and sudden death involving a microorganism of inherently low pathogenicity for mammals? The owner initially reported that the puppies were born at term. According to the structure of the kidney cortex and the diffuse, extensive ex-tramedullary haemopoiesis in the spleen and liver at the age of seven days, we estimated that they were underdeveloped. Therefore immaturity could be a possible factor contributing to susceptibility for the infection and the fatal exit (17). Common canine viral infections, which can be an underlying factor for the secondary infection were ruled out as a predisposing factor because there were no morphological or histological lesions characteristic of parvoviral, distemper or herpes virus infections, and besides that the dam was vaccinated against them on a regular basis. The pathomorphological lesions in both puppies were predominantly located in lungs. The diagnoses of pneumonia in children with A. hydrophila septicaemia were made after clinical and x-ray examinations only. There is no information on pathomorphological and histopathological changes in the lungs of children with septicaemia (15, 16). Pneumonia is frequently diagnosed in cases of A. hydrophila septicaemia (11, 18). In a study of fifteen cases of A. hydrophila septicaemia, extensive bilateral pulmonary lesions were found in more than half of the patients (13). The pneumonia in our case is, according to the histopathological lesions, similar to A. hydrophila pneumonia described in man (11). No pulmonary lesions were noted in dogs with A. hydrophila septicaemia (8, 9). Histopathological lesions in the lung can be partly the consequence of the aspiration of milk that occurs during supplemental feeding. But in our case that happened after the puppies devel- 28 I. Zdovc, T. Švara, P. Juntes, T. Kotnik, M. Ocepek, M. Černe, M. Pogačnik oped clinical signs and was just a factor contributing to the quicker course of the disease. Infections of the neonate may be acquired from a vaginal flora during parturition, through penetrated skin, a contaminated umbilicus or from the environment. No skin or umbilical lesions were found in any of puppies. Due to the A. hydrophi-Za-negative results from all the samples taken from the dam, we don't believe that the dam was a carrier of this bacterium. On the other hand, it should be noted that the dam was on a prolonged antibiotic therapy, which could have changed intestinal bacterial flora by the time the rectal swabs were taken for bacteriology. This could explain why the A. hydrophila cultural examinations were unable to determine the presence of the bacterium. Looking for other sources of infection we checked for the presence of other animals in their household that are known as carriers of A. hydrophila. The owners had no other animals or reptiles at all. Contaminated water or food, or even the hospital environment might have been the source of infection in this case but unfortunately we were not able to check into these possibilities. Aeromonas hydrophila has been implicated as a cause of gastroenteritis in humans and ingestion of contaminated water is another possible point of access for A. hydrophila into the intestine. The source of the puppies infection seems unlikely to be determined now since the material, taken from the dam, was negative to A. hydrophila. But the reason for the negative results could also be the preventative antibiotic treatment of the bitch after the caesarean section. The dam (but not the puppies) was treated with a combination of amoxycillin + clavulanic acid and gentamicin, the combination that is usually successfully used when dealing with A. hydrophila in fish (19). But in the present case the rapid progress of the disease was unfortunately fatal for all the puppies. According to the owner's data the dam had no further problems with infection. Six months after parturition, another rectal swab was taken and was also negative for Aeromonas. To the best of our knowledge this is the first well document case of severe pneumonia caused by Aeromonas hydrophila in newborn puppies and their consequential sudden death with almost no chance of a successful treatment. Infection or contamination of domestic animals with A. hydrophila can be considered as a health risk for animals as well as humans, especially those in the early (neonatal and perinatal) and most-sensitive periods of life, particularly underdeveloped or premature individuals, and those with an impaired immune system. References 1. Bergey's manual of systematic bacteriology. Vol 1. Baltimore: Williams & Wilkins, 1984: 546-8. 2. Pearson TA, Mitchell CA, Hughes WT. Aeromonas hydrophila Septicemia. Amer J Dis Chil 1972; 123: 579-82. 3. Austin B, Allen-Austin D. Bacterial pathogens of fish. J Appl Bacteriol 1985; 58: 483-506. 4. Gold WL, Salit IE. Aeromonas hydrophila infections of skin and soft tissue: report of 11 cases and review. Clin Infect Dis 1993; 16: 69-74. 5. Paniagua C, Arguello-Villares JL, Arias MA, Herreros M. Aeromonas hydrophila associated with a severe outbreak of infection in farmed rabbits. Zentralbl Hyg Umweltmed 1998; 201: 423-30. 6. Deodhar LP, Saraswathi K, Vardukar A. Aeromonas spp. and their association with human diarrhoeral disease. J Clin Microbiol 1991; 29: 853-6. 7. San Joaquin V, Pickett DA. Aeromonas associated gastroenteritis in children. Pediat Infect Dis J 1988; 7; 53-7. 8. Pierce RL, Daley CA, Gates CE, Wohlgemuth K. Aeromonas hydrophila Septicemia in a Dog. J Am Vet Med Assoc 1973; 6: 469. The role of Aeromonas hydrophila bacterium as a causative agent of septicaemia in dogs 29 9. Andre-Fontaine G, Monfort P, Buggin-Daubie M, Filloneau C, Ganiere JP. Fatal disease mimicking lep-tospirosis in a dog, caused by Aeromonas hydrophila. Comp Immun Microbiol Infect Dis 1995; 18: 69-72. 10. Monfort P, Baleux B. Haemolysin occurrence among Aeromonas hydrophila, Aeromonas caviae, and Aeromonas sabria strains isolated from different aquatic ecosystems. Res Microb 1991; 142: 95-102. 11. Goncalves JR, Brum G, Fernandes A, Biscaia I, Correia MJS, Bastardo J. Aeromonas hydrophila fulminant pneumonia in a fit young man. Thorax 1991; 47: 482-3. 12. Krovacek K, Conte M, Galderisi P, Morelli G, Postiglione A, Dumontet S. Fatal septicaemia caused by Aeromonas hydrophila in a patient with cirrhosis. Comp Immunol Microbiol Infect Dis 1993; 16: 267-72. 13. Picard B, Arlet G, Goullet P. Aeromonas hydrophila septicemia. Epidemiologic aspects: 15 cases. Presse Med 1984; 19: 1203-5. 14. Ghenghesh KS, Abeid SS, Jaber MM, Ben-Taher SA. Isolation and haemolytic activity of Aeromonas species from domestic dogs and cats. Comp Immunol Microbiol Infect Dis 1995; 22: 175-9. 15. Gupta P, Ramachandran VG, Seth A. Early onset neonatal septicemia caused by Aeromonas hydrophila. Indian Pediatr 1996; 33: 703-4. 16. Verghese SL, Raju BB, Ramanath A, Raghu MB, Siyakumar A. Septicaemia caused by Aeromonas hydrophila in a neonate. J Assoc Phys India 1994; 42: 909-10. 17. Quinn PJ, Donnelly WJC, Carter ME, Markey BKJ, Torgeson PR, Bteathnach RMS. Male and female reproductive systems. In: Microbial and parasitic diseases of the dog and cat. London: WB Saunders Company, 1997: 207-12. 18. Takano Y, Asao Y, Kohri Y, Oida K, Imanaka T. Fulminant pneumonia and sepsis due to Aeromonas hydrophila in an alcohol abuser. Intern Med 1996; 35: 410-2. 19. Quinn PJ, Carter ME, Markey B, Carter GR. Aeromonas, Plesiomonas and Vibrio species. In: Clinical veterinary microbiology. London:Mosby-Year Book Europe limited, 1994: 243-7. VLOGA BAKTERIJE AEROMONAS HYDROPHILA KOT POVZROČITELJICE SEPTIKEMIJE PRI PSIH I. Zdovc, T. Švara, P. Juntes, T. Kotnik, M. Ocepek, M. Černe, M. Pogačnik Povzetek: Aeromonas hydrophila je mikroorganizem, ki ga običajno najdemo v vodi, predvsem v toplejših obdobjih, zelo redko pa ga izoliramo kot povzročitelja bolezni pri domačih živalih in ljudeh. V prispevku je prvič natančno predstavljen primer patogenega delovanja bakterije Aeromonas hydrophila pri pasjih mladičih. Opisani so klinični potek septikemije, spremembe na notranjih organih in diagnostični postopki v vseh fazah preiskave. Osem novorojenih mladičev, ki so bili po porodu klinično zdravi, je v naslednjih desetih dneh postopoma poginilo z enakimi kliničnimi znaki bolezni. Poginule živali smo pregledali patoanatomsko in patohistološko, notranje organe pa tudi bakteriološko. Ugotovlili smo difuzno akutno fibrinozno in nekrotično bronhopneumonijo. V notranjih organih (jetrih, vranici, pljučih in črevesju) smo z mikroskopsko preiskavo ugotovlili številne identične gramsko negativne paličke, z gojiščno preiskavo pa je bila izolirana bakterija Aeromonas hydrophila. Izolacija čiste kulture, difuzna rast iz vseh pregledanih notranjih organov ter ujemanje izolata s histološkimi in kliničnimi spremembami pomeni potrditev diagnoze, da je bila bakterija povzročiteljica septikemije in posledičnega pogina mladičev. Zaradi ugotavljanja vira okužbe smo pregledali še vaginalni in rekatalni bris ter vzorec mleka psice. V nobenem vzorcu nismo ugotovili bakterije Aeromonas hydrophila, kar pripisujemo dejstvu, da je bila psica po porodu zdravljena z antibiotikom, za katerega je bila bakterija zelo dobro občutljiva. Najbolj verjeten vzrok za razvoj sepse pri mladičih je bila njihova nerazvitost ob porodu in velika dovzetnost za okužbo v zgodnjem obdobju, ko imunski sistem še ni opravljal svoje vloge. Ključne besede: pes; septikemija; bakterijska pljučnica; Aeromonas hydrophila Slov Vet Res 2004; 41 (1): 31-4 UDC 619:618.19-002:615.33 Original research paper TREATMENT OF SUBCLINICAL STAPHYLOCOCCAL MASTITIS Ožbalt Podpečan 1*, Andrej Pengov 2, Uroš Hrastnik 3 Addresses of authors: 1 Savinjska Veterinary Station, Celjska c. 3/a 3310 Žalec, Slovenia; 2 Veterinary Faculty, Gerbičeva 60, 1000 Ljubljana, Slovenia; 3 Veterinary Surgery, Farma d.o.o., Podlog 1, 3311 Šempeter, Slovenia ""Corresponding author. E-mail: svp.zalec@siol.net Summary: The purpose of this study was to evaluate the efficacy of a treatment of subclinical mastitis in dairy cows, caused by the Staphylococcus aureus bacteria strain. In both Europe and the USA S. aureus is the most frequently isolated pathogen present in dairy cows suffering from subclinical mastitis. Studies have shown that amoxycillin, in its own right, is not sufficiently effective in combating S. aureus, however, when used in conjunction with clavulanic acid its efficacy improves significantly. The infected animals were treated with Synulox®, which contains amoxycillin and clavulanic acid. In accordance with the manufacturer's instructions, the treatment was applied intramuscularly and intramammarily. In total, 61 mammary glands of 37 cows were treated. On average, the bacteriological efficacy of the treatment was 51.3 %. In animals with only one infected mammary gland the efficacy was 69.9 %. Considering that such animals represented 56.7 % of all the animals included in the study, we suggest that treating animals with only one or two infected mammary glands with Synulox® is sensible and economically justifiable. Key words: veterinary medicine; mastitis-treatment; Staphylococcus aureus; amoxycillin; clavulanic acid Introduction The subclinical form of mastitis in dairy cows represents a significant problem in contemporary milk production, as it is associated with lower productivity and an increase in the somatic-cell count (SCC) in milk. Milk with an elevated SCC is of a lower quality due to an alteration in the quantity of the single ingredients (fat, protein, lactose, and minerals). The efficacy of subclinical-udder-infection treatments mainly depends on the species of the infectious agent and the duration of the infection. Other factors, such as the age of the animal, the preparation of the udder before milking and nutrition, also play an important role in the outcome of a mastitis treatment (2, 4). While it is true that the percentage of animals suffering from subclinical mastitis in Slovenia is decreasing (31.3 % in 1990 compared to 21.9 % in 1997), the efficacy of treatments of such infections is also diminishing. This is as a result of the increasing resistance of microorganisms to antibiotics, and to the weaker immune systems of the animals due to increased milk production and management failures (1). Received: April, 2004 Accepted for publication: June, 2004 Staphylococcus aureus (S. aureus) is, besides coagulase-negative Staphylococcus species, the most frequently isolated pathogen present in dairy cows suffering from subclinical mastitis in both the USA and Europe (2). On average, such infections in Slovenia account for half of all cases, however, this varies significantly within individual herds (1). In this study the S. aureus bacteria was more prevalent in the mammary glands of the older animals. The outcome of a treatment of a mammary gland infection caused by S. aureus is very uncertain and is influenced by the age of the animal, the clinical type of infection (acute vs. chronic), the sensitivity of the infectious agent to certain antibiotics and the duration of the treatment. In cases of chronic infection, the ability of the antibiotics to penetrate the affected tissue is poor. The fact that S. aureus can also be present intracellularly and in micro-abscesses represents an additional problem. In those areas it is particularly hard to achieve the appropriate concentrations of the antibiotic (3, 4). The sensitivity of S. aureus to various antibiotics has diminished over time (1, 5), while the share of p-lactamase-positive strains has increased. The percentages vary significantly from country to country and ranges from 4 % in Norway to 76 % in Ireland. Very early in the devel- 32 O. Podpecan, A. Pengov, U. Hrastnik opment of antibiotics an enzyme, which destroys penicillin, was described. The enzyme was termed "penicillinase" (now referred to as p-lactamase) and was found to be produced by a wide variety of bacteria. This still remains the most important method of bacterial defence to the p-lactam antibiotics (penicillins, cephalosporins etc.) (5, 6, 7, 8). The duration of a therapy can also significantly influence the efficacy of the treatment. Acceptable results are achieved through a combination of intramuscular and intramammary applications of appropriate antibiotics over a 3 to 5 day period (9, 10). According to other authors, the rate of success of bacteriological cures for clinical and subclinical mastitis caused by S. aureus, ranges from 15 % to 70 % (2, 7, 9, 10), which indicates the level of difficulty and complexity involved in the approach to treatment. Amoxycillin in combination with clavulanic acid, which is a p-lactamase inhibitor, is one of the antibiotics that are being used with increasing regularity in the treatment of subclinical bovine mastitis. This combination was first successfully applied in human medicine in the treatment of infections caused by p-lactamase-positive strains of S. aureus. Authors agree that amoxycillin in its own right is not sufficiently effective in combating S. aureus (7, 10, 12), however, when used in combination with clavulanic acid the efficacy of a treatment improves significantly. Considering all the aforementioned facts we decided to test the efficacy of the amoxycillin-clavu-lanic acid combination in the treatment of subclin-ical mastitis in dairy cows caused by S. aureus. Material and methods Selection of Animals The study included 37 dairy cows from 11 different herds, of different ages (Graph 1) and breeds, each with an increased SCC in their milk. In all cases, a microbiological test of the milk sample revealed the presence of S. aureus. The average SCC in milk from the infected udder quarters of the selected animals was 1428x103/ml before treatment, and ranged from 210x103/ml to 4057x103/ml. Twenty-one animals had one infected mammary gland, 9 had two, 6 had three and in one case all four mammary glands were infect- ed. In total, 61 mammary glands of 37 cows were included in the study. Treatment Synulox®, which is manufactured by Pfizer Animal Health, was used to treat the infection as follows: • an injector with an intramammary solution containing 50 mg of clavulanic acid in the form of potassium clavulanate, 200 mg of amoxy-cillin in the form of Amoxycillin-Trihydrate and 10 mg of prednisolone. • a solution for an intramammary application where 1 ml contains 35 mg of clavulanic acid in the form of potassium clavulanate and 140 mg of amoxycillin in the form of Amoxycillin-Trihydrate. All the animals included in the study were treated in accordance with the following predetermined protocol: • an application of the Synulox® injector into the affected udder quarter every 12 hours at 6 consecutive milkings; • a parenteral application of the Synulox® solution, in the amount of 8.75 mg/kg, on the first and second days of the treatment with an interval of 24 hours between the doses. Eleven days after the final application of the drug, another milk sample was collected from each of the animals. These samples were subjected to microbiological analyses and used as a control of the treatment's efficacy. Treatment Efficacy Assessment Criteria The efficacy of the treatment was assessed on the basis of a bacteriological examination. For the purpose of the study, a treatment was regarded as having been successful when the result of the bacteriological examination was negative for S. aureus. Results In Table 1 the distribution of cows in relation to the number of S. aureus infected udder quarters is presented. In 57 % (n = 21) of the selected cows only one udder quarter was infected. In 9 cows the infection was present in two, and in 6 cows in three udder quarters. There was only one case where all four quarters were infected. In Table 2 the results of the treatment are pre- Treatment of subclinical staphylococcal mastitis 33 Table 1: Distribution of animals in relation to the number of infected udder quarters before treatment No. of infected quarters per animal 1 2 3 4 TOTAL No. of animals 21 9 6 1 37 No. of infected quarters 21 18 18 4 61 Table 2: Distribution of bacteriologically negative (successfully treated) animals in relation to the number of infected udder quarters before treatment No. of infected quarters per animal 1 n = 21 2 n = 9 3 n = 6 4 n = 1 TOTAL Bacteriological cure rate No. of cured animals 13 5 1 0 19 51.3 No. of cured quarters 14 13 9 0 36 59.0 % of cured animals 61.9 55.5 16.7 0 % of cured quarters quarters m.g 66.6 72.2 50.0 0 12 10 CO 75 8 E c 6 n 0 4 1 2 0 y /— /— /— 11 9 /— 7 • /— 3 4 s—?| ? 1 / HP/ 4 5 6 No of lactation 8 Graph 1: Distribution of infected cows per lactation sented in relation to the number of initially infected udder quarters. In our study we found a significant correlation between the number of successfully treated animals and the number of infected quarters. However, the difference in the percentages of successfully treated udder quarters was less significant. Discussion Synulox® was used in the treatment of 37 animals with one or more mammary glands infected by S. aureus. The treatment protocol described above was selected in accordance with the claims of most authors that an efficient and effective treatment of subclinical and clinical mastitis caused by S. aureus takes at least 3 to 5 days (9.10). The amoxycillin-clavulanic acid combination in the therapy was selected due to the high probability of the presence of ^-lactamase-positive strains of S. aureus in the infected animals (5, 6, 7). There were differing opinions amongst authors regarding the best number of control samples to collect and when to collect them (7, 9, 10), hence, given our circumstances, we settled on one sample collection 14 days following the first application of the drug. While the overall level of success using this bacteriological treatment (51.3 %) does not deviate substantially from claims made in other publications, it is clearly among the more successful methods, particularly as only subclinical types of bovine mastitis were treated (2, 7, 9, 10). The level of success that was achieved in treating animals with only one infected mammary quarter 34 O. Podpečan, A. Pengov, U. Hrastnik (61.9 %) was very promising as they represented the majority (56.7 %) of all the treated animals. The comparative success of the treatment in our case is slightly diminished by the fact that in certain cases (n = 4), despite a successful bacteriological cure, the SCC did not fall below the level set down in the regulations governing the health and hygiene of milk (< 400,000 cells/ml). While most of the infected animals were in their 2nd or 3rd lactation, which represents a lower average compared to some authors (2), we could not confirm a link between the number of lactations and the efficacy of the treatment. Given our results we believe that the use of a combination of amoxycillin and clavulanic acid lived up to our expectations, and that the therapy of animals with only one or two infected mammary glands is sensible and economically justifiable. In our opinion however, it makes no sense to treat animals with three or four infected udder quarters irrespective of the type of therapy. In such cases culling should be seriously considered. So far only a few studies of this type have been conducted in Slovenia. However, these previous studies considered both the clinical and subclini-cal forms of bovine mastitis together and therefore the results are not comparable (13, 14). References 1. Pengov A. The mastitis situation in Slovenia - An overview. In: VI Congress FeMeSPRum. Postojna, 1998: 455-7. 2. Sol J, Sampimon OC, Barkema HW, Schukken YH. Factors Associated with cure after therapy of clinical mastitis caused by Staphylococcus aureus. J Dairy Sci 2000; 83: 278-84. 3. Nickerson SC. Eliminating chronic Staphylococ-cus aureus mastitis. Vet Med 1993; 88: 375-81. 4. Sears PM, Smith BS, English PB, Merer PS, Gonzales RN. Shedding pattern of Staphylococcus aureus from intramammary infections. J Dairy Sci 1990; 73: 2785-9. 5. Watts JL, Salmon SA. Activity of selected antimicrobial agents against strains of Staphylococcus aure-us isolated from bovine intramammary infections that produce p-lactamase. J Dairy Sci 1997; 80: 788-91. 6. Jones TO, Heath PJ. P-lactamase production in Staphylococcus aureus isolated from bovine mastitis milk. Vet Rec 1985; 117: 340-44. 7. Cracknell VC, Flogdell K. Clavulanate-potentiat-ed amoxycillin with Prednisolone in the intramammary treatment of clinical mastitis. 15th World Buiatrics Congress Palma de Mallorca, 1988: 505-9. 8. De Oliveira AP, Watts JL, Salmon SA, Aarestrup PM. Antimicrobial susceptibility of Staphylococcus aureus isolated from bovine mastitis in Europe and the United States. J Dairy Sci 2000; 83: 855-62. 9. Pyorala SHK, Pyorala EO. Efficacy of parenteral administration of three antimicrobial agents in treatment of clinical mastitis in lactating cows: 487 cases (1989-1995). J Am Vet Med Assoc 1998; 212: 407-12. 10. Owens WE, Watts JL, Boddie RL, Nickerson SC. Antibiotic treatment of mastitis: comparison of intra-mammary and intramammary plus intramuscular therapies. J Dairy Sci 1988; 71: 3143-7. 11. Eliopoulos GM, Moellering RC. Antimicrobial combinations. In: Antibiotics in laboratory medicine. Baltimore, 1996: 330-6. 12. Wilson DJ, Gonzalez RN, Case KL, Garrison LL, Grohn YT. Comparison of seven antibiotic treatments with no treatment for bacteriological efficacy against bovine mastitis pathogens. J Dairy Sci 1999; 82: 1664-70. 13. Klinkon Z, Klinkon M. Successful cure of bovine mastitis - S. aureus: In: VI Congress FeMeSPRum. Postojna, 1998: 364-6. 14. Pengov A Treatment of clinical and subclinical mastitis cases with COBACTAN®. Vet Nov 2001; 27: 145-8. ZDRAVLJENJE SUBKLINIČNIH STAFILOKOKNIH MASTITISOV O. Podpečan, A. Pengov, U. Hrastnik Povzetek: Namen raziskave je bil oceniti uspešnost zdravljenja subkliničnih mastitisov pri kravah molznicah, povzročenih z bakterijsko vrsto Staphylococcus aureus. S. aureus je najpogosteje izolirana patogena bakterija pri živalih s subkliničn-im mastitisom v ZDA in v Evropi. Podatki iz literature kažejo, da amoksicilin sam ni dovolj učinkovit proti S. aureus, v kombinaciji s klavulansko kislino pa se učinek zdravljenja bistveno izboljša. Okužene živali smo zdravili s preparatom Synulox®, ki vsebuje amoksicilin in klavulansko kislino. Preparat smo po navodilu proizvajalca aplicirali v mišico in v mlečno žlezo. Skupno je bilo zdravljenih 61 vimenskih četrti pri 37 kravah. Uspešnost bakteriološke ozdravitve je bila v povprečju 51,3 %, pri živalih z eno okuženo vimensko četrtjo pa 61,9 %, in sicer ob dejstvu, da so le-te predstavljale 56,7 % vseh živali v raziskavi. Glede na rezultate menimo, da je uporaba kombinacije amoksicila in klavulanske kisline upravičila naša pričakovanja in da je zdravljenje živali z eno ali dvema okuženima četrtma smiselno in ekonomsko opravičljivo. Ključne besede: veterinarska medicina; mastitis - zdravljenje; Staphylococcus aureus; amoksicilin; klavulanska kislina Slov Vet Res 2004; 41 (1): 35-9 UDC 619:615.28:638.154 Original research paper USING OXALIC ACID FOR VARROA MITE CONTROL IN HONEYBEE COLONIES DURING THE BEEKEEPING SEASON Aleš Gregorc 1*, Ivo Planine 2 Addresses of authors: 11nstitute for Breeding and Health Care of Wild Animals, Fishes and Bees, Veterinary Faculty, Gerbičeva 60, 1000 Ljubljana, Slovenia; 2 National Veterinary Institute, Pri hrastu 18, 5000 Nova Gorica, Slovenia *Corresponding author. E-mail: ales.gregorc@vf.uni-lj.si Summary: Twenty-one Apis mellifera carnica honeybee colonies were used to record the levels of Varroa mite mortality in the periods before and after oxalic acid (OA) treatments, which were conducted after the honey harvesting. The colonies each received from four to seven OA treatments. During the pre-treatment period, the daily natural mite mortality was estimated at 0.56 (±0.74). A high correlation was established between the daily pre-treatment mite mortality and the cumulative total of dead mites after each of the consecutive OA treatments (R = 0.92387). In the colony with the lowest daily mite mortality (0.08 ±0.05) there was no correlation with the number of mites that fell after the initial OA treatments conducted on August 1. The relative mite mortality ranged from 7.78 % (±1.68) during the brood period to 88.87 % (±8.41) in colonies without brood. Reducing a colony's mite population by employing OA treatments and a mite-control programme are discussed in this study. Key words: honeybee; Apis mellifera; varroa control; oxalic acid Introduction Varroa destructor, a parasite of Apis mellifera, has to be controlled by the regular use of acaricides in order to maintain honeybee colonies. These are usually synthetic and their lipophilic and persistent characteristics result in a build-up of their residues in the wax and honey (1). Acaricide resistant mites have appeared in several European countries (2, 3). Natural, non-toxic substances to control varroa mites, such as organic acids (4) and essential oils (5, 6), have been developed and are increasingly being used by beekeepers (7). Oxalic acid is a natural constituent of honey and EU regulations permit its use in biological beekeeping (EU Council Regulation, No. 1804/1999). Because of its high efficacy, OA is widely used in most Western European countries (7). Research has been conducted into the efficacy of oxalic acid applications (OA) as a method for controlling the mite in colonies, both with and without brood (8, 9, 10, 11). Received: December, 2003 Accepted for publication: February, 2004 Experiments have been conducted in honeybee colonies by spraying (9), evaporating (12) and trickling an OA-water solution into the hive (13, 14, 15). During broodless periods, Radetzki (16), Nanetti et al. (17) and Imdorf et al. (18) found it to be highly effective in killing the mite and they estimated the elimination level at 97.3 %, 98.3 % and 99.5 %, respectively. When a capped brood was present, Mutinelli et al. (15) achieved 95 % efficacy after three treatments of a 5-%-OA solution and Brrdsgaard et al. (10) reported a 24 % efficacy of one spring treatment administered by trickling. In our previous experiments 50 millilitres of an OA solution was used to treat one normally developed colony. Three OA treatments had an efficacy of 39.2 % when a brood was present and 99.4 % when there was no brood (13). This paper presents data from the periodic checking of the number of mites that had fallen onto the bottom of hives to determine the natural mite-fall. The aim was to establish the effectiveness of OA as a single substance for controlling varroa in honeybee colonies by using a sucrose-in-water solution (14). We also aimed to establish the optimal strategy for using oxalic acid applications to control mites during the 36 A. Gregorc, I. Planinc 2002 season in colonies with capped broods and for winter treatments of broodless colonies. Materials and methods Twenty-one Apis mellfera carnica honeybee colonies, populated in national standard AZ "back load" hives (19) with nine combs (41 x 26 cm) in each brood and honey compartment, were located at one site near Vipava. In the spring of 2002, metal sheets (38 x 29.8 cm) were placed on the floor of each of the hives in order to record the hives' natural mite mortality. On the sampling dates, the numbers of mites were recorded. The pre-treatment natural mite fall of each of the colonies was recorded on 6 different occasions for those colonies whose initial OA treatment was on August 1 and on 9 different occasions for those colonies whose treatment began on August 8. The mite mortality after each of the consecutive OA treatments was also recorded. The number of OA treatments each colony received was determined after establishing the mite mortality before and after each treatment. The treatments were performed as follows: • Group A, which consisted of five colonies (Nos. 1, 2, 8, 14, 19), received 7 OA treatments; on August 1, 8 and 20, September 7, October 10, November 14 and December 24; • Group B - four colonies (6, 9, 17, 31) - received 6 OA treatments; on August 1, 8 and 20, September 7, November 14, and December 24; • Group C - four colonies (11, 20, 21, 24) -received five OA treatments; on August 1, 8 and 20, November 14, and December 24; • Group D - three colonies (4, 23, 29) - received five OA treatments; on August 8 and 20, September 7, November 14, and December 24; • Group E - five colonies (3, 5, 7, 27, 28) -received four OA treatments; on August 8 and 20, November 14, and December 24. The treatments were applied to each colony by trickling the OA solution over the combs, in situ, and squirting the bees in the brood compartment using a syringe. Respiration masks, protective glasses and rubber gloves were worn while applying the solution. The mite fall during the treatment period was recorded after each OA application. The outside temperatures during the August OA treatments ranged from 30 to 33 0C. On September 7 the outside temperature was 24 0C, on October 10 it was 22 0C, on November 14 it was between 10 and 12 0C and on December 24 it was 6 0C. The colonies received 50 millilitres of a 2.9 % OA and 31.9 % sucrose-in-water solution (w/w), using oxalic acid dihydrate (Riedel-de Haën), sucrose (sugar) and de-mineralised water (Gregorc and Planinc, 2001). The percentages of mites killed by the experimental treatments (FTB) were estimated using the formula: FTB = FOA1 / (FOA1 + FOA2) x 100 (Gregorc and Planinc, 2001). FOA1 is the total number of mites that dropped during the consecutive treatments of colonies with capped broods and FOA2 is the number of mites that fell during the December treatment of the broodless colonies. The efficacy of the treatments was also estimated by comparing the numbers of mites that fell before and after the treatments and the mite mortality between the consecutive OA treatments. The data analyses were performed by ANOVA (analysis of variance) with the use of the Statgraphic (20) programme. Results During the pre-treatment observation periods, a total of 39 days performed in intervals between May 2 and August 1, 2002, the average daily natural mite-death was estimated at 0.56 (±0.74). The average mortality per colony during the total observation period was estimated at 21.71 (±29.01) mites. In this period 1.45 % (±0.83 %) of the total varroa mite population died naturally. The average numbers of mites that dropped onto the bottom boards of each group are shown in Table 1. Mite mortality after the first OA treatment was significantly higher (P<0.01) in the colonies of group A than it was in groups B, C, D and E. Statistically significant differences were also found between these groups. The number of mites that fell after the first OA treatment of the highly-infested colonies of group A correlated with the daily mite mortality prior to treatment (R = 0.81398) (Fig. 1). A high correlation was found when the total number of dead mites (R = 0.8851387) and the daily mite mortality (R = 0.92387) during pre-treatment period were compared to the cumulative total of dead mites observed after each of the consecutive OA treatments. Using oxalic acid for varroa mite control in honeybee colonies during the beekeeping season 37 Table 1: The average (±SD) mite mortality after consecutive OA treatments. The number of mites that fell naturally prior to treatment and the % of natural mite mortality prior to the treatments compared to the total number of mites that fell during the experiment. The data relate to the five groups of colonies, each of which were exposed to a different number of OA treatments Group Mite mortality after the OA treatments (±SD) Mite mortality prior to the OA treatments (±SD) Mite mortality (%) prior to the OA treatments (±SD) A 3107.80 (±1622.57) 60.00 (±40.84) 1.94 (±0.81) B 1248.00 (±553.12) 1.00 (±1.15) 0.06 (±0.08) C 754.25 (±331.82) 13.25 (±5.56) 1.77 (±0.66) D 831.00 (±176.55) 14.33 (±4.04) 1.68 (±0.13) E 520.40 (±271.14) 0.00 (±0.00) 0.00 (±0.00) The cumulative mite mortality after the OA treatments of the different groups and the natural daily mite mortality prior to the OA treatments are shown in Figure 2. In the colony that had the lowest average daily mite mortality (0.08 ±0.05) there was no statistically significant correlation with the number of mites that fell after the initial OA treatment conducted on August 1. In the colonies of group A, the average mite mortality after the first OA treatment was 204.4 (±81.60). The mite mortality after each OA treatment ranged from 114.4 (±150.80) mites after the seventh treatment conducted on December 24 to 1065 (±605.38) mites after the fifth treatment on October 5 (Figure 3). The number of mites that fell after the first OA treatment on August 1 was, on average, 341.48 (±352.17) times higher than the average daily natural mite mortality (0.55 ±0.78) monitored in the pre-treatment period. The relative mite mortality during the brood period ranged from 7.78 % (±1.68) recorded after the first OA treatment conducted on August 1 to 88.87 % (±8.41), which was recorded after the November 11 OA treatment when the colonies were without brood. The relative values of mites that fell during the OA treatments are shown in Figure 4. All the colonies had queens throughout the experiment and the normal death rate of the worker bees was not altered by the treatments. The development of the colonies over winter and the spring of 2003 were normal and comparable to the development of other colonies in the region. Discussion Counting the mites that drop onto the bottom board is a reliable diagnostic method (21). The correlation between the high natural mite mortal- ity and the number of mites that fell after the OA treatments indicates the importance of measuring the natural mite mortality prior to treatments in order to accurately establish the degree of a colony's infestation. The natural mite-mortality per day, which until the August OA treatments was estimated at an average of 0.56, increased up to a maximum of 1.54 mites per day. The colonies of the experimental apiary showed variations in the levels of mite infestation. The high daily natural mite-mortality correlates with the numbers of mites that fell after the first and subsequent OA treatments. The total mite mortality after several OA treatments also correlates with the pre-treatment natural mite mortality. A relationship between the mites in the hive debris and the mite population (22) is evident in colonies with approximately one mite "drop down" per day. In colonies with a low mite mortality (approximately 0.5 mite per day), the correlation is not as evident and considerable differences in mite mortality after the OA treatments are found. In these colonies the cumulative mite mortality during the OA treatments averaged 701 (±161) (Fig. 2). It is important to continuously monitor the number of mites dropping in bee colonies in order to establish the appropriate timing and sequence of the summertime OA treatments. It seems that for colonies with approximately 1.5 natural mite-deaths per day a suitable programme of OA treatments should be established. In our experiment the reduction of the mite population by 7.78 % after the first, and 9.2 % after the second OA treatment ensured that the colonies remained viable. The efficacy of further OA applications increased up to 53.4 % and 88.9 % after the October and November OA treatments, respectively. The results of our experiment confirmed that using OA to treat 250 > 200 2 150 o E a) 100 50 0 - — - A: 7X OA B: 6X OA C: 5X OA D: 5X OA H1. OA 1.8. —♦— average mite mortality/day E: 4X OA 1,8 1,6 1,4 1,2 1 0,8 0,6 0,4 0,2 0 > re T3 > re t o E a) 2 D re Figure 1: Diagram showing treatments groups (A-E) and the number of mites that fell after the first OA treatment conducted on August 1. The average natural mite mortality/day during the pre-treatment observation period is also shown to illustrate the high correlation between the two sets of figures (R = 0.81398) o > +3 J2 3 E 3 o 3500 3000 2500 2000 1500 1000 500 0 1,8 1,6 1,4 1,2 1 0,8 0,6 0,4 0,2 0 A: 7 X OA B: 6X OA C: 5X OA D: 5X OA E: 4X OA No. of OA treatments I I Cumulative No. fallen mites Natural mite mortality/day Figure 2: The cumulative mite mortality following the oxalic acid applications during the experiment is shown separately for each group. The average daily natural mite mortality during the pre-treatment observation period is also shown. As indicated, the colonies in each group received between four and seven treatments 1800 1600 1400 1200 £ 1000 o £ e 800 600 400 200 0 T I i I 1 II TT ÎI„ k h Iri II Mi Xi