THE ROLE OF THE BURSA OF FABRICIUS
AND HIGHLY DILUTE BURSIN
IN IMMUNONEUROENDOCRINE INTERACTIONS
IN THE CHICKEN

Biological effects of highly dilute bursin

 

B.J. YOUBICIER-SIMO*, F. BOUDARD*, M. GUELATTI**, M. MEKAOUCHE**, J.D. BAYLÉ**, AND M. BASTIDE*

* Laboratoire d'Immunologie & Parasitologie

Faculté de Pharmacie, Université de Montpellier-1

15 Av. Ch. Flahault, 34060, Montpellier cedex 1, France

** Laboratoire de Physiologie Générale

Faculté des Sciences, Université de Montpellier-2

Pl. E. Bataillon, 34090 Montpellier cedex 5 (France)

 


1 - Introduction
2 - Material and Methods
3 - Results
4 - Discussion
5 - References

 

1. Introduction

               

A huge set of data, pertaining to Mammals namely, supports the existence of two-way communication between the immune and neuroendocrine systems. For example, glucocorticoids released by the adrenals are immunodepressive; in turn, immune products act at different levels of the hypothalamo-pituitary-adrenal axis (HPA) (for review see Lilly and Gann, 1992). Also, the chief pineal hormone melatonin (MLT) is immunostimulatory (Maestroni et al., 1989) whereas interferon-g stimulates the release of MLT by cultured pineal glands (Withyachumnarkul et al., 1990). The concept that products of the immune system act on the neuroendocrine system was suggested by Besedovsky and Sorkin (1977), based on their observation that in rats, serum corticosterone (CORT) increases during the course of immune response. However, in Mammals, most studies working out immunoneuroendocrine interferences deal with the T immune component mainly, therefore neglecting the involvement of the B lymphoid compartment, because the latter is rather anatomically diffuse. Therefore, it appears experimentally difficult to identify specific B immune signal (s) endowed with neuroendocrine competence.

                Aves possess a species specific and anatomically distinct organ termed the bursa of Fabricius (Toivanen et al., 1981). The latter is a diverticulum of the hindgut and it is also the primary site for B-cell differentiation (Toivanen et al., 1981). The bursa anlage arises by the 5 th day of embryonic development as a swelling of the cloacal plate and it is invaded by B stem cells between the 8 th and 14 th day of embryonic life (Le Douarin et al., 1975; Houssaint et al., 1976; Lassila et al., 1978). In the bursa follicles, the B precursor cells differentiate into mature lymphocytes which seed to the periphery by the end of the embryonic phase (Le Douarin et al., 1975; Weill and Reynaud, 1987). Early removal of the bursa anlage leads to permanent B-deficient chickens (Fitzsimmons et al., 1973; Jankovic et al., 1977; Granfors et al., 1982; Jalkanen et al., 1983). This surgical procedure provides a biological preparation suitable for the study of possible involvement of the B immune component in immunoneuroendocrine crosstalks. The function of the bursa of Fabricius is mediated by cellular and soluble factors that have been identified as integral parts of the bursa microenvironment (Heller and Friedman, 1979; Glick, 1984ab; Kuznik et al., 1988). One of these factors is a low-molecular weight inducing agent termed bursopoietin that has been extracted from the chicken bursa (Brand et al., 1976). This agent has been isolated, purified and termed bursin (Lys-His-Gly-NH2) by the US team of Goldstein (Audhya et al., 1986). Bursin has been immunohistochemically identified in the bursal epithelium (Viamontes et al., 1989), as well as in avian and bovine bone marrow and in the intrahepatic bile ducts of bovidæ (Audhya et al., 1990). A tetradecapeptide precursor named probursin and containing the sequence of bursin, tuftsin and the active site of somatostatin has also been discovered (Audhya et al., 1991). Bursin induces B cells from their avian and mouse precursors in vitro (Goldstein et al., 1977; Audhya et al., 1986; Lassila et al., 1989). However, it is still unclear whether the neuroendocrine function of the bursa of Fabricius also operates through bursin-dependent mechanisms.

The present study was an attempt to disclose the possible involvement of the B immune component in immunoneuroendocrine interactions, using permanent B-deficient chickens (bursectomized chickens). This aim was achieved by testing:

                1. the ability of bursectomized chickens to respond hormonally and immunologically to various environmental agents (ether vapour, antigen, photoperiod). The following parameters were assessed:

                - the pituitary (corticotrophin: ACTH) and adrenal (corticosterone: CORT) responses to ether vapour exposure,

                - the specific antibody (IgM, IgG), as well as the pituitary-adrenal (ACTH, CORT) and pineal (melatonin: MLT) responses to immunization against porcine thyroglobulin (Tg),

                - the nature of the endocrine responses displayed by porcine Tg-sensitized chickens,

                - the circadian rhythms of both pituitary-adrenal (ACTH, CORT) and pineal (MLT, N-acetyltransferase: NAT, hydroxyindole-o-methyltransferase: HIOMT) activities.

                 2. the effectiveness of the bursa-derived signal (bursin) in reversing the effects of bursectomy, by means of in ovo administration of different amounts of bursin (100 µg, 100 pg (5CH), 100 fg ( 7CH), 5 x 10-27 g (a pool of 15CH to 20 CH dilutions) to bursectomized embryos.

 

 

2. Materials and Methods

 

2.1. INCUBATION OF EGGS AND CHICKS  MANAGEMENT

 

Eggs from New Hampshire strain were purchased from Couvoir des Cévennes (Ledenon 30, France) and incubated in a Maino Ladi France incubator (38 ± 1° C, 45-50% humidity, permanent darkness). An incorporated electric fan allowed homogeneity of the temperature and air inside the incubator. The eggs were permanently turned on automatically rotating drawers with 45° maximal inclination, so as to avoid sweating of the eggs opened during surgical bursectomy.

                Newly-hatched chicks were housed under controlled photoperiod [12 h light (0700-1900), 12 h dark] and the environmental temperature was progressively reduced 1° C per day from 38° C to 22° C. Feed and water were available ad-libitum.

 

2. 2. SURGICAL METHODS

 

2. 2. 1.  Embryonic bursectomy

The removal of the bursa anlage was performed at 80 h of incubation (4 th day of embryonic life) according to Fitzsimmons et al. (1973). The surgical operation was carried out under aseptic conditions. Briefly, the embryo was localized into the egg by candling and the shell was opened above it using fine forceps. Then a small slit was made in the chorion and amnios at the level of the caudal bud. A sterilized fine loop was passed over the caudal bud and tightened posterior to the leg buds, caudally to the cloacal membrane. The tail was cut off posterior to the nod using microsurgical scissors and the free fragment was discarded. The lips of the chorion and amnios were closed with forceps and the shell opening was overlaid with parafilm M and sealed with paraffin.

 

2. 2. 2. Removal of  the pineal glands and eyes

The chickens were sacrificed by decapitation and the skull was opened dorsally between the cerebral hemispheres and the cerebellum. The meninges were slit above the pineal gland along the line of suture. The lips of the opening were drawn aside by means of slender forceps. The pineal body gripped by its stalk, withdrawn from the skull, immediately frozen in liquid nitrogen and stored at -80° C until assayed for the pineal enzymatic activities (NAT, HIOMT). The eyes from each bird were dissected out, wrapped individually in aluminium foil, immediately frozen in liquid nitrogen and stored at -20° C until time of assay. Then the retinae were isolated by the procedure described by Hamm and Menaker (1980).

 

2. 3. TRIPEPTIDES AND IN OVO  TREATMENT

 

Bursin (Lys-His-Gly-NH2) was synthesized de novo  by fragment peptide coupling (Le Nguyen, 1987). The control tripeptide (Trp-Leu-Leu-NH2) was a gift from Dr. Martinez, University of Montpellier, France). The injected solutions were prepared by sequential centesimal dilution steps including potent vertical stirring between two successive steps. On the 6 th and 9 th day of incubation, the embryos were treated in ovo by single administration of 100 µl of solution each day in the yolk sack , as follows: sham-bursectomized embryos (N) were given 100 µl of saline (N+S). The surgical treatment group (Bx) could receive either 100 µl of saline (Bx+S) or 100 µl of saline containing 5 CH of the control peptide (Bx+Pf) or 100 µl of saline containing different doses of bursin: 100 µg (Bx+Bµ) or 5 CH (Bx+Bp) or 7 CH (Bx+Bf) or a 15-20 CH pool (Bx+B-27).

 

2. 4. EXPERIMENTAL GROUPS

 

The birds were distributed as follows:

                N: Sham-bursectomized (the surgical handling was limited to opening and closing the chorion and amnios in 80 h old embryos),

                N+S: sham-bursectomized birds injected in the yolk sack by single administration of 100 µl of saline on the 6 th and 9 th day of embryonic life,

                Bx: chickens embryonically bursectomized at 80 h of incubation,

                Bx+B: chickens embryonically bursectomized at 80h of incubation and supplemented on the 6th and 9th day of embryonic development by in ovo administration of 100 µl of saline containing different amounts of bursin: 100 µg (Bx+Bµ) or 5 CH (Bx+Bp) or 7 CH (Bx+Bf) or the 15-20 CH pool (Bx+B-27),

                Bx+Pf: chickens embryonically bursectomized at 80 h of incubation and receiving 100 µl of saline containing 5 CH of the control tripeptide (Trp-Leu-Leu-NH2) on the 6 th and 9 th day of embryonic development.

 

2. 5. ANTIGENS

 

Porcine Tg or keyhole limpet hemocyanin (KLH) purchased from Sigma was injected subcutaneously. The dose of immunogen to be injected was determined in a preliminary experiment (data not shown). Chickens received three doses of 125 µg/100 g B.W.  (750 µg/ml of saline) mixed (v/v) with complete (first immunization) or incomplete (second and third immunizations) Freund adjuvant.

 

2. 6. EXPERIMENTAL DESIGN

 

Four experimental protocols including a total of 10 experiments have been carried out during six years.

 

2. 6. 1. First protocol

Experiment 1. Five chicken samples were used: N (4), Bx (4), Bx+Bµ (4), Bx+Bp (4), Bx+Bf (3). In brakets is the number of animals. Eight week-old chickens were placed in an ether vapour saturated box for 1min between 0900 and 1000. Blood samples were withdrawn from the brachial vein 7 and 14 min after ether vapour application, and assayed for plasma ACTH and CORT respectively.

 

2. 6. 2. Second protocol

This protocol was designed to measure the hormonal and specific humoral immune responses to antigen challenge. Four expriments were performed (experiments 2, 3, 4, 5). In experiments 2, 3 and 4, young chickens were repeatedly immunized by injecting porcine Tg at 21, 30 and 39 days of age. Blood samples were withdrawn from the brachial vein, within 30 s, between 0900 and 1000, the day before immunization (d20) and at 9 day intervals after the first (d29), the second (d38) and the third (d47) antigen challenge. All the sera were blinded before hormones and antibody evaluation. The specificity of anti-Tg antibodies was also determined.

Experiment 2. The following groups of animals were used: N (8), Bx (7), Bx+Bµ (6), Bx+Bp (6) and Bx+Bf (7). The plasma levels of ACTH, CORT and MLT, as well as the serum titres of specific anti-Tg antibodies (IgM, IgG) were measured.

Experiment 3. It was carried out with 4 groups of chickens: N (8), Bx+S (6), Bx+Bf (9), Bx+B-27 (7). The plasma levels of CORT, as well as the serum titres of anti-Tg IgG antibodies were determined. 

Experiment 4. Seven chicken samples were studied: N (7), N+S (6), Bx (11), Bx+S (7), Bx+Pf (4), Bx+Bf (7), Bx+B-27 (8). The following parameters were assessed: the plasma levels of ACTH, CORT and MLT, as well as the serum titres of specific anti-Tg IgG. 

Experiment 5. This experiment was designed to determine if the hormonal responses of bursa-intact chickens sensitized to porcine Tg derived from metabolic reactions involving Tg or if these answers were linked to the immunization phenomenon. Tg is the endogenous prohormone of thyroid hormones which themselves modulate the pituitary-adrenal activity during the course of immune response (Besedovsky et al., 1975; Trout et al., 1988). Keyhole limpet hemocyanin (KLH) is a T-dependent antigen without any pharmacological activity. To attempt disclosing the nature of the hormonal response to porcine Tg, 17 bursa-intact chickens immunized against porcine Tg were compared with their KLH-sensitized counterparts (18 chickens) for CORT, MLT and IgG responses. The chickens were immunized at 21, 30 and 36 days of age by subcutaneous inoculation of either porcine Tg or KLH. Blood samples were collected between 0900 and 1000 at the age of 20 (d20), 29 (d29), 35 (d35) and 38 (d38). Then the plasma levels of CORT and MLT, as well as the serum titers of anti-Tg or anti-KLH antibodies (IgG) were measured.

 

2. 6. 3. Third protocol

Twelve week-old chickens raised since hatching under a 12L00700-1900-12D alternating light-dark regime underwent four experiments (experiments 6, 7, 8, 9).

Experiment 6. In this preliminary experiment, bursa-intact chickens (N, 13) were decapitated and their circulating, ocular and pineal MLT contents were measured between 1200 and 1300 and between 0000 and 0100 (under red light).

Experiment 7. Five samples were studied: N (9), Bx (3), Bx+Bµ (3), Bx+Bp (3) and Bx+Bf (9). The chickens were bled at two-point time, between 1200 and 1300 and between 2000 and 2100 (under red light) to assess plasma ACTH and CORT levels. MLT levels were measured in blood samples collected between 1200 and 1300 and between 0000 and 0100 (under red light).All the sera were blinded before hormones  evaluation.

Experiment 8. Blood samples were withdrawn from the N (7), Bx (5), Bx+Bf (4) and Bx+B-27 (4) groups between 1200 and 1300 and between 0000 and 0100 (under red light) to assess the plasma levels of MLT. After blood sampling, the animals were immediately sacrificed by decapitation and their pineal tissues assayed for NAT and HIOMT activities.

Experiment 9. Blood samples were collected from the brachial vein at 4 h intervals around the clock (1000, 1400, 1800, 2200, 0200, 0600) to measure the plasma levels of ACTH, CORT and MLT. The chicken samples used were as follows: N (13), N+S (6), Bx (12), Bx+S (7), Bx+Pf (4), Bx+Bf (7), Bx+B-27 (7). All the sera were blinded before hormones evaluation.

 

2. 6. 4. Fourth protocol

Experiment 10. This experiment was aimed at checking whether a succussed highly dilute solution of bursin was structurally different from the same dilution unsuccussed. To answer this question, we decided to measure the concentration of labelled bursin in  serial dilutions of this compound. Since labelled bursin was not available on the market place, we used 3H-Thymidine. 3H-Thymidine (cpm) was measured in different concentrations of succussed thymidine solutions ranging from 10-7 to 10-41 M, and compared with equal dilutions of unsuccussed 3H-thymidine and succussed solvent. Unsuccussed 3H-thymidine solutions were prepared from an initial 3H-thymidine solution (1 mCi/ml, specific activity 25 Ci/mmol., Amersham), by serial centesimal dilution in pure water. Succussed 3H-thymidine solutions and the solvent (pure water) were prepared in the same way, but in addition, potent vertical shaking was performed between successive dilution steps. Finally, 200 µl from each solution were added to 3 ml of scintillation solution (Pico-Fluor 15) and counted in a ß-liquid scintillation analyzer (Packard).

 

2. 7. DETERMINATION OF SPECIFIC ANTI-Tg IgM AND IgG TITRES

 

The titres of specific antibodies against porcine Tg were determined in sera by indirect enzyme-linked immunosorbent assay (ELISA) technique. Microtitre plates (Immunoplates II, NUNC, Roskilde, Denmark) were coated with a solution of porcine Tg at 10 µg/ml in phosphate buffered saline (PBS) at pH 7.2, and the plates were incubated overnight at 4° C. The plates were washed three times with PBS containing 0.1% Tween 20 (V/V); then 100 µl of chicken plasma diluted by half from 1/20 were added to the plates and incubated at 37° C for 1 h. The plates were washed three times. For the determination of anti-Tg IgM, 100 µl of horseradish peroxidase-labelled anti-chicken IgM conjugate (Bethyl Laboratories, Montgomery, AL) was added and the plates were incubated at 37° C for 1 h. After the last washing, the enzyme substrate, o-phenylenediamine (Sigma) was added and the plates were icubated for 15 min at room temperature. The absorbance was measured at 450 nm using a Multiskan photometer (Flow Laboratories, Marseille, France). For the determination of anti-Tg IgG, 100 µl of alkaline phosphatase-labelled anti-chicken IgG conjugate (Sigma) were added and the plates were incubated for 1 h at 37° C. After the last washing, 4-nitrophenylphosphate (Sigma) was added and the plates were incubated for 30 min at 37° C. The absorbance was measured at 405 nm. The titres of anti-Tg IgG or IgM were defined as the reciprocal of the plasma dilution giving a absorbance equal to 1.0 by indirect ELISA. Results were expressed in titre log.

The same technique was used to measure anti-KLH antibodies. Briefly, microtitre plates were coated with 100 µl/well of a KLH solution (10 µg/ml PBS) and incubated overnight at 4°C. Then, the plates were washed three times with PBS containing 0.1% Tween 20 (V/V). Afterwards, 100 µl of chicken plasma dilute at 1/40 were added to each well and the mixture held under 37° C for 1 h. The plates were rinced three times. Then 100 µl of peroxydase-conjugated affinipure rabbit anti-chicken IgG Fc fragment specific (Jackson ImmunoResearch Laboratories) dilute at 1/1000 in PBS-0.1% Tween 20 were added to the wells. The plates were incubated again for 1 h at 37° C. After three washings, 100 µl of the enzyme substrate termed o-phenylenediamine (Sigma) were added to the wells and the plates were allowed to incubate 5 min at room temperature. The enzymatic reaction was stopped by adding 50 µl/well of H2SO4 and the absorbance was measured at 490 nm. The titres of anti-Tg IgM or IgG antibodies were defined as for anti-Tg antibodies and the results were expressed in the same way.

The specificity of anti-Tg antibodies was checked by either an inhibition technique or a cross reaction test, using self-chicken (ovalbumin: OVA, ß-actin: ACT, myosin: MYO) or foreign (BSA, porcine insulin: INS, porcine Tg: Tg) antigenic proteins (Sigma). For the inhibition technique, the sera collected from 38 day-old chickens were dilute according to their titres and they were incubated with 2 µg/ml of the different proteins overnight at 4° C. Then an ELISA test using either porcine Tg or KLH as the antigen was performed on each treated serum in comparison with the corresponding untreated serum. The cross reaction consisted in testing the binding capacity of the above-listed antigens to sera collected at d20, d29, d38 and d47, by means of ELISA test.

 

2. 8. DETERMINATION OF HORMONES

 

The plasma concentrations of CORT (ng/ml) were measured by competitive protein-binding assay. This technique was previously made suitable for chicken plasma samples (Buckland et al., 1974). Intra and inter-assay coefficients of variation were 3.1 and 5.6% respectively. The sensitivity of the assay was 0.05 ng/ml. ACTH was assessed using a RIA kit (ACTH K-PR) purchased from CEA (Paris, France) (Ramade et al., 1985). Intra- and interassay coefficients of variation were 7.5 and 9.0 % respectively ; the sensitivity of the assay was 10 pg/ml. Plasma MLT concentrations (pg/ml) were determined by RIA according to the method developed by Rollag and Niswender (1976) and adapted to chicken by Cogburn et al. (1987). Intra and inter coefficients of variation were 14 and 12% respectively. The sensitivity of the assay was 3 pg/ml.

 

2. 9. ENZYMATIC ACTIVITIES

 

NAT and HIOMT activities were measured according to the techniques of Voisin and Colin (1986) and Voisin et al. (1988), respectively. Individual pineal glands were homogenized by sonication in 60 µl sodium phosphate buffer (0.05 M, pH 7.9). The enzyme substrates [20 µl tryptamine (20 mM), 20 µl of 3H-acetylcoenzyme A and H-acetylcoenzyme A mixture (specific activity: 4 Ci/mole)], 20 µl of sodium phosphate buffer (0.05 M, pH 7.9) and 40 µl of pineal homogenate, for a total volume of 100 µl were introduced in Eppendorf tubes. The blank Eppendorf were filled with 20 µl of tryptamine (20 mM), 20 µl of a mixture of 3H-acetylcoenzyme A and H-acetylcoenzyme A (specific activity: 4 Ci/mole) and 60 µl of sodium phosphate buffer (0.05 M, pH 7.9), total volume 100 µl. The Eppendorf contents were mixed and incubated for 30 min at 37° C. Thereafter, 1 ml chloroform was added into the tubes at 4° C to stop the reaction. Then the tubes were shaken, centrifuged (1000 g, 1 min) and the upper phase discarded. The chloroform phase containing the reaction product(s) was washed with 200 µl sodium phosphate buffer , centrifuged (10,000 g, 1 min) and the upper phase discarded. The chloroform phase (500 µl) was evaporated (2 h, 50° C), 7 ml of scintillant added to the residual powder and the radioactivity measured in a ß scintillation counter.

For HIOMT activity, 100 µl containing the substrates [25 µl N-acetylserotonine (25 mM), 25 µl of a mixture of 3H-S-adenosyl-methionine and H-S-adenosyl-methionine (specific activity: 25 Ci/mole)], 40 µl borate buffer (0.3 M, pH 10) and 10 µl of pineal homogenate. In blank tubes, 25 µl of N-acetylserotonin (25 mM), 25 µl of a mixture of 3H-S-adenosyl-methionine and H-S-adenosyl-methionine (specific activity: 25 Ci/mole) and 50 µl borate buffer (0.3 M, pH 10) were mixed. The tubes were treated similarly to those used to determine NAT activity, except that the reaction was stopped using 1 ml of chloroform and 20 µl of borate buffer.

 

2. 10. STATISTICAL ANALYSIS

 

The data represent mean ± S.E.M and were processed by two-way ANOVA, followed by one-way ANOVA and Mann-Whitney tests.

 

2. 11. HISTOLOGICAL CONTROLS

 

The birds were sacrificed at the end of the experiments. After autopsy, the absence of bursal remnants was ascertained by thorough ocular inspection and Cleveland-Wolff stained serial sections of the whole cloacal region were examined to verify completeness of bursectomy. Only the animals which met this creteria were used.

 

 

3. Results

 

3. 1. PITUITARY-ADRENAL RESPONSES TO ETHER STRESS

 

In the first protocol (Figure 1), ACTH and CORT levels were at resting values prior to ether stress application, irrespective of experimental groups. Bursa-intact chickens (N) exhibited strong hormonal responses after ether vapour exposure. No ACTH response was noticed in bursectomized birds (Bx) who rather raised a significant CORT response. The effectiveness of bursin treatment in reversing the effects of bursectomy appeared inversely proportional to the amount of the tripeptide administered: 100 µg induced a moderate increase in CORT but not ACTH levels and only after stress application, while either 5 CH or 7 CH restored quite normal pituitary-adrenal responses.

 

Figure 1.  Plasma ACTH (A) and Corticosterone (B) levels before (white columns) and after (dark columns) ether stress application to sham-operated (N), bursectomized (Bx) and bursectomized chickens supplemented with decreasing amounts of bursin: 100 µg (Bx+Bµ) or 100 pg (Bx+Bp) or 100 fg (Bx+Bf) . +P<0.01 vs Rest (white columns);  *P<0.01 vs Bx

 

 

3. 2. HORMONAL AND SPECIFIC HUMORAL IMMUNE RESPONSES TO IMMUNIZATION

 

In the second protocol, 20 to 47 day-old chickens underwent an immunization program to check the hormonal (ACTH, CORT, MLT) and specific antibody (IgM, IgG) responses. The results are outlined in Figures 2, 3 4. Regardless of experimental groups, hormonal levels (Figures 2A, B; 3A; 4A, B, C) were at resting values prior to immunization (d20) and hardly shifted 9 days after the first antigen challenge (d29). However, in sham-operated birds (N or N+S), hormonal responses reached a zenith at d38 and finally dropped to basal values after at d47. Assessed in parallel to endocrine

 

Figure 2.  Hormonal (A, B) and specific antibody responses (C, D) after immunization with porcine Tg. Sham-operated birds (N) or bursectomized chickens (Bx) treated in ovo with different amounts of bursin (Bx+Bµ; Bx+Bp; Bx+Bf) were sequentially immunized at 21, 30 and 39 days of age. Then plasma levels of ACTH (A) and corticosterone (B), as well as the serum titers of anti-Tg IgM (C) and anti-Tg IgG (D) were measured the day before (d20: white columns) and at the age of 29 (d29: dotted columns), 38 (d38: hatched columns) and 47 (d47: dark columns) days. +P<0.01 vs d20;  *P<0.01 vs Bx.

 

parameters, the specific humoral immune response (Figures 2C, D; 3B, C; 4 D) presented a different feature: especially in sham-operated samples, not only antibody production arose earlier, but it varied increasingly from the first to the third immunization.

 

Figure 3.  Corticosterone (A) and specific anti-Tg IgM (B) and IgG (C) responses to immunization against porcine Tg as measured in sham-operated (N) and bursectomized chickens supplemented with either the saline (Bx+S) or 100 fg (Bx+Bf) or a high dilution (5 x 10-27 g) of bursin. The birds were sequentially immunized against porcine Tg at the age of 21, 30 and 39 days and Corticosterone and antibody levels were evaluated the day before the first immunization (d20: white colums) and at the age of 29 (d29: dotted columns), 38 (d38: hatched columns) and 47 (d47: dark columns). +p<0.01 vs d20; *p<0.01 vs Bx. ND: Non-determined antibody titers (d20 for IgM,  d20 and d29 for IgG).

 

On the other hand, hormonal levels remained steadily low in bursectomized chickens (Figures 2A, B; 3A; 4A, B, C) who also failed to mount specific antibody production in spite of repeated immunization (Fig. 2C, D; Figures 3B, C; 4D) and whether they had received the saline or the control tripeptide (Figure 4D).

 

 

Figure 4 a:.   ACTH (A), corticosterone (B), melatonin (C) and anti-Tg IgG (D) responses to immunization against porcine Tg. Seven experimental groups were used (N, N+S, Bx, Bx+S, Bx+Pf, Bx+Bf, Bx+B-27. The animals were immunized at the age of 21, 30 and 38 days. Blood samples were collected the day before the first immunization (d20: white columns) and at the age of 29  (dotted columns), 38 (hatched columns) and 47 (dark columns) to check the plasma levels of hormonal (ACTH,corticosterone, melatonin) and specific antibody (IgG) responses.

+P<0.01 vs Bx; *P<0.01 vs Bx+S.

                The administration of 100 µg of bursin could no longer reverse of the alterations caused by bursectomy. Conversely, either discrete amounts (5 CH, 7CH) or a high dilution (15-20 CH pool) of bursin induced the recovery of normal endocrine and specific humoral immune performances in bursectomized recipients.

The specificity of anti-Tg antibodies was ascertained by an inhibition technique (Figure 4E) or a cross reaction test (Figure 4F) using either chicken self-proteins (OVA, ACT, MYO) or xeno-antigens (BSA, INS, Tg). Clearly, at the exclusion of the other proteins used, porcine Tg specifically reacted with the tested sera.

 

 

 

Figure 4 b.   The specificity of anti-Tg IgG was assessed either by an inhibition technique (E) or a cross reaction test (F) using self-chicken (ovalbumin: OVA, myosin; MYO, ß-actin: ACT ) or foreign (BSA, porcine insulin: INS, porcine Tg: Tg) proteins. In the inhibition technique, the sera collected from 38 day-old chikens were tested. The cross reaction test included sera harvested at the age of 20, 29, 38 and 47 days. .P<0.01 vs other proteins.

 

 

In the study designed to elucidate the nature of endocrine responses triggered by porcine Tg (experiment 5), bursa-intact chickens exhibited the same profile of CORT response (Figure 5A), irrespective of the sensitizing agent (porcine Tg or KLH): CORT levels increased significantly at d35 and crested at d38. Even though antibody responses (IgG) started earlier (d29) than endocrine responses (d35), KLH- and porcine Tg-challenged chickens displayed the same feature of humoral immunity: IgG levels rose from prime to third immunization (Figure 5B). It is worth noting that in each case (KLH or porcine Tg), the humoral immune response (IgG) was contemporary of the endocrine response (CORT), both cresting simultaneously.

 

         

 

Figure 5.   CORT (A) and IgG (B) responses of bursa-intact chickens sensitized to either keyhole limpet hemocyanin (KLH) or porcine Tg. Both parameters were measured the day before immunization (d20: white columns) and at the age of 29 (dotted columns), 36 (hatched columns) and 38 (dark columns) days.*P<0.01 vs d20. The samples comprised 18 and 17 chickens for KLH and porcine Tg respectively.

 

3. 3. CIRCADIAN RHYTHMS OF THE PITUITARY-ADRENAL AND PINEAL ACTIVITIES

 

In the third protocol (Table 1; Figures 6; 7), birds experienced a 12L-12D light-dark regime.Sham-operated birds (N) exhibited diurnal rhythms of plasma ACTH (Figure 6A), CORT (Figure 6B) and MLT (Figures 6C; 6D), with low light-time levels and significantly higher night-time values, in phase with light-dark cycle. Hence, broadpeaks were recorded during the dark phase (ACTH: 75.5 ± 3.42 pg/ml and CORT: 8.9 ± 0.25 ng/ml at 2100; MLT: 154.84 ± 8.21 pg/ml at 0000), whereas trough levels occurred during photophase (ACTH: 46 ± 3.49 pg/ml; CORT: 3.46 ± 0.28 ng/ml;

MLT: 14.5 ± 3.06 pg/ml at 1200). Pineal and ocular MLT contents (Table 1) were also higher at night (13960 ± 2514 pg/pineal and 612 ± 102 pg/eye at 0000) than during daytime (1662 ± 349 pg/pineal and 292.5 ± 74.8 pg/eye at 1200).

 

 

TABLE 1. Day/night differences in melatonin levels

________________________________________________

 

                                Organ                                                      Time

________________________________________________

 

                                                                               1200                         0000

                                                               _______________________________

 

                               Pineal gland                             1,662 ± 349               13,960± 2514*

 

                               Retina                                         292 ± 75                     612 ± 10*

 

                               Plasma                                     14.54 ± 3.06              154.84 ± 8.21*

________________________________________________

Melatonin concentration measured at 1200 and 0000 in the pineal

gland (pg/pineal), retina (pg/retina) and plasma (pg/ml) of 13 intact chickens  P < 0.01 vs 1200.

 

                This pattern of variation was in keeping with the rhythm of pineal NAT (Figure 6E) activity (1.3 ± 0.14 nmol/pineal/h at 1200 versus 24 ± 2.2 nmol/pineal/h at 0000). On the other hand, HIOMT (Figure 6F) did not vary significantly between the photo- and scotophase (13.62 ± 1.43 nmol/pineal/h at 1200; 18.05 ± 1.64 nmol/pineal/h) and irrespective of experimental groups. After bursectomy (Bx), the daytime levels of plasma ACTH and CORT. (Figure 6A, B) were no longer significantly different from nighttime ones (ACTH: 38 ± 0.46 to 47 ± 2.82 pg/ml; CORT: 4.24 ± 0.32 to 4.44 ± 0.29 ng/ml). Persistent diurnal rhythms were noticed in plasma MLT (5.7 ± 0.38 pg/ml at 1200 versus 65.77 ± 2.78 pg/ml at 0000) as well as NAT activity (0.73 ± 0.07 nmol./pineal/h at 1200 versus 5.13 ± 0.28 nmol./pineal/h at 0000), even though reduced by 50% in magnitude. In our attempt to reverse the effects of bursectomy, different amounts of bursin were administered in the yolk-sack of 6 and 9 days old bursectomized embryos.

                

 

 

 

Figure 6. Pituitary (A), adrenal (B) and pineal (C, D, E, F) daily activities measured at daytime (white columns) and at night (dark colmuns). Six groups of chickens (N, Bx, Bx+Bµ, Bx+BP, Bx+Bf, Bx+B-27) were submitted to a 12L-12D light-dark cycle. Blood samples were collected during daytime (1200-1300) and at night (0000-0100) to measure plasma ACTH (A), corticosterone (B) and melatonin (C, D).The pineal glands were assayed for enzymatic activities of NAT (E) and HIOMT (F).  + P<0.01 vs 12.00; * P<0.01 vs Bx.

 

Figure 7. Circadian rhythms of the pituitary (A, B), adrenal (C, D) and pineal (E, F) hormones. The 7 groups of animals listed above were submitted to a 12L-12D alternating light-dark cycle, with lights off from 0700 to 1900 (black bar below the figures) .The plasma samples were assayed for ACTH (A, B), corticosterone (C, D) and melatonin (E, F). +P<0.01vs Bx; *P<0.01 vs Bx+S.

 As results, the highest dose of bursin (100 µg) could not induce the recovery of the daily rhythms of plasma ACTH and CORT, whose levels remained steadily low and close to values observed in Bx animals (Fig. 6A, B). Likewise, 100 µg of bursin could no longer restore the normal amplitude of plasma MLT (Figure 6C). Conversely, lower amounts of bursin (5 CH, 7 CH, 15-20 CH pool) elicited significant daily hormonal and enzymatic rhythms, in an inverse dose-dependent manner (Figure 6A, B, C, D).

The hormonal parameters were also assessed every 4 h during a single light-dark cycle (Figure 7). Bursa-intact birds displayed pronounced circadian rhythms of plasma ACTH (Fig. 7A, B), CORT (Figure 7C, D) and MLT (Figure 7E, F), phase-locked with photoperiod and reaching a sharp peak at midscotophase. Bursectomy completely abolished circadian rhythms of the pituitary-adrenal hormones (Figure 7A, C), whereas MLT rhythm persisted, though markedly reduced in height (Figure 7E). Once more, neither the control tripeptide (Figure 7A, C, E) nor the saline (Figure 7B, D, F) could correct the effects of bursectomy. Administration of 100 µg of bursin has been reported also harmless in doing so (Youbicier-Simo, 1990). On the other hand, lower amounts of bursin (5 CH, 7 CH, 15-20 CH pool) allowed restoration of quite normal hormonal rhythms, in an inverse dose-dependent manner (Figure 7B, D, F).

 

3. 4. SERIAL DILUTIONS OF 3H-THYMIDINE

 

In the fourth protocol, we demonstrated that in sequentially dilute 3H-Thymidine solutions, the radioactivity decayed linearly, without showing any difference between succussed and unsuccussed solutions (table 2).

 

                __________________________________________________________

3H-Thymidine (cpm) was quantified in different concentrations of succussed thymidine solutions ranging from 10-7 to 10-41 M and compared to unsuccussed thymidine and solvent. The data represent means ± SEM of three independent experiments. As result, a plummet in 3H-Thymidine concentration was recorded and no residual tritiated compound could be detected above 10-15 M.

 

4. Discussion

 

4.1. BURSECTOMY AND EMBRYONIC MORTALITY

 

Our experimentation lasted 6 years during which more than 10.000 chicken embryos have been studied. In the sham-bursectomized groups, fetal loss (5-10%) was essentially restricted to the end of the incubation period. A possible explanation to this time-limited mortality is that the end of incubation corresponds to the preparation of hatching and represents a critical period during which the conditions of an extremely severe stress are gathered, weakening the fetus (Bauman and Bauman, 1977). Besides, two main critical periods characterized by a high death rate were identified during the development of the bursectomized chicken embryo: 20 to 30% of the operated embryos died between the day of bursectomy (4 th day of incubation) and the first day of in ovo  treatment (6 th day of incubation), whereas 40 to 65% mortality occurred around hatching. A third critical phase arised during the couple of weeks following hatching, due to either cloacal malformations or immunodepression (Belo et al., 1985). Finally, we recorded 8-10% survival two weeks after hatching. Performing surgical bursectomy at 60 h of embryonic life, Fitzsimmons et al. (1973) reported 6% survival after hatching; Belo et al. (1985) and Corbel et al. (1987) reported 5 and 7 % survival respectively, operating on 5 day-old chicken embryos. Grafting of 9.5 day-old chicken bursae (period of colonization of the bursa by B stem cells) does not improve the survival of bursectomized recipients (Ramade et al., 1985; Abdul-Karim et al., 1987), whose mortality rather got worse after the twofold administration (6 th and 9 th days of incubation) of bursin in the yolk sack. Finally, embryo handling appears as the prominent factor of mortality.

 

4.2. BURSECTOMY AND HUMORAL IMMUNITY

 

In order to evaluate the physiological consequences of both embryonic bursectomy and in ovo treatment with bursin, we tested the ability of bursectomized and bursin-treated bursectomized chickens to respond immunologically and hormonally to some environmental stimuli such as ether vapour, antigen challenge and photoperiod.

Control birds exhibited strong and increasing immunoglobulin production (IgG) during the course of immune response. These antibodies were specific to porcine Tg, since they reacted with neither self-chicken proteins (OVA, ACT, MYO) nor foreign antigens (INS, BSA). Unlike their bursa-intact counterparts, bursa-lacking chickens failed to raise specific anti-Tg antibodies, in spite of repeated immunization. These data are consistent with earlier reports indicating that bursa-lacking animals are able to produce immunoglobulins of the IgM, IgG and IgA classes, but fail to mount specific antibody responses against various antigens, despite iterative immunization (Jankovic et al., 1977; Granfors et al., 1982; Jalkanen et al., 1983). This lack of specificity is correlated with decreased number of Ig bearing cells: IgG positive cells in the spleen, thymus and blood (Jalkanen et al., 1984). The finland team of Toivanen also demonstrated that the number of B-cell clones is strikingly reduced in bursectomized chickens (Mansikka et al., 1990). This oligoclonality of the B-cell compartment is paralleled by a 10 fold decrease in serum IgG titres, IgM and IgA levels remaining unchanged (Jalkanen et al., 1983). It is worth noting that our results match these data, at least as far as the serum titres of IgG are concerned. By contrast, in bursectomized chickens, the IgM titres assessed by us appeared rather lessened. Nevertheless, our analysis must be shaded, because the level of Ig depends on age: in fact, if the levels of IgM, IgG and IgA are markedly low in the 10 day-old bursectomized chickens, the latter recover normal Ig titres by the age of 10 weeks (Eerola et al., 1983). Now, our measurements were performed half-way (3 to 7 weeks), between 10 days and 10 weeks. Evidence has been presented that at protein level, the isoelectric spectrum of immunoglobulins derived from bursectomized chickens displays normal gamma chains, but altered light chains, due to predominance of basic versus deficit in acid amino acids, which results in reduced antigen-antibody binding affinity (Granfors et al., 1982, Jalkanen et al., 1984). In addition, a low Ig gene conversion rate (V-J and V-D-J gene rearrangements) leading to a poorly diversified immunoglobulin repertoire has been reported in bursectomized chickens (Mansikka et al., 1990). Collectively, these data suggest that the bursal microenvironment plays a crucial role in the elaboration and diversification of the specific antibody repertoire. These bursa-dependent events might occur very early during embryonic life, probably at the beginning of the colonization of bursa follicles by B precursors cells (Weill and Reynaud; 1986, Reynaud et al., 1992). The bursa anlage develops on the 5 th day of embryonic life and is invaded by a single wave of B stem cells between the 8 th and the 14 th day of incubation (Houssaint et al., 1976; Lassila et al., 1978; Le Douarin et al., 1985). We performed surgical bursectomy at 80 h of embryonic development, so as to obtain complete and permanent B-deficient animals.

 

4.3. BURSECTOMY AND IMMUNO-NEUROENDOCRINE RESPONSIVENESS

 

If the bursa of Fabricius is undoubtlessly a key organ of the B immune component in Birds, it has been for a long time suspected to also serve various neuroendocrine functions: reduced CORT levels and enlarged adrenals, have been observed in newly-hatched chicks bursectomized at 68 h of embryonic life (Pedernera et al., 1980); Bursectomy performed in 2 week-old chicks increases the adrenal ascorbic acid response to ACTH (Perek and Eckstein, 1959, Perek and Eilat, 1966), whereas the hyperglycemic response following intramuscular injection of ACTH is lessened (Freeman, 1971); CORT levels increase after bursectomy (Mashaly, 1984). Analogous results have been previously obtained in our Laboratory. We reported that early surgical bursectomy (4 th day of embryonic life) results in various alterations in pituitary and adrenocortical functioning: ether stress-induced ACTH and CORT stimulations can be detected more precociously in bursectomized than in bursa-intact embryos; the non-stress responsive period of newly-hatched bursa-intact birds does not arise in bursa-lacking chicks (Guelatti, 1990; Guelatti et al., 1991); young bursa-lacking chickens exhibit a smaller CORT response to stress than controls (AbdulKarim et al., 1987); the delayed pituitary-adrenal responses of adenohypophysectomized chickens to Brucella abortus does not occur when embryonic bursectomy preceedes hypophysectomy (Baylé et al., 1991). In line with our previous results, we now present evidence of inability for bursa-deprived young chickens to cope with ether stress.

                Thyroglobulin (Tg) which was used by us to sensitize the chickens is also the endogenous precursor of thyroid hormones (Teppermann and Teppermann, 1987), and the latter are correlated to CORT during the course of immune response (Besedovsky et al., 1975; Trout et al., 1988). Unlike Tg which is a bioactive compound, KLH is a T-dependent antigen devoid of any pharmacological activity. We found that bursa-intact chickens sensitized to porcine Tg displayed the same profiles of pituitary-adrenal and humoral immune responses as their KLH-treated counterparts. The same feature of MLT response was recorded when immunizing bursa-intact chickens against porcine Tg or KLH (data not shown). Accordingly, it can be reasonably inferred that the changes of ACTH, MLT and CORT levels subsequent to porcine Tg treatment were not modulated by thyroid hormones derived from the sensitizing agent (porcine Tg), but were strictly linked to the immunization phenomenon. The fact that porcine Tg failed to induce noticeable endocrine answers in bursa-lacking chickens supports this point of view.

                Therefore, the physiological meaning of the concomitance of hormonal and antibody responses to porcine Tg should be discussed in the context of two-way communication between the immune and neuroendocrine systems. The hormonal response to antigen challenge is always contemporaneous with humoral immunity: in rats or mice primed intraperitoneally with three antigens, concomitant peaks of CORT and antibody responses occurred 5 to 6 days after immunization (Besedovsky et al., 1975); in the same way, both responses crested simultaneously 3 h after intravenous administration of Brucella abortus to immature chickens (Trout et al., 1988). We recorded analogous pituitary-adrenal and IgG responses when immunizing chickens against porcine Tg: ACTH and CORT levels, as well as IgG titers crested 18 days after prime immunization (d38). It is worth noting that in the aforementioned studies (Besedovsky et al., 1975; Trout et al., 1988), the delay of onset between hormonal and antibody responses was short (several minutes to three days). This was not the case in our experiments: the antibody response arose earlier (d29) than the hormonal response which occurred only 9 days later (d38). This discrepancy probably comes from methodological differences, leading to specific response kinetics. In the experiments of Besedovsky et al. (1975) or Trout et al. (1988), the antigens were injected free of adjuvant and directly into the general circulation (intraperitoneally or intravenously); hence they were rapidly conveyed into the secondary lymphoid organs to induce antibody production and cytokine release. In our study, not only porcine Tg was mixed with either complete (prime immunization) or incomplete (second and third immunizations) Freund adjuvant, but it was applied subcutaneously. As a result, the antigen molecules diffused slowly and had to pass through the lymphatic circulation before reaching the site of immune response. Some cytokines such as IL-1 or IL-2 have potential to increase blood glucocorticoid level via stimulation of the HPA axis (Glick, 1984b; Besedovsky et al., 1981; Lilly and Gann, 1992; Hu et al., 1993). Considering above arguments and despite induction of humoral immunity at d29, it is likely that the level of cytokines reached at d29 was not high enough to feedback on the HPA axis and elicit noticeable ACTH and CORT responses. Accordingly, we assume that after prime immunization, these responses had been initiated later than d29, at a moment when the effective concentration of cytokines was attained. A few days after subcutaneous administration of porcine Tg, we observed inflammatory nodules at the site of immunization. Evidence has been presented suggesting that cytokines released by inflammatory cells play a major role in HPA axis immunoregulation (Lilly and Gann, 1992; Besedovsky and Del Rey, 1987). Therefore, an alternative explanation to the lack of endocrine responses at d29 is that these responses were missed, because they had been induced earlier than d29 by inflammatory cytokines.

                Another distinctive feature of our results comes from the fact that endocrine and antibody responses were associated to repeated rather than unique immunization as reported by other authors (Besedovsky et al., 1975; Trout et al., 1988). Especially, CORT crested at d38, whether this sampling point was assessed after the second (experiments 2, 3, 4, 5) or the third (experiment 6) antigen challenge. Moreover, we noticed either a decline (experiment 2, 3, 4, 5) or a peak (experiment 6) of CORT following the third immunization. Also, the CORT level measured at d36 was half-way between the values at d29 and d38. For these reasons, the hormonal profiles might depend, not on iterative immunization, but on the time elapsed after prime immunization. This explanation seems more in agreement with the assumption of retarded arousal of ACTH and CORT responses (after d29). The concomitance of the peaks of CORT and IgG (d38) can be intrepreted as part of an immuno-neuroendocrine regulatory loop: the activation of B-cells, in conjunction with T-cell activation, induces the release of cytokines which increase glucocorticoid level via the HPA axis (Besedovsky et al., 1981; Lilly and Gann, 1992; Hu et al., 1993); this is a possible explanation to the peaks of ACTH and CORT observed at d38. This circuit is enhanced by MLT (Maestroni, 1993) which have been demonstrated by us to crest contemporaneously with pituitary-adrenal hormones. In turn, increased glucocorticoid levels inhibit MLT rise (Besedovsky et al., 1975; Poon et al., 1994), and also impede cell-mediated immunity, leading to both reduction of cytokine levels and decline of CORT level (Lilly and Gann, 1992; Hu et al., 1993; Besedovsky et al., 1985) as observed by us at d47. Finally, iterative immunization seemingly played the role of amplifying endocrine and immune responses.

                However, the HPA axis and the pineal gland not only mediate adaptation to nociceptive situations, but they are also sensitive and responsive to light stimuli which acts as zeigteber on biological rhythms in most Vertebrates (Binkley, 1993). Therefore, we wondered if the bursa of Fabricius possibly influences the circadian rhythms of both HPA and pineal gland in the chicken.

 

4.4. BURSECTOMY AND CIRCADIAN RHYTHMS OF HPA AXIS AND PINEAL GLAND

 

The circadian activity of the pituitary-adrenal axis is governed by the rhythmic activity of CRF-containing neurons of the hypothalamus (Peczely and Antoni, 1984; Jozsa et al., 1984). These CRF-neurons project in the median eminence, in the vicinity of pituitary portal capillaries where CRF is rhythmically released (Peczely and Antoni, 1984; Jozsa et al., 1984) and triggers the circadian secretion of ACTH and downstream CORT. In the present study, bursa-intact chickens exhibited broad circadian rhythms of ACTH and CORT under a 12L-12D light-dark cycle. Some reports indicate that lesioning the hypothalamic paraventricular nucleus in the rat (Makara et al., 1981) or the suprachiasmatic nucleus in the pigeon (Bouillé and Baylé, 1973) abolishes the circadian variations of plasma ACTH and CORT. Now the withdrawal of the bursa anlage from 4 day-old chicken embryos was also accompanied by the suppression of the daily rhythms of plasma ACTH and CORT in young chickens, despite exposure to a 12L-12D light-dark regime.

                The daily synthesis and release of MLT is catalyzed in vivo  by two key enzymes, NAT and HIOMT (Binkley et al., 1973). In 8 week-old chickens raised under 12L-12D light-dark cycle, pineal NAT and MLT night-time levels are respectively 27 and 10 fold higher than light-time values, whereas HIOMT activity varies by 20 % only (Binkley et al., 1973). Likewise, ocular and blood MLT levels display prominent daily rhythms with a midscotophase crest (Cogburn et al., 1987). Our control chickens exhibited similar patterns of enzymatic and hormonal rhythms: night/daytime ratios were 20, 8, 2 and 10 for pineal NAT activity, pineal MLT content, and ocular and circulating MLT, respectively. Little change occurred in HIOMT activity, probably due to a slow turnover of this enzyme (Voisin et al., 1988). Very striking is the observation that in bursectomized chickens, the magnitude of NAT and MLT rhythms was markedly reduced, in spite of persistent cyclicity. Phototransduction is anatomically supported by a complex network comprising several elements: photoreceptors (retinal, hypothalamic, pineal) perceive light information and convey it either directly or via the suprachiasmatic nucleus (SCN) to the pineal gland where the message is converted into MLT (Binkley, 1993). The visual suprachiasmatic nucleus (equivalent of mammalian SCN) is connected to the pineal gland through sympathetic catecholaminergic fibers and governs the circadian release of noradrenaline in the vicinity of pinealocytes (Cassone et al., 1990). Noradrenaline turn-over is higher during the photo- compared to scotophase (Cassone et al., 1990). The activation of a-2 adrenergic receptors on pinealocytes upon daytime increase in noradrenaline level inhibits the synthesis of NAT molecules and, therefore, MLT production, via an intracellular increase of cAMP (Voisin and Colin, 1986; Bylund et al., 1988). Since the amplitude of pineal NAT activity and MLT production is controlled by the rhythmic delivery of noradrenaline, itself driven by the circadian function of the vSCN (Cassone et al., 1990), the vSCN-SCG-pineal complex appears as the most probable locus of interferences between bursal signals (or the absence of bursal signals) and pineal rhythms. This hypothesis is strengthened by the fact that the excision of the bursa anlage (4 th day of embryonic life) preceeded the beginning of pineal sympathetic innervation which occurs by day 20 of embryonic life (Wight, 1971), as well as the onset of pineal NAT and MLT rhythms which are observed in 18 day-old embryos (Binkley and Geller, 1975). Similarly, the chronology of the onset of bursal and pituitary-adrenal functions might help understanding how the precocious influence of the bursa of Fabricius on pituitary-adrenal functioning occurs: the bursa anlage arises by the third day in the embryo (Toivanen et al., 1981); it was excised (4 th day of embryonic life) prior to the onset of the HPA axis functioning which in the developing chicken embryo occurs during the second half of embryonic phase (Scanes et al., 1987). Therefore, it is likely that as a result of precocious bursectomy, the normal development of pituitary-adrenal activities was hindered, due to the lack of adequate inductive bursa cue (s).

                However, the exact locus of the interferences between bursa signals and both the HPA and the pineal gland have yet to be identified. The studied systems (bursa of Fabricius, HPA and pineal gland) arise during the embryonic phase which in chickens is also the critical period of sexual differentiation of the brain (Wilson and Glick, 1970). Accordingly, the mechanisms of brain development are likely to be involved in the functional maturation of bursa-pituitary-adrenal-pineal interrelationships. In fact, during the critical period, sex steroids can imprint parts of the brain that are associated with the cyclic control of hormonal release in such a way as to prevent their rhythmic function (Pilgrim et al., 1994). Since embryonic bursectomy leads to significantly increased plasma testosterone levels (Pedernera et al., 1980; Ramade et al., 1986), such a rise might affect the cyclic activity of hypothalamic pacemakers. Among the latter is the SCN neurally coupled to CRF-containing nuclei which govern the circadian release of ACTH and CORT (Peczely and Antoni, 1984; Jozsa et al., 1984) and to the pineal gland which secretes MLT in a cyclic manner (Binkley, 1993). As working hypothesis, it can be assumed that similar mechanisms occur in bursectomized embryos in the absence of specific bursal cue (s); in intact embryos, such signal (s) must normally prevent the unsettling of the hypothalamic and pineal pacemakers.

                These dyschrogenic effects emphasize the crucial role held by signals originating in the bursa microenvironment in the ontogeny and functioning of immune, HPA and pineal activities. Further support to this assertion comes from the fact that bursectomized chickens embryonically grafted with 9 day-old embryonic bursae recover normal immune responsiveness  and normal levels of plasma CORT (Abdul-Karim et al., 1987) and testosterone (Pedernera et al., 1980). In this regards, we demonsrated that the bursa-derived signal termed bursin is an important mediator of bursa-dependent functions.

 

4.4. BURSIN AS MEDIATOR OF BURSA-DEPENDENT FUNCTIONS: EFFECTIVENESS OF HIGHLY DILUTE BURSIN

 

Differents concentrations of bursin were tested, aimed at reversing the effects of bursectomy. In ovo  administration of very low doses (5 CH, 7 CH) of bursin to bursectomized chickens allowed recovery of the ability to cope with ether stress, antigen challenge and restored normal pituitary-adrenal and pineal rhythmicity. Chorio-allantoic grafting of bursal rudiment from 9.5-day-old donor embryos has been reported to alleviate gonadal (Pedernera et al., 1980) and adrenal (Abdulkarim et al., 1987) changes caused by bursectomy. Also bursa extracts have ben reported to restore specific humoral immunity in bursectomized chickens (Baba and Okuno, 1976). Kuznik et al. (1988) demonstrated that peptides from bursal origin correct both immunity and homeostasis in bursectomized chickens. Increasingly, the data strengthen the assumption of specific inductive properties for minute doses of bursin on immune and neuroendocrine functions during embryonic life: native as well as synthetic bursin (Lys-His-Gly-NH2) stimulates an intracellular increase of cAMP and cGMP in the human Daudi B cell line, at threshold concentrations of 0.1 to 1 µg/ml (Audhya et al., 1986). On the other hand, either the inverse sequence of bursin (Gly-His-Lys-NH2) or other variants of this tripeptide (Gly-His-Lys; Lys-His-Gly) appear poorly inductive, and only when administered in massive amounts (Audhya et al., 1986; Lassila et al., 1989), which makes this action physiologically irrelevant. Using either the saline or a control tripeptide with unknown biological activity (Trp-Leu-Leu-NH2), we could not overcome the effects of bursectomy; in ovo administration of 100 µg of bursin was no longer effective. Hence, bursin worked in an inverse dose-dependent manner, the lower doses being the most effective. Underlying mechanisms remain unraveled. Bursin triggers intracellular increase of cAMP and cGMP (Goldstein et al., 1977; Audhya et al., 1986). Like homologous tripeptides (Gly-His-Lys, His-Gly-Lys) known to enhance cell growth, bursin contains histidine which binds to and promote copper uptake in a variety of cells (Pickart and Thaler, 1973; Pickart, 1981). Bursin is likely a growth factor (Lassila et al., 1989).

                Much more striking is that highly dilute bursin (15-20 CH pool) improved the performances of bursectomized chickens, to the same extent as did 7 CH of bursin. Highly dilute bursin seemed to transmit specific "information" to bursectomized recipients. Bursin's specific inductive properties were ascertained by replicating the data pertaining to the immunization protocol: the previous results (Youbicier-Simo et al., 1993) were exactly reproduced (Youbicier-Simo et al., 1996b). Furthermore, highly dilute bursin wholly restored normal circadian rhythms of the pituitary-adrenal axis (Youbicier-Simo, 1994; Youbicier-Simo et al., 1996b) and pineal gland (Youbicier-Simo, 1994; Youbicier-Simo et al., 1996a). Another striking observation is that the pharmacological effect induced by the saline in saline-supplied chickens (N+S) was suppressed after embryonic bursectomy: in figure 4 B, the bursa-intact chickens supplimented with the saline (N+S) exhibited a slightly higher CORT response than their untreated counterparts; on the other hand, this effect did not occur when bursa-lacking chickens (Bx) received the saline (Bx+S). How do highly dilute solutions work remains enigmatic, since they are theoretically devoid of active material compounds. It has been shown by nuclear magnetic resonance that the structure of high dilutions is different from that of solvents (Demangeat et al., 1992). Several theories have been put forward, but a deciding factor of the efficiency of high dilutions seems to lie in the way such media are prepared, using sequential centesimal dilution steps with potent vertical stirring between successive steps. We demonstrated that in sequentially diluted 3H-thymidine solutions, radioactivity decayed linearly, without showing any difference between succussed and unsuccussed solutions (table 2). This testifies to a homogeneous molecular dispersion as result succussion, therefore precluding the assumption of heterogeneous molecular distribution sustained by the theory of clusters. We suggest that a non-molecular specific information signal corresponding to the original molecule (bursin) is conveyed to the recipient organism (Bastide and Lagache, 1992). Indeed, the activity of high dilutions is sensitive to electromagnetic fields (Hadji et al., 1992) and had been transferred to unshaken water through a coil (Endler et al., 1995) and may be probably electromagnetically transmitted. Currently, an increasing number of experiments aimed at substantiating the biological activities of high dilutions are under way, and some reports are compelling (Bastide et al., 1985; Bastide et al., 1987; Daurat et al., 1988; Jacobs et al., 1994; Reilly et al., 1994; Bastide and Boudard, 1995).

            Finally, using a relevant experimental model, we have provided evidence that a compound found in the bursal microenvironment can functionally mimic in vivo  some biological functions of the bursa Fabricii. We assert that in chickens, the bursa-derived signal (bursin) which is probably an ontogenic organizer, mediates early bursa-dependent functions. The originality and novelty of the present work lie in the underscoring of the inverse dose-dependent effectiveness of high dilutions, in contrast with the classical dose-dependent effect which underlies the molecularist theory. We introduce a new conceptual trend of the information processing termed the "paradigm of signifiers"; the latter is rather based on the encoding of the "specific information" contained in bioactive compounds by highly dilute media.

 

 

Aknowledgement.

The present work was supported by grants from Dolisos Laboratories (Paris).

 

 

5. References

 

Abdul-Karim, A.K., Ramade, F., and Baylé, J.D. (1987) Development of basal and stress-induced adrenocortical function in early bursectomized embryos and chickens, Life Sci. Adv.  6, 121-128.

Audhya, T., King, R., and Goldstein, G. (1991) Bovine probursin tetradecapeptide contains amino acid sequence from somatostatin, tuftsin and bursin, Life Science  48, 773-780.

Audhya, T., Kroon, D., Heavner, G., Viamontes, G., and Goldstein, G. (1986) Tripeptide structure of bursin, a selective B-cell differentiating hormone of the bursa of Fabricius, Science  231, 997-998.